diving for science 2007 - Rubicon Research Repository

Transcription

Rubicon Research Repository (http://archive.rubicon-foundation.org)
PROCEEDINGS
DIVING FOR SCIENCE
2007
DIVING FOR SCIENCE 2007
AAUS
Proceedings of the
. Pollock & Godfrey
American Academy of Underwater Sciences
26th Scientific Symposium
Diving For Science 2007 Proceedings Of The American Academy Of Underwater Sciences
Diving For Science 2007
Proceeding of the
American Academy of Underwater Sciences
26th Scientific Symposium
Neal W. Pollock and Jeff M. Godfrey
Editors
University of Miami
March 9-10, 2007
i
The American Academy of Underwater Sciences (AAUS) was formed in 1977 and incorporated in
the State of California in 1983 as a 501c6 nonprofit corporation. Visit: www.aaus.org.
The mission of AAUS is to facilitate the development of safe and productive scientific divers through
education, research, advocacy, and the advancement of standards for scientific diving practices,
certifications, and operations.
Acknowledgments
Thanks to Rick Riera-Gomez, Alma Wagner and the following staff and students of the University of
Miami and the Rosenstiel School of Marine and Atmospheric Science for hosting the 2007
symposium of the American Academy of Underwater Sciences:
Luke Logan
Marina Nunes
Monte Shalett
Joe Tomoleoni
Marylin Brandt
Viktor Brandtneris
Adam Chasey
Megan James
Special thanks to Kathy Johnston for her original artwork which continues to fund AAUS
scholarships and appears on the cover of this publication (www.kathyjohnston.com).
Sea Grant Publication # CTSG-07-14.
ISBN# 0-9800423-1-3
Copyright © 2007 by the American Academy of Underwater Sciences
Dauphin Island Sea Lab, 101 Bienville Boulevard, Dauphin Island, AL 36528
ii
Table of Contents
Bottomfish Variability in the Proposed Marine Reserves of Skagit County, Washington
J. Henry Valz and Paul A. Dinnel ................................................................................ 1
Baseline Survey Protocol
Hannah L. Markham and Nicola K. Brown ............................................................... 13
AAUS Diving Officer and Scientific Diver Certifications: The Need for Quality Control
Michael A. Lang, Glen H. Egstrom, and Christian M. McDonald............................. 23
NOAA Test and Evaluation of Two, Commercial-Off-the-Shelf, Multi-Gas Dive Computers for
Providing Accurate Depth Measurements and Acceptable Mixed Gas and Air
Decompression Schedules
J. Morgan Wells and David A. Dinsmore .................................................................. 27
Physical Fitness of Scientific Divers: Standards and Shortcomings
Alison C. Ma and Neal W. Pollock ............................................................................ 33
Scientific Diving Safety: Integrating Institutional, Team and Individual Responsibility
Neal W. Pollock ......................................................................................................... 45
Using SCUBA and Snorkeling Methods to Obtain Model Parameters for an Ecopath Network
Model for Calabash Caye, Belize, Central America
Rebecca A. Deehr, Deirdre B. Barry, David D. Chagaris,
and Joseph J. Luczkovich........................................................................................... 51
Submerged Cultural Resource Management on the Last Frontier: Reconnaissance, GIS
Mapping,and Biotic/Geochemical Characterization of Threatened Shipwreck Sites in
Southeast Alaska
J. David McMahan, John O. Jensen, Stephen Jewett, John Kelley,
Sathy Naidu, Hans Van Tilburg, and Michael Burwell.............................................. 69
Video iPod Instructional Design Considerations for Dive Training and Underwater Subject
Matter
Michael Dermody and Calvin Mires .......................................................................... 83
Closed-circuit Rebreathers in the Forensic Study of the Rouse Simmons Shipwreck
Gregg Stanton, Keith Meverden, Tamara Thomsen and James Garey ...................... 89
Rebreather Fatality Investigation
Richard D. Vann, Neal W. Pollock and Petar J. Denoble ........................................ 101
Pressure Related Incidence Rates in Scientific Diving
Michael R. Dardeau and Christian M. McDonald.................................................... 111
When Everything Goes Right: Implications for Scientific Diving Safety Programs
Vallorie Hodges........................................................................................................ 117
iii
Behavior and Sound Production by Longspine Squirrelfish Holocentrus rufus During Playback of
Predator and Conspecific Sounds
Joseph J. Luczkovich and Mark Keusenkothen ........................................................127
Deploying Benthic Chambers to Measure Sediment Oxygen Demand in Long Island Sound
Prentiss H. Balcom, Jeff M. Godfrey, Diane C. Bennett, Gary A. Grenier,
Christopher G. Cooper, David R. Cohen, Dennis A. Arbige
and William F. Fitzgerald .........................................................................................135
An Evolution of Scientific Mixed Gas Diving Procedures at the National Park Service Submerged
Resources Center
Jeffrey E. Bozanic.....................................................................................................143
Diving at Extreme Altitude: Dive Planning and Execution During the 2006 High Lakes Science
Expedition
Robert Morris, Randy Berthold and Nathalie Cabrol ...............................................155
Long Term Monitoring of a Deep-water Coral Reef: Effects of Bottom Trawling
John K. Reed.............................................................................................................169
Measuring Structural Complexity on Coral Reefs
Anders Knudby and Ellsworth LeDrew....................................................................181
Reef Status Protocol (RSP): A Prognostic Reef Survey Methodology, Rapid Yet Comprehensive
Hannah L. Markham and Nicola K. Browne ............................................................189
Speaker Schedule ................................................................................................................................196
iv
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Of Underwater Sciences
Proceedings of the American Academy of Underwater Sciences 26th Symposium.Dauphin Island, AL: AAUS; 2007.
Bottomfish Variability in the Proposed Marine Reserves of
Skagit County, Washington
J. Henry Valz* and Paul A. Dinnel
Shannon Point Marine Center, Western Washington University
1900 Shannon Point Rd., Anacortes, WA 98221
henryvalz@yahoo.com; padinnel@aol.com
*corresponding author
Abstract
Among groundfishes, rockfishes (genus Sebastes) are some of the most over-fished species in
the United States. Populations in Puget Sound have declined rapidly over the last century; some
rockfish species have 10% or less of their historical reproductive output. This population
decrease is paralleled by declines in physical size. To address this problem, the Skagit County
Marine Resources Committee has proposed establishing one or more no-take marine reserves to
protect rockfishes and other groundfishes. The majority of studies of groundfish populations in
the northeast Pacific have focused on summer distributions. The goal of this study was to relate
intra-annual dynamics of groundfish, especially rockfishes, to physical and biological factors to
assess the suitability of individual sites as reserves. Twenty-four dives were performed over one
year at six sites under consideration as Skagit County marine reserves. Each dive consisted of
eight 25-m transects where fish number and size, plus bottom composition were measured.
Initial results show that changes in season were associated with changes in densities of the most
abundant groundfishes (copper rockfish, kelp greenling, and lingcod). Depth and site also had a
significant effect on densities of copper rockfish, which was the most common rockfish
observed. Because of the large seasonal trends and effects of physical factors, more diverse
sampling regimes should be used rather than the common summer sampling schedules.
Additionally, larger-sized and deeper reserves should be considered to compensate for the
seasonal migration of groundfishes.
Introduction
Fishery stocks have declined to the point where 29% of worldwide stocks have collapsed (Worm et
al., 2006). Using different measurements, the United Nations' Food and Agricultural Organization
suggested that 69% of the world's marine stocks were either fully exploited or depleted (Tuya et al.,
2000).
Fish that live on reefs have been especially affected by overfishing and alterations to the reefs
(Parrish, 1999). Benthopelagic and benthic fish that have moderate to little movement have been
particularly affected by trawl-based fisheries (reviewed in Parris, 1999; Weispfenning, 2006).
Fish of the genus Sebastes, commonly called rockfish, are some of the most overfished and depleted
species in the United States (Love et al., 2002). Formerly commercially viable species such as
boccacio and blue rockfish have had their catches diminished by 99 and 95%, respectively (Love et
al., 2002). Over the last four decades the average size of several species of rockfish have fallen
significantly (Mason, 1998).
Rockfish populations in Puget Sound have declined to where some rockfish species have less than
10% of their historical reproductive output (Palsson, 1998). Associated with this decrease in
1
reproductive output are declines in size and abundance (Palsson, 2001). In response to these declines,
the Washington Department of Fish and Wildlife (WDFW) gradually reduced bag limits from 15 to a
bag limit of one rockfish per day (Palsson, 1998). However, because rockfish are often bycatch for
the recreational lingcod fishery, and do not recover from barotrauma, they are likely far more affected
by fishing than this bag limit represents.
While rockfish populations have suffered over the last 20 years, lingcod (Ophiodon elongatus) and
kelp greenling (Hexagrammos decagrammus) populations in Skagit County have increased (Moulton,
1977; Weispfenning, 2006). These shorter-lived fish are also predators of both larval and adult
rockfish. Their recovery may be hampering the recovery of the long-lived and slower reproducing
rockfish. While even the shortest-lived rockfish take several years to reach reproduction, male lingcod
are reproductive after two years, at the size of 50 cm. Kelp greenlings also grow and mature quickly,
reaching reproductive age in two to three years (Eschmeyer et al., 1983). Both of these fish prey on
immature rockfish, especially in their pelagic and young-of-the-year (YOY) forms.
Marine Reserves
In order to protect marine species governments around the world have instituted areas where harvest
of organisms is prohibited or severely limited. These marine reserves are designed to conserve marine
species and habitat for research, posterity, and replenish non-reserve areas for future commercial and
recreational fisheries. In some areas the restrictions are limited to a few species, while other
governments have expanded them to include all living organisms in the designated area.
Theoretically, these reserves will protect fish and invertebrates that live within the boundaries,
creating a large broodstock that will reach, or approach, pre-exploitation levels. These broodstocks
will then export excess larvae into areas still fished, causing increases in available stocks outside of
the marine reserves. This is referred to as the 'reserve effect' (reviewed in Yoklavich 1998,
McConnell et al., 2001).
The effects of marine reserves have been tested in tropical ecosystems. Studies have shown both an
increase in fish density inside the reserves, as well as increased fisheries catches outside of the
reserves (Roberts et al., 2001; Russ and Alcala, 1996)
In the northeast Pacific few studies have been completed to show if marine reserves are producing a
reserve effect. Palsson (1998) showed much higher fish densities in five protected areas in Puget
Sound. Also observed was a 55-fold increase in reproductive output of the area, compared to
unprotected areas with similar bottom composition. Paddack and Estes (2000) compared three
reserves and three unprotected areas in kelp forests in coastal California. They found that fish in the
reserves were larger and that densities increased 12-35% in the protected areas. Both studies showed
that older reserves have larger effects than newer reserves.
Different marine regions and ecosystem types are protected to varying degrees. Spalding and others
(2006) estimated that 23% of coral reefs receive some level of protection, while only four percent of
the areas of the sea 200 m or shallower receive some protection. However, Mora et al. (2006)
described only 0.01% of the world's coral reefs as fully protected from poaching, overfishing, coastal
development, and pollution.
By comparison less than three percent of north Puget Sound and Strait of Juan de Fuca shorelines are
protected. When looking at subtidal area, far less than one percent of that area is protected. Within
Skagit County less than a hundredth of the shoreline is protected, with no subtidal areas specifically
designated for protecting groundfish.
2
Recommended sizes for the reserves depend on the goal of each reserve system. For minimal
protection, in order to allow the perpetuation of species for heritage and research, Yoklavich (1998)
recommended a size up to five percent of the habitat. The recommendations for protecting target
species for insurance against overfishing with current management rules ranges from five to 20%
(Yoklavich, 1998). Palsson (2001) sets a target of 20 percent for Puget Sound. In order to create an
alternative to the current management strategies, marine reserves would have to be 20 to 50% of the
available habitat of the fished species (Yoklavich, 1998). However, Parrish (1999) argued that these
large reserves may actually have deleterious effects for those species the reserves are supposed to aid.
It should be noted that the author used a spawning biomass that is less than the current biomass of
Puget Sound groundfish. The species used are also less sedentary, therefore making them less ideal
for protection than those in most need of help in Puget Sound.
Marine reserves must also occur in diverse enough habitats and areas to protect the majority of the
life stages of the organisms they are trying to protect. Yoklavich (1998) recommended protecting the
vast majority of the movement of the organisms that a marine reserve is intended to protect. This
means that for several species which migrate either seasonally, or at some point during their life, a
large area, which is also diverse, must be included. For example, yellowtail rockfish (S. flavidus) in
Washington State migrate from the north Puget Sound to the open coast as they reach maturity (Love
et al., 2002). If a marine reserve were to only protect the adult phase, the protection would likely be
inadequate, due to fishing pressure on the juvenile form in Puget Sound.
Research Objectives
The primary goal of this research is to describe the variability of rockfish in the proposed marine
reserves of Skagit County, and how that variability relates to season, bottom composition, and site.
Additionally, by comparing the fish densities from previous studies done in Skagit County (Moulton,
1977; Weispfenning, 2006), estimations of inter-annual variability can be refined.
Season has been shown to affect the movement of rockfish as well as other groundfish (Moulton,
1977; Love et al., 2002). It is believed that the fish migrate to avoid changes in water temperature,
and to also avoid the large storm surge that affects shallower water. However, it is not well
established how great the seasonal migration is in protected areas such as Skagit County, where the
storm surge is not as great.
There have been very few studies done throughout an entire year, even though previous publications
have mentioned that seasonal studies need to be completed (Jagielo et al., 2003).
This study is significant, especially at the local level, where the sizes and locations of the proposed
marine reserves might be reconsidered. If there is found to be a large effect of season, where species
move to greater depths during parts of the year, additional areas might need to be considered, in order
to protect the intended species. While a small seasonal study has been completed in the area, the study
did not include the proposed marine reserves. Additionally, this study was completed before a large
population shift occurred. Formerly, long-lived prey dominated, however now the more common
species are short-lived predators (Moulton, 1977; Weispfenning, 2006).
Our hypotheses for this research are:
1. Groundfish densities will change throughout the year in the surveyed areas. Groundfish densities
will be lowest in winter, when the large surge and colder temperatures of winter storms will drive
fish out of the relatively shallow depths of the survey areas.
3
2. Groundfish will be found in higher densities at the deeper transects. This will confirm previous
research locally (Weispfenning, 2006) and in California (Love et al., 2002).
3. The effect of depth on the density of groundfish species will change with the seasons. The lowest
densities in shallow water will be recorded during the winter, and the highest densities will be
recorded during the summer.
4. Bottom composition, intra-annual (season), YOY habitat, bottom algal composition, and visibility
will describe the majority of the variance seen in rockfish populations.
Methods
Seasonal Surveys
Twenty-four underwater visual strip transect surveys following the methods of Weispfenning (2006),
adapted from Eisenhardt (2001), were used to determine changes in groundfish densities at six sites in
Skagit County, Washington (Figure 1). A pair of divers completed an 800 m2 survey at each site
during each season starting in the fall of 2005 and ending in the summer of 2006. The six sites existed
within five of the eight marine reserves proposed by the Skagit County Marine Resources Committee
(McConnell et al., 2001; McConnell and Dinnel 2002). The six sites surveyed were Burrows Channel
(BC), northwest Allan Island (AI), Dennis Shoal (DS), South Cone Island (CI), east Strawberry Island
(SI), and Towhead Island (TI) (Figure 2). Sites were chosen from within the proposed marine reserves
on the basis of similar topography (boulder/wall with a moderate slope) and the known presence of
bottomfish based on previous research by Weispfenning (2006).
Figure 1. Diagrams of survey technique. Visual strip transects using open-circuit SCUBA were employed to
determine groundfish densities along a 25 m wide, 4 m tall, 4 m wide transect. The transects were separated into
four shallow (30-45 ft) and four deep (45-65 ft) transects. The first diver recorded groundfish counts as well as
total length to 5 cm accuracy using a dual laser system. The second diver recorded bottom composition and
coverage data.
4
The Burrows Channel site surveys started at 48°29.451'N and 122°41.872'W and progressed west
along the south-facing slope of Fidalgo Island near Washington Park. This site exists within the
proposed Burrows Channel marine reserve. It is heavily fished during the lingcod season from boats
and shore, but is closed for salmon fishing from July 1 to September 30, during the peak season.
The northwest Allan Island site
surveys started at 48°28.097'N
and
122°42.552'W
and
progressed northwest along the
underwater reef at the tip of the
island. This site exists within the
proposed
Allan
Island,
Williamson Rocks and Dennis
Shoal marine reserve. The area is
fished during the lingcod season
from May 1 to June 15 by anglers
and spearfishers, but is closed to
salmon fishing from July 1 to
September 30, during the peak
salmon fishing season. It is also
close to a small seal haul-out on
the west side of Allan Island
(Banks, WWU, pers. comm.).
The Dennis Shoal site surveys
started at 48°27.545'N and
122°48.880'W and progressed
north along the west-facing slope
of the underwater reef, eventually
curving towards the east. Like the
northwest Allan Island site, this
site is also within the proposed
Allan Island, Williamson Rocks
and Dennis Shoal marine reserve.
This site is fished during the
lingcod season but is closed to
salmon fishing. It is also near
Williamson Rocks, which is part
of the San Juan National Wildlife
Refuge (Site #81), which is part
of the San Juan Islands National
Wildlife Refuge and is a large
haul-out area for seals and stellar
sea lions.
Figure 2. Sites sampled among the eight marine reserves proposed by
Skagit Marine Resources Committee. Five of the eight proposed
reserves were sampled, three near Cypress Island, and three near
Fidalgo Island. Sites were selected for similarity of bottom composition
and availability for diving. Map altered by author and original courtesy
of the Skagit Marine Resources Committee.
The South Cone Island site surveys started at 48°35.437'N and 122°40.921'W, and progressed south
along the west-facing depth contour of the underwater reef south of the island, eventually turning east
and then north. This site is within the proposed Cone Islands reserve and also lies within the
Washington Department of Natural Resources (DNR) Cypress Island State Aquatic Reserve. The
designation as a DNR reserve give it no protection from fishing in subtidal areas. The site is fished
during lingcod and salmon seasons.
5
The east Strawberry Island site surveys started at 48°33.684'N and 122°44.079'W and continued south
along the east-facing slope of the island. This site is within the proposed Strawberry Island reserve
and also lies within the DNR Cypress Island State Aquatic Reserve. This site is fished during lingcod
and salmon seasons.
The Towhead Island site surveys started at 48°36.776'N and 122°42.916'W, and continued east,
progressively turning south, along the north-east facing curve of the island. This site exists within the
proposed North Cypress Island, Towhead Island, and Cypress Reef reserve. It also exists within the
DNR Cypress Island State Aquatic Reserve. This site is heavily fished during lingcod and salmon
seasons.
The surveys were completed seasonally at least 22 days after the change of season, and at least 16
days before the start of the next season. This gave at least 38 days between each set of surveys, in
order to maximize the differences observed between each season. Fall surveys were conducted from
October 24, 2005 to December 2, 2005. Winter surveys were conducted from January 12, 2006 to
February 27, 2006. Spring surveys were conducted from May 2, 2006 to June 5, 2006. Summer
surveys were conducted from July 18, 2006 to August 17, 2006.
Dive surveys were only performed when current was predicted to be less than 0.2 knots for 45
minutes or longer, as predicted by Tides and Currents Pro 3.0 (Nobeltec). Surveys were completed on
high-slack tides, after 1200 Pacific Standard Time and 1300 during Pacific Daylight Savings Time.
Surveys usually took between 25 and 35 minutes.
The surveys were 200 m long, 4.0 m wide, and 4.0 m in height (Figure 1). Each survey was divided
into eight transects of 25 m. The first four transects completed were between 45 and 65 ft (13.7 and
19.8 m). These were identified as the 'deep' transects. The second four transects completed were
between 30 and 45 ft (9.1 and 13.7 m). These were identified as the 'shallow' transects. In each 25 m
transect, the first diver would count and measure groundfish, using a 20-cm wide dual-laser
(Lasermate, Inc.) measuring device, to the nearest 5.0 cm.
The second diver followed the first diver and estimated bottom composition and biotic bottom
coverage. Bottom composition was separated into four categories: Sand and Gravel (under 3 cm in
diameter), Rock (3-30 cm), Boulder (30 cm-3 m), and Wall (greater than 3 m in diameter). Bottom
coverage of all organisms (algae and fauna) was visually estimated as a percentage of the 4.0 m wide,
25 m long transect. All data were recorded using an underwater slate and pencil.
During the summer of 2006, short young-of-the-year (YOY) rockfish transects were used following
the general strategy of Hayden-Spears (2005) in water 3-6 m deep. The first diver swam
approximately 30 fin strokes disturbing shallow flat-bladed brown macroalgae while looking for
YOY rockfish. The second diver estimated the percent coverage of different types of brown algae
during the swim. These data were also recorded underwater on dive slates.
Additional data were also recorded during and after the dive: time in and out of the water, perceived
currents by the divers, surface and at depth underwater visibility, weather conditions, swell size and
timing, cloud cover, marine life observed near site, estimated wind speed, tidal heights, predicted
currents, precipitation, and measured wind speed at a nearby site. These were recorded for future
analysis as covariates that may have affected rockfish densities.
Analysis
Analysis of variance with repeated measures (ANOVAR) was completed using SPSS version 12.0
(SPSS, Inc.). Huynh-Feldt corrections were made for all p values, when departures from sphericity
6
occurred. A posteriori contrasts were made controlling alpha at the family level using a Scheffé
correction.
Results
The average density of all groundfish observed over all sites was 284.4 fish⋅ha-1. Of this, 47.4 fish⋅ha-1
were lingcod (Ophiodon elongates), 83.3 fish⋅ha-1 were kelp greenling (Hexagrammos
decagrammus), 46.9 fish⋅ha-1 were copper rockfish (Sebastes caurinus), 95.3 fish⋅ha-1 were Puget
Sound rockfish. The remaining fish were a small number of cabezon, quillback rockfish, wolfeel and
red Irish lords.
140
Copper rockfish density (fish/ha)
Seasonal Effects
The effects of season on
groundfish species were
different
for
different
species. The most common
rockfish in this study, the
copper rockfish densities
was significantly affected
by season (Huynh-Feldt
corrected
ANOVAR,
p=0.007). During fall the
average density of copper
rockfish was 52.1 fish⋅ha-1
(Figure 3). The density then
decreased five-fold to 10.4
fish⋅ha-1 in winter. In spring,
densities increased to 35.4
fish⋅ha-1, and in summer the
density increased to its
highest density of 89.6
fish⋅ha-1. Season did not
interact with either of the
two other factors (depth and
location).
120
100
80
60
40
shallow
deep
20
0
Fall
Winter
Spring
Summer
Season
Figure 3. Densities of copper rockfish (S. caurinus) separated by season and
depth. Both depth and season had a significant effect on copper rockfish
density. Densities in deeper water (45-65 ft) averaged 90% higher than those
in shallow water (30-45 ft). The greatest density of rockfish was during the
summer, when densities reached 96 fish⋅ha-1. Error bars represent ±1
standard error.
The most common groundfish observed in the study was kelp greenling. The effect of season on
density was significant for kelp greenling as well (Huynh-Feldt corrected ANOVAR, p<0.001).
Density of kelp greenling started at 122.9 fish⋅ha-1 in fall (Figure 4). Like copper rockfish, the density
fell in winter, to 43.5 fish⋅ha-1. The density then rose in spring to 58.3 fish⋅ha-1. The density finished
at 108.3 fish⋅ha-1 in summer. Season did not interact with either of the other two factors
.
Lingcod had a similar trend during the year as kelp greenling. The effect of season was significant for
lingcod as well (Huynh-Feldt corrected ANOVAR, p=0.021). Density of lingcod was highest in fall,
at 70.8 fish⋅ha-1 (Figure 5), but then decreased in winter to 29.2 fish⋅ha-1. It then increased to 35.4
fish⋅ha-1 in spring, and then increased further to 54.2 fish⋅ha-1 in summer. There was an interaction
between season and depth (p=0.027); however, this was a result of large decreases in densities from
fall to winter at northwest Allan Island and Dennis Shoal.
7
Depth effects
Of the three common groundfish, depth had only a consistent effect on copper rockfish (Figure 3).
For this species, deeper sites consistently had higher densities (between-subject ANOVAR, p=0.035)
through all seasons. Transects in deeper water had a 90% greater density of copper rockfish than
shallower water (32.3 vs. 61.4 fish⋅ha-1). For kelp greenling and lingcod, depth did not have a
significant effect on density (Figure 4). There was no significant interaction for depth with location or
season for all species. However, for both kelp greenling and lingcod, there were much higher
densities in shallower water compared to deep water in spring (Figures 4 and 5).
Kelp greenling density (fish/ha)
180
160
140
120
100
80
60
40
shallow
20
deep
0
Fall
Winter
Spring
Summer
Season
Figure 4. Density of kelp greenling (H. decagrammus) separated by season and depth. Season had a significant
effect on kelp greenling density (p<0.001), but depth did not (p>0.05). While depth was not significant as a
whole, an A posteriori pairwise contrast between the deep and shallow sites during spring showed a significant
change in density (p=0.0033, Sheffé corrected alpha = 0.0125). Error bars represent ±1 standard error.
100
Lingcod density (fish/ha)
90
80
70
60
50
40
30
20
shallow
10
deep
0
Fall
Winter
Spring
Summer
Season
Figure 5. Density of lingcod (O. elongates) separated by season and depth. Season had a significant effect on
kelp greenling density (p=0.021), but depth did not (p>0.05). An a posteriori pairwise contrast between the
deep and shallow sites during spring showed non-significant change in density, however there is a strong
trend, with the lack of power in this initial analysis (p=0.07441, Sheffé corrected alpha = 0.0125). Error bars
represent ±1 standard error.
8
Location (site) effects
Location was only found to cause a significant difference in density for copper rockfish (Figure 6).
The three sites near Burrows Bay and Fidalgo Island were found to have significantly lower densities
than the three sites that surrounded Cypress Island (between-subject ANOVAR, p=0.004). Both kelp
greenling and lingcod were not significantly affected by location effects (p=0.225 and p=0.058,
respectively).
Copper rockfish density (fish/ha)
120
100
80
60
40
20
0
Burrows Pass
NW Allan
Dennis Shoal Strawberry Is
Cone Is.
Towhead Is.
Site
Figure 6. Density of copper rockfish (S. caurinus) at six sites in five of the proposed reserves. Sites with
unshaded bars were those closest to Fidalgo Island, the location of high human populations. Sites with
shaded bars are those closest to Cypress Island, which is further away from heavily inhabited areas. Site
had a significant effect of copper rockfish densities (p=0.004), and an A posteriori contrast between sites
near Fidalgo and Cypress Island showed significantly different densities (p=0.00025, Sheffé corrected
alpha=0.0033). Error bars represent ±1 standard error.
Discussion
We observed seasonal migration of the three most common species of groundfish in the proposed
marine reserves of Skagit County. Our initial analysis shows that season had a significant effect on
densities of kelp greenling, lingcod and copper rockfish (Figures 3, 4 and 5, respectively). This has
strong implications for marine reserve design in Skagit County, and throughout Puget Sound. The
movement of groundfish during seasons, especially the movement of rockfish to water deeper than
the transects we used (Figure 3), shows that reserves should be designed to include large amounts of
deep water habitat, which the rockfish are likely to inhabit during the winter and spring months. Most
of the proposed reserves in Skagit County are associated with an island (Figure 2), which improves
ease of enforcement and avoidance by fishers: however, this tends to include shallower water than a
reserve which may be placed in a channel, or include an entire deep water environment.
While seasonal migration has been seen in other rockfish species, especially into deeper water during
the summer (Love et al., 2002), this is one of the few studies to show this seasonal migration of
copper rockfish in protected waters like Puget Sound.
9
Copper rockfish were found in greater densities in deeper transects, confirming observations in Skagit
County by Moulton (1977) and in nearby San Juan Island County by Eisenhart (2001).
Changes in density with changes in depth were not significant for kelp greenling and lingcod
throughout the year (Figures 4 and 5), with the exception of spring. In the spring, densities of kelp
greenling and lingcod in shallow water became much higher than those in deep water. For kelp
greenling, densities in shallow water were 200% higher, and densities were 225% higher for lingcod
in the shallow water transects. While a cause for this movement into shallow waters was not
investigated in this study, it may be that these two predators are following newly available food
sources into shallow waters, as the prey ascends into the photic zone. Included in these prey are newly
hatched larvae from rockfish, as well as lingcod and kelp greenling (Love et al., 2002). Additionally,
kelp greenling and lingcod also nest in shallower water during the late winter and early spring,
however, the vast majority of these nests would be hatched and abandoned before the spring surveys
(Wright et al., 2000; King and Beaith, 2001; King and Withler, 2005).
While changes in densities with changes in season and depth were expected, a large effect of site was
not anticipated. The six sites had relatively similar bottom composition (unpublished data): however,
there was a large difference in densities between the sites (Figure 6). This difference separates based
on the site's region, with sites near Fidalgo Island having far less density of copper rockfish than those
sites near Cypress Island (Figure 2 and 6). There are several possible explanations for this large
difference. Cypress Island is closer to a better supply of available larvae, with the San Juan Islands to
the west and Strait of Georgia to the north. The San Juan Islands have over a dozen established
marine reserves, which have shown increased fish densities, as well as increased larval production
from the reserves (Eisenhart, 2001; Weis, 2004). Additionally, the San Juan Islands are believed to
offer better habitat for rockfish (Palsson, WDFW, pers. comm.). However, the distance from Cypress
to the San Juans is not much less than the distance from the Fidalgo Island sites to the San Juans.
Another possible explanation for this difference in rockfish density may be the proximity of the sites
to local marinas used by recreational fishers for the lingcod and salmon fisheries. The three sites near
Fidalgo Island are all less than four miles from a large marina, and the passage to these sites is in
waters protected from high wave action and heavy winds, and is easy to access for anglers. These
anglers often have rockfish as part of the bycatch from the recreational fisheries. The Cypress Island
sites, by comparison, are much further, through less protected waters.
Acknowledgments
I thank my advisor and co-author, Paul Dinnel for his tireless work to get the presentation, as well as
this publication completed. I extend my gratitude to my dive buddies Nate Schwarck, Brandon
Jensen, Rich Hoover, Pema Kitaeff, Andy Weispfenning, Jason Hall, and Nick Wenzel. I also
acknowledge the help of Gene McKene and Steve Sulkin in supporting my underwater research.
Additionally, this manuscript was improved and clarified greatly due to the help of an anonymous
reviewer, and we thank them. This work was made possible through grants from Skagit Marine
Resources Committee, Northwest Straits Commission, NOAA, Western Washington University's
Bureau of Faculty Research, Huxley College, and a Ross Travel Grant. This paper is dedicated to the
memory of Ralph Barr of Tumwater High School, who inspired thousands of students, including
myself. Finally I would like to thank my wife, Erin, for putting up with me through eight years of
higher education.
10
References
Berkeley SA, Chapman C, Sogard SM. Maternal age as a determinant of larval growth and survival in
a marine fish, Sebastes melanops. Ecology. 2004; 85(5): 1258-1264.
Eisenhardt EP. Effects of the San Juan Islands marine preserves on demographic patterns of nearshore
rocky reef fish. M.S. Thesis, University of Washington. Seattle, Washington, 2001.
Eschmeyer WN, Herald ES, Hamann H. A Field Guide to Pacific Coast Fishes of North America.
Peterson Field Guide Series. Houghton Mifflin Co., Boston, MA, 1983.
Hayden-Spears J. Minutes of the San Juan Marine Resources Committee, November 16, 2005.
Jagielo TJ, Wallace FR, Cheng YW. Assessment of lingcod (Ophiodon elongatus) for the Pacific
Fishery Management Council in 2003. Washington Department of Fish and Wildlife, Olympia, WA,
2003.
King JR, Beaith BW. Lingcod (Ophiodon elongates) nest density SCUBA survey in the Strait of
Georgia, January 16-April 26, 2001. Can Tech Rep Fish Aquat Sci. 2001; 2374; 21pp.
King JR, Withler RE. Male nest fidelity and female serial polyandry in lingcod (Ophiodon elongates,
Hexagrammidae). Molec Ecol. 2005; 14: 653-660.
Love MS, Yoklavich M, Thorsteinson L. The rockfishes of the northeast Pacific. University of
California Press, Berkeley, California, 2002.
Mason JE. Declining rockfish lengths in the Monterey Bay, California, recreational fishery, 1959-94.
Mar Fish Rev. 1998; 60: 15-28.
McConnell ML, Dinnel PA, Dolph J, Robinette J, Semrau D. Rocky reef bottomfish recovery in
Skagit County: Phase I Final Report: Marine Protected Areas – Preliminary Assessment and public
input. Skagit County Marine Resources Committee, Mount Vernon, WA, 2001.
McConnell ML, Dinnel PA. Rocky reef bottomfish recovery in Skagit County: Phase II Final Report:
Assessment of eight potential marine reserves sites and final site recommendations. Skagit County
Marine Resources Committee, Mount Vernon, WA, 2002.
Mora C, Andrefouet S, Costello MJ, Kranenburg C, Rollo A, Veron J, Gaston KJ, Myers RA. How
protected are coral reefs? Letters to Sci. 2006; 314: 758-760.
Moulton LL. An ecological analysis of fishes inhabiting the rocky nearshore regions of northern
Puget Sound, Washington. Ph.D. Dissertation, University of Washington, Seattle, Washington, 1977.
Paddack MJ, Estes JA. Kelp forest fish populations in marine reserves and adjacent exploited areas of
central California. Ecolog Applic. 2000; 10: 855-870.
Palsson WA. Monitoring the response of rockfishes to protected areas. Pp. 64-73 in Marine harvest
refugia for west coast rockfish: A workshop, ed. M. Yoklavich. NOAA Tech. Mem. NMFS-SWFSC
255, 1998.
Palsson WA, Northrup TJ, Barker MW. Puget Sound groundfish management plan (revised). State of
Washington Department of Fish and Wildlife, Marine Resources Division, Olympia, WA, 1998.
11
Palsson WA. The development of criteria for establishing and monitoring no-take refuges for
rockfishes and other rocky habitat fishes in Puget Sound. In: Puget Sound Research. Puget Sound
Water Quality Authority, Olympia, WA, 2001.
Parrish R. Marine reserves for fisheries management: why not. CalCOFl Rep. 1999; 40: 77-86.
Roberts CM, Bohnsack JA, Gell F, Hawkins JP, Goodridge R. Effects of marine reserves on adjacent
fisheries. Science. 2001; 30: 1920-1923.
Russ GR, Alcala AC. Marine reserves: long-term protection is required for full recovery of predatory
fish populations. Oecoloia. 2004; 138: 622-627.
Skagit County Marine Resources Committee. www.skagitcounty.net/Common/Asp/
Default.asp?d=PublicWorksMRC&c=General&p=smrcmain.htm. Accessed: April 6, 2005.
Spalding M, Kelleher G, Boucher T, Fish L. How protected are Coral Reefs? Letters to Sci. 2006;
314: 758.
Tuya FC, Soboil ML, Kido J. An assessment of the effectiveness of marine protected areas in the San
Juan Islands, Washington, USA. ICES Journal of Marine Science. 2000; 57: 1218-1226.
Weis LJ. The effects of San Juan County, Washington, marine protected areas on larval rockfish
production. M.S. Thesis, University of Washington. Seattle, Washington, 2004.
Weispfenning AJ. Survey of nearshore demersal fishes within candidate marine reserves in Skagit
County, Washington. M.S. Thesis, Western Washington University, Bellingham, Washington, 2006.
Worm B, Barbier EB, Beaumont N, Duffy JE, Folke C, Halpern BS, Jackson JBC, Lotze HK, Micheli
F, Palumbi SR, Sala E, Selkoe KA, Stachowicz JJ, Watson R. Impacts of biodiversity loss on ocean
ecosystem services. Science. 2006; 314: 787-790.
Wright N, Kum J, King C, Knaggs E, Leos B, Perez C. Kelp greenling, Hexagrammos decagrammus,
rock greenling Hexagrammos lagocephalus, whitespotted greenling, Hexagrammos sterelli 19931999 commercial catch by ports and blocks. Marine Fisheries Profiles: Volume 1: Nearshore.
California Department of Fish and Game, 2001.
Yoklavich MM. Marine harvest refugia for west coast rockfish: a workshop. NOAA Technical
Memorandum NMFS, Pacific Fisheries Environmental Laboratory, Pacific Grove, CA, 1998: 64-73.
12
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Baseline Survey Protocol
Hannah L Markham* and Nicola K Browne
Society for Environmental Exploration, 50-52 Rivington Street, London EC2A 3QP
*corresponding author: hannah.markham@gmail.com
Abstract
Baseline Survey Protocol (BSP) is a quantitative coral reef survey methodology recently
developed in order to conduct ecoholistic assessments of marine habitats. The technique
combines well-established survey protocols with newly introduced methods in order to fulfil
comprehensive biological and physical evaluations of each chosen reef site. Four in-water and
one surface personnel conduct the survey in a multidisciplinary team. The biological assessment
comprises of: Line Intersect Transect (LIT) to assess benthic cover and coral diversity to genre
level; Underwater Visual Census (UVC) data for target ichthyofaunal species abundance and
biomass; invertebrate diversity and abundance, and algal species composition. The physical and
environmental assessments include sea surface temperature and salinity, horizontal and vertical
visibility levels, and human activities on the surface. In addition, water and sediment samples
are collected for further chemical analysis. The strengths of the census centre from the
identification of reef fauna to species level and the incorporation of bathymetric data, resulting
in compatible data sets that allow for analysis of cause and effect relationships. The
methodology can be applied to the majority of reef community habitats resulting in a
representative data set, which can be assessed to determine biological trends on both a spatial
and temporal scale. Furthermore, socially and economically important species are used to
highlight the importance of specific reefs which can then be targeted for management initiatives.
BSPs provide a firm scientific grounding to allow marine scientists to develop site specific
management recommendations.
Introduction
Assessment of the biological status of marine ecosystems can be carried out using one or more of a
number of well established techniques ranging from broad scale to fine scale methods (English et al.,
1997; Hill and Wilkinson, 2004). There have been a number of studies advocating the use of certain
methodologies (Loya, 1978; Weinberg, 1981; Marsh et al., 1984; Dartnall and Jones, 1986;
Montebon, 1992; Sale, 1997) but ultimately choosing a technique that best suits the surveys
requirements will rely on the objectives of the survey and resources available. Limitations often
imposed include time, money and expertise. Scientists, if not financially constrained, will still be
limited by 'down time,' thus restricting personnel on the number of survey dives using SCUBA. This
imposes time constraints on studies, and may deter scientists from embarking on the project or
forcing them to significantly curb replicates. Hence, conducting surveys is more often than not a
compromise between the quantity and quality of the data collected and the availability of resources.
There is a current trend within non-governmental organisations conducting environmental
conservation (i.e., Coral Cay Conservation, Global Vision International, Operation Wallacea,
Greenforce, Earth Watch, Blue Ventures, Reef Doctor) to use non-specialist volunteers to aid in the
collection of data. However, said organisations tend to be overlooked by the scientific community as
a consequence of a lack of expertise which leads to questions regarding the quality of data collected.
As a direct consequence, a number of studies have been conducted in order to validate the use of nonspecialist volunteers (Chou, 1994; Mumby et al., 1995; Hodgson, 1999; Roxburgh, 2000). Although
13
some scientists express their concerns over the quality of data (Darwell and Dulvy, 1996), differences
in data collected by specialists versus non-specialists have shown to be insignificant (Semmens and
Semmens 1998; Harding et al., 2000). Furthermore, such organisations tend not to suffer from factors
that constrain non-governmental organizations (NGOs), and hence fill a certain niche in the
conservation/scientific world. Grants often limit a project's capacity, particularly with regards to time
frames, thus survey periods are short. In addition, funding bodies/private investment for scientific
organisations require significant evidence that monies invested will be put to good use. This often
restricts research to new areas. Hence, organisations which are not heavily reliant on grant
applications maintain the ability to move into regions that have yet to be studied for extended time
periods. However, their most valuable resource is the high man power with the potential to collate
large amounts of data to contribute both to the scientific community and to the formation of local
marine management initiatives.
In order to maximise data collection and effectiveness, a technique should aim to fill a number of
criteria. It needs to incorporate a number of biological, physical and chemical variables in order to
provide an ecoholistic assessment of the area; the technique must be relatively rapid with the ability
to be easily repeated to allow assessment of large areas potentially containing a diverse range of
marine habitats; and finally large data sets have to be adapted to allow quick analysis for data
dissemination. Thus, a technique incorporating all of the above factors, to be utilised by non-specialist
volunteers and to be applied to a remote unknown region, was developed by the Non-Governmental
Organisation Frontier-Madagascar/Society for Environmental Exploration (SEE). The technique,
termed the Baseline Survey Protocol (BSP), was used in the initial assessment of Diego Suarez Bay,
situated in Northern Madagascar.
State of Coral Reefs of Madagascar
Madagascar lies in the West Indian Ocean, in the waters of the Southern Equatorial Current (SEC)
forming part of the Agulhas large marine ecosystem (LME). Madagascar is renowned for the endemic
nature of its flora and fauna due to its evolutionary isolation from mainland Africa. As a result, its
terrestrial biology has been well documented. However, knowledge on the coral reef systems of
Madagascar is poor, as the majority of the Madagascan coastline has yet to be scientifically explored
(McClanahan et al., 2000). However, it has been estimated that coral reefs cover an area approximately
2000 km2, in excess of 20% of the islands coastline and studies conducted have indicated an increasing
threat from anthropogenic activities. It has been noted that without a considerable change in the
management of the coral reefs within this region of the western Indian Ocean, it is predicted that by
2014 coral reef cover in the region will decline to less than 20% of current live hard coral cover
(Wilkinson, 2004).
Political Climate
Historically, Madagascar has been at the forefront of conservation, with the first terrestrial protected
areas established in the early twentieth century. In 1995, the first 'Priority Setting' workshop was held
from which the 'National Environmental Action Plan' (NEAP) was developed. The creation of NEAP
resulted in a network of protected areas to be managed by the Association Nationale pour la Gestion
des Aires Protégées (ANGAP), a government body supported by legislation whose sole purpose was
to establish newly protected areas. In 2003, the Durban Declaration came into affect which stated that
the volume of protected areas was to increase from 1.7 million hectares to 6 million hectares by 2008.
Unfortunately there is a gap in the marine environmental sector and capacity building is required at
all levels in order to create a network of marine protected areas, thus maintaining reef ecosystems that
could potentially include some of the most biological diverse reefs in the world.
14
Frontier Madagascar/Society for Environmental Exploration
Established in 1989, The Society for Environmental Exploration (SEE) is an international
environmental research and conservation NGO (www.frontier.ac.uk). Frontier-Madagascar was
initiated as collaboration between the Institute of Marine Sciences in Tuléar (IHSM) and the Society
for Environmental Exploration (SEE), with the objectives of:
• Conducting high quality scientific research in to date understudied areas.
• Building capacity of local stakeholders
• Contributing directly to regional and national conservation
In 2005 Frontier Madagascar relocated its volunteer based projects to the Antsiranana region of
Madagascar, with the intention of providing support to existing organisations within this biologically
exciting area. The marine project has continued its collaboration with IHSM and has established links
with the local University of Antsiranana.
Study Rationale
Biological assessment of Diego Suarez Bay began in May 2005. Prior to Frontier's arrival, there had
been no sub-aqua assessments in the region and hence no scientific data was available on the
condition of the marine ecosystems both within the borders of the bay and regions situated on the east
coast directly outside the bay. However, the area had been identified as a potential site for high
marine biodiversity and endemism following a rapid marine assessment by Conservation International
(2002) in the Nosy Hara region, a region that lies directly west of Diego Suarez Bay in the
Mozambique Channel (McClanahan et al., 2000). Hence, Frontier-Madagascar initiated their marine
survey work within the region with the aim of assessing the biological status of the of Diego Suarez
bay.
This aim was identified as the central theme for the projects first phase of work within the region that
was scheduled to last for approximately one year (March 2005 – March 2006). This central aim was
identified after consultation with local stakeholders, who identified knowledge of the Bay's marine
ecosystems as poor, and highlighted a need for this information prior to the addition of marine related
management initiatives within the communes surrounding the bay. A series of objectives were
identified in order to obtain a comprehensive overview of the bay's biological status, while helping to
train the next generation of marine conservation workers: 1) gather data on fish assemblages and
biomass, invertebrate abundance, algal cover, coral biodiversity and cover; 2) gather data on chemical
and physical parameters; 3) provide training to both overseas and national students; and 4) assess the
social and economic importance of Diego Suarez Bay. Hence, Frontier-Madagascar developed the
BSP technique in order to meet the objectives outlined above.
Methodology
Study Region
The Diego Suarez Bay is located in the north of Madagascar between the tropic of Capricorn and the
equator 12°8-12'S, 049°8-12'E. The mouth of the bay opens out onto the Indian Ocean to the east by a
relatively narrow opening with a number of shallow islands on the seaward side (Figure 1). The bay is
subject to strong south-easterly winds for eight to nine months of the year, there are distinct windy
period from July to September (local knowledge). This is also the coolest time of year with water
temperatures dropping to around 23°C. In addition to the cool, dry season there is a very distinct rainy
15
season over the summer months from December to April. During the warm, wet season the average
maximum temperatures reach 35ºC with a maximum rainfall of 350 mm per month (World Weather
Information Service). During these warmer months the sea surface temperature can rise to >30°C.
The benthic topography of the bay is characterised by a deep central section with an average depth of
approximately 40 m, extending from the opening of the bay to the north and to the west. The bays to
the south and east of this section are all characterised by shallower topographies ranging from 20-10
m. The bay's large size relative to its seaward opening reduces water exchange and circulation, making
it particularly susceptible to pollution. The absence of any major rivers entering the bay prevents large
influxes of freshwater, sediment and land born pollutants. However, its convoluted coastline, and the
situation of the regional capital, Antsiranana, on a peninsula protruding into the centre of the bay,
indicates the potential for high levels of terrestrial influences on the marine environment.
Figure 1: Map of Diego-Suarez Bay, northern Madagascar (12°10'S, 049°15'E)
The bay itself supports a number of coastal communities, including two large communities in the
region; the town of Antsiranana and the tourist resort of Ramena. The main town in the Northern
region of Madagascar; Antsiranana is Madagascar's fifth largest town and has a population of
approximately 80,000 people. The town has a large deep-water port for container transportation
vessels, fishing vessels and cruise ships all year round. Aside from the smaller fishing and tourist
town, Ramena, other settlements are generally in the form of villages with less that 200 people. Diego
is fast becoming a popular town for tourists with an airport 6 km from the centre. As well as the
beaches to the south, the Bay is close to a number of national parks, which are a major tourist
attraction.
16
Industry within the immediate vicinity of the bay is limited to a large desalination plant, the main salt
producer in northern Madagascar, within one of the inlets. Potential negative impacts include
alterations to sea surface temperature, salinity levels and input of waste products. Information
regarding the level of impact this has on the bay is limited and needs further investigation. Threats to
the bay through human actions include deforestation, pollution, over-fishing, and anchor damage
(pers. obs.). Destructive fishing practises such as cyanide poisoning and bombing are negligible.
Survey Protocol
The BSP technique involves five surveyors - four in-water surveyors and a boat marshal. The boat
marshal's primary responsibility is to ensure diver safety throughout the survey. Their secondary role
is collect surface data once the divers have descended. They should note the following parameters:
cloud cover; wind direction and strength; air temperature; sea surface temperature; surface salinity;
vertical visibility and visible boating activity. Furthermore, a surface water sample is taken for
chemical analysis.
Studies have indicated that in regions where spatial heterogeneity of habitats is high, a long transect
will incorporate a high level of spatial variation, thus reducing the statistical power of the survey
(Brown et al., 2000). Thus, the survey is carried out along a 45 m transect line, comprising of 2x20 m
sections separated by a 5.0 m gap (Figure 2), thus incorporating the value of data replication within a
site. The survey can be carried out at any depth depending on the objective of the study, hence the
methodology can be adapted for snorkelling studies of the site at depths <1.5 m. In all cases, the line
is laid parallel to shore at a constant depth. The four surveyors are allocated a role prior to
descending: Physical surveyor; fish surveyor; benthic surveyor; invertebrate and algae surveyor
(A&I). On reaching the sea bed, the two buddy teams (Physical / Fish, Benthos / A&I) start the
survey in the formation represented in Figure 2.
Figure 2: Diagrammatic representation of the Baseline Survey Protocol (BSP) indicating
transect ariel view and surveyor positions in relation to transect line and buddy pairs.
Physical Surveyor - responsible for health and safety on the dive. They should ensure that buddy
checks are carried out before entering the water, at the 20 m and 45 m mark of the transect and on
ascent at the end of the dive. In addition, they should check air levels at the 20m mark and send up
any divers who do not have sufficient air to complete the next stage of the survey. At the start of the
transect they choose where to lay the transect line, following a depth contour at a speed of no faster
than 12 m⋅min-1, keeping slightly behind fish surveyor and ensuring that the tape is secured under
rocks or rubble and facing upwards. They are also responsible for reeling in the line once the survey
is complete. At specific points on the survey (0 m, 20 m, 25 m, 45 m) the following parameters
should be noted: water temperature; horizontal visibility; estimation on current strength and depth. In
addition a water sample should be collected at the start and rugosity measurements taken at 20 m and
17
45 m. Throughout the survey they should sketch an aerial view of the site highlighting significant
coral forms, topography, human impacts, etc. Equipment required includes: 50 m tape measure and
weight, two water sample pots, 5.0 m rugosity chain, dive computer with a depth gauge, timer and
thermometer, pencil and dive slate.
Fish Surveyor - records the species and size of every fish that they see along the survey. The surveyor
should count the fish within a 5.0 m box (5.0 m in front, 2.5 m either side of the transect line, 5.0 m
into the water column). They should be half a body length in front of the physical surveyor at all
times to reduce disturbance to fish populations. In addition, they carry a surface marker buoy (SMB)
and strict buddy contact must be maintained at all times. Equipment required includes: an SMB, a
dive slate and pencil.
Benthic Surveyor – follows the Line Intercept Transect Method (LIT) as described by English et al.
(1997). They record changes in substrate that occurs beneath the transect line (one centimetre or
greater). Therefore, they swim directly above the tape measure recording substrate type (rock, sand,
sediment, rubble), coral form/genera/species, algae genera/species and sea grass genera/species.
Equipment required includes: blank dive slate and pencil
Algae and Invertebrate Surveyor - records all species of algae, takes samples and approximates
cover/number of thallus. They also record the abundance of each species of invertebrate present on
the transect line, if unknown taking photographs for further identification. They move back and forth
across the transect line to a distance of 1m either side of the tape measure. The A&I surveyor also
carries a surface marker buoy and is responsible for collecting sediment samples at the start of each
transect. Equipment required includes: two sediment sample pots; blank slate and pencil; SMB;
digital underwater camera; small mesh algae-collecting bag.
Data Validation
Validating data collected by non-specialist volunteers is a central component of FrontierMadagascar's marine programme, as questions are commonly raised as to its accuracy (Darwell and
Durvey, 1996). As a direct consequence such organisations have to self-analyse and regulate their
volunteers to ensure competency. They must also ensure that the supporting methodologies are both
within the capabilities of the volunteer and do not allow opportunity for error in collection. Hence, a
degree of validation is incorporated into a one month training period prior to surveying.
The initial two weeks focus on marine species identification. The level to which identification of
marine species (family/genus/species) are identified varies between data sets. Fish training begins
with 47 common families found in the region and progresses to species level (~170 species are
taught). Invertebrates species within six phylum and 18 classes are taught straight to species level. A
total of 49 coral genera are taught with the most common and/or important corals identified to species
level, and a total of 39 marine plant genera are taught. Training is conducted both in the class room
and during point out snorkels/dives, and is completed with repeatable tests for statistical analysis.
Tests are carried out both in situ and using computer images with a 95% pass rate required for fish
and invertebrates, and an 80% pass rate for corals. Marine plants are assessed differently with
individuals being tested on three aspects of the plant; colour, form and species. The examinee is
expected to pass the first two with 100% and to identify 95% of the species. Tests are conducted
using real samples and with computer images. The focus of the third week is on the BSP, which is
broken down into component parts to include swim pace, compass work, size estimation exercises,
and dry-runs in order to ensure diver competency in the water. Size estimation exercises are
conducted in-water. The examinee has to correctly estimate the length of 50 strips of plastic which are
18
attached to a rope laid out on the sea bed at a depth of 2.0 m. A pass in three concurrent size
estimation exercises are required. A successful exercise requires the sum of estimates to be within
10% of the actual total. However, they must also obtain a standard deviation of <8% in order to
ensure that over-estimations are not compensating for under-estimations. The final week of the
training period combines all elements and volunteers carry out supervised surveys paired with a staff
member and results are compared to assess for errors. Data validation continues throughout the data
collection period with students sitting 'check-tests' at a period of one month intervals after completion
of the training programme. An additional form of data validation is currently underway, comparing
data collected by non-specialists volunteers to specialists from: i) the same survey, and ii) the same
site.
Data Analysis
The data collected from the BSP is transferred to an Access Data base forming five data sets (fish,
coral/benthic, algae, invertebrate and environmental factors). Statistical analysis of the data sets can
be conducted at a number of levels within a transect/site/region in order to evaluate similarity indices.
Intra and interspecies interactions can be assessed in order to assess specific relationships and
ecological processes, or could simply be used to generate species inventories or species richness
values for an area. Thus, questions regarding specific species versus communities can be addressed
and hierarchical cluster analysis can be applied to identify faunal patterns of communities.
Alternatively individual data sets can be crossed. For example, site specific data can be evaluated in
order to determine relationships between biological/physical/chemical parameters or used to correlate
benthic habitat variables with species composition, abundance and size. Such analysis may provide a
basic model for establishing habitat affinities of species, and thus act as a diagnostic tool for the
establishment and assessment of marine protected areas. Furthermore, temporal changes can be
tracked and the impact of anthropogenic activities, bleaching events and disease can be assessed.
Discussion
Quantitative assessment of marine ecosystems is required in order to determine current status of reefs,
track changes, identify problem sources and assess the sustainability of resource use. The latter aspect
is becomingly increasingly important as marine environments are over-exploited, and hence effective
management of coastal habitats and deep waters is essential. The design of management strategies is
based on both biological and socio-economic data, and hence methods which assess the marine
environment need to be employed.
There are a number of survey methodologies that are currently used and have been developed as a
means of meeting specific objectives and resources available. Transects are commonly used as a
medium scale method for assessing reef environments. They are easy to position and are a framework
from which a number of biological and physical parameters can be assessed. For example, the benthic
environment can be quantitatively measured using one of two commonly used techniques; Point
Intercept Transect (PIT) or the Line Intercept Transect (LIT) (Montebon, 1992; Brown et al., 2000;
Hodgson et al., 2003). The LIT technique, incorporated into the BSP, is a continuous measurement of
all abiotic and biotic substratum that lines directly underneath the tape measure and hence is
commonly regarded as the more accurate of the two techniques. However, previous studies have
indicated there are disadvantages to the technique in that it can be time consuming and potentially too
detailed, particularly for non-specialist volunteers (Erdmann et al., 1997). However, the BSP
technique incorporates a four week training period which was found to provide sufficient training to
enable individuals to carry out the survey effectively. A fine scale survey technique for the study of
19
benthic environments involves the use of quadrates which quantitatively assess the number of sessile
organisms, calculate percentage cover of benthic substratum and determine relationships between the
two variables (Hill and Wilkinson, 2004). However, the transect method allows assessment of fish
populations by underwater visual census (UVC) (Bohnsack, 1996; Cheal and Thompson, 1997;
Semmens and Semmens, 1998), and hence data collection on fish abundance, size and diversity is
carried out in conjunction with benthic assessment. This forms one of the central aspects of the BSP
technique, i.e., all data is collated within the same area at one point in time in order to produce
compatible data sets that provide cause and effect relationships. Additional components assessed are
invertebrates and algal cover. During the training period, algal cover is highlighted as an important
variable and quantitative assessment is provided by the LIT component. However, as the benthic
surveyor is limited to recording species that lie directly under the tape measure, small/cryptic species
are often missed. Therefore, in order to supplement algal analysis, the invertebrate surveyor is
responsible for compiling an algal species list as they swim back and forth across the transect line
searching a larger proportion of the reef area.
Previous assessments of marine environments have used a similar combination of quantitative
techniques. However, they tend to target one or two aspects of the marine community on each dive
before combining data sets from different dives. Hence, time over which data collection is carried out
increases, reducing the efficiency of the monitoring scheme. The BSP technique is a combination of
well established survey techniques (LIT, UVC) and additional components (algal searches, rugosity,
chemical analysis, surface observations) carried out simultaneously. Therefore, the combination of
the quantity of data, the number of data sets, and the reduced level of resource use compared to
previous monitoring schemes are the key strengths of the BSP technique. Furthermore, the technique
incorporates a detailed assessment of the benthic environment, as training is conducted to coral
genre/species and can be applied to the majority of reef communities.
The BSP technique was originally developed as a means of comprehensively assessing reef
environments whilst maintaining a comparatively low level of resource use. The use of non-specialist
volunteers aided in the collection of large quantities of data over short periods of time, which in turn
helped to reduce running costs and time frames. As data quantity and quality were maintained, the
level of compromise often undertaken during the design of a monitoring scheme was reduced.
Maintaining low resource use levels is particularly poignant in developing countries where a common
short fall is the lack of money, knowledge and expertise required to carry out effective monitoring on
which management decisions and their effectiveness are founded. However, the BSP technique would
provide a comprehensive yet basic technique from which such decisions could be made with a high
degree of confidence. The training programme has proven to be successful with educating individuals
with no previous marine background and thus would aid to build local capacity promoting long term
sustainability of the monitoring schemes. The technique would be suited for the assessment of reefs
prior to management implementation by providing baseline data from which the success of the
initiatives can be determined. In addition, biological trends can be assessed on both a spatial and
temporal scales, and the socio-economic value of reefs can be determined through the quantitative
assessment of economically important species.
The biological assessment of Diego Suarez Bay using the BSP technique has highlighted the area as a
potential candidate for marine protection. The biological status of reefs within the bay range from
high to poor depending on proximity to populations and coastal development, and temporal data
analysis suggest that conditions are declining. Frontier-Madagascar's comprehensive research in this
area will allow technical knowledge of the bay's ecological profile to be coalesced with the
administrative expertise of local planners and managers, a committee within which Frontier
Madagascar looks forward to playing a fundamental role.
20
References
Bohnsack JA. Two visually based methods for monitoring coral reef fishes. In: Crosby MP, Gibson
GR, Potts KW, eds. Coral Reef Symp on Practical, Reliable, Low Cost Monitoring Methods for
Assessing the Biota and Habitat Conditions of Coral Reefs, 26–27 Jan 1995. Office of Ocean and
Coastal Resource Management, NOAA, Silver Spring, 1996: 31-36.
Brown EK, Cox EF, Tissot BN, Jokiel PL, Rodgers KS, Smith WR and Coles SL (2004)
Development of benthic sampling methods for the Coral Reef Assessment and Monitoring Program
(CRAMP) in Hawai‘i. Pacific Science 7: 145-158.
Cheal AJ, Thompson AA. Comparing visual counts of coral reef fish: implications of transect widths
and species selection. Mar Ecol Progr Ser. 1997; 158(1-3): 241-248.
Chou LM. Living coastal resources of southeast Asia: management through continuing education by
institutions of higher education. Aquatic Conserv: Freshwater Marine Ecosystems. 1994; 4: 179-184.
Dartnall A, Jones M. A manual of survey methods: living resources in coastal area. ASEAN-Australia
Cooperative Program on Marine Science Handbook. Townsville: Australian Institute of Marine
Science, 1986; 167 pp.
Darwall WRT, Dulvy NK. An evaluation of the suitability of non-specialist volunteer researchers for
coral reef fish surveys. Mafia Island, Tanzania – a case study. Biolog Conserv. 1996; 78: 223-231.
English S, Wilkinson CR, Baker V. Survey manual for tropical marine resources. Australian Institute
of Marine Science: Townsville, Australia, 1997; 390 pp.
Erdmann MV, Mehta A, Newman H, Sukarno. Operation Wallacea: low-cost reef assessment using
volunteer divers. In Lessios HA, Macintyre (eds.) Proc. 8th Int. Coral Reef Sym. Smithsonian
Tropical Research Institute, Panama 1997; 2: 1515-1519.
Harding S, Lowery C, Oakley S. Comparison between complex and simple reef survey techniques
using volunteers: is the effort justified? Paper presented at the Ninth International Coral Reef Symp.
Bali, Indonesia, 2000.
Hill J, Wilkinson CR. Methods for Ecological Monitoring of Coral Reefs. Australian Institute of
Marine Science, Townsville, Australia, 2004; 117 pp.
Hodgson G. A global assessment of human effects on coral reefs. Marine Pollution Bulletin. 1999;
38: 345-355.
Hodgson G, Maun L, Shuman C. Reef Check Survey Manual for Coral Reefs in the Indo-Pacific,
Hawaii, Red Sea, Atlantic/Caribbean, and Arabian Gulf. Reef Check-UCLA, Institute of the
Environment, Los Angeles, California, USA, 2003; 33 pp.
Loya Y. Plotless and transect methods. In: Stoddart DR, Johannes RE, eds. Coral Reefs: Research
Methods. UNESCO Monographs on Oceanographic Methodology. UNESCO, 1978: 197-217.
Marsh LM, Bradbury RH, Reichelt RE. Determination of the physical parameters of coral
distributions using line transect data. Coral Reefs. 1984; 2: 175-180.
McClanahan TR, Sheppard CR, Obura DO, eds. Coral reefs of the Indian Ocean: Their ecology and
conservation. Oxford University Press, 2000: 411-444.
21
Montebon A. Use of the Line Intercept Technique to Determine Trends in Benthic Cover Proceedings
of the Seventh International Coral Reef Symposium, Guam, 1992: 1: 151-155.
Mumby PA, Harborne AR, Raines RP, Ridley JM. A critical appraisal of data derived from Coral Cay
Conservation volunteers. Bull Marine Sci. 1995; 56: 737-751.
Roxburgh T. Checking Reef Check in Tanzania. Reef Encounter . 2000; 27: 21.
Sale PF. Visual census of fishes: how well do we see what is there? Proc 8th Int Coral Reef
Symp.1997; 2: 1435-1440.
Semmens CP, Semmens BX. Fish census data generated by non-experts in the Flower Garden Banks
National Marine Sanctuary. J Gulf Mexico Sci. 1998; 2: 196-207.
Weinberg S. A comparison of coral reef survey methods. Bijdragen tot de Dierkunde. 1981; 5 1: 199218.
Wilkinson CR, Ed. Status of Coral Reefs of the World: 2004. Global Coral Reef Monitoring Network
(GCRMN) and Australian Institute of Marine Science (AIMS). Townsville, Australia, 2004; 547 pp.
22
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
AAUS Diving Officer and Scientific Diver Certifications: The Need for
Quality Control
Michael A. Lang1, Glen H. Egstrom2, and Christian M. McDonald3
1
Smithsonian Institution, PO Box 37012-MRC 715, Washington, DC 20013-7012.
University of California Los Angeles, 3440 Centinela Avenue, Los Angeles, CA 90066-1813.
3
Scripps Institution of Oceanography, 8602 La Jolla Shores Drive, La Jolla, CA 92037-0210
2
Abstract
A roadmap is outlined as a model for an American Academy of Underwater Sciences (AAUS)
'Diving Officer' and 'Scientific Diver' certification program. AAUS historically has relied on
nationally recognized diver training agencies for its scuba instructor and entry-level diver
certifications, yet neither commercial nor military diving entities have done so. The specificity
of training requirements necessary to becoming a qualified scientific diver has been
fundamental to the evolution of AAUS. Indeed, the concept of reciprocity is based upon a
verifiable level of standardized scientific diver training. As AAUS has evolved, it has become
increasingly obvious that a clear definition and implementation of scientific diving training
standards is fundamental to its continued successful development in support of scientific diving
programs. Further, there is a need to develop quality control mechanisms due to an often
transient scientific diver population. A reduction of the multiple mechanisms for organizational
member program scientific diver certification through a standardized approach will help meet
this need. The AAUS Diving Officer and Scientific Diver certification program would result in:
a) imposition of a measure of quality control; b) reduction of reliance on recreational diver
training agencies; c) generation of additional revenue for AAUS; and, d) retention of control of
scientific diver certification standards. An implementation plan is offered for AAUS
consideration, including promulgation of certification standards and procedures, training
materials, liability exposure, quality control, and program roll-out to the diving community. If
AAUS is indeed 'self-regulated,' it is necessary to demonstrate the implementation mechanics
for such self-regulation.
Introduction
In March of 2006, a position paper on 'Diving Officer' and 'Scientific Diver' certifications was coauthored by a random selection of a dozen Diving Officers and presented at the annual AAUS
symposium at Friday Harbor Marine Labs for discussion. There appeared to be minimal disagreement
among the 100+ AAUS members present based on the concept as it was presented. Notwithstanding,
a number of uncertainties with regards to liability exposure, cost, and implementation of AAUS
certifications were identified and are currently being evaluated. The paper's objective was to serve as
a concise statement towards the promulgation of 'AAUS Diving Officer' and 'AAUS Scientific Diver'
certification standards.
Background
The precursor of modern-day AAUS scientific diving programs began in 1950 at the Scripps
Institution of Oceanography (SIO) when graduate students Andreas Rechnitzer and Conrad Limbaugh
acquired newly imported scuba gear from University of California, Los Angeles (UCLA). The first
formalized diver training at Scripps was organized by Rechnitzer, Limbaugh, and Jim Stewart in 1952
23
under mandate from the University of California (UC) Office of the President in response to a fatal
diving accident at another UC campus. The following year a small group from Los Angeles County
Parks and Recreation came to SIO to participate in dive training. This group, Al Tillman, Ramsey
Parks, and Bev Morgan went on to provide the first recreational diver training nearly a decade prior to
the birth of NAUI, PADI, SSI, and YMCA. AAUS has historically relied on nationally recognized
training agencies for its scuba instructor and entry-level diver certifications with little concern for the
wide variations in quality control in these organizations. These recreational training organizations
have developed divergent training standards both within and between their organizations. Specialty
training was initially intended as a supplement to basic training, used to cover the training gaps
between the specific functional needs of the groups with different operational objectives. At this time,
there are independent training organizations for nearly all of these specialties and in most cases each
application has several agencies claiming to be the best. There are multiple training organizations for
public safety divers, police and fire groups, technical divers, etc. These special interest groups
continue to proliferate and we will very likely see this trend include scientific diving in the near
future. This trend appears obvious and either AAUS will develop the criteria and the programs or
someone else will. Any shift of control over setting scientific diving standards to other training
agencies will further complicate the status of the current national scientific diving organization, the
AAUS. It is important to note that neither commercial nor military diving entities rely on recreational
diver training organizations to certify commercial or military divers and the public safety diving
sector is rapidly developing in the same manner.
Current Status
The issues that confront AAUS with regard to certification are neither unique nor are they new. As
each specialized sector of the diving community has evolved, it has had to face the inevitable
dilemma associated with growth and implementation of specific operational standards. The specificity
of training requirements necessary to becoming a qualified scientific diver has been fundamental to
the evolution of AAUS. Indeed, the concept of reciprocity is based upon a verifiable level of
standardized scientific diver training. As AAUS has evolved, it has become increasingly obvious that
a clear definition and implementation of scientific diving training standards is fundamental to its
continued successful growth and development.
The recognition that the 1980 (and pre-cursor) AAUS open-circuit, compressed air, single-hose
regulator scuba diving standards are rapidly becoming inadequate for our needs is clearly manifest.
There are many new and worthwhile advances in equipment and procedures that must be incorporated
into our diving programs if they are to remain state of the art. There is a further need to develop
quality control mechanisms for an often transient scientific diver community. A reduction of the
multiple mechanisms for organizational member (OM) scientific diver certification through a
standardized approach will help meet this need.
Impact of Diving Officer Certification
The AAUS Diving Officer and Scientific Diver certification program would result in: a) imposition of
a measure of (currently non-existent) quality control; b) reduction of reliance on recreational diver
training agencies for setting standards for instructor and entry-level scientific diver; c) generation of
additional revenue for AAUS in concert with its strategic and business plans, including the
consideration of development, fundraising, and external grants; and, d) retention of control and
revisions of future certification standards for Diving Officer and Scientific Diver.
24
AAUS Diving Officer qualification criteria (AAUS, 2006) consist of being: a) trained as a scientific
diver; b) a full member as defined by AAUS; and c) an active underwater instructor from an
internationally recognized certifying agency. We offer that these 'certification criteria' are minimal at
best, while the expectation is for AAUS, as an academia-based organization, to raise the bar for
scientific diving and safety. AAUS Scientific Diver certification criteria are specified in AAUS
Standards Sections 4 and 5.
The implementation of these certifications rests with the AAUS Board. We offer the following plan
for consideration:
1.
2.
3.
4.
5.
6.
Promulgation and adoption of an 'AAUS Diving Officer' certification standard. This process
should start by careful review and modification of current recreational scuba instructor
certification standards (RSTC, 2004). This should be followed by replacement of the purely
recreational aspects of diver training topics with specific scientific diving topics (theoretical,
practical and administrative).
Promulgation and adoption of a 'Diving Officer Trainer' certification standard.
Initial certification procedure for current AAUS Diving Officers (approximately 100). This
process should be done through one or more minimum three-day Diving Officer certification
workshops. AAUS must make the effort through the acquisition of grant support to provide
airfare/lodging for AAUS Organizational Member Diving Officers to participate. We recommend
against an 'across the board' Diving Officer 'grandfathering' clause. In this next phase of AAUS
evolution, it will be important to establish a platform where uniform training will be consistently
implemented. Evolution rather than revolution argues for a carefully phased approach to the
necessary changes. It might be necessary to develop a three to five year plan in order to ensure a
progressive change for research institutions, university programs with recreational scuba
components (who may need to retain recreational scuba instructor credentials in addition to
AAUS Diving Officer certification) and 'other OMs.'
Adoption of an AAUS 'Scientific Diver' certification standard. Certification cards would be
specifically restricted to scientific divers (as defined by the Occupational Safety and Health
Administration) and Diving Officers of AAUS organizational member programs. These
certifications would not be applicable to recreational, technical, commercial, military, or public
safety diving purposes. Scientists should have the ability to receive technical applications (e.g.,
trimix, rebreathers, and surface-supplied diving) training and certification from within the
scientific diving community as well. If scientific diver certification is recognized externally, then
perhaps these 'specialties' should be recognized as well.
Initial certification of current AAUS Scientific Divers (approximately 4,000), retroactively for
those individuals meeting the certification criteria. Recertification requires further consideration.
Consideration of liability exposure to AAUS for both certifications. Initial discussion with
industry professionals indicates that AAUS would be subject to no additional liability by
providing certification of 'Diving Officers' than it already incurs through the promulgation of
consensual standards and their imposition on research organizations as a condition of
membership. Liability for the training and certification of 'Scientific Divers' has and will remain
the responsibility of the sponsoring OM. Yet, AAUS derives no benefit (financial or otherwise)
from this exposure and furthermore has no mechanism in place to assure quality control and
adherence to standards. AAUS must implement a formal quality control mechanism for
'complaints, standards violations, investigations,' etc.
Rationale:
a. Diving Officers relying on recreational instructor insurance policies may not be covered for
'commercial ventures' (i.e., scientific diver training/certification using AAUS standards);
b. AAUS would certify Diving Officers who would certify scientific divers for their OM; and,
25
c. These considerations would allow AAUS to secure insurance coverage for itself as the
certifying agency and for its Diving Officer members.
7. Establishment of 'Scientific Diver' training curricula per AAUS Standards and development of
standardized training materials in collaboration with NOAA. We further suggest that the
proposed 'AAUS Diving First Aid Training for Scientists' (Lang et al., in press) be formally
adopted by AAUS.
8. Press release to the diving community at large and specific discussion with chief executive
officers of PADI, NAUI, YMCA, and SSI regarding recognition of AAUS Scientific Diving and
Diving Officer certifications through their retail outlets.
9. AAUS infrastructure and business plan development is a responsibility of the AAUS Board,
which is best positioned to determine costs and set revenue targets. Consideration of a vendor to
provide these certification card services for AAUS must be explored.
Conclusion
If AAUS is indeed 'self regulated' it is necessary to demonstrate the implementation mechanics for
such self-regulation. The above recommendations represent the outline for a mechanism that should
permit AAUS to gain further control over its future.
References
American Academy of Underwater Sciences. Section 1. Standards for Scientific Diving Certification,
2006. www.aaus.org
Lang MA, Marsh AG, McDonald C, Ochoa E, Penland L. Diving First Aid Training for Scientists. In:
Godfrey JM, Pollock NW, eds. Diving for Science 2006. Proceedings of the American Academy of
Underwater Sciences 25th Symposium. Dauphin Island, AL: AAUS, 2007: 85-102.
Recreational Scuba Training Council. Minimum Course Content for Recreational SCUBA Instructor
Certification. RSTC, Jacksonville, FL, 2004; 9 pp.
26
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
NOAA Test and Evaluation of Two, Commercial-Off-the-Shelf, Multi-Gas
Dive Computers for Providing Accurate Depth Measurements and
Acceptable Mixed Gas and Air Decompression Schedules
J. Morgan Wells1 and David A. Dinsmore2
1
2
Undersea Research Foundation, P.O. Box 696, North, VA 23128
NOAA Diving Program, 7600 Sand Point Way, NE, Seattle, WA 98115-0070
Abstract
Dive computers offer the potential to increase both the safety and efficiency of diving
operations. Numerous models have been in existence for several decades. Such devices use a
variety of hardware and software, and produce decompression profiles that can be significantly
different, making a comparison of the various units, with respect to safety and efficiency, a
difficult and confusing task. In 2005, the NOAA Diving Center contracted the Undersea
Research Foundation to test and evaluate two, commercial-off-the-shelf (COTS), multi-gas dive
computers: the Delta P Technology Ltd., VR3, and the HydroSpace Engineering, Inc., HS
Explorer. The study, which involved the evaluation of software generated by the dive computers
and pressure tests in hyperbaric chambers, was initiated to provide data and analysis to assist
NOAA personnel in making informed decisions related to the selection of dive computers and
decompression tables to increase the safety and efficiency of NOAA diving operations. This
paper describes the tests performed and the results achieved by the two dive computers
evaluated in providing accurate depth measurements and producing acceptable mixed gas and
air decompression schedules.
Introduction
In 2005, the NOAA Diving Center awarded a contract to the Undersea Research Foundation to
evaluate the reliability of selected mixed gas dive computers in producing acceptable mixed gas and
air decompression schedules and providing accurate depth measurement. The purpose of the study
was to provide data and an analysis to allow NOAA personnel to make informed decisions related to
the selection of decompression computers and decompression tables that can increase the safety and
efficiency of NOAA diving operations.
The dive computers evaluated in this study were the Delta P Technology Ltd., VR3 (2004) and the
HydroSpace Engineering, Inc., HS Explorer (2000). Both units are capable of use in an open-circuit,
semi-closed-circuit, and closed-circuit mode, with air, nitrox, heliox, and trimix gas mixtures. This
study was limited to the open-circuit mode with air and trimix. Adequacy of decompression profiles
produced by the software packages, and by the computers during pressure exposure were the primary
functions studied. The accuracy of the depth sensing system was also determined.
The software used in dive computers is normally available to the user to compute decompression
tables for predicting in-water decompression, or to produce backup decompression tables for use
during a dive. During an actual dive, this software receives real time input from sensors in the
computer, and computes a real time decompression profile based on this input.
Decompression profiles generated by the respective dive computer software packages and dive
profiles produced by the computers during actual pressure exposures were compared to two
27
computation methods that are considered 'industry standards,' the US Navy model (Workman, 1965),
and the Hamilton (2004) method, as well as to each other. This comparison of 'apples and oranges' is
accomplished via the 'Multiplex' analysis program developed by the JM Wells. This program
'dissects' decompression profiles and computes compartment (tissue) inert gas pressures in a selected
series of half-time compartments. These half-times, and the computation model, are not necessarily
those used by the respective original dive computer software systems, but rather, they normalize any
decompression profile to a common basis, thus allowing comparison. The normalizing compartment
half-times used in this study are the nine used in the US Navy model (Workman, 1965): 5, 10, 20, 40,
80, 120, 160, 200 and 240 minutes.
Sensors in the dive computers provide inputs to the software required to compute the real time dive
profile. Pressure, time, breathing gas composition, and in some cases temperature data is used for the
computation. The computer display provides the diver with information including depth, time of the
dive, temperature, ascent rate, minimum acceptable depth, depth of decompression stops, time at
decompression stops, estimated total decompression time, and the breathing gas used for
computations.
Procedures
Ten dive profiles were selected for use in this study. The profiles were provided by the NOAA Diving
Center based on current or projected operational requirements. Each profile required breathing gas
changes during ascent. The air dives involved a switch to oxygen at 20 ft while the trimix dives
required a switch to a nitrogen/oxygen mixture containing 36% oxygen, NN36, at 110 ft and to
oxygen at 20 ft. The dive profiles are shown in Table 1.
Table 1: Dive profiles tested in study trials
Depth
(fsw)
110
130
150
170
180
200
240
270
300
300
Time
(min)
35
35
30
20
30
25
25
15
15
20
Breathing Gas
Air
Air
Air
Air
Tx18/50 (18% O2, 50% He, 32% N2)
Tx18/50
Tx18/50
Tx14/60 (14% O2, 60% He, 26% N2)
Tx14/60
Tx14/60
Two dive computers of the same brand were used on each actual dive. They were immersed in a
water bath inside a pressure chamber. A video camera was positioned in the chamber viewport in a
position that allowed a clear view of both computer displays. A video 'quad processor' and two
additional video cameras allowed the recording of the view of a digital 'standard pressure gauge' (3D
Instruments, Inc., DPG-6600), and a digital stopwatch all in the same video frame. These views were
recorded on VHS tape during the chamber dives.
Data from the videotapes were transcribed into hard copy dive profiles using the 'forward' and 'pause'
controls of the VCR. The depth calibration data were obtained using a small bench-top chamber in the
28
Undersea Research Foundation lab. The actual dives were conducted in a double-lock hyperbaric
chamber at the NOAA Diving Center in Seattle, WA. Both brands of computers required manual
manipulation to switch to different breathing gases during ascent. This was accomplished by
pressurizing a person in the outer lock of the chamber to the depth at which the switch was to be
made, and holding them there until the inner lock arrived at that depth. The person then entered the
inner lock and rapidly performed the gas switch on the dive computers. The person was then returned
to surface pressure in the outer lock while the dive proceeded in the inner lock.
The software packages provided with the dive computers were used to produce decompression
schedules that were analyzed in the same fashion as those produced during the actual pressure
exposures. Both computers have software options for making decompression computations more
conservative. The VR3 uses a derivative of the Buhlmann ZHL 16 algorithm (VR3 Manual, 2004).
On decompression dives it added deep-water (microbubble) stops. Safety factors in 10% increments
up to 50% can be applied at the user's discretion. Every 10% increase adds 2% to the inert gas content
of the gas selected, and this value is then used for computing decompression requirements. A safety
factor of '0' was used on the VR3 for all chamber dives and software computations.
The HS Explorer system for adding conservative adjustments to the computations involves what they
call the 'computation formula' (CF). The CF levels go from 0-9, and in general, the computed
decompression time increases with the CF level (HS Manual, 2000). The CF levels are based on
various combinations of the Buhlman, US Navy, and Reduced Gradient Bubble Model (RGBM)
models. CF values of '0, 3, and 5' were used for the HS Explorer.
Findings
The accuracy of the pressure sensors in dive computers is critical to the production of valid
decompression profiles. Tables 1 and 2 show each of the computer's depth readings in relation to the
master gauge. Note that the VR3 depth readings were all within ±1% of the 'standard' reading. The
HS Explorer readings were within ±2%.
The full results of the Multiplex analysis of the software produced decompression profiles, as well as
those produced during the chamber dives, are not included in this paper due to space limitations;
however, the full report is available on the NOAA Diving Center website at: www.ndc.noaa.gov.
Conclusions
VR3 Dive Computer
1. The depth readings produced by the dive computer were all within 1 foot of seawater of the
pressure standard. None of the readings were shallower than the standard.
2. There was excellent agreement between the pressure generated profile compartment inert gas
pressures and those generated by the software.
3. At a 'safety factor' of '0,' all compartment inert gas pressures were well within US Navy 'M'
values and very similar to those of the NOAA/Hamilton values.
4. There was very good agreement between the two computers during the pressurization tests.
29
HS Explorer Dive Computer
1. The depth readings produced by the dive computer were all within –1 and +4 fsw of the standard.
2. There was good agreement between the pressure generated profile compartment inert gas
pressures and those generated by the software.
3. At CF0, compartment inert gas pressures were significantly greater than those of
NOAA/Hamilton, and in some cases approached the USN 'M' values.
4. At CF3, compartment inert gas pressures remained significantly higher than those of
NOAA/Hamilton.
5. At CF5, compartment inert gas pressures were, in general, slightly higher than those of
NOAA/Hamilton, but were similar.
6. There was very good agreement between the computers during the pressurization tests.
Table 1. Depth comparison of VR3
dive computers with master gauge.
Master
0
10
20
40
50
60
70
90
100
110
120
130.1
139
148.9
160
170
180
190
200
210.2
220
229
239
239.5
250
260.8
270
280
290
300
VR3 #1
0
10
20
40
50
60
70
90
100
110
120
130
140
150
161
170
180
191
200
210
221
230
240
240
250
260
270
280
290
300
Table 2. Depth comparison of HSE dive
computers with master gauge.
VR3 #2
0
10
20
40
50
60
70
90
100
110
120
130
140
149
161
170
180
191
200
210
221
230
239
240
250
261
270
280
290
300
Master
0
10
20
30
40
50
60
70
80
90
110
120
130
140
150
160
170
180
190
200
210
220
230
240
250
260
270
280
290
300
30
HSE #1
0
10
19
29
40
51
62
71
81
92
112
122
132
142
152
163
172
183
193
204
212
222
233
242
252
263
273
283
293
303
HSE #2
0
9
20
30
41
51
61
71
81
91
113
122
132
142
152
162
171
182
192
202
212
223
232
243
252
262
272
282
292
303
Acknowledgments
The authors would like to thank the following individuals who played a key role in acquisition of data
for this project: Jim Bostick, Bill Gordon, Eric Johnson, and Steve Urick.
References
Hamilton RW. NOAA-Hamilton Trimix Decompression Tables. Hamilton Research, Ltd., 80 Grove
Street, Tarrytown, NY 10591-4138, 2004; 188 pp.
HS Explorer Dive Computer Owner's Manual. Copyright 2000-2003, HydroSpace Engineering, Inc.
VR3 Dive Computer Operators Manual. Copyright Delta P Technology Ltd, 2004: V2.1CXR
Workman RD. Calculation of Decompression Schedules For Nitrogen-Oxygen and Helium-Oxygen
Dives, U.S. Navy Experimental Diving Unit Research Report 6-65, 1965.
31
32
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Physical Fitness of Scientific Divers: Standards and Shortcomings
Alison C. Ma1, 2 and Neal W. Pollock1, 2*
1
Center for Hyperbaric Medicine and Environmental Physiology, Duke University Medical Center,
Durham, NC 27710, USA
2
Divers Alert Network, 6 West Colony Place, Durham, NC 27705, USA
alisoncma@gmail.com, neal.pollock@duke.edu
Abstract
Scientific diving standards provide guidelines for program oversight, crew and equipment
requirements, and diver selection. Diver selection is primarily based on medical fitness and
physical competency criteria. Physical competency is generally evaluated through certification
records and initial tests of swimming skills and in-water diving abilities. Continuation of active
status requires periodic medical evaluation, current emergency care certifications, and
documentation of diving activity exceeding minimum requirements. While individual
institutions and programs may require additional evaluations, formal requirements for periodic
assessment of physical fitness are notably absent from most parent standards. This may have
evolved from an expectation that medical evaluation would adequately address physical fitness
or that inadequate physical fitness was not an issue in the scientific diving community.
Problematically, medical evaluation may not provide an effective evaluation of physical fitness,
and physical fitness within the scientific diving community cannot be assured given societal
trends toward decreasing fitness. This paper reviews the physical fitness-related content found
in a cross-section of institutional, national and extra-national scientific diving standards and
offers suggestions for formalizing periodic evaluation of physical fitness.
Introduction
The evolution of labor-saving and communication technologies has increased the prevalence
of sedentary lifestyles in developed countries. Increased access to a wide range of convenient
and calorically dense foods has challenged efforts at personal restraint. Despite the known
benefits of physical activity, more than 50% of American adults are not active enough to
provide health benefits, and 24% are classified as completely inactive (CDC, 2007).
Assessing physical fitness on a large scale is impractical, but surrogate measures that provide
estimates on a population scale are useful. Body mass index (BMI) is one such estimator. BMI is not a
measure of body composition but a simple integration of height and weight (BMI = weight in kg /
[height in m]2) used to assign individuals to categories presumed to reasonably describe fatness.
Categories include 'normal' (18.5-24.9 kg·m-2), 'overweight' (25.0-29.9 kg·m-2) and several
subcategories of 'obesity' (≥30.0 kg·m-2). While BMI categorizations penalize individuals with welldeveloped muscle mass, the technique is useful for large-scale studies when more sophisticated
measures are unavailable.
According to BMI estimates, the number of American adults considered obese has more than doubled
from 1960 to 2004 (Figure 1) (CDC, 2006). Even allowing for some miscategorization, this is a
disturbing trend.
33
40
30
Percent
20
of
Population
10
0
1955
1965
1975
1985
1995
2005
Year
Figure 1. Trends in BMI of adults categorized as obese (30.0 kg·m-2 or greater) among
American adults 20-72 years of age from 1960 to 2004 (CDC, 2006).
The preponderance of high BMI values is also evident in mortality data from the North American
recreational diving population (Vann et al., 2004; 2005; 2006). Data from 2002-2004 (n=199)
identify 48% of deceased divers as obese and an additional 32% as overweight (Figure 2).
70
Normal
Overweight
Obese
60
18.5 - 24.9 kg·m-2
25.0 - 29.9 kg·m-2
30.0 kg·m-2 or greater
50
Percent
of
Fatal
Cases
40
30
20
10
n=49
n=61
n=89
2002
2003
2004
0
Year
Figure 2. Classification of DAN recreational diver fatalities by BMI for 2002, 2003 and 2004
(Vann et al., 2004; 2005; 2006)
Similar trends, if present within the scientific diving population, might threaten divers' physical
capabilities to safely perform under normal and emergent conditions. Since scientific divers
experience similar societal pressures, concern is warranted. Problematically, minimal information is
34
available regarding physical fitness of scientific divers. This paper will consider existing standards
and discuss possible alterations in practice and regulation concerning diver physical fitness.
Methods
Physical fitness is frequently defined in terms of cardiovascular and muscular endurance, muscular
strength and flexibility. A range of proficiency tests can address these elements of physical fitness.
A review of major institutional, national and extra-national diving standards was conducted to
identify the physical fitness-related components. Documentation was found for: American Academy
of Underwater Sciences (AAUS) (2006); Australia/New Zealand Occupation Diving: Scientific
Diving (AS/NZ) (2002); Canadian Association for Underwater Sciences (CAUS) (1998); Canadian
Coast Guard (CCG) (S. Simms, pers. com.); US Coast Guard (USCG) (2004); US Environmental
Protection Agency (EPA) (2004); US National Oceanic and Atmospheric Administration (NOAA)
(2003); and the World Underwater Federation (CMAS) (Flemming and Max, 1996).
Requirements for physical fitness testing were considered in four areas:
1. Initial certification
2. Return to dive after illness or injury
3. Renewal of active status after lapse in minimum diving activity
4. Recurrent evaluation
Physical fitness elements were found in various sections addressing swimming skills, rescues and
miscellaneous relevant clauses.
Results
Initial certification
Swimming skills for six of the eight reviewed standards are summarized in Table 1. Six specific skills
included length swimming, surface kick, underwater swim, treading water/survival float, head-first
surface dive, and entry or exit from water.
Swim distances ranged from 219 yd (200 m) to 550 yd (500 m). AAUS (2006) was the only scientific
standard incorporating a time component for a surface swim: 400 yd (366 m) in less than 12 min (33
yd·min-1or 30 m·min-1). EPA (2004) offered the option of completing either a 250 yd (229 m) swim or
a 440 yd (402 m) swim in scuba gear.
Coast Guard documents provide relevant information regarding physical fitness testing but are
considered separately since they are not scientific diving standards. CCG (S. Simms, pers. com.)
required a fin swim of 656 yd (600 m) in 15 min (Table 2). The USCG (2004) required a finless 500
yd (457 m) swim in 14 min, completed as part of a continuous sequence with push-ups, sit-ups, and a
run (Table 3).
Surface kicking requirements varied from either using snorkels or wearing full scuba gear. The
distance obligation ranged from 219 yd (200 m) (AS/NZ, 2002) to 875 yd (800 m) (Flemming and
Max, 1996). Surface kick distances for AAUS (2006) and CAUS (1998) were nearly double that of
AS/NZ (2002) (Table 1).
35
Underwater breath-hold swims were without gear for AAUS (2006) and NOAA (2003) standards.
EPA (2004) listed a slightly shorter distance but with the candidate wearing scuba gear with a closed
air supply (Table 1).
Treading water/floating requirements ranged from 10 to 30 min, with one standard providing an
option of two minutes of treading water with both hands remaining out of the water (AAUS, 2006)
(Table 1).
A head-first surface dive to 10 ft (3 m) was included in three standards. CAUS (1998) and EPA
(2004) required that an object or weight, respectively, be retrieved from the bottom as part of the
effort; AAUS (2006) did not specify retrieval (Table 1).
The method of water entry and exit in open water or surf ranged from completing the skill either
donning full gear (AAUS, 2006; EPA, 2004) or exiting a boat by using a ladder (EPA, 2004) (Table
1).
Rescue performance was mentioned in five documents (Table 4), although the requirements were
vague for AAUS (2006), CAUS (1998) and AS/NZ (2002). The detail provided by AAUS (2006), for
example, was limited to 'rescue and transport victim.' Obligatory rescue tow distance ranged from 25
yd (23 m) to 109 yd (100 m). EPA (2004) specified a 25 yd (23 m) tow or 50 yd (46 m) in full scuba
gear. Coast Guard requirements for rescue performance were extensive and not reviewed in this
paper.
Table 1. Swimming skill standards required for initial certification¹
Skill
Swim
units
yd
(m)
CMAS
-
AAUS
4003
(366)3
CAUS²
219
(200)
NOAA
550
(500)
EPA
250
440
(229) or
(402)
in gear
AS/NZ
219
(200)
Surface
Kick
yd
(m)
8754
(800)4
snorkel
400
(366)
in gear
437
(400)
snorkel
-
-
219
(200)
in gear
UW BreathHold Swim
yd
(m)
-
25
(23)
-
25
(23)
-
Tread/Float
min
-
20
30
Head-first
Surface
Dive
yd
(m)
-
10, or 2
no hands
tread
10
(3)
17
(15)
in gear
15
float
10
(3)
p/u object
float
-
float
10
(3)
p/u weight
Entry/Exit
n/a
in gear
using boat ladder
¹ AAUS - American Academy of Underwater Sciences; CAUS, Canadian Association for
Underwater Sciences; NOAA - US National Oceanic and Atmospheric Administration; EPA US Environmental Protection Agency; AS/NZ - Australia/New Zealand Occupational Diving:
Scientific Diving
² Only one of the four skills listed is required to fulfill the CAUS standard, in addition to a
required rescue tow of 109 yd (100 m)
3
Requirement = 400 yd (366 m) in less than 12 min (minimum 33 yd·min-1 or 30 m·min-1)
4
Guideline = 875 yd (800 m) in 16 min (minimum 55 yd·min-1 or 50 m·min-1)
36
10
tread
-
-
Table 2. Canadian Coast Guard (CCG) physical fitness requirements1
(S. Simms, pers. com.)
1
Skill
656 yd
(600 m)
Fin Kick
Requirement
<15 min
(44 yd/min or 40 m/min)
40 Sit-ups
2 min
20 Push-ups
Equal rate up and down
5 Chin-ups
Overhand
Minimum standard required for initial certification and annual requalification
Table 3. US Coast Guard (USCG) physical fitness skills to be completed in continuous
sequence1
(USCG, 2004)
Skill
500 yd
(457 m)
Swim
10 min rest
Requirement
<14 min
(36 yd·min-1 or 33 m·min-1)
42 Push-ups (minimum)
2 min rest
50 Sit-ups (minimum)
2 min rest
1.5 mi
<12:25 min:sec
(2.4 km)
(8.3 min·mi-1 or 5.3 min·km-1)
Run
1
Minimum standard required for initial certification and annual requalification
Table 4. Rescue standards required for initial certification¹
Skill
Rescue
Obligation
Transport
AAUS
Self;
Diver
Rescue and
transport
victim
CAUS
Self;
Diver at
surface and
UW
Accident
management
and evacuation
NOAA
EPA
AS/NZ
Science divers
none1
2
Diver
recovery to
surface
Science divers
none1
2
Recovery of
diver from
water
25 yd (23 m)
109 yd (100 m) Science divers 25 yd (23 m) or
50 yd (46 m)
person of
none1
in gear
equal size
¹ Working Diver: rescue skills are covered in a three-week Working Diver course
2
No distinction between scientific and working divers: rescue skills were to be covered in
either a one-week EPA course or a three-week NOAA Working Diver course
Rescue Tow
37
-
NOAA (2007) recently established age- and gender-indexed fitness tests (Table 5). Push-up
requirements decreased with age, from a maximum of 37 to 10 for males and from 16 to 4 for
females. Sit-up requirements decreased with age from a maximum of 45 to 25, irrespective of gender.
NOAA (2007), USCG (2004; Table 2) and CCG (S. Simms, pers. com.; Table 3) were the only
standards reviewed that included separate strength/endurance tests in addition to tests of swimming
skill.
Table 5. NOAA push-up and sit-up criteria (D. Dinsmore, pers. com.)
Age
(y)
<25
26-30
31-40
41-50
51+
Male
37
32
25
20
10
Push-ups
Female
16
13
9
6
4
Sit-ups
All
45
40
34
27
25
Returning to dive after illness or injury
Physical fitness testing after illness or injury was not formally specified in any of the standards
reviewed. EPA (2004) stated that a diver may be asked to requalify or that physical examination may
be required 'after a serious accident, injury or illness at the discretion of the [unit diving safety
officer].' Several standards required physical examination after any major injury or illness (CAUS,
1998; NOAA, 2003; AAUS, 2006). AAUS (2006) also specified the need for evaluation after any
condition requiring hospital care. AS/NZ (2002) allowed for an increase in frequency of examinations
at the discretion of the medical practitioner.
Renewing active status after lapse in diving activity below minimum requirements
Physical fitness testing following a lapse of diving activity was not formally specified in any of the
standards reviewed. AAUS (2006) left renewal requirements to the discretion of the institutional
diving control board. CAUS (1998) did not address renewal. AS/NZ (2002) required divers to
complete a checkout program at the discretion of the diving safety officer after a six-month hiatus
from diving. NOAA (2003) called for checkout dives and any other requirements prescribed by the
unit diving safety officer after a six-week to six-month hiatus, a line office diving officer/fleet diving
officer approved requalification program for a six- to 12-month hiatus, and completion of a NOAA
Diving Program-approved refresher training program following breaks longer than 12 months. EPA
(2004) stated that divers not diving for more than 12 months may be required, at the recommendation
of the diving safety board or the unit diving officer, to attend a diver certification course in order to be
requalified for diving activities.
Recurrent evaluation
Formal requirements for ongoing physical fitness evaluation were minimal or vague in several of the
standards reviewed (CAUS, 1998; AS/NZ, 2002; EPA, 2004; AAUS, 2006). AAUS (2006) required
divers over 40 years to complete an exercise stress test during periodic medical evaluation if
considered at risk for heart disease. AS/NZ (2002) stated that divers should ensure that they are fit to
dive – with fitness being maintained by exercise and regular diving. EPA (2004) declared that divers
were to dive only if physically and mentally fit and that they were to maintain a level of fitness
38
compatible with safe diving operations and be willing to retake the swim test any time. CAUS (1998)
did not state requirements.
Other standards provided more specific requirements for ongoing physical fitness testing (Flemming
and Max, 1996; NOAA, 2003; USCG, 2004; CCG, S.Simms, pers. com.). CMAS (Flemming and
Max, 1996) provided the most detail, recommending that divers be able to complete an 875 yd (800
m) snorkel swim in 16 min, and that "…the senior diver in charge should consider the fitness of
personnel involved, taking into account the recent diving operations or sporting activities…," and that
divers should be given a series of swimming and snorkeling exercises in the weeks proceeding field
work if they are unfit. NOAA (2003) required an annual refresher course which included in-water
rescues. USCG (2004) and CCG (S.Simms, pers. com.) required divers to successfully complete the
initial certification tests annually. In addition, CCG (S. Simms, pers.com) required a recertification
course every three years, and declared that divers were expected to maintain fitness compatible with
safe diving operations. Divers could be evaluated, and refusal could result in restricted diving activity.
Discussion
The normal demands of diving may well be met by a modest level of physical fitness. Every diver,
however, must be prepared to meet exceptional physical demands in emergent conditions. The
inability to predict the magnitude of such demands requires preparation for the worst of
circumstances, in part by maintaining a superior level of physical fitness. Good physical fitness can
improve the outcome of many situations. Successful rescue of self or a teammate may depend on an
individual's physical fitness reserve - the difference between any given effort and maximal capacity.
The paucity of recurrent evaluation requirements within the diving community is notable. Programs
with military origins demonstrate the most stringent oversight. Those with historical ties to the
recreational community, such as scientific diving, mandate somewhat less oversight. The recreational
community provides lifetime certification to divers with no recurrent evaluation of skills, medical
health or physical fitness. While the scientific diving community addresses medical health with
recurrent evaluation, and requires minimum diving activity requirements, programs generally fall
short in physical fitness evaluation.
The dearth of physical fitness assessment may have been based on an historic perception that selfgovernance within the community was adequate. Societal trends, however, suggest that selfgovernance may not be sufficient. Decreasing levels of physical activity and physical fitness threaten
performance capabilities of current and future divers.
The effects of aging are faced by even the most health-conscious of individuals. Maximal aerobic
capacity, a benchmark of physical fitness, may decline at an average rate of approximately one
percent per year beyond age 25 (Rosen, 1998). While dedicated training efforts may postpone and
reduce the rate of decline, a decrement of 0.5 percent per year may still be expected (Marti and
Howald, 1990).
Recognition of the age-related decline in physical fitness may have contributed to the recent NOAA
decision to adopt an age-indexed performance scale (D. Dinsmore, pers. com.). Practically, an
expectation of lower physical fitness in an older versus a younger diver may be compensated for by
greater experience, skill and economy. This assumes, however, that the older diver has greater
experience. Another societal trend is the pursuit of multiple occupations over a lifetime and delayed
retirement. Professionals who are mid-life or later and starting work in scientific diving may not have
the accumulated skills of the longtime diving professional.
39
Physical performance standards should establish at least a minimum capability to complete relevant
work. Reasonably, one standard should be applied to all persons. Age- and gender-indexed physical
fitness tests do not function in this way. The NOAA (D. Dinsmore, pers. com.) push-up standard, for
example, allows a 51st birthday to halve the requirement for males (from 20 to 10). Using a wellestablished standard for reference (Table 6; Pollock et al., 1978), the NOAA scale requires the
youngest individuals to achieve push-up performance classed as 'average' and 'fair' for males and
females, respectively. The required performance is 'fair' for both genders in the 50 year old range. The
NOAA requirements do not engender faith in the physical preparedness of candidates. The lack of
demand for absolute strength in women is most marked. It should be remembered that the strength
required to complete a push-up is proportional to mass. Smaller individuals already have an
advantage for this reason.
Table 6. Standard values for push-up endurance (Pollock et al., 1978).
Rating
Men
Excellent
Good
Average
Fair
Poor
Women
Excellent
Good
Average
Fair
Poor
20-29
30-39
Age (y)
40-49
50-59
60+
>54
45-54
35-44
20-34
<20
>44
35-44
25-34
15-24
<15
>39
30-39
20-29
12-19
<12
>34
25-34
15-24
8-14
<8
>29
20-29
10-19
5-9
<5
>48
34-48
17-33
6-16
<6
>39
25-39
12-24
4-11
<4
>34
20-34
8-19
3-7
<3
>29
15-29
6-14
2-5
<2
>19
5-19
3-4
1-2
<1
Designing tests for initial or recurrent evaluation of physical fitness requires careful consideration.
The available standards provide a range of elements to draw from to develop effective evaluation.
The most appropriate assessments will create a reasonable approximation of the working
environment. Field-based tests are generally most effective. A continuous, serial skill assessment is
probably most representative of real-world demands, a concept evident in the USCG (2004)
standards. The specifics, however, may fall short of optimal. Allowing a 10 min rest after the swim is
inconsistent with the demands of a real rescue situation where one 'skill' after another may be required
to succeed in an emergent condition with minimal rest throughout. Immediate transition from one
phase into the next may be more appropriate.
Accidents normally manifest when an accident chain - the series of sometimes individually small
issues - grows to the point of failure. Testing should present a chain of events that challenge
participants and provide reminder that successful efforts to break the chain can render non-events out
of situations that could otherwise end in disaster.
Physical fitness testing in open water may be most appropriately administered at the end of a standard
working dive, to start the process with a normal level of fatigue. If the dive and physical fitness tests
are being conducted strictly for evaluation, the dive phase could include a review of basic diving
skills or relevant operational task activity. We propose a test sequence for consideration (Table 7).
The specific elements, distances and durations could be accepted or modified as appropriate.
40
The open water physical fitness test sequence would begin with a 219 yd (200 m) surface swim in full
gear (AS/NZ, 2002), followed by a 109 yd (100 m) rescue tow (CAUS, 1998), with both victim and
rescuer in full gear. The tow would be followed by an unassisted removal of the victim onto a boat,
dock or shore as appropriate for typical operations. Adding to the realistic nature of the simulation,
the tow would be accompanied by verbalized decision-making regarding victim management and
simulated effort to request aid. The removal would be followed by simulation of basic life support
checks of the victim and deployment of emergency oxygen equipment. After deployment of the
oxygen equipment, the diver would complete 15 full deflection, military-style push-ups without
stopping.
Table 7. Proposed physical fitness test for scientific divers to be completed in continuous sequence.
Open Water Testing Scenario1
219 yd (200 m) surface swim in full gear
109 yd (100 m) rescue tow (both in full gear)
Beach/Dock/Boat removal of victim
Basic life support simulation
30 (male) / 20 (female) military-style push-ups
1
immediately following working dive
2
immediately following underwater skill drills
Pool Modifications2
328 yd (300 m) surface swim (full gear; no suit)
220 yd (200 m) rescue tow (full gear; no suit)
Poolside removal of victim
Victim/Rescuer wear 15 lb (7 kg) weight belts
Same
Same
The strengths of the above series are the relevance to the normal working environment, the integrated
use of standard emergency equipment, and the test of physical capacity. A diver complaining about
the relevance of push-ups after completing the rescue can easily be reminded that the need for postremoval evacuation of the victim may reasonably require additional exertion. The point to drive home
is the importance of an adequate physical reserve to deal with the any potential demands of an
emergent situation. Eliminating the age-indexing on the push-up requirements will certainly be
challenging to some, but this is also a component that individuals can easily prepare for
independently. Establishing a challenging basic standard can encourage divers to keep track of their
fitness.
Field-based tests will not be appropriate for all programs. Appropriate field conditions may not be
available in all locations, such as the land-locked home base of a dive team. Seasonal restrictions may
also render suitable sites unavailable. For these reasons, alternative physical fitness tests should be
available.
Pool tests can provide a convenient environment to evaluate physical fitness. Completing the test
scenario in the pool, however, reduces the applicability to the normal working environment by
removing some of the natural stressors. The absence of current and minimal wave activity are two
examples. Temperate-water divers will also not be able to wear the normal protective suits because of
excessive thermal loading. Since a normal working dive will not precede the test scenario, the
poolside test could follow a series of underwater skill drills. Where possible, the test should remain
relevant to operational diving. Length swimming may not be suitable since it is more of a test of
watermanship than physical fitness. While superior watermanship is important and should be
promoted, its place is in building or maintaining, rather than evaluating, physical fitness.
The test sequence for the pool could be modified by increasing both the surface swim distance to 328
yd (300 m) and the rescue tow distance to 220 yd (200 m) (Table 7). The poolside removal could be
41
completed with both victim and rescuer wearing 15 lb (7 kg) weight belts in order to partially
simulate the load of equipment normally worn or the physical stress of environmental conditions that
a diver may experience during a rescue. The weight belts would be donned immediately prior to the
rescuer exiting the water and removed after both are on deck. Requiring the rescuer to exit the water
using a pull-up, not by using a ladder, would provide a better test of upper body strength (note: the
weight belt could only be used with decks close to flush with the water). The standard basic life
support simulation and push-up test would follow.
Another option is the exercise stress test, a test that is well established in the scientific diving
standards to evaluate divers over 40 years of age if considered at risk for heart disease (AAUS, 2006).
While normal physician examination does not evaluate physical fitness, the exercise stress test could
be used to assess exercise capacity.
Re-evaluation would reasonably be required on an annual basis in order to establish that physical
fitness has not fallen below the minimum standard. An ancillary benefit of scheduled testing is that
individuals will have additional motivation to maintain their physical fitness. Recommendations for
maintenance programs should be provided to individuals where possible.
Conclusions
An excellent record of safety has been enjoyed by the scientific diving community. The lack of
recurrent physical fitness evaluation, however, is a weakness in the current system of oversight.
While alternative forms of evaluation such as clinical exercise stress tests and pool tests should be
permitted, the preferred standard would be an open-water field-based test completed as a continuous
sequence of skills following a typical working dive. The test sequence proposed in this paper includes
a surface swim, surface rescue tow, water removal, basic life support and emergency communications
and a final measure of strength and endurance with push-ups. The test sequence can be used to
evaluate current physical fitness and skill levels, promote fitness consciousness and provide an
opportunity for health and safety-related dialogue within the community. Each of these elements is
believed to be important to ensure continued readiness and responsible oversight.
References
Australian/New Zealand Standard. Occupational Diving Operations Part 2: Scientific Diving. Sydney,
Australia and Wellington, New Zealand: Standards Australia/Standards New Zealand, 2002; 73 pp.
Coast Guard Diving Policies and Procedures. United States Coast Guard, 2004.
EPA Region 10 Diving Safety Manual, rev 2. U.S. Environmental Protection Agency, 2004: 1-6.
Flemming NC, Max MD, eds. Scientific Diving: a General Code of Practice, 2nd ed. Flagstaff, AZ:
Best Publishing, 1996; 278 pp.
Marti B, Howald H. Long-term effects of physical training on aerobic capacity: controlled study of
former elite athletes. J Appl Physiol. 1990; 69(4): 1451-1459.
National Center for Health Statistics, 2006 with Chartbook on Trends in the Health of Americans.
Hyattsville, MD: US Centers for Disease Control and Prevention 2006: 54-57.
42
NOAA Administrative Order 209-123. Seattle, WA: National Oceanic and Atmospheric
Administration Diving Program, 2003; 24 pp.
Physical Activity and Good Nutrition – Essential Elements to Prevent Chronic Diseases and Obesity.
US Centers for Disease Control and Prevention 2007; 4 pp.
Pollock ML, Wilmore JH, Fox SM. Health and Fitness through Physical Activity. New York, NY:
John Wiley & Sons, 1978.
Rosen MJ, Sorkin JD, Goldberg AP, Hagberg JM, Katzel LI. Predictors of age-associated decline in
maximal aerobic capacity: a comparison of four statistical models. J Appl Physiol. 1998; 84(6): 21632170.
Standard of Practice for Scientific Diving, 3rd ed. Canadian Association for Underwater Science,
1998; 38 pp.
Standards for Scientific Diving. Nahant, MA: American Academy of Underwater Sciences, 2006; 76
pp.
Vann RD, Denoble PJ, Dovenbarger JA, Freiberger JJ, Pollock NW, Caruso JL, Uguccioni DM.
Report on Decompression Illness, Diving Fatalities, and Project Dive Exploration. Durham, NC:
Divers Alert Network, 2005: 70.
Vann RD, Denoble PJ, Dovenbarger JA, Freiberger JJ, Pollock NW, Caruso JL, Uguccioni DM.
Report on Decompression Illness, Diving Fatalities, and Project Dive Exploration. Durham, NC:
Divers Alert Network, 2004: 79.
Vann RD, Freiberger JJ, Caruso JL, Denoble PJ, Pollock NW, Uguccioni DM, Dovenbarger JA, Nord
DA, McCafferty MC. Annual Diving Report. Durham, NC: Divers Alert Network, 2006: 45-47.
43
44
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Scientific Diving Safety: Integrating Institutional, Team and
Individual Responsibility
Neal W. Pollock1,2
1
Center for Hyperbaric Medicine and Environmental Physiology, Duke University Medical Center,
Durham, NC 27710, USA
2
Divers Alert Network, 6 West Colony Place, Durham, NC 27705, USA
neal.pollock@duke.edu
Abstract
Scientific diving enjoys an enviable safety record. An unusual cluster of accidents occurred over
an eight-day period in August 2006, when three incidents associated with disparate science
operations resulted in four fatalities. The first involved a 37-year-old graduate student in
Florida, freediving to hand-capture an estimated 300-lb green sea turtle. The second involved a
56-year-old researcher in New Jersey, diving from a research vessel to install a sensor on an
underwater observatory platform positioned 50 ft underwater. The third involved two US Coast
Guard divers (31 and 22 years-of-age), diving from a Coast Guard icebreaker, 500 miles north
of Barrow, Alaska. The collection demonstrates how easily seemingly reasonable situations can
lead to incidents. Safety depends upon appropriate regulations, crew training, operational
practices, individual capability and individual responsibility. Review of the three fatal incidents
is used to frame the need for active strategies to reinforce the complementary function of
institutional, team and individual responsibilities in scientific diving.
Introduction
Scientific diving in the United States currently enjoys a relative autonomy from external regulation.
The exemption from commercial diving standards was issued in 1982 after extensive lobbying by the
scientific diving community. The success of the petition was due, in large part, to the excellent safety
record of scientific diving. Continued freedom relies upon the continuation of that enviable record.
This requires the commitment of every individual involved in or responsible for any aspect of
scientific diving.
An unusual cluster of accidents occurred over an eight day period in August 2006. The three
incidents, each related to a science operation, resulted in a total of four fatalities. The collection
provides a powerful example of how easily seemingly reasonable situations can lead to incidents. The
cases are used to discuss strategies to reinforce the complementary nature of institutional, team and
individual responsibility in operational safety discussed.
August, 2006 Case Reports
August 10 - 3 mi (5 km) south of Sebastian Inlet, Florida.
A 37-year-old male graduate student failed to surface after a breath-hold dive attempting to handcapture an estimated 300-lb (136 kg) green sea turtle to bring it to the surface for study.
45
Background: A group of turtle researchers was conducting a turtle survey from small vessels. The
goal of one graduate student was to develop a project to study male sea turtles that, unlike females, do
not come into shore for nesting. The technique for hand capture of turtles was well established: A
swimmer grabs the carapace fore and aft from above and behind, levers the aft end down, and directs
the swimming turtle upwards. Additional swimmers are usually available to provide assistance. Scuba
is not worn, as it makes the diver too slow to complete the capture. In this case, the experienced
victim jumped into the water from the boat when he saw the turtle just below the surface. He held on
and was taken underwater. Emergency services were contacted when the victim did not surface within
four minutes.
Outcome: The events occurring underwater are unknown. There was not a second swimmer in the
water watching the events. The victim's body was found four days later, with no wounds evident.
August 14 - offshore of Atlantic City, New Jersey.
A 56-year-old male diver died after entering the water from a research vessel to install a sensor on an
underwater observatory platform positioned 50 ft (15 m) underwater. The problem developed shortly
after the diver entered the water. He was unresponsive during the initial communication check.
Outcome: The events occurring underwater are unknown. Cause of death was attributed to a preexisting medical condition.
August 17 - 500 mi (800 km) north of Barrow, Alaska.
A 31-year-old female and a 22-year-old male died while conducting a familiarization dive from the
surface of the ice alongside a 420-ft (128-m) US Coast Guard icebreaker. The vessel was conducting
science missions for the US National Science Foundation.
Background: The dive coincided with a period of ice liberty, when normal operations were
suspended. It was approved for familiarization purposes. The tenders supporting the dive were largely
inexperienced, with most of the training they received provided immediately prior to the dive.
The female had conducted only seven polar dives, all surface-supplied. The male had completed no
polar dives. Each diver wore approximately 60 lbs (27 kg) of weight. Each was tended on a separate
line. Communications with the surface were to be maintained by line signals. The plan was to dive to
a depth of 20 fsw (6 msw). A third diver experienced a suit leak and left the site to get changed
aboard the ship.
Once below the ice, both divers descended extremely rapidly, reaching maximum depths of 187 and
220 fsw (57 and 67 msw), respectively. The pull on the tether generated by each overweighted diver
was mistakenly accepted as repeated 'okay' signals. Concern of surface personnel was voiced but not
acted upon until most of the tether line had run out and the third diver returned to the site. The divers
were pulled to the surface, initially at a rate of approximately one foot per second, and then as rapidly
as possible. The divers reached the surface after approximately 10 min underwater.
Outcome: The divers' air supply was found to be exhausted or close to exhausted. Cause of death was
designated as asphyxia and pulmonary barotrauma. Critical errors were attributed to both command
and the dive team. Command errors included inadequate oversight and lack of familiarity with Coast
Guard and Navy rules. Crew errors included inadequate crew complement, inadequate crew
training/experience, deviation from plan, and gross overweighting.
46
Case Insights
While each of the three incidents described above were associated with scientific research programs,
none fell under the classic definition of scientific diving. The first incident involved freediving,
distinct due to the lack of a pressurized gas supply, and unregulated by most research diving
programs. The second situation was more appropriately classified as a working dive. The third
problem arose on a dive conducted for training that was completely independent of the scientific
mission of the cruise.
The problem, though, is not of definition, but responsibility. Aquatic activities conducted in support
of science, direct or indirect, require oversight by diving control boards and diving safety officers.
This makes sense if for no other reason than to provide the benefit of the experience of scientific
diving leaders to the user group. Techniques to hand-capture turtles are well tested. Proposals to do so
as a solo activity, however, would have led to discussions that might have altered the outcome in this
case. While few enjoy having their activity regulated, planning that includes those with professional
safety experience can be beneficial. It is far better to have this communication before an accident
occurs and with an authority trying to help researchers achieve their goals. In this case, the presence
of a backup swimmer and a deployable buoyancy device may have made a critical difference to the
outcome.
The second incident was attributed to a medical fitness issue. While fitness issues can be exacerbated
by exceptional conditions, they can also be exacerbated by relatively benign ones. The shift of blood
to the thorax resulting from immersion and the added resistance associated with breathing through a
regulator are both physiological stressors faced by divers. The speed with which the problem
developed in this case suggests that these normal stressors may have played the dominant role.
Medical screening identifies many at-risk persons, but serious problems can arise even after problemfree comprehensive physical exams. The risks of unexpected collapse can be reduced by ensuring that
participants remain physically active and fit. Realistically, however, this risk will never be eliminated.
It may even increase since the population of divers in general is seen to be aging. Problematically,
patterns of medical fitness within the scientific diving community are not commonly reported and
physical fitness is generally not well monitored after initial evaluation (Ma and Pollock, 2007).
The third incident demonstrates how things can still go wrong in a system with well-established rules
that include multiple levels of oversight and responsibility. Violations could have stopped the fatal
dive operation at several points before it was carried out. The fact that it was not stopped may reflect
an assumption of protection given the other levels of approval. To its credit, the Coast Guard
conducted a full investigation of the incident and openly released the findings. The findings confirm
that having regulations is not enough. In this case, the chain of command likely obfuscated individual
responsibility and led to calamity. Too many individuals were not sufficiently aware to ask the right
questions, and a faith in the system deterred action.
Programming Safety
Once solely the concern of the individual, societal norms have shifted much of the responsibility for
safety to higher authorities. This is generally effective for product safety and public safety but in no
case should it release individual responsibility. Safety programs are most effective when each level,
from institution to individual, actively support initiatives (Figure 1). For scientific diving, it is the
umbrella organization, such as the American Academy of Underwater Sciences (AAUS), that brings
together experts in operations and safety to develop general standards of practice appropriate for the
work and able to meet all mandatory regulatory codes. The resultant documents serve as the parent
47
standard. Institutions review the parent standard and make modifications appropriate to the demands
of their operating circumstances. Since most in the administration will not have the necessary
expertise, the diving control board, typically acting through the diving safety officer and sometimes
with the help of outside experts, is tasked with ensuring that the right protections are in place. The
key responsibilities of the institution are to ensure that appropriately trained personnel are available to
develop the institutional standards and that adequate support is provided so that the standards can be
met effectively by all workers.
Professional Organization
Regulatory Authority
Institution
Diving Control Board
Diving Officer
Dive Team
Expert
Team Leader
Individual
Figure 1. The hierarchical structure of scientific diving regulation.
The diving control board, again generally acting through the diving safety officer and possibly
additional experts, is responsible for ensuring that dive teams and individual divers are both capable
and adequately trained to understand and follow the standards. The dive team leader assumes the role
of the safety supervisor operationally, when the diving safety officer will likely not be present. The
dive team works collectively to support the team leader and each other.
The ability to perform safely ultimately comes down to the capabilities, training and support of the
individual diver and his or her awareness and motivation. Having rules is not enough to ensure safety:
The individual must know the rules and have the capacity to follow them appropriately.
Understanding the rationale behind any rule is important to having it followed. Regulations deemed to
be of little personal value will typically be the first ones ignored. Each person must be committed to
the system to achieve adherence. The individual is influenced by a range of resources (Figure 2). One
of the most critical roles of the diving safety officer is to help the individual diver reconcile input
from these sometimes disparate resources.
Hazards to Safety
Hazards to safety begin with the educational process. A generational loss of knowledge can
compromise a rising group. Problems that had to be overcome by an earlier generation may not reach
the level of awareness of a subsequent generation. This can leave the less informed at a disadvantage
in some circumstances. Reductionism can create a similar threat. Diving education has progressively
been streamlined as equipment has become more reliable and easy to use and optimization of
teaching/learning techniques has become a subject of study. While streamlining can reduce the time
and effort necessary to master basic skills, it may also eliminate some of the struggle that serves to
develop a capacity for problem-solving in the student.
48
Time pressure is another hazard faced by scientific divers. The demand for high productivity in
multiple professional areas makes it increasingly difficult to allow as much time as desired for either
training or task accomplishment. While educational streamlining helps in terms of meeting paper
goals, a vicious circle can be established that leaves individuals less than well prepared.
Independent
Learning
Expert
Trainers
Regulatory
Authority
Diving
Safety
Officer
Individual
Institution
Dive
Team
Professional
Organization
Diving
Control
Board
Figure 2. Influences contributing to individual scientific diver safety
Loss of mentorship can be a problem arising from the streamlining of training, programs, staff
support and collegial interactions. Mentorship allows individuals to learn from the skills and mistakes
of others. This can have a great positive impact on the level of performance of the mentee and can
help to reduce the problem of generational loss of knowledge. Mentorship in scientific diving is
critical since so many of the challenges that can arise are idiosyncratic. The ability to effectively
problem-solve is often increased through mentorship relations.
The final hazard to be discussed is the faith in indestructibility most pronounced in (young) males and
those who gravitate to activities classified by society as (relatively) high-risk. Risk-takers frequently
know the rules and may even believe them appropriate for general application, but they also believe
that their personal abilities preclude the necessity to be restricted. The hazards are compounded when
risk-taker mentality is passed on to someone who may not share the same capabilities. Those in
leadership positions must be extremely sensitive as to how their attitudes and actions may be
interpreted by those around them. Safety, or lack thereof, starts with attitude. Professionalism may
dictate masking personal proclivities to reduce the risk of negative impact on teammates or programs.
Educating for Safety
Knowledge is critical to support an attitude for safety. Opportunities must be provided to ensure that
all personnel can develop and maintain appropriate skills and awareness. Written standards are of
little value without the means to ensure that both the letter and spirit of the rules are adopted.
The standard core for personnel training is the clinic format. A modular structure of clinics can
provide valuable regular contact. Topic-specific information is exchanged and the opportunity for
opportunistic exchange is created. Issues or questions that may otherwise lie fallow will often be
49
brought out. This can be valuable for the individual and the group. Workshops and symposia provide
similar benefits in facilitating exchange within the wider professional group. Collaboration with
related organizations to develop special-topics workshops provides an excellent opportunity to pool
experience.
Alternatives are needed to meet the evolving reality of current professional life. The computer-based
focus of many professionals makes electronic communications desirable and cost-effective. Brief
summaries of primary or recapped information can keep individuals thinking about scientific diving
even if in phases of minimal activity. Including a variety of items - news, incident reports, insights,
events, links to related sites - can make the reading interesting to most. Releasing the summaries on a
monthly basis may provide the best balance between timely delivery and parsimonious use of
distribution lists.
Web-based training (WBT) programs offer another powerful modern tool. Continuing education
opportunity can be delivered to the individuals. While online training should not replace the
workshop structure in a skills-oriented field such as scientific diving, it provides a great means to
augment training. Obligatory training can be scheduled, graded and tracked automatically. Online
training can also be used to provide background prior to clinics. This can help to ensure that the time
spent in the clinic environment is appropriately spent.
WBT modules can also be useful in promoting awareness across the institutional hierarchy.
Administrators that might not attend clinics or symposia might be willing to review computer-based
materials. This can help to keep the needs of the program within the focus of upper administration.
There is a cost, primarily time, to develop WBT modules. The advantage, however, comes in the
minimal forward cost. Access can be provided to advertise or help to justify the program,
modifications can be made at little or no expense and, in some cases, and creation of a revenue
generating stream is possible.
Conclusions
We are fortunate to have an excellent safety record for scientific diving. Protecting this record
requires vigilance in the face of evolution and inevitable pressures on time and budgetary support.
Technology and collaborations can facilitate opportunities for individual learning and integration
throughout the organizational hierarchy.
Reference
Ma AC, Pollock NW. Physical fitness of scientific divers: standards and shortcomings. In: Pollock
NW, Godfrey JM, eds. Diving for Science 2007. Proceedings of the American Academy of
Underwater Sciences 26th scientific meeting. Dauphin Is, AL: AAUS, 2007: 33-43.
50
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Using SCUBA and Snorkeling Methods to Obtain Model Parameters for
an Ecopath Network Model for Calabash Caye, Belize, Central America
Rebecca A. Deehr1, Deirdre B. Barry2, 4, David D. Chagaris2, 5 and Joseph J. Luczkovich1, 2, 3
1
Coastal Resources Management Program, East Carolina University, Greenville, NC 27858
Department of Biology, East Carolina University, Greenville, NC 27858
3
Institute for Coastal & Marine Resources, East Carolina University, Greenville, NC 27858
4
Present address: 6064 Steeplechase Dr., Roanoke, VA 24018
5
Present address: Florida Fish and Wildlife Research Institute, St. Petersburg, FL 33703
RAD: rad0708@ecu.edu; DBB: deirdre@deirdreandalex.com; DDC: dave.chagaris@myfwc.com;
JJL: luczkovichj@ecu.edu
2
Abstract
We have been collecting biological data for the construction of a food web model of Calabash
Caye, Belize using Ecopath with Ecosim software. Characterization of the mangrove, seagrass
and coral reef environments surrounding Calabash Caye will provide baseline data for a
network model of the trophic relationships in this area, which is currently being considered as a
marine protected area. Our Ecopath model follows a previously published Puerto Rico – Virgin
Islands 50-compartment model; 27 of the compartments represent fishes. The collection of fish
biomass and abundance data was done using visual surveys with both SCUBA and snorkel.
Important dietary information was gathered from fish collected by spearfishing, beach seine, gill
nets and hook and line. Smaller benthic invertebrates were collected using a small suction
dredge and larger invertebrates were counted along visual transects using SCUBA and
snorkeling. Additional data for various compartments that were not sampled, including birds,
phytoplankton, benthic autotrophs and detritus, come from other Belizean or Caribbean studies
and literature values. Abundance and biomass of all compartments were determined and
converted to g wet weight⋅m-2. Selected Ecopath outputs, such as effective trophic levels and
mixed trophic impacts, allow us to gain insight into trophic structure and how changes in
biomass travel through the food web to affect other species. The model will enable Belizean
scientists and managers to monitor and formulate predictions about potential ecosystem changes
that are associated with the marine reserve.
Introduction
The Mesoamerican Barrier Reef system is the second largest barrier reef in the world and provides a
220 km off-shore boundary that runs the length of Belize. As is common throughout the Caribbean,
mangroves and seagrass beds are intimately connected to the reef system (Bossi and Cintron, 1990;
Kitheka, 1997; Moberg and Rönnbäck, 2003; Mumby et al., 2004; Harborne et al., 2006; Mumby,
2006) and line Belize's eastern shoreline, supporting a variety of organisms that may eventually
inhabit the reefs. Each of these ecosystems provides valuable functions and services to the natural
environment as well as to humans (Costanza et al., 1997).
The Belize reef system is home to many commercially and recreationally important species, such as
groupers, snappers, tarpon, conch and lobster (Gardiner and Harborne, 2000). Marine target species
are sought after by artisanal, recreational and commercial fishermen. In many cases, the fisheries are
self-regulated; there are 13 registered cooperatives owned, operated and managed by the fishers
themselves (Gillet, 2003). Behind tourism and agriculture, fisheries represent Belize's third most
51
important industry (McField et al., 1996). In 1995, Belize exported $20 million BZ alone, with
lobster and conch contributing greater than 90% of the total value of exported seafood (McField et
al., 1996). Shrimp and finfish are also important, but to a much lesser extent.
Belize has been pro-active in protecting its coastal marine resources. Since 1982, Belize has created
thirteen marine protected areas (MPAs) (CZMAI, 2003). These MPAs exist along the coastline, but
several areas remain unprotected. The Turneffe Islands atoll has been identified as an area necessary
for the completion of a national protected area system (Meerman and Wilson, 2005). It has also been
nominated as a United Nations Educational, Scientific, and Cultural Organization (UNESCO) Man
and the Biosphere (MAB) Reserve (CZMAI, 2003). It is expected that the formation of an MPA or
MAB reserve at the Turneffe Islands will cause significant shifts in the trophic structure of the
ecosystem (Polunin and Roberts, 1993; Roberts, 1995). Without baseline data, it will be difficult to
track changes in trophic flows and food web structure. The collection of such data and construction of
a network model of trophic flows are the objectives of this study.
A network analysis model can be useful to predict multi-species interactions that may occur due to
changes in fishery harvests should the MPA be established. Ecopath is a network modeling software
package that was first developed by Polovina (1984a, 1984b) to mathematically estimate the standing
stock and production budget of the French Frigate Shoals ecosystem in the northwest Hawaiian
Islands. Ecopath has since been developed to its present form of Ecopath with Ecosim 5.1, which
combines the biomass budget approach of Polovina with network analysis theory (Ulanowicz, 1986)
for analyzing flows between compartments. It also includes time (Ecosim) and space (Ecospace)
dynamic simulation modules to investigate how the system will change over time in response to
policies on fishing and MPAs (Christensen and Pauly, 1992; Christensen and Walters, 2004;
Libralato et al., 2006).
Previously, a network model was developed for a Caribbean coral reef ecosystem, and we used this as
a base model. Opitz (1993, 1996) created a 30-year comprehensive network model of the Puerto Rico
– Virgin Islands (PRVI) coral reef ecosystem using Ecopath II software. The Opitz model served as a
starting point for the construction of our Calabash Caye food web model. We wanted to collect sitespecific data for Calabash Caye coral reefs, seagrass beds and mangroves to measure the abundance
and biomass of benthic invertebrates and fishes and to collect dietary data on some of the fishes. Our
goal was to provide managers with the data necessary to help make informed policy decisions
regarding Calabash Caye marine resources. The construction of this Ecopath network model will
represent a baseline condition of the ecosystem prior to establishment as a protected area, describe
trophic interactions among members of the Calabash Caye ecosystem, and identify any major gaps in
knowledge that should be addressed with further research.
Methods
Site Description
The Turneffe Islands are located on the second submarine ridge on the Belize continental shelf
(Figure 1a). The Turneffe Islands atoll is about 48 km long and 16 km wide at its widest point. This
formation of mangrove islands is 10–16 km east of the Mesoamerican Barrier Reef, with a nearly 300
m deep channel between barrier reef and the Turneffe Islands (Garcia and Holtermann, 1998). The
University of Belize research station, the Institute for Marine Studies (IMS), is located on Calabash
Caye, a small mangrove caye located on the southeastern side of the atoll, nearly 51 km off the coast
of Belize (Figures 1b and 1c). Researchers from East Carolina University (ECU) have been utilizing
the facilities at IMS since 1999 for a tropical marine ecology course, where several student research
52
projects have been conducted. Specific data from these projects (especially Rueter, 2004) have been
combined with new data collected in this study to create an Ecopath network model of the Calabash
Caye system.
Figures 1a, b and c. Maps of study areas: 1a) Map of Belize, Central America (from Magellan Geographix);
1b) Turneffe Islands Atoll (from BelizeNow Network); 1c) Calabash Caye examined from the northeast
(photo from Coral Caye Conservation).
A shallow central lagoon, filled with Thalassia and Halimeda spp., dark organic muds and some
small patch reefs, is about 500 m long and up to 8 m deep (Garcia and Holtermann, 1998). There is
very little flushing or circulation that occurs within the lagoon (Gischler, 2003). Mangroves cover
most of the caye, especially red mangrove Rhizophora mangle, black mangrove Avicennia germinans,
and white mangrove Laguncularia racemosa. Seagrass beds are dominated by Thalassia testudinum
with lesser amounts of Syringodium filiforme, Halodule wrightii and a variety of algal species, and
can be found between the mangrove-lined coast and the coral reefs. The coral reefs show typical West
Indian windward coral reef zonation patterns (Garcia and Holtermann, 1998).
Ecopath Model Parameters
Ecopath models are composed of compartments, or biomass pools, that can be represented by a single
species, life stage of a single species, or group of similar species. The basic input requirements for
each compartment are biomass, consumption, production, diet and unassimilated diet to consumption
ratio. The 50 compartment PRVI model created by Opitz (1996) using Ecopath II served as the
template for the Calabash Caye Ecopath model. Compartments were identified by group number
accompanied by the common names of representative members of that group. For a more complete
listing, see Table 1 for the representative members of each compartment. Specific information about
the compartments can be found below in the section on Biological Data Collections and in Opitz
(1996), Barry (2006) and Chagaris (2006). Site-specific measurements for fish and invertebrate
biomasses and fish dietary data were collected in mangrove, seagrass and coral reef habitats, and all
data were combined for the construction of one model of Calabash Caye. This model, constructed
53
using Ecopath with Ecosim 5.0, emphasizes an inclusive ecosystem rather than three independent
systems. Compartments not sampled at Calabash Caye (such as seabirds, turtles, phytoplankton) were
filled with data from the Opitz (1996) model or the diet study conducted by Randall (1967).
Table 1. Selected characteristics of the 50-compartment Calabash Caye ecosystem model, arranged by
effective trophic level.
Group
Number
28
2
4
18
29
12
1
5
13
15
16
17
7
14
27
3
6
8
31
34
19
21
32
30
36
45
39
46
35
40
43
20
22
33
38
37
41
9
11
42
10
23
24
25
26
44
47
48
49
50
Model Compartment Name
Representative Members
birds
Sea birds
Sharks/scombrids, carnivorous
sharpnose shark, bonnethead, mackerel, tunny
Intermediate jacks, carnivorous
palometa, bar/yellow/horse-eye jack, blue runner
Large groupers, carnivorous
black grouper, itajara
Squids
squids
Large reef fish, carnivorous
hogfish, barracuda, cubera snapper, moray eel
Large sharks/rays, carnivorous
southern stingray, eagle ray, blacktip shark, nurse shark
Small jacks, carnivorous
scad, ladyfish
Intermediate reef fish, carnivorous 3 trumpetfish, flounder, creole wrasse, tilefish, coney,
graysby, rock/red hind
silversides
Small schooling fish, pelagic
Engraulidae, herbivorous
anchovies
Small reef fish, carnivorous 2
fairy basslet, clown wrasse, harlequin bass
Large-intermediate schooling fish,
needlefish, herrings
pelagic
yellowhead wrasse, chromis, hamlet, razorfish
Small reef fish, carnivorous 1
Small Gobiidae, carnivorous
yellowline goby, sharknose goby
Large jacks, carnivorous
jacks, permit
Intermediate reef fish, carnivorous 1 bonefish, mutton snapper, schoolmaster, black margate,
angelfish
Intermediate reef fish, carnivorous 2 butterflyfish, mojarra, grunts, squirrelfish, porgy,
goatfish, sergeant major, lane/gray/mahogany snapper
Octopuses
octopi
Shrimps/hermit crabs/ stomatopods all shrimps, hermit crabs and stomatopods
Intermediate reef fish, carnivorous 4 filefish, angelfish, rock beauty
Small reef fish, omnivorous 2
beaugregory, yellowdamselfish, puffer
Lobsters
spiny lobsters
Sea turtles
sea turtles
Asteroids
sea stars
Corals/sea anemones
all corals and anemones
Chitins/scaphopods
chitons and tusk shells
Zooplankton
zooplankton
Small benthic arthropods
small crustaceans
Polychaetes/priapuloids/ophiuroids
segmented worms, penis worms, brittle stars
Ascidians/barnacles/ bryozoans
tunicates, barnacles, moss animals
damselfish, orange spotted filefish
goldspot goby
Crabs
all crabs
Gastropods
conch, snails and slugs
Echinoids
sea urchins
Holothuroids/sipunculids/
sea cucumber, peanut worms, echiuroid worms
echiuroids/hemichordates
Hemiramphidae, herbivorous
halfbeaks
Intermediate reef fish, herbivorous
surgeonfish, doctorfish, tang, durgeon
Bivalves
bivalved mollusks
Kyphosidae, herbivorous
sea chubs
Large Scaridae, herbivorous
rainbow/midnight/queen parrotfish
Intermediate Scaridae, herbivorous
princess/redfin/stoplight parrotfish
Small Scaridae, herbivorous
striped/redband parrotfish
Blenniidae, herbivorous
redlip blenny
Sponges
all sponges
Decomposers/microfauna
decomposing bacteria
Phytoplankton
photosynthetic plankton
Benthic autotrophs
seagrasses and algae
Detritus, POM, DOM
nonliving organic material
54
Effective
Biomass
Trophic Level (gWW/m2)
4.5
0.003
4.2
2E-09
3.9
1.233
3.9
2E-09
3.9
0.8
3.8
3.072
3.7
0.247
3.6
0.015
3.6
4.894
3.6
3.6
3.6
3.5
3.529
4.33
1.257
7.15
3.5
3.5
3.4
3.4
4.886
0.0000072
0.027
2.394
3.4
6.75
3.3
3.1
3
2.9
2.8
2.7
2.7
2.7
2.6
2.6
2.5
2.5
2.5
2.4
2.4
2.4
2.4
2.2
2.2
3.025
5.199
1.967
0.0664
1.986
0.00614
5.4
109.51
32.997
31.2
5.949
23.509
45
1
1.152
26.6
26.789
17.1
30.978
2.1
2.1
2.1
2
2
2
2
2
2
2
1
1
1
2E-09
7.003
43.345
0.00000001
0.00000001
28.135
7.026
1.067
95
15
42
2000
2000
Biological Data Collection
Field surveys of mangrove, seagrass and coral reef sites were completed in June 2005 using snorkel
and SCUBA gear. This sampling time frame was based on previous studies in Belize that found little
or no seasonal variation in fish abundance and biomass (Sedberry and Carter 1992) and the
Mesoamerican Barrier Reef System (MBRS) recommendation that fish, coral, algae and mangrove
surveys be completed between June 1 and July 31 to account for variations in environmental
conditions (Almada-Villela et al., 2003).
Three habitat types were sampled – mangroves, seagrasses and coral reefs. Seventy random x and y
(UTM) points were created in Excel and locations were transformed into a sampling map using
ArcView. Survey site information was mapped over a base map using a Landsat 7 satellite image
from May 23, 2000. Eight sites within each habitat (24 sites total) were selected for estimating
biomass of fish and benthic invertebrates, while fishes collected for the diet matrix were sampled at
different locations where various gear could be more efficiently deployed (Figure 2). Areas in water
deeper than 11 m were not sampled due to light absorption problems associated with visual surveys,
gear limitations and difficulties associated with repetitive diving.
Figure 2. Calabash Caye study sites identified by habitat type and
method of sampling. Landsat 7 satellite image from May 23, 2000.
Fish Biomass Measurement
Transect lines and radial point counts were used to visually survey the fish populations. These
methods are similar to techniques used by the MBRS (Almada-Villela et al., 2003) and Reef Check
(www.reefcheck.org), groups that also collect data about Belizean reefs. The radial point count
requires a stationary diver and was designed so that non-mobile, cryptic species could be observed
55
(Bohnsack and Bannerot, 1982; Buxton and Smale, 1989; Polunin and Roberts, 1993; Roberts, 1995).
Because the transect line and radial point systems may select for fish with different behaviors, the two
methods were combined to reduce bias from each method.
Surveys were made within one hour of sunrise and sunset, periods of maximal crepuscular activity.
Fish sightings were recorded on a dive slate with prepared data sheet and attached T-bar unit (a 1.27
cm diameter PVC pipe of 60 cm length). At each site, fish were counted along a 30 m x 2.0 m
transect line followed by a 7.5 m radial point count at the end, and returned along the same 30 m
transect performing another census. A second transect/radial point survey was conducted at least 15 m
away from and parallel to the first (Figure 3). The total area surveyed was 593 m2. Only fish observed
along the transects and within the radial point surveys were counted. At mangrove sites, only one
transect/radial survey was performed (area sampled = 297 m2).
Figure 3. Diagram of fish abundance and biomass sampling survey. A 7.5 m radial point count was conducted at
the end of each 30 m transect.
To estimate individual fish total length, we modified procedures used by Bell et al. (1984), Bohnsack
and Bannerot (1986) and MBRS (Almada-Villela et al., 2003). Length-weight relationships published
by Bohnsack and Harper (1988), Opitz (1996), Cleveland and Montgomery (2003) and Fishbase
(Froese and Pauly, 2006; www.fishbase.org) were used to calculate biomass of each species
identified. We calculated fish weight based on the midpoint of its size class (e.g., for size class 6-10
cm, biomass was calculated using 8.0 cm). Mean biomasses were used for all reporting and statistical
analyses.
Benthic Invertebrate Biomass Measurement
Benthic samples are often difficult to collect using only one method because substrates can vary
substantially from organic-rich, soft-bottomed mangrove sites to rhizome- and root-dense seagrass
beds to carbonate rock and coral rubble. To eliminate sampling bias associated with using different
sampling techniques, a modification of Brook's (1979) suction dredge was used for this study. A
gasoline-powered irrigation pump was used, providing 3.5 hp and 145 gal/min to pump water through
hoses with an inside diameter of 7.5 cm. At the suction head, the opening was reduced to 3.75 cm,
56
creating the Venturi vacuum effect necessary to cause sufficient suction. Collection bags, attached at
the outlet of the suction head, had a mesh width of 2 mm (or 3 mm stretch) and easily contained the
7,065 cm3 samples. Triplicate samples were taken at sites corresponding to the fish biomass surveys
(eight locations each in mangrove, seagrass and coral reef areas).
Most samples were collected using SCUBA. A bottomless 20 L bucket with 30 cm inside diameter
was placed on the substrate to delineate sample surface area and to contain mobile species. Samples
were dredged to a depth of 10 cm. Any seagrasses or algae present in the sample were manually
removed from the substrate inside the bucket and sucked into the sample bag. All samples were
collected mid-day, usually between 1500 and 1800 hours local time, and fixed with 10% formalinseawater mixture. Despite the 2.0 mm mesh collection bag, large shells and benthic autotrophs
prevented particles smaller than 2.0 mm from passing through the mesh bag. Thus, samples were
passed through a 500 µm sieve to remove all benthic invertebrates that were alive at the time of
collection. A microscope was used to identify benthic invertebrates and classify them into the
corresponding Opitz (1996) non-fish groups. Specimens were weighed to the nearest 0.00001 g wet
weight using a Mettler H51 precision balance.
Additional visual surveys of macroinvertebrates were conducted by a snorkeler along thirty-meter
transects at five mangrove, six seagrass and eight coral reef locations. All macroinvertebrates
encountered within a meter along either side of the transects were counted and sizes estimated. A
size-weight relationship for each species was determined from the published literature (Cary, 1916;
Doran, 1958; Horn, 1982; Hunte et al., 1989; Pauly, 1993; Pomory, 1998; SEAMAP-SA, 2002;
Muthiga and Jaccarini, 2004). Biomasses (g wet weight⋅m-2) of each group observed during the visual
surveys were then added to the biomass values from the microinvertebrate (suction dredge) sampling.
Fish Diet Study
Fish were collected using experimental gill nets that were 38.10 m long, 1.83 m deep, and composed
of five 7.62 m sections with increasing mesh size from 2.5-12.7 cm stretch mesh monofilament. A 15
m bag seine with 0.64 cm ace mesh netting was used to sample the fish community present in the
seagrass areas along the shoreline. A 1.8 m pole spear and hook and line were used to sample the
fishes within the reef. All spear fishing was done while snorkeling.
A sieve fractionation technique was used as described by Carr and Adams (1972), Luczkovich and
Stellwag (1993), Christian and Luczkovich (1999), and Luczkovich et al. (2002) to examine the
stomach contents of each individual. Weight was measured to 0.00001 gram using a Mettler H51
precision balance. Fish with empty stomachs were not analyzed.
A 137 by 137 square diet matrix consisting of 96 fish species and 41 non-fish groups was constructed
using dietary data from this study (Chagaris, 2006), Randall (1967), Fishbase (Froese and Pauly,
2006; www.fishbase.org), and Opitz (1996). Within each cell of the matrix, the proportion of each
prey to the overall diet of its predator was entered so that the diet for each consumer summed to one.
The taxonomic level of organisms identified from stomach contents in this study and Randall (1967)
varied from species to broader categories, with most invertebrates grouped under broader taxonomic
categories. The non-fish prey aggregations used by Opitz (1996) allowed for representation of all prey
items in the diet matrix. Dietary data for non-fish taxa were taken directly from the diet composition
matrix in PRVI Ecopath model (Opitz, 1996). Diet information of coral reef invertebrates is scarce
thus the quantitative diet composition of each non-fish group was estimated from qualitative reports
and/or was transferred from ecosystems other than coral reefs (Opitz, 1996).
57
Model Balancing and Analysis
In order for Ecopath to formulate outputs, the model must first be mass-balanced, accounting for all
mortality within the system and providing enough prey biomass to support their associated predators.
The model was mass-balanced by first manually adjusting parameters with least confidence in its
source data followed by running the auto-mass balancing routine, a feature of Ecopath with Ecosim
5.0.
Ecopath generates up to 37 different outputs ranging from basic analysis of each compartment to
ecosystem-level indicators to network analysis of pathways (Christensen et al., 2005). For simplicity,
we have chosen only to examine those outputs that reflect system-wide trophic information (effective
trophic levels, trophic level diagram and mixed trophic impacts). The Ecopath approach to
determining effective trophic levels uses a routine that assigns primary producers and detritus a
trophic level of 1. Consumers' trophic levels (TLs) are calculated as 1 + the weighted average of the
preys' trophic level (Christensen et al., 2005). For example, if 40% of a consumer's diet comes from
that of TL 1 and 60% comes from TL 2 then it will have a TL = 1 + [0.4(1) + 0.6(2)] = 2.6. Effective
trophic levels (ETLs) are calculated using diet matrix information and visualized in the trophic level
diagram. The mixed trophic impacts matrix and diagram show how small increases in the biomass of
one compartment can impact all other compartments. This is done by computing the material flow
into and out of each compartment, and scaling these flows by the total flows from all compartments.
Thus, if one compartment comprises the bulk of the material flow into another compartment, it will
have a large impact on that compartment if it is removed or increased. This is summed over all
possible pathways to estimate the total mixed trophic impact of any compartment on another.
Results and Discussion
Interpretation of Model Outputs
Biomass in Balanced Compartments
The greatest biomass was associated with benthic producers (seagrasses and algae) and organic matter
(detritus, particulate organic matter or POM, dissolved organic matter or DOM), as would be
expected at the base of a food web (Table 1). The next largest standing stocks were associated with
large invertebrate groups that provide structure on the reef (corals/anemones = 109.5 g wet wt⋅m-2;
sponges = 95 g wet wt⋅m-2). Other invertebrate groups associated with the benthos were next most
important in biomass (ascidians/barnacles/bryozoans = 45 g wet weight⋅m-2; bivalves = 43.3 g wet
wt⋅m-2; chitins/scaphopods = 33.0 g wet wt⋅m-2; holothuroids and related taxa = 31.0 g wet wt⋅m-2).
Phytoplankton (42 g wet wt⋅m-2) and zooplankton (31.2 g wet wt⋅m-2) were next most dominant in
biomass. Among fishes, herbivorous parrotfishes (Intermediate Scaridae = 28.1 g wet wt⋅m-2) had the
largest biomass, while carnivorous fishes (large groupers, jacks, scombrids, rays and sharks) were less
than 1 g wet wt⋅m-2 each.
Effective trophic levels
In the Calabash Caye model, seabirds (Group 28; ETL = 4.5) represented the apex predators in this
system because they consumed only fishes and had no predators; the biomass of this group was
relatively small (Table 1 and Figure 4). Sharks and scombrids (Group 2) were representative of the
second highest trophic group with an ETL of 4.2 (also with relatively low biomass). There were 19
compartments with ETLs from 3.0 to 3.9, the largest of which were three carnivorous reef fish groups
dominated by different species of jacks (Groups 6, 8 and 13). There were 26 compartments with
58
ETLs from 2.0 to 2.9, the largest of which represented parrotfish, and three compartments [detritus
(Group 50), benthic autotrophs (Group 49) and phytoplankton (Group 48)] were classified at ETL 1.0.
Mixed trophic impacts
Direct and indirect tropic impacts are shown as a matrix of impacting groups (rows) and impacted
groups (columns) (Figure 5). The mixed trophic impacts were largely positive (black upward bars)
when lower trophic levels (detritus, benthic producers, decomposers/microfauna, zooplankton) were
the impacting groups. These are the 'bottom-up' effects that would be expected if production
increased at lower trophic levels. One exception is that Group 19 (dominated by angelfishes) would
undergo a negative trophic impact. In contrast, mixed trophic impacts were largely negative (gray
downward bars) when top-level groups [intermediate jacks (Group 3), schoolmasters and yellowtail
snappers (Group 6) and hogfishes/barracudas/snappers (Group 12)] were the impacting groups. These
are the 'top-down' effects. If these top predators were to increase, decreases would be expected in
most other groups.
If targeted species such as large sharks/rays (Group 1), sharks/scombrids (Group 2), intermediate
jacks (Group 4) and hogfishes/barracudas/snappers (Group 12), were to experience an increase in
biomass (possibly due to reduction of fishing pressure after the establishment of an MPA), then there
would be negative impacts on 18 fish compartments due to increased predation and/or competition
and positive impacts on 12 fish compartments as a result of reduction of their predators and/or
competitors (Figure 5). A closer look at the large sharks/rays (Group 1), present only in very low
biomass in the Calabash model, revealed important trophic information. Only two species within this
compartment (compared to 13 species in the PRVI model) were observed or collected at Calabash
indicating that the species may be overfished, not selected by the sampling gear, avoided divers, or
existed in deeper water outside of our sampling boundaries. The absence of these large sharks and
scombrids may allow a greater abundance of herbivorous reef fish around Calabash Caye (Table 1
and Figure 4).
Figure 4. Trophic level diagram for the 50-compartment Calabash Caye ecosystem model. The vertical axis
represents the trophic level of the compartment and the horizontal axis is dimensionless.
59
60
IMPACTED GROUP
IMPACTING GROUP
60
61
61
Figure 5. Mixed trophic impact diagram for the Calabash Caye ecosystem. Black bars represent positive impacts and gray bars represent negative impacts.
IMPACTING GROUP
Some commonly targeted species, especially snappers, can play a significant role in the Calabash
Caye ecosystem (Figure 6). Of the three snapper compartments (solid box, Figure 6), a small increase
in the biomass of only one compartment (hogfishes/barracudas/snappers, Group 12), would have
large negative impacts on several other fish compartments in the system. In terms of the abundant
herbivorous fishes (dashed box, Figure 6), particularly the parrotfishes (Groups 24 and 25), small
changes in their biomasses would not yield significant changes in most fish groups, although
hogfishes/barracudas/snappers (Group 12) would be positively impacted by such a change.
Figure 6. Mixed trophic impacts for Calabash Caye ecosystem with
emphasis on snappers (solid box) and parrotfishes (dashed box).
The mixed trophic impact output showed the importance of small benthic arthropods (ETL = 2.51) in
this system (Figure 6). The small benthic arthropods compartment had a biomass of only 6.0 g wet
weight per square meter (Table 1). A small increase in the biomass of this compartment produced
small positive or negative impacts on nearly every compartment in the Calabash model. This may
indicate that these organisms are keystone group of species (Libralato et al., 2006) within the
Calabash Caye ecosystem. These organisms were found only at mangrove and seagrass sites using the
suction dredging method. This suggests the critical importance of coral reef-associated habitats and
the necessity to manage coastal development to conserve these habitats (Mumby et al., 2004).
62
Examination of Model Construction
Our fish hbiomass data suggest that three separate models could be constructed to show the
differences between mangrove, seagrass and coral reef habitats. Separate models might be beneficial
to managers who can focus on 'smaller' networks, providing specific fish and non-fish compartmental
data for each habitat. Ecopath with Ecosim 5.0 can be used to partition habitats within the ecosystem.
However, due to the interconnectedness of the three sub-systems, it may be difficult to demarcate
strict boundaries between the habitats.
The diets of the fish examined during this study were found to be similar to Randall (1967). Any
differences in the diets between the two studies are likely attributable to the small sample sizes of
some fishes from Calabash. Small differences in the diets of fish aggregated in our model may be
masked by having multiple species in each compartment. For example, a fish at Calabash Caye may
prey upon one species of xanthid crab while the same species of fish examined by Randall (1967)
may eat a different species of xanthid crab, but in the model they are both feeding from the 'crabs'
compartment. Because organisms are grouped into functionally or taxonomically similar
compartments, dietary data reported in the literature from nearby or similar systems may prove to be
sufficient for most ecosystem models. Refer to Chagaris (2006) for additional information regarding
the diet matrix.
Ecopath uses a top-down approach to mass-balancing models. However, our data revealed large
imbalances at the lowest trophic levels largely because many of those compartments were not
sampled directly at Calabash Caye (especially detritus, benthic autotrophs, phytoplankton and
zooplankton). As such, our manual mass-balancing method involved adjusting compartments and
their parameters from the bottom-up. Higher trophic level compartments were deemed more reliable
than the lower trophic level compartments. Care was taken to keep our values within reasonable
ranges by frequently referring to the Opitz (1996) PRVI model. See Barry (2006) and Chagaris
(2006) for additional information regarding the construction and balancing of the Calabash Caye
model.
Applications for Use in Management
Our Ecopath model represents the most complete model to date of the Calabash Caye ecosystem.
However, many assumptions were made to provide complete parameterization of the model. Dame
and Christian (2006) call into question four major sources of uncertainty in ecological network
analysis models: natural variability of input parameters, data collection methods, model construction,
and algorithm assumptions. Should we have used the same 50 compartments as Opitz (1996), or
should we have grouped our species according to our own assumptions? By using the same species
aggregations, we can easily compare the Puerto Rico-Virgin Islands and Calabash Caye models (see
Barry [2006] and Chagaris [2006] for comparisons). However, there were some differences in species
groups between the two models. Fortunately, the data still exist, and new Calabash models can be
created. Are the model outputs valid, and can they be trusted for making management decisions? To
test validity, it may be helpful to create a similar Ecopath model for a current Belize MPA, such as
the Hol Chan Marine Reserve, and compare outputs. Christensen and Walters (2004) warn against
making management decisions based solely on the outputs of these models, but it is a starting point.
Perhaps this preliminary Calabash Caye model is most useful in making qualitative predictions about
the effects that species removals or additions will have on the community structure. Important data
that can improve the model include fisheries harvest data and life history data. These two sets of
information can allow for model simulations using Ecopath with Ecosim, which may provide another
method to predict ecosystem-wide changes. It may also be important to conduct lobster and conch
63
surveys at Calabash Caye (not done in this study) to provide accurate biomass data, since these
fisheries are so important to the Belize economy. With or without the creation of a new marine
protected area at Calabash Caye, our model provides managers with an important tool that may allow
them to prevent devastating changes in the ecosystem that may be caused by human interactions.
Acknowledgments
The authors thank the staff members at the IMS field station at Calabash Caye for logistical
assistance, Molly Kilpatrick for conducting visual benthic surveys, the students of the Belize Marine
Ecology class and our dive master/dive safety officer Mark Keusenkothen.
References
Almada-Villela PC, Sale PF, Gold-Bouchot G, Kjerfve B. Manual of Methods for the MBRS
Synoptic Monitoring Program. Mesoamerican Barrier Reef Systems Project, 2003; 146 pp.
Barry DB. Using visual censuses to estimate biomass of marine fishes for an Ecopathmodel of coral
reef, seagrass and mangrove habitats surrounding Calabash Caye, Belize. East Carolina University
Masters Thesis, 2006; 126 pp.
Bell JD, Craik GJS, Pollard DA, Russell BC. Estimating length frequency distributions of large reef
fish underwater. Coral Reefs. 1985; 4: 41-44.
Bohnsack JA, Bannerot SP. A random point census technique for visually assessing coral reef fishes.
The visual assessment of fish populations in the Southeastern United States: 1982 Workshop. South
Carolina Sea Grant Technical Report, 1982; 1: 5-7.
Bohnsack JA, Bannerot SP. A stationary visual census technique for quantitatively assessing
community structure of coral reef fishes. NOAA Technical Report. NMFS 41, 1986.
Bohnsack JA, Harper DE. Length-weight relationships of selected marine reef fishes from the
Southeastern United States and the Caribbean. NOAA Technical Memorandum NMFS-SEFS-215,
1988.
Bossi R, Cintron G. Mangroves of the wider Caribbean: toward sustainable management. UNEP
(Nairobi), CCA (Barbados) and Panos Institute (Washington), 1990.
Brook IM. A portable suction dredge for quantitative sampling in difficult substrates. Estuaries. 1979;
2(1): 54-58.
Buxton CD, Smale MJ. Abundance and distribution patterns of three temperate marine reef fish
(Teleostei: Sparidae) in exploited and unexploited areas off the Southern Cape Coast. J Appl Ecol.
1989; 26: 441-451.
Carr WES, Adams CA. Food habits of juvenile marine fishes: Evidence of cleaning habit in the
leatherjacket, Oligoplites saurus and the spottail pinfish, Diplodus holbrooki. Fishery Bulletin, U.S.
1972; 70: 1111-1120.
64
Cary LR. The influence of the marginal sense organs on metabolic activity in Cassiopea xamachana
bigelow. Proceedings of the National Academy of Sciences of the United States of America 1916;
2(12): 709-712.
Chagaris DD. Diet analysis of selected fishes in a Belizean coral reef ecosystem and development of
an Ecopath food web model. East Carolina University Masters Thesis, 2006; 125 pp.
Christensen V, Pauly D. Ecopath II – a software for balancing steady-state ecosystem models and
calculating network characteristics. Ecol Model. 1992; 61: 169-185.
Christensen V, Walters CJ. Ecopath with Ecosim: methods, capabilities and limitations. Ecol Model.
2004; 172: 109-139.
Christensen V, Walters CJ, Pauly D. Ecopath with Ecosim: a User's Guide. Fisheries Centre,
University of British Columbia, Vancouver, Canada. November, 2005; 154 pp.
Christian RR, Luczkovich JJ. Organizing and understanding a winter's seagrassfoodweb network
through effective trophic levels. Ecol Model. 1999; 117: 99-124.
Cleveland A, Montgomery WL. Gut characteristics and assimilation efficiencies in two species of
herbivorous damselfishes (Pomacentridae: Stegastes dorsopunicans and S. planifrons). Mar Biol.
2003; 142: 35-44.
Costanza R, d'Arge R, de Groot R, Farber S, Grasso M, Hannon B, Limburg K, Naeem S, O'Neill
RV, Paruelo J, Raskin RG, Sutton P, van den Belt M.The value of the world's ecosystem services and
natural capital. Nature. 1997; 387: 253-260.
Coastal Zone Management Authority and Institute (CZMAI). State of the Coast 2001-2002 Belize.
Coastal Zone Management Authority and Institute, Belize City, Belize, 2003.
Dame JK, Christian RR. Uncertainty and the use of network analysis for ecosystem-based fishery
management. Fisheries. 2006; 31(7): 331-341.
Doran E. The Caicos conch trade. Geograph Rev. 1958; 48(3): 388-401.
Froese R, Pauly D, eds. FishBase. World Wide Web electronic publication, 2006. www.fishbase.org,
version (02/2006).
Garcia E, Holtermann K. Calabash Caye, Turneffe Islands Atoll, Belize. Coastal Region and Small
Island Papers 3. Environment and Development in Coastal Regions and in Small Islands, 1998.
Gardiner R, Harborne A. Summary of Coral Cay Conservation's data on commercially important reef
fish and invertebrates from Turneffe Atoll, Belize. Turneffe Atoll Commercial Species Report. Coral
Cay Conservation Limited, UK, 2000; 66 pp.
Gillet V. The fisheries of Belize. Fisheries Center Research Reports. 2003; 11(6): 141-147.
Gischler E. Holocene lagoonal development in the isolated carbonate platforms off Belize. Sediment
Geol. 2003; 159: 113-132.
65
Harborne AR, Mumby PJ, Micheli F, Perry CT, Dahlgren CP, Holmes KE, Brumbaugh DR. The
functional value of Caribbean coral reef, seagrass and mangrove habitats to ecosystem processes. Adv
Mar Biol. 2006; 50: 57-189.
Horn P. Adaptations of the chiton Sypharochiton pelliserpentis to rocky and estuarine habitats. NZ J
Mar Freshwater Res. 1982; 16: 253-261.
Hunte W, Marsden JR, Conlin BE. Habitat selection in the tropical polychaete Spirobranchus
giganteus: effects of coral species on body size and body proportions. J Mar Biol. 1989; 104(1): 101107.
Kitheka JU. Coastal tidally-driven circulation and the role of water exchange in thelinkage between
tropical coastal ecosystems. Estuar Coastal Shelf Sci. 1997; 45(2): 177-187.
Libralato S, Christensen V, Pauly D. A method for identifying keystone species in food web models.
Ecol Model. 2006; 195: 153-171.
Luczkovich JJ, Stellwag EJ. Isolation of cellulolytic microbes from the gastrointestinal tract of the
pinfish, Lagodon rhomboides: size-related changes in diet and microbial abundance. Mar Biol. 1993;
116: 381-388.
Luczkovich JJ, Ward GP, Johnson JC, Christian RR, Baird D, Neckles H, Rizzo WM. Determining
the trophic guilds of fishes and macroinvetebrates in a seagrass food web. Estuaries. 2002; 25(6A):
1143-1163.
Meerman J, Wilson JR, eds. The Belize national protected areas system plan. Ministry of Natural
Resources and the Environment, 2005; 61 pp.
McField M, Wells S, Gibson J, eds. State of the Coastal Zone Report Belize 1995. Coastal Zone
Management Program, Government of Belize. Sustainable Development and Management of
Biologically Diverse Coastal Resources – Belize Project No. Bze/92/G31, 1996.
Moberg F, Rönnbäck P. Ecosystem services of the tropical seascape: interactions, substitutions and
restoration. Ocean Coast Management. 2003; 46: 27-46.
Mumby PJ. Connectivity of reef fish between mangroves and coral reefs: algorithms for the
design of marine reserves at seascape scales. Biol Conserv. 2006; 128(2): 215-222.
Mumby PJ, Edwards AJ, Arias-Gonzalez JE, Lindeman KC, Blackwell PG, Gall A, Gorczynska MI,
Harborne AR, Pescod CL, Renken H, Wabnitz CCC, Llewellyn G. Mangroves enhance the biomass
of coral reef fish communities in the Caribbean. Nature. 2004; 427(6974): 533-536.
Muthiga NA, Jaccarini V. Effects of seasonality and population density on thereproduction of the
Indo-Pacific echinoid Echinometra mathaei in Kenyan coral reef lagoons. Mar Biol. 2004; 146: 445453.
Opitz S. A quantitative model of trophic interactions in a Caribbean coral reef ecosystem. In:
Christensen V, Pauly D, eds. Trophic Models of Aquatic Ecosystems. ICLARM Conference
Proceedings, 1993: 259-267.
Opitz S. Trophic interactions in Caribbean coral reefs. ICLARM Tech. Rpt. 43, 1996; 341 pp.
66
Pauly D, Sambilay V, Opitz S. Estimates of relative food consumption by fish and invertebrate
populations, required for modeling the Bolinao reef ecosystem, Philippines. In: Christensen V, Pauly
D, eds. Trophic Models of Aquatic Ecosystems. ICLARM Conf. Proc. 26; 1993: 236-251.
Polovina JJ. Model of a coral reef ecosystem: The Ecopath model and its application to French
Frigate Shoals. Coral Reefs. 1984a; 3: 1-11.
Polovina JJ. An overview of the Ecopath model. ICLARM Fishbyte. 1984b; 2(2): 5-7.
Polunin NVC, Roberts CM. Greater biomass and value of target coral-reef fishes intwo small
Caribbean marine reserves. Mar Ecol Progress Series. 1993; 100: 167-176.
Pomory C, Lawrence M. Energy content of Ophiocoma echinata (Echinodermata: Ophiuroidea)
maintained at different feeding levels during arm regeneration. J Experiment Mar Biol Ecol. 1998;
238(1): 139-150.
Randall JE. Food habits of reef fishes of the West Indies. Studies Trop Oceanog. 1967; 5: 665-847.
Reefcheck. www.reefcheck.org. Accessed September 15, 2005.
Rueter J. A mutivariate approach to establish habitat classifications in a coral reef ecosystem using
LandSat 7 imagery at Turneffe Atoll, Belize, Central America. East Carolina University Masters
Thesis, 2004.
Roberts CM. Rapid build-up of fish biomass in a Caribbean marine reserve. Conserv Biol. 1995; 9(4):
815-826.
SEAMAP-SA Shallow Water Trawl Survey. Cruise Report, Spring 2002. Atlantic States Marine
Fisheries Commission, 2002.
Sedberry GR, Carter J. The fish community of a shallow tropical lagoon in Belize, Central America.
Estuaries. 1993; 16(2): 198-215.
Ulanowicz RE. Growth and Development: Ecosystem Phenomenology. New York, NY: SpringerVerlag, 1986; 203 pp.
67
68
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Submerged Cultural Resource Management on the Last Frontier:
Reconnaissance, GIS Mapping, and Biotic/Geochemical Characterization
of Threatened Shipwreck Sites in Southeast Alaska
J. David McMahan1, John O. Jensen2, Stephen Jewett3, John Kelley3, Sathy Naidu3, Hans Van
Tilburg4, and Michael Burwell5
1
Office of History and Archaeology, Alaska Department of Natural Resources, 550 W. 7th Ave.,
Suite 1310, Anchorage, AK 99501; davemc@dnr.state.ak.us
2
Sea Education Association / University of Rhode Island / Jensen Research and Heritage Services,
708 South Road, Wakefield, Rhode Island 02879; jensenheritage@cox.net
3
School of Fisheries and Ocean Sciences, 245 O'Neill Bldg., University of Alaska Fairbanks
Fairbanks, Alaska 99775-7220; jewett@ims.uaf.edu / ffjjk@uaf.edu / ffsan@uaf.edu
4
NOAA/NOS, Pacific Islands Region, National Marine Sanctuary Program, 6600 Kalaniana`ole
Highway, suite 302, Honolulu, HI 96825; Hans.VanTilburg@noaa.gov
5
U.S. Department of the Interior, Minerals Management Service (MMS), Alaska OCS Region, 3801
Centerpoint Drive, Suite 500, Anchorage, AK 99503-5823; michael.burwell@mms.gov
Abstract
In April 2006, Alaska's Office of History and Archaeology collaborated with the University of
Alaska Fairbanks, the University of Rhode Island, the U.S. Minerals Management Service, and
the National Oceanic and Atmospheric Administration (NOAA) National Marine Sanctuary
Program to collect information on five historic shipwrecks in Southeast Alaska. Under a grant
from the NOAA Office of Ocean Exploration, the project team documented the shipwrecks
through dives, interviews, the use of a DIDSON sonar unit, and recordation of biota. Sediment
samples adjacent to wrecks are being analyzed to detect changes in soil chemistry due to wreck
degradation. These data sets will allow researchers to track changes to the sites as a result of
vandalism and natural decay processes. Public education components of the project included
public talks, radio and newspaper interviews, and the development of websites.
Introduction
Recent geographical information system (GIS) data suggests that around 44,000 miles or nearly half
the United States coastline falls within the boundaries of Alaska. The U.S. Minerals Management
Service (MMS), a federal agency that maintains a database of shipwrecks and other sub-sea
obstructions, estimates that Alaskan waters contain more than 3,000 shipwrecks. Even though many
of the wrecks hold archaeological or historical significance, their systematic exploration and
documentation have been precluded by high costs, vast distances, and seasonally cold violent
weather. In recent years, however, new and inexpensive remote sensing and diving technologies have
removed many of these barriers, resulting in increased incidents of unauthorized disturbance. Most of
the wrecks are on or embedded in state submerged lands and are protected under state law. While
diving on the wrecks is allowed, disturbance or collection of historic artifacts requires a permit. The
State of Alaska, with no dedicated underwater archaeology staff position and limited expertise has
been unable to generate the data needed to manage its underwater heritage sites. As an alternative
strategy, the state has built partnerships with experienced federal agencies and academic institutions
outside Alaska. States with active submerged cultural resource programs such as Maryland,
Michigan, Florida, North Carolina, South Carolina, and Wisconsin have demonstrated the importance
69
of archaeological surveys, not only as a means for gathering data, but as a mechanism for developing
support for preservation within the dive community and the general public.
Another important line of defense in the protection of Alaska's submerged heritage is the
development of data which will contribute to a better understanding of decay processes, corrosion,
and biotic relationships specific to coldwater heritage resources. This has implications for the
development of management and conservation plans. Wooden structures of shipwrecks in tropical
waters, along with associated metal artifacts, are typically subjected to intense submarine weathering,
corrosion, and biological encrustation, which results eventually in poor preservation. Conversely, it is
generally believed that colder waters support relatively less dense benthic biological populations and
contribute to better wreck preservation. In another context, the shipwrecks could be local microcosms
for focused kelp and benthos colonization. To achieve a better understanding of these processes
submerged cultural resource managers need baseline information on sediments, trace metals,
organics, and benthic biota. This is best achieved through multidisciplinary collaboration with marine
biologists, geologists, chemical oceanographers, and other scientists with appropriate expertise.
In 2005, the State of Alaska and collaborating organizations applied for and were awarded a National
Oceanic and Atmospheric Administration (NOAA) Office of Ocean Exploration grant
(OE_2005_078) to document several threatened historic shipwreck sites in Southeast Alaska. The
field phase of the project was carried out in April 2006. Three submerged wrecks were targeted: the
Princess Sophia, the Princess Kathleen, and the Clara Nevada (formerly Hassler). These particular
wrecks were chosen due to their popularity as recreational dive sites, good accessibility in fair
weather, and reports of vandalism. Additionally, two historically important intertidal shipwrecks, the
Islander and Griffson, were targeted. They were fortuitous inclusions in the survey due to their
proximity to the route used to access the submerged wrecks. This modest pilot project is considered a
first step towards the collection of baseline data that will allow for submerged cultural resource
management planning and the initiation of community outreach.
Methods
Prior to the field-phase of the project, team members (Burwell, Jensen) conducted archival research
on the wrecks, relying heavily on published materials. In Juneau and Haines, this information was
supplemented by interviews with local historians, museum staff, and divers. During the field phase, a
six-person team (including four divers) operated off the U.S. Fish and Wildlife Service (FWS) vessel
Curlew, chartered through an agreement between FWS and the State (Figure 1). The field phase
included two days for local interviews and dive preparation, six days at sea (dive days), and two days
for post field phase archival research in Juneau. The Curlew, with its field team, staged its operations
from the NOAA docks at Auke Bay Laboratory outside Juneau in order to test equipment with the
Dive Safety Officer (Jewett) prior to going to sea. Once at sea, dive logistics were planned around
slack tides to minimize depths and currents whenever overall scheduling allowed. Wrecks were
located using existing, albeit not always accurate, coordinates coupled with verbal or published
descriptions, and depth finder observations. Once a suspected wreck site was located, a skiff with an
operator and divers was deployed. The skiff team further defined the wreck location via transects with
a depth finder, then dropped a weighted descent line with a buoy. An initial pair of divers would
descend the line and conduct a general site reconnaissance, then surface and report their observations.
Armed with information from the reconnaissance, a second team continued the effort. The divers
characterized and recorded the condition of the wrecks by use of digital still photos and video clips
with the intent that these 'snapshots in time' could be compared against future data to document
change at the sites. Photographic images were supplemented by a rough inventory and description of
the salient structural elements, features, and associated materials at each site. The team also attempted
70
to ascertain if damage had occurred to the wrecks as a result of anchors from recreational dive boats.
An important project component was the addition of a high frequency DIDSON sonar unit (Kelley),
loaned by the Alaska Department of Fish and Game. This new technology holds promise for future
shipwreck work, especially in poor dive conditions. The DIDSON, designed for port security, is
relatively compact. It can produce good images even in 0 visibility conditions and can be used either
stationary or moving at slow speeds (via attachment of a mounting arm to a rail).
In addition to basic archaeological recordation (Jensen, Van Tilburg, McMahan), a marine biologist
(Jewett) inventoried biota in the vicinity of the wrecks and collected sediment samples for trace
element analysis (Naidu). The latter included a control sample from Berger Bay, away from any
shipwreck sites. These data will potentially contribute to a better understanding of decay processes,
corrosion, and biotic relationships specific to coldwater heritage resources. This has implications for
the development of management and conservation plans.
To ensure that visitors to the sites are aware of their protected status, engraved brass survey caps were
placed at the visited wreck sites, along with engraved plastic tags, identifying the wrecks as heritage
sites protected under state law (Alaska Statute 41.35) (Figure 2). Finally, the team collected GPS data
for each wreck site, deploying pelican buoys from key features when possible. The data were used to
update the Alaska Heritage Resources Survey (AHRS) database, the official inventory of Alaska's
archaeological and historical sites. AHRS data are available to land and resource managers, as well as
to authorized researchers, but are restricted from public access due to potential for site vandalism. The
coordinates will also be used to update the MMS shipwreck database and incorporated into their GIS
product. Like the AHRS, the MMS database withholds specific coordinates from public access to
help protect the resource. Data is also being provided to NOAA, who plans to conduct a hydrographic
survey at the Clara Nevada / Hassler site in support of archaeological recordation efforts.
Figure 1. The project's field team (L to R):
Dave McMahan, John Kelley, John Jensen,
Hans Van Tilburg, Stephen Jewett, and Mike
Burwell. Not shown: Ed Grossman.
Figure 2. Example of one of the brass survey cap
placed on the wreck sites, inscribed with the AHRS
site number and proclamation of the site's protected
status under State law.
71
Results
Princess Kathleen
Most of the historical background for the Princess Kathleen is derived from an earlier inventory (City
and Borough of Juneau 1992) and from information contained in the MMS and AHRS databases. The
Princess Kathleen was built for the Canadian Pacific railway by John Brown and Company Ltd., at
Clydebank, Scotland in 1925. It was 352 ft long, with a beam of 60 ft and a depth of 26 ft (Figure 3).
Modern in design with three distinctive stacks, its twin screws were powered by four oil-fired steam
turbines. It was licensed to carry 1,500 passengers and had a gross tonnage rating of 6,000 tons. The
Princess Kathleen served in the Vancouver-Victoria-Seattle coastal service until taken over during
the war in 1939 for troop transport. In 1947, it resumed service on her old route, and two years later
was transferred to the Canadian Pacific's Vancouver-Alaska cruise service. The Princess Kathleen
sank after running aground bow-first on Lena Point, 17 miles north of Juneau, on September 7, 1952,
with no loss of life.
Figure 3. Starboard view of the Princess Kathleen, from a photo postcard in the private collection of
Dave McMahan.
Even with better navigational aids than in earlier times, the wreck of the Princess Kathleen reemphasized the dangers of the waters of Southeast Alaska. Even though no lives were lost with the
wreck of the Princess Kathleen, the incident struck a responsive chord with residents of the
community, who were relieved that this incident had not ended in tragedy like so many before. The
wreck, which is relatively intact, reflects the history of the ship and events surrounding its sinking.
Because of the intact nature of the wreck, there is a possibility that more information could be
extracted from the site in the future.
The Princess Kathleen lies with its bow in about 50 ft of water near Lena Point and the stern away
from shore in 120 ft of water. Its hull lists to port approximately 80 degrees and its bow is pointed
generally in a northerly direction. The ship's three stacks are still in place although the wooden decks
are badly rotted and collapsed. Much of the steel superstructure is twisted and fallen. The starboard
72
cargo bay door is wide open, as well as some of the passenger stateroom doors. Much of the wooden
structure within the staterooms has rotted and collapsed. Some of the ship's larger windows are still
intact.
The 2006 project team dove
only once on the Princess
Kathleen, descending to the
vessel's bow via a buoy line,
then explored along the deck
to about amidships before
backtracking. The dive was
plagued by poor visibility
(±10 ft) and moderately strong
currents. While visibility
precluded usable photographs,
divers were able to inventory
the biota (Table 1), collect a
sediment sample, and place
both a brass monument and
plastic tag at the wreck site.
On the return trip to Auke
Bay, the team attempted a
second dive on the Princess Figure 4. Comparison of a 2006 DIDSON sonar image with a historic
Kathleen, but failed to locate photo of the Princess Kathleen's stern.
the wreck due to faulty depth
finder readings caused by a loose transducer. Scheduling constraints did not allow for a second dive.
Instead, the team made several passes over the wreck site with the DIDSON sonar (Figure 4). This
ultimately provided better data on the overall condition and position of the wreck than could have
been obtained via low visibility dives.
Princess Sophia
Most of the historical background for the Princess Sophia derives from an earlier inventory (City and
Borough of Juneau 1992) and from information contained in the MMS and AHRS databases. The
Princess Sophia was built by Bow, McLachlane & Co., Ltd. at Paisley, Scotland in 1911. She was
constructed for the Canadian Pacific railway Company at a cost of around $250,000. The vessel was
245 ft long with a beam of 44 ft and a depth of 24 ft (Figure 5). It featured a wooden deck with a
single stack and bow mast. The Princess Sophia arrived in Vancouver under the command of Captain
Lindgren in February of 1912. Soon after its arrival, it was converted from coal burning to oil. The oil
fired tripled-expansion steam engine powered a single screw. It was licensed to carry 250 passengers,
and if necessary, could carry up to 500 with special permission and with additional floatation devices.
It displaced 2,320 gross tons and was run by a crew of 61. Its first run into Southeast Alaska and
Juneau was in June of 1912. The Princess Sophia was lost with all on board on October 23, 1918,
when it grounded on Vanderbilt Reef in a storm, then slid off before passengers could be rescued.
With the deaths of around 350 people, the incident may be the most tragic in terms of loss of life of
any shipwreck on the North American West Coast. Ironically, the tragedy received little attention
from the press, whose attention was focused on the end of World War I.
73
Table 1. Benthic reconnaissance dive survey of algae, epifaunal invertebrates and fish from the shipwreck
Princess Kathleen, off Lena Point, Auke Bay, Alaska.
OVERVIEW: Divers descended down line to the top of the wreck at 45 ft. Vessel lay on its port side
with the bow orientated N-NW. After noting moderate cover of flora and fauna on starboard side, we
descended perpendicular to the keel to the substrate at 73 ft. The covering of flora and fauna was less
on the port side of the hull than on its starboard side. Sediment sample #1 was collected within 6 ft
of the vessel.
TAXON
ALGAE - Rhodophyta
Coralline
ALGAE - Phaeophyta
Laminaria sp. (juv.)
Agarum clathratum
CNIDARIA – Hydrozoa
Sertulariidae
ANNELIDA – Polychaeta
Crucigera sp.
Serpula sp.
MOLLUSCA – Gastropoda
Fusitriton oregonensis
Neptunea lyrata
Unidentified nudibranch
MOLLUSCA – Polyplacophora
Cryptochiton stelleri
MOLLUSCA – Bivalvia
Chlamys sp.
Pododesmus macroschisma
ARTHROPODA – Crustacea
Hyas lyratus
Oregonia gracilis
Elassochirus tenuimanus
E. gilli
Unidentified shrimp
ECTOPROCTA
Microporina borealis
Dendrobeania murrayana
Microporina borealis
Dendrobeania murrayana
BRACHIOPODA
Terebratalia transversa
Terebratulina unguicula
ECHINODERMATA – Asteroidea
Evasterias troschelii
Pycnopodia helianthodes
ECHINODERMATA – Ophiuroidea
ECHINODERMATA – Echinoidea
Strongylocentrotus droebachiensis
ECHINODERMATA – Holothuroidea
Parastichopus californicus
CHORDATA – Urochordata
Halocynthia aurantium
CHORDATA – Pisces
Myoxocephalus polyacanthocephalu
74
COMMON NAME
Red algae
Encrusting coralline algae
Brown algae
Kelp
Seive kelp
Hydroid
Polychaete
Polychaete
Oregon triton
Snail
Nudibranch
Gumboot chiton
Bay scallop
Rock jingle
Lyre crab
Decorator crab
Hermit crab
Hermit crab
Shrimp
Bryozoan
Bryozoan
Bryozoan
Bryozoan
Lamp shell
Lamp shell
Sea star
Sunflower sea star
Brittle star
Green sea urchin
Sea cucumber
Sea peach
Figure 5. Starboard view of the Princess Sophia in front of Taku Glacier, from a photo postcard in the private
collection of Candy Waughaman (Fairbanks).
At the time of its demise, the Princess Sophia was engaged in the transportation commerce of the last
stages of the gold rush. Its wreck re-emphasized the dangers of the waters of Southeast Alaska and
the need for better navigational aids.
By all accounts from local divers, the Princess Sophia is now hardly recognizable until one gets to
within very close proximity to it. The hull is leaning to port about 20 degrees, and about 25% of the
bow is broken off and appears storm battered. According to local accounts, it is believed that this
happened between 1982 and 1992. The stern is broken up as well, with the boilers and side doors still
evident. The wooden decks are badly rotted and are beginning to collapse onto each other. The
foremast is lying perpendicular to the ship and the single stack is not visible. In 1992, 60% of the hull
was estimated to be intact.
The 2006 team arrived at Vanderbilt Reef at high tide and made several passes over the wreck site
with the DIDSON sonar. The DIDSON images, which were acquired at the beginning of our learning
curve with the system, lacked clarity. This probably also relates to the depth and scattered condition
of the wreck relative to that of the Princess Kathleen. Due to the depth of the wreck, the team decided
not to attempt no-decompression dives at high tide. This was not a major setback to the project,
because the wreck was included in an earlier investigation by a professional marine archaeologist as
part of a Canadian documentary. During the summer of 2006, a Juneau diver placed a brass
monument on the site at the request of the project team.
Clara Nevada / Hassler
Historical background for the Clara Nevada / Hassler derives from an earlier inventory (City and
Borough of Juneau, 1992), using information contained in the MMS and AHRS databases, and a
recent manuscript document on the vessel's earlier history (Jensen, 2007). The Clara Nevada was
75
built in Camden, New Jersey in 1872, as a survey vessel for the U.S. Coast and Geodetic Survey
(Figure 6). Originally christened the USS Hassler, it was involved with deep sea research before
spending more than twenty years surveying remote stretches of the Alaska and Northwest Pacific
coast. Decommissioned in 1895, it was purchased by the Pacific & Alaska Transportation Company
in 1897 and renamed the Clara Nevada. While the name selection was supposedly to honor a
Hollywood movie star, no actress by that name has been identified. The vessel measured 154 ft in
length and had a 24-foot beam. The hull was iron and the deck was constructed of wood. It was
originally powered by sail, and the three prominent masts were retained after its conversion to steam
power. The converted ship was powered by a steam engine that drove a single screw propeller. The
Clara Nevada had a capacity for 100 passengers in first class accommodations and 100 passengers in
steerage. It had a gross tonnage of 464 tons and could carry 300 tons of freight.
Figure 6. Port view of the Clara Nevada. Courtesy of the Alaska State Library, Winter and
Pond Collection, No. P87-1594.
The Clara Nevada was lost in 1897 when it ran into the reef on the north side of Eldred Rock, around
20 miles south of Haines, and exploded into a fireball seen by witnesses. At the time of its demise, it
was carrying passengers on their way to make their fortunes in the gold fields of the Yukon. All on
board (around 75 people), were reported to have lost their lives on the ship, but the incident is
shrouded in mystery. Some accounts describe sightings of the captain plying his trade at other places
after the wreck, prompting speculations of survivors and stolen gold. Like the other wrecks, this
incident emphasized the dangers of the waters of Southeast Alaska and reinforced the need for better
navigational aids. Largely in response to the tragedy, the U.S. government constructed a lighthouse
on Eldred Rock in 1905. The oldest standing lighthouse in Alaska, it is now on the National Register
of Historic Places. The wreck site and lighthouse are reminders of the difficulties and dangers of life
at sea during this period of Southeast Alaska's history. Although the wreckage is scattered, it bears
witness to the important history of the ship and events surrounding the wreck. It is possible that future
investigation of the Clara Nevada's in situ wreckage may provide further insights into the vessel's
history and demise.
76
The remains of the Clara Nevada are in about 30 ft of water at low tide, on the north side of Eldred
Rock. The wreckage is broken up over at least a 200 yard area, but many large pieces of the ship's
structure remain. Contemporaneous accounts indicate that the six-foot diameter propeller, shaft, and
boiler were salvaged a few years after the wreck for their brass content.
Early in the 2006 project, the team decided to focus efforts on the Clara Nevada. This wreck, while
more exposed, is relatively shallow and easily accessible in good weather. Because it is further from
Juneau, recreational diving activity at the site has not been as intense. Consequently, it was hoped that
the wreck would offer more information potential than the better known Princess Kathleen and
Princess Sophia wrecks. The project team made multiple dives at the Clara Nevada site while in
route to Haines, then again upon the return voyage the following day. Visibility was moderate (±20
ft) initially, but had improved (±50 ft) by the following afternoon. The team conducted extensive
photo documentation at the site, inventoried local biota (Table 2), and collected a sediment sample.
Remnants of the ship's double hull are present, testimony to ongoing corrosion problems within the
cellular iron structure during the life of the vessel. Most of the vessel's propeller shaft is intact, with
only its aftermost section missing. Bolts had been removed at the point of attachment with the extant
forward sections of the shaft to remove the propeller during salvage. Remnants of the engine, frames,
firebox, and stack were noted, along with fragments of ceramic plates and other shipboard artifacts.
The windlass was observed in about 50 ft of water, along with segments of anchor chain, which
together indicate the position of the bow (Figure 7). Additional investigations by NOAA and State
archaeologists were conducted at the Clara Nevada site in May 2007, resulting in supplemental
mapping and photography. This investigation produced maps and images that are being incorporated
into websites under development to commemorate NOAA's 200th anniversary celebration.
Figure 8. Stephen Jewett (L) and Ed Grossman (R) by the steamer Islander's
sternpost in April 2006.
77
Table 2. Benthic reconnaissance dive survey of algae, epifaunal invertebrates and fish from the shipwreck
Clara Nevada, off the north side of Eldred Rock Lighthouse, Lynn Canal, Alaska.
OVERVIEW: Observations of the biota were based mainly on the three dives conducted for this study;
two other dives were made by non-project divers to obtain additional photos and videos. The wreckage
was scattered between 15 and 50 ft, with the majority at 15 to 30 ft. Sediment samples 2 and 3 were
collected at 45 ft within the wreck area.
TAXON
ALGAE - Rhodophyta
Coralline
ALGAE - Phaeophyta
Desmarestia sp.
Laminaria bongardiana
Agarum clathratum
Costaria costata
CNIDARIA – Hydrozoa
Sertulariidae
CNIDARIA – Anthozoa
Unidentified anemone
ANNELIDA – Polychaeta
Crucigera sp.
Serpula sp.
MOLLUSCA – Gastropoda
Fusitriton oregonensis
Dendronotus sp.
MOLLUSCA – Polyplacophora
Cryptochiton stelleri
Tonicella _iliate
Mopalia _iliate
ARTHROPODA – Crustacea
Hyas lyratus
Oregonia gracilis
Elassochirus tenuimanus
BRACHIOPODA
Terebratalia transversa
ECHINODERMATA – Asteroidea
Henricia sp
Crossaster papposus
Evasterias troschelii
Leptasterias hexactis
Pycnopodia helianthodes
ECHINODERMATA – Ophiuroidea
ECHINODERMATA – Echinoidea
Strongylocentrotus droebachiensis
CHORDATA – Pisces
Hexagrammos decagrammus
COMMON NAME
Red algae
Encrusting coralline algae
Brown algae
Acid kelp
Split kelp
Seive kelp
Seersucker
Hydroid
Anemone
Polychaete
Polychaete
Oregon triton
Nudibranch
Gumboot chiton
Lined chiton
Hairy chiton
Lyre crab
Decorator crab
Hermit crab
Lamp shell
Blood star
Rose star
Sea star
Sea star
Sunflower sea star
Brittle star
Green sea urchin
Kelp greenling
Islander / Griffson
The steamer Islander wrecked and sunk in around 250 ft of water on August 15, 1901, while carrying
passengers to the Yukon gold fields. The ship is best known for the 1934 monumental salvage effort
in which most of the iron-hulled vessel was moved to a beach on Admiralty Island.
78
The ship's bow broke off during salvage, and has recently received renewed interest as a possible
source of lost Klondike gold the vessel was reportedly carrying. The Islander was salvaged by
extending cables beneath the vessel and securing their ends to salvage vessels floating on either side
of the Islander's hull. As each rising tide lifted the vessel, it was moved closer to the beach and the
cables tightened. One of the salvage vessels was the wooden schooner barge Griffson, abandoned on
the beach near the Islander in 1934. The intertidal remains of the Islander now comprise only the iron
hull with missing bow, along with scattered pottery and glass fragments (Figure 8). Some pottery
sherds with the Islander name have been documented, but are now rare at the site. The 240-foot long
ship was built in Scotland in 1888. Its sinking was said to have been the result of an iceberg. A large
portion of the Griffson's heavily-constructed wooden hull remains on shore, and is accessible at low
tide (Figure 9). Less has been written about the Griffson than the Islander, yet the vessel is deserving
of further research to more fully document its history. In 1952, the Islander was dismantled and now
only portions of its steel frames are still visible on shore.
The 2006 project team took advantage of a low tide to visit the site of these two intertidal wrecks
while bringing the Curlew from Juneau around Douglas Island to the NOAA Auke Bay dock. The
wrecks were found to be as described above. The team conducted extensive photo documentation of
the two wrecks, supplemented by basic measurements and observations. No biotic data were
recorded, and no sediment samples were collected.
Trace Element Analysis
Results from the analysis of acid-digested
solutions of five sediment gross samples (all
sandy) are shown in Figure 10. The
concentrations of all elements, with the
exception of one (lead [Pb] in sample #2) are
generally similar or lower than in unpolluted
world coastal sediments of comparative
granulometry. The concentrations of the
individual metals in the 'wreck' samples are
also significantly lower than the mean
concentrations of the corresponding metals in
glaciomarine muds of Valdez, Prince William
Sound and Chatham Strait, Alaska (Sharma,
1979; Naidu and Klein, 1988). The relatively
lower concentrations of metals in our samples
could be due to a combination of factors such
as sandy nature of the sampled sediments and
lower natural input from the local hinterland,
which is in contrast to the muddy composition
of the sediments compared and their
terrigenous source rich in metal deposits.
Contrary to expectation, the ship wrecks have
not been a local source of metal contaminants.
Figure 9. Hans Van Tilburg measuring the schooner barge
Griffson wreckage in April 2006.
79
Table 3
UAF
SE Alaska NOAA/OE
Reported by Frontier Geosciences, Inc., 414 Pontius Avenue North, Seattle, WA 98109
July 19, 2006
Conclusions
The shipwrecks documented during this project exemplify the current situation regarding Alaska's
historic shipwrecks. Using archival records and reports of obstructions, MMS and the State are
developing an excellent, although not yet complete, inventory of vessels lost in Alaskan waters.
Locational data for Alaska's shipwrecks, however, is often incomplete and inaccurate. Very few of
Alaska's historic shipwrecks have been inspected, much less evaluated archaeologically. Currently,
Alaska's wrecks remain the province of, and, in sad cases, the private property of recreational and
commercial divers. Despite strong laws, implementation of effective shipwreck management and
preservation policy is in its infancy and a strong marine preservation ethic has yet to develop. This
project followed in the footsteps of other successful state programs by beginning the process of
gathering accurate shipwreck site location data, assessing the condition of important shipwrecks that
are immediately threatened or sustain high levels of diver visitation, and developing outreach
programs with a strong marine preservation ethic. In conjunction with data collection activities,
project participants conducted public outreach in the Juneau and Haines areas through public lectures
and interaction with local dive communities. The success of the project was enhanced through
coordination with local museums, resource agencies, and the media. The project team hopes that this
multidisciplinary pilot project represents the beginning of a strong cooperative relationship between
local, state, and federal agencies, and academic institutions to help Alaska manage and preserve its
unique and historically significant submerged maritime cultural heritage.
More information is available at from the Alaska Office of History and Archaeology and the NOAA
Office of Ocean Exploration:
www.dnr.state.ak.us/parks/oha/index.htm
http://snake.nos.noaa.gov/explorations/06alaska/welcome.html
www.sanctuaries.noaa.gov/maritime/expeditions/hassler/welcome.html
80
References
City and Borough of Juneau. Inventory and Survey of Historic Shipwreck Sites. City and Borough of
Juneau, Community Development Department, document on file at the Office of History and
Archaeology, Division of Parks and Outdoor Recreation, Alaska Department of Natural Resources,
Anchorage, AK, 1992a.
City and Borough of Juneau. Supplement to the Inventory and Survey of Historic Shipwreck Sites
(confidential location data). City and Borough of Juneau, Community Development Department,
document on file at the Office of History and Archaeology, Division of Parks and Outdoor
Recreation, Alaska Department of Natural Resources, Anchorage, AK, 1992b.
Naidu AS, Klein LH. Sedimentation Processes. In: Shaw DG, Hameedi MJ, eds. Environmental
Studies in Port Valdez, Alaska. New York, NY: Springer-Verlag, 1988: 69-91.
Sharma GD. The Alaskan Shelf: Hydrography, Sedimentary, and Geochemical Environment. New
York, NY: Springer-Verlag, 1979; 498 pp.
81
82
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Video iPod Instructional Design Considerations for Dive Training and
Underwater Subject Matter
Michael Dermody1 and Calvin Mires2,3*
1
Department of Communications, East Carolina University, Greenville NC 27858, USA,
dermodym@ecu.edu
2
Program in Maritime Studies, Admiral Eller House, Greenville NC 27858, USA, miresc@ecu.edu
3
Coastal Resources Management, East Carolina University, Greenville NC 27858, USA
Abstract
The latest video iPod, and similar mobile video casting technology puts true mobile technology
in the hands of the masses. For only $299 US dollars you can purchase a fully functional 30 gig
handheld audio/video device that will provide up to 150 hours of full motion, 30 frames per
second, video and audio. However, designing audio and video content for a mobile 2.5 inch
color display screen presents design considerations on three fronts – content creation, video
production, and environmental usage. This paper will discuss these considerations, using current
marine projects: one for a maritime heritage trail in Wilmington, NC, and the other on the
proper use of the Hookah system for East Carolina University's Scientific Diving Course.
Introduction
The iPod is a new and unique medium for presenting educational and training content. Designing this
content for iPod presentations, however, is not the same as designing content for the average
television or computer. When a new medium, such as an iPod, is introduced, it is often thought of in
terms of old medium applications. This often creates the problem of adapting learning theory to
accommodate new learning technology (Stanton et al., 2001). This paper is designed to identify some
of the changes in the development and use of educational material that should be considered in order
to create effective instructional and training materials in the new video iPod medium.
To illustrate the unusual characteristics of the iPod learning technology, this paper presents two ongoing projects involving the development of video iPod content. The first case involves the
development of educational material for visitors to the waterfront of Wilmington, NC. Utilizing the
video iPod's mobility factor and leveraging its audio and video storytelling potential, modern visitors
can connect to a submerged and largely forgotten part of Wilmington's maritime heritage. The second
example examines the development of training materials and modules for use by students,
participating in East Carolina University's Scientific Diving Course.
Public Outreach – Maritime Heritage Trail
The first project involves the public dissemination of a maritime archaeology project investigating
submerged vessels in Wilmington, NC. One professor in East Carolina University's Program in
Maritime Studies is researching use and re-use patterns of historic vessels and cultural remains of
Wilmington's pre-WWII maritime industry. These remains have been reclaimed by the Cape Fear
River and by overburden and vegetation along the shoreline across from Wilmington's modern
waterfront. The professor has partnered with the authors to create a public outreach component of the
project with the intent to enhance visitors' personal experience as each individual moves along the
boardwalk of Wilmington's historic waterfront. The authors proposed something new--an iPod tour.
83
The authors believe that this learning environment is a perfect place to showcase the iPod's unique
capabilities. An interactive experience, using iPod technology, will provide the user with a more
personal experience, one that the user controls, and one that allows the power of audio and video to
tell a richer, more exciting story. The authors are designing the video content to create a more 'human
story' than static signage (the traditional medium of outdoor trails) alone can convey.
Visitors who want to use the iPod tour will have the opportunity to check-out the video iPod
hardware from the Information Booth at the beginning of the riverwalk in exchange for some
collateral, such as a driver's license. In this way, device maintenance and installation of the most
current software can be controlled. In future projects, however, the authors expect to mirror a tour on
a robust web site with accompanying set of videos – the same that are used for the video iPod tour.
On this web site, a user would be able to download the videos for their personal iPods.
Training – Hookah System
The second project involves the use of iPod technology to augment and reinforce training modules for
East Carolina University's Scientific Dive Training. For this project, the authors have chosen to use a
blended training approach, mixing the use of learner-controlled technology (iPod content), print
medium training, and instructor-led training for the different modules recommended by the Dive
Safety Officer. The purpose is to evaluate whether this multi-layered pedagogical approach allows
students increased access for reinforcement of skills presented by their instructor, and whether it
improves performance during evaluation. The use of the Hookah system, in and out of the water, was
explored as a test case primarily due to the relative ease of logistics for production of video content
and different layers of knowledge. For instance, a student needs to learn how to operate the Hookah
both on land and in the water. On land, students need to know how to operate the air-compressor, set
up hoses, attach hoses to regulators, break the unit down properly, and other relevant information. In
the water, students need to understand operating procedures, buddy skills, and safety procedures
particular to the Hookah system.
Video content will be accessible for download on the website of ECU's Dive Safety Office. These
Hookah videos were developed with instructor input and consent and were designed to support a very
detailed on-line training module for the Scientific Diver training at East Carolina University. Because
the type of internet connection (e.g., DSL, high-speed cable) used by students could not be
guaranteed, it was decided to make sure the files were small and therefore easily downloadable no
matter what kind of internet connection students might have. Different formats of the videos
(discussed below) also allow students, who do not have video iPods, to view the training material online on their computers.
Discussion
The iPod has three distinct characteristics that make it a completely new and unique medium. First,
the iPod's video capacity is based on H.264 750-Kbps video at 320-by-240 resolution combined with
128-Kbps audio and displayed on a 2.5 inch screen. This means that while the fidelity of the image is
very high the small screen size dramatically impacts the visual storytelling capabilities available to
the producers. Second, the iPod leverages advanced audio production techniques, by delivering the
audio through 'earbuds' more commonly known as headphones. This allows the producers to deliver
audio messages directly to the user with minimal environmental interference thereby enhancing the
effects of music, voice, and sound effects. And finally, the iPod is completely mobile, allowing the
user to be un-tethered from a dedicated, static communication environment. This means that because
the user dictates where and when the information will be conveyed, producers must be aware that the
video messages are competing with 'real world' surroundings.
84
Every learning design, regardless of medium, must be developed according to certain accepted
standards of development. The following factors should be considered by producers when developing
learning technology: 1) needs and goals for learning; 2) learning objectives; 3) physical and/or virtual
space; 4) tasks and interactions; 5) assessment methods; 6) audience and their characteristics; 7)
domain area; 8) community of learners and practices; and 9) technological capabilities and
possibilities (Ausburn, 2004; Kirkley et al., 2005).
Using this development process we begin to see how the attributes unique to this delivery medium
factor into the design and implementation of the learning message. Due to its three salient
characteristics of mobility, aural message delivery, and small visual display screen, content for video
iPods, such as dive training information, will not rely heavily on text for dissemination of
information. Instead audio and video will play the essential roles in delivery of the messages. With
this in mind, creators of content must constantly be aware that in some instances the iPod message
will be competing with the distractions of the external physical environment in which it is viewed.
Content Considerations
As with all media design, a solid design methodology is essential. This methodology follows six core
production phases: Analysis, Design, Development, Validation, Implementation, and Evaluation.
While many of the constraints of linear video learning transfer to the iPod, the unique iPod attributes
impact the content design in a number of significant ways. First, the small screen makes the use of
text rich learning difficult. For that reason text use will be minimized. Similarly, video segments
should be short since the iPod can be accessed in a variety of environments that will compete for the
learner's attention. Content delivered in short bursts will allow information to be presented in small
segments controlled by the viewer. For these reasons the authors are designing the content to
minimize text usage, and all video segments are designed to be under ninety seconds in length.
Additionally, there is a question regarding how these segments of content will be presented to the
learner. Most iPod users are familiar with its linear presentation of audio and video. For instance,
video content may be watched on iPod easily, but what happens at the end of the video? Normally,
you have to navigate through a series of menus to watch another video or listen to audio. However,
one of the iPod's least lauded and under utilized features is its ability to present information
interactively and out of sequence. This nonlinear ability allows visitors and students to pick and
choose what information they want to access. It also presents the opportunity to present simple
methods of information reinforcement. For instance, iPods can be used to ask students multiplechoice, true/false, or yes/no questions at any point in the learning process, which reinforce previous
training material. This is accomplished through simple and limited markup language. Markup
languages add code to text files to establish connections between files. Examples of markup
languages include HTML or XML, found on web pages. Limited markup language for iPods provides
for simple directives that create interactive links between the iPod's audio, video, images, and text.
The authors are developing their educational content with this nonlinear capability in mind in order to
provide clear and simple directions for users with a minimum amount of text (Sadun, 2006; Sande,
2006).
Environmental Considerations
While the flexibility of eMobile communication is obvious, one of the greatest concerns for designers
of content for the video iPod is that the user may have significant distraction around them in the place
they choose to view the material. The authors are less concerned about this occurrence for the
85
scientific dive training modules because the user can and is more likely to choose a quiet place to
receive the message. Additionally, learning is supported by alternative mediums – instructors and
written material.
On the other hand, the maritime heritage trail project adds a component of interactivity with the
environment. The learner will be holding the iPod and watching the video at the same time they are
looking at the exact geographical location of the submerged vessel. The outdoor viewing environment
will be a separate, as well as, a collective experience. Also the viewer is expected to stop their
physical movement on and off. They are expected to switch their attention from his or her external
surroundings to the video screen and the audio components of the learning module. In this case study,
it must be remembered that the video is simply supporting a display in the 'real world.' This means
that not every video segment will require a beginning, middle, and end – a strict narrative structure.
The iPod experience must be created as one that is integrated into the 'real world' of Wilmington's
waterfront.
Video Production Considerations
The authors are anticipating two production considerations that make the iPod experience different
from traditional video production development. First, because each user will have a headset, the audio
production will play a more significant role in the storytelling than it might in other open
environments. Because the user will be wearing the 'earbuds', there will be a greater opportunity to
speak directly to the user and to leverage the power of sound effects to create a virtual environment.
To that end, the scripts should be written to speak to the individual with one-on-one visual and audio
grammatical constructs.
Visually, the small screen presents some production considerations of its own. The small screen will
not display fine text, defined as anything under 36 points. Consequently, designers should develop
content that is video/audio based and not text based. Additionally, the authors are storyboarding
segments with limited use of variant shots. For example, the authors are designing for very few wide
shots. Most of the segments will be shot in medium shots, close ups, and extreme close ups and
almost all scenes will be void of text.
Video Production Process
During the pre-production phase the producer determines how the video message will be
disseminated. Usually there are three choices: full resolution DVD/broadcast, the web, and smallscreen mobile video. Each of these outlets brings with it certain production considerations regarding
video/audio shooting and editing. However, a compromise in production techniques can be reached if
the client intends to leverage all three means of distribution.
To that end, the video will be captured in the highest quality format allowed within the budget
constraints. Once captured and digitized for edit, it will be edited in the highest fidelity possible. With
the project completed and approved by the client it will now be reproduced for one or each of the
formats. The difference is the compression algorithm used. For example – DVD/broadcast will be
output at the same format the program was edited and with minimal compression to the video.
Conversely, the web version of the program will be severely compressed so that it can be passed
quickly between servers and the home user's computer. The compression drastically affects the
quality of the video when viewed by the user.
86
Conclusions
It is the authors' contention that the video iPod is a unique learning technology based on its small
screen, enhanced audio, and its mobility. These attributes require considerations for content design,
environmental usage, and video production. It is with great anticipation that the authors will
document and study the outcomes of our case studies using a variety of assessment tools including
pre and post testing, surveys and interviews to identify the effectiveness of the design related to the
specific learning goals proposed by each project.
References
Ausburn L. Course Design Elements Most Valued by Adult Learners in Blended Online Education
Environments: An American Perspective. Educ Media Internat. 2004; 41(4): 327-337.
Kirkley SE, Kirkley JR. Creating Next Generation Blended Learning Environments Using Mixed
Reality, Video Games and Simulations. Tech Trends. 2005; 49(3): 42-54.
Sadun E. Building Interactive iPod Experiences. MacDavCenter, 2006. www.macdevcenter.com.
Sande S. Take Control of Your iPod: Beyond the Music, TidBits Electronic Publishing, 2006.
www.takecontrolbooks.com
Stanton N, Porter L, Stroud R. Bored with Point and Click? Theoretical Perspectives on Designing
Learning Environments. Innovat Educ Teach Internat. 2001; 38(2): 175-182.
87
88
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Closed-Circuit Rebreathers in the Forensic Study of the Rouse Simmons
Shipwreck
Gregg Stanton1*, Keith Meverden2, Tamara Thomsen2 and James Garey3
1
Wakulla Dive Center; Inc., 147 White Oak Drive, Crawfordville, FL 32327
Maritime Archaeology and Preservation, Wisconsin Historical Society, 523 Atlas Ave., Madison,
WI 53714
3
Division of Cell Biology, Microbiology and Molecular Biology, University of South Florida, 4202
E. Fowler Ave. SCA110, Tampa, FL 33620
* corresponding author
2
Abstract
The Rouse Simmons is a three mast wooden schooner that sank in 165 ffw (50 mfw) in 1912
during a winter storm near the Wisconsin shoreline of Lake Michigan. The wreck is largely
intact and was surveyed during two weeks of the summer of 2006. Four dive team members
used closed-circuit rebreathers while three team members used technical open-circuit rigs. This
allowed a direct comparison of the two technologies under the challenging conditions of deep
cold water diving. Water temperature appeared to be the limiting factor in dive time for both
rebreather and open-circuit divers. A temporary filling station was set up at the boat dock with
oxygen, helium, booster pumps and a limited output air compressor. Expendable supplies for
rebreathers (gas and sorb) cost one-half that of the open-circuit rigs but the main benefit was the
decreased fill time and simpler logistics of the rebreathers. The rebreathers allowed a very
detailed survey of the hull, the rigging and the debris field around the wreck. Preliminary
forensic analyses suggest that the Rouse Simmons had steerage and was heading for shelter
when it sank. The mizzen mast snapped off just above the deck line and the upper portion was
not found. The main mast lies forward and to the port side of the hull and the base appears to be
missing. The foremast is intact and lies nearly parallel but on top of the main mast suggesting at
least one of these masts fell out of the mast step as the ship went down.
Introduction
Closed-circuit rebreathers (CCR) technology has been used in underwater survey work for many
years. By the 17th century Geovanni Borreli and Cornelius Drebbel had proposed and built rebreather
technology that permitted man extended investigations in refreshed air environments. Schwann and
Fluess, in the mid-1800s built functional personal rebreathers that were used for research,
investigations and rescue. Siebe and Dragger developed the technology for commercial and military
applications in the late 1800s. Hans and Lotte Haas were known in the early 1900s for their
underwater photography and studies using rebreathers. Clandestine applications of CCR by the
military expanded their availability through and after WWII (Bozanic, 2002).
By 1964, NASA was using a General Electric CCR in support of the Tektite underwater habitat. This
rebreather permitted numerous research scientists extended underwater time studying reef community
and structure while working from the habitat. Dr. Fred Parker used this experience with a company
called Biomarine that later supplied the US military with rebreathers. The Biomarine CCR influenced
many rebreather designs in use today. In the mid 1980s Dr. Bill Stone built a CCR for cave
exploration and mapping. Martin Parker of AP Valves in the 1990 brought CCR technology into the
civilian market (after the popular Drager semi-closed rebreathers (SCR) proved their limitations) with
89
the Inspiration CCR. Others soon followed. Today, there are dozens of CCR companies and more on
the market every year.
To date, many universities do not fully utilize rebreather technology within their scientific diving
programs. This may be due to the elevated initial cost, complexity of maintenance and training or the
additional hazards of this technology. But when compared to open-circuit diving (OC), the improved
diver comfort, options in an emergency and increased efficiency of operations (what we call the
footprint) of CCR technology are appealing to many underwater researchers (Gurr, 2002).
In November 1912, the 124 ft (38 m) schooner Rouse Simmons departed from Thompson, Michigan,
on northern Lake Michigan with a cargo of Christmas trees bound for Chicago (Pennington 2004).
The Rouse Simmons encountered a fierce November gale, and was last seen flying a distress signal off
Kewaunee, Wisconsin. A surf boat from the Two Rivers Life-Saving Station searched several hours
for the Rouse Simmons, but found no trace of the vessel or its crew. Discovered by recreational divers
in 165 ffw (50 mfw) in 1971, the 2006 survey was the first systematic documentation of her
wreckage.
The Rouse Simmons shipwreck site is located 12 mi (19 km) northeast of Two Rivers, WI, and six
miles (10 km) offshore. Project goals were to conduct a Phase II archaeological survey to the level
necessary for nomination to the State and National Registers of Historic Places, as well as document
any evidence of how and why the Rouse Simmons sank.
Methods
Diving: The estimated budget to support an eight person OC team for two weeks was nearly $10,000
for breathing gases alone, therefore, CCR technology was chosen to minimize breathing gas expense
and logistics. It was estimated that CCRs would reduce breathing gas costs by half, with the added
benefit of improving diver safety and comfort. There was insufficient time to train all staff on CCR,
thus providing the opportunity to compare CCR to OC performance while divers completed similar
tasks at the same site.
Field work was conducted over two five-day surveys, with one month between each survey. Nine
divers volunteered their diving services to the Wisconsin Historical Society (WHS) under the
direction of Keith Meverden, Nautical Archaeologist. Five OC divers participated in teams of two and
three, depending on their availability. Four CCR divers participated in all dives with one exception at
the end of the project. Lake Michigan's surface temperature is near freezing during winter months, but
rises during the summer with an upper thermocline between 60-70°F (16-21°C) that can reach to a
depth of 30 ft (9 m). Below the uppermost thermocline, the temperature drops to 48-60°F (9-16°C),
and can reach a depth of 60 ft (18 m). Below 60 ft in depth the temperature ranges from 38-42°F (36°C).
Candidates had to provide medical evaluations, diving credentials and health insurance, attend
appropriate training and follow the Wisconsin Historical Society's diving manual (based upon the
AAUS model). Several of the staff were newly trained on CCR technology in Florida and took time
to gain experience prior to the project. Closed-circuit rebreather and OC staff were required to have at
least Normoxic Trimix training. All had experience in shipwreck survey techniques. One member was
an accomplished UW photographer, one a former master rigger of sailing vessels, and several were
senior diving instructors. Males and females participated as divers. Three CCR models were used
during the project: Kiss Classic by Jetsam, Ouroboros by CC Research and Megalodon (APECS 2.0)
90
by Innerspace Systems Corp. All staff provided their own life support equipment except that two
Megalodon units were provided by Wakulla Diving Center, Inc. (WDC).
Boats with operators, lodging and food were provided by the WHS. A temporary dive locker and
blending station was set up at the dock in Two Rivers, WI. All blending technology (dedicated Haskel
booster pumps with controls), and CCR gasses were provided by WDC. All OC gasses and
compressors were provided by the WHS.
Divers were organized into OC or CCR teams. Daily briefs prior to departure defined research
objectives for each team, which included identification and description of components, measuring and
orientation, and photographing for later documentation on a master site plan drawing. Several things
helped the team do their work. A photomosaic (Figure 1) was provided to assist in the planning of
each dive. Each team was assigned an area of the site to document.
Dive profiles were based upon expected gas supplies and thermal stress. Visibility was very good
(Figure 2) and not a factor unless silt was disturbed during the dives. With a bottom temperature of
40°F (4°C), most divers did not stay beyond 30-40 minutes, with maximum bottom times of 44
minutes. Cold water acclimatization was not possible for two divers that traveled from Florida. All
divers used argon for dry suit inflation. Deep stops were utilized during decompression, but the
majority of decompression was spent in the warmer thermoclines. Including decompression, run
times exceeded 60 minutes, and often approached 90 minutes. Surface support was available for
entering and exiting the water from small boats.
OC configuration included back-mounted, large-capacity double cylinders using a 23/36 blend.
Stages were 30-50 ft3 aluminum or steel cylinders, one dedicated to 50% nitrox and one dedicated to
100% oxygen.
Figure 1. Photomosaic of the Rouse Simmons indicating the major features of the wreck
91
CCR configuration included three liter (20 ft3) steel RB cylinders (with the exception of the
Ouroboros using two liter (13 ft3) aluminum cylinders), and two 40 ft3 aluminum stages with bailout
gases in case of emergency. The diluent gas on all CCR rigs was 18/36 (18% oxygen/36%
helium/46% nitrogen). The bailout stages were 23/36 (23% oxygen/36% helium/43% nitrogen) and
70% nitrox. All participants used VR3 computers and followed the Buhlman decompression
algorithm. All divers carried either a redundant VR3 computer or a redundant depth and timing
devices and back-up decompression tables.
Each diver conducted one dive per day, after which all cylinders were recharged for the next diving
day. CO2 absorbent canisters were refilled after every dive, regardless of bottom time, with fresh
Grace Sodasorb 6-12. Staff archaeologists conducted post-dive debriefings with each dive team
member while reviewing the photo mosaic and emerging site drawing.
Survey
The length of the spars near and around
the wreck was measured underwater. In
some cases the ends were buried in silt
and had to be excavated or identified by
feel. Circumference and diameter of
each spar was measured at key positions
and taper of the diameter noted. The
ends of each spar were documented by
drawings or photographs when possible
to help identify the function of the spar,
and to determine if it was intact. All
measurements were done with a
measuring tape, ruler and calipers as
needed. Fittings and standing rigging
were documented by drawings,
measurements and photographs. The
relative position of each item to the hull
was noted on a site plan based on a
mosaic photograph of the site (Figure 1).
Spar, fitting and standing rigging
components were compared to historic
photographs of the intact Rouse
Simmons (Hirthe and Hirthe 1986;
Historical Collections of the Great
Lakes, Bowling Green State University).
Figure 2. Two CCR divers descend to the wreck showing the
generally good visibility at the site during the survey. The
bow of the wreck is clearly visible as is an open-circuit diver
to the right of the wreck. (photo by T. Thomsen)
Results
Diving:
Three OC scientists participated on the first week of the project. During the second week one returned
and two were replaced with new OC staff. The same four CCR staff participated in both weeks of the
survey. The day consisted of briefing, loading, transit to the site, research on the bottom (BT),
decompression, return to port, recharging and debriefing, supper, and retire.
92
Time to support the three OC staff at the blending station averaged a total of 2.5 hours
(blending/filling cylinders) per day. The four CCR staff required a total of 30 minutes to complete
recharging their two 20 ft3 cylinders each (base taken from a premix bank, but augmented as
required). Bail out cylinders did not require recharging during the project. Canister packing time must
be added to the equipment recharging picture. CCR staff repacked at home every day as part of their
setup procedures. Each reported taking 30-45 minutes to set up their rigs prior to diving, 10-15
minutes of which is used for canister repacking. Therefore CCR recharging takes 70 minutes total
when compared to the OC recharge of 150 minutes (since both must still set up their rigs).
Dive statistics over two one week periods:
OC dives (five divers)
Number of person dives: 25
Average BT: 31 minutes
Average In-water time: 62 minutes
Average gas consumption:
Oxygen 1,300 ft3 total /3 = 433.3 ft3 per diver or /25 = 52 ft3 per dive
Trimix: 3,500 ft3 total/3 = 1166.7 ft3 total or /25 = 140 ft3 per dive
Nitrox: 960 ft3 total (used in blending 70%)/3 = 320 ft3 per diver or /25 = 38.7 ft3/dive
CCR dives (four divers)
Number of person dives: 34
Average BT: 31 minutes
Average In-water time: 91 minutes
Average Sorb consumption: 5.5 pounds/dive
Average gas consumption:
Oxygen: 350 ft3 total /4 = 87.5 ft3 per diver or /35 = 10 ft3 per dive
Diluent: 470 ft3 total /4 = 117.5 ft3 per diver or /35 = 13.43 cf per dive
Trimix: 320 ft3 total /4 = 70 ft3 per diver (never used – refilled on 2nd wk)
Nitrox: 320 ft3 total /4 = 70 ft3 per diver (never used – refilled on 2nd wk)
Cost of the life support materials (not counting fuel, labor, rent, transport much of which was donated
or volunteer):
OC gasses: He 2,500 ft3
O2 1,300 ft3
Air 2,500 ft3
Total cost is $677.70 /3 = $225.90 per diver or $25.10 per day or $8.37 per dive
CCR gasses: He dil 470 ft3
O2 350 ft3
bail 680 ft3 (never used)
sorb 193 pounds
Total cost is $758.20 /4 = $189.60 per diver or $21.06 per day or $5.27 per dive
Survey
The Rouse Simmons was a three mast wooden schooner built in 1868 in Milwaukee, Wisconsin and
was approximately 160 ft (49 m) in length overall. The hull is unusual in that it was a double
centerboard design, but the rigging is typical of ships built for the great lakes trade at the time. The
standing rigging was wire rope that had been largely parceled and served, and fastened to the hull
93
with dead-eyes and lanyards attached to metal chain plates. Each of the three masts was made of two
spars: the lower mast and the topmast. The lower and top masts were typically connected together in
an overlapping fashion known as the doubling. Nearly all the standing rigging and spars were found
just forward of the bow of the wreck (Figures 1 and 3). The following is an overview of the survey
results.
The bowsprit was broken at the point it protruded from the bow of the ship and was found just
forward of the hull. The base of the bowsprit was still within the hull where it fits into the Sampson
post below the forecastle deck. The external portion of the bowsprit was still connected to the intact
jib boom. Much of the bowsprit and jib boom standing rigging was still attached, composed of chain
and wire. The martingale was not found.
The lower foremast was intact but found lying forward of the hull with the base near the port side
bow and angled slightly to the starboard with the mast largely parallel to the hull. A top (platform)
was found in place on the lower foremast. The fore lower yard was still attached to the lower mast by
the metal truss, which was severely bent on one side. The yard was broken at one end and the piece
missing. The fore topmast appeared to be intact, its base lying near the top end of the lower mast at a
90o angle. The upper end was buried deeply in the silt and could not be observed. The hardware that
connects the topmast to the lower mast (fid and cap) were intact and in place. The lower end of at
least some of the wire rope topmast shrouds appeared to be still connected to the futtock shrouds. The
futtock shrouds were made of solid round metal stock approximately 1.0 inch (2.54 cm) in diameter
and appeared to have turnbuckles forged onto them. They were deformed and bent. The main lower
mast was found on the forward port side of the hull, slightly aft and generally parallel to, and
underneath the foremast. The base of the main lower mast could be examined only by touch because
it is buried in silt appeared to be incomplete. Crosstrees were found and a top (platform) was not
present. The main topmast was found lying over both the fore lower mast and the main lower mast. It
was completely intact as were the fid and mast cap. Heavy wire standing rigging was found loosely
looped around the topmast and appeared to extend under the silt to the underside of the buried main
lower mast (Figure 3).
The mizzen lower mast was broken above the deck with the stump in its original location in the mast
step. The remainder of the mizzen lower mast was not found, nor was the mizzen topmast. A variety
of gaffs, booms and other spars were found on and around the hull. Most appeared to be standard
gaffs and booms with wooden jaws as expected on a gaff-rigged schooner. Three smaller spars found
on the site were not positively identified.
The Rouse Simmons's hull is largely intact and lies on an even keel. With the exception of the port
and starboard quarters, all outer hull planking was intact with caulking. Outer hull planks not
colonized by zebra and quagga mussels exhibit traces of green and blue paint. The only visible
damage to the outer hull was located at the stern. Both port and starboard quarters had several outer
hull planks missing immediately forward of the transom (Figure 5). The missing planks were located
above the water line. The transom is dislodged on the port side, pushed several inches outward from
the hull.
The vessel's stern cabin is no longer extant. Very little weather deck planking remains intact. Much of
the cargo hold is visible through the deck beams, filled with silt and stacked evergreen trees. Many of
the trees retain their needles. Nearly all deck beams exhibit a salt channel cut lengthwise along their
uppermost surface that was used to salt the deck as a preservative. Both fore and aft centerboard
trunks are visible on the vessel's centerline, with the centerboards intact within the trunks. A deck
winch to raise the aft centerboard remains intact, with a single-acting bilge pump intact aft of the
94
centerboard winch. The fore and main chainplates on the port side have been forcibly pulled toward
the stern.
Figure 3. Rigging deposited off the bow (upper left) of the vessel, showing top masts perpendicular to the
foremast. The top on the lower foremast can be clearly seen. The main mast is buried below the foremast
and cannot be seen in this photo. The main top mast is shown held perpendicular to the foremast in the
background. The cap and fid of each topmast can be seen. The diagonal spar in the left foreground is the
broken yard used for the Raffee sail. (photo by T. Thomsen)
The weather deck is intact from the forecastle scuttle to the stempost, as well as the windlass and
forecastle deck. Both the port and starboard anchor chains have been heaved from the chain locker
and are piled on either side of the weather deck aft of the windlass. Both port and starboard anchor
chains take four turns around the wooden patent windlass before exiting the hawse pipes.
There is evidence that the port anchor was being prepared for deployment. The four turns of the
starboard anchor chain have been loosened around the windlass barrel, gathered, and tied to the
strongback that was affixed atop the windlass. The anchor chain has rusted into its tied-up position,
and fragments of the fiber line that tied the chain is visible protruding from the chain links. One of the
four turns of the port anchor chain is run through an iron Norman Pin that is embedded into the
windlass barrel (Figure 4). The Norman Pin (shaped like a large iron staple) was used as a chain
stopper prior to modern iron stoppers with a swinging iron gate. Once an anchor was deployed, the
Norman Pin was driven into the windlass to lock down the anchor chain, preventing further
deployment of chain, and took the strain off the wooden windlass pawl. The Norman Pin is in place at
the ready, but has not been driven into windlass to lock down the chain.
The starboard anchor was salvaged by recreational divers, and the severed starboard anchor chain
hangs freely from the hawse pipe. The port anchor chain hangs loosely from the bow, is draped over
95
the broken bowsprit on the lakebed, and disappears into the sand forward of the bow. The port
anchor's location is unknown, but there is no record of diver salvage. Early underwater footage
recorded soon after the vessel's discovery suggests the port anchor chain has not been disturbed, and
that the port anchor may have been deployed prior to sinking. A search for the port anchor off the
immediate wreck site was not conducted, but will be the focus of 2008 efforts.
Figure 4. The windlass showing the Norman Pin
(inverted u-shaped metal in lower left). (photo by
T. Thomsen)
Figure 5. The stern of the ship showing damage to
the port side of the transom. The stump of the
mizzen mast can be seen. (photo by T. Thomsen)
Discussion
Diving
From our data, it appears that the bottom time, or effective time at the research site, for both OC and
CCR was virtually the same on this project. Both OC and CCR divers were faced with the same cold
water stress that may have normalized this data, and not due to life support efficiency. Two cold
water acclimatized CCR divers spent an average of 37 min/dive bottom time across two weeks of the
project, while the two non- acclimatized CCR divers spent an average of 24 min/dive of bottom time.
All OC divers were cold water acclimatized.
Time at decompression for OC folks was found to be half that for CCR folks. The VR3 does
accommodate the more time consuming gradual reduction of trimix possible with the CCR when offboard gas shifting is not followed (as in this case). OC people did perform gas shifts during
decompression.
The cost of supporting a CCR dive was found to be 63% that of the OC dive when factoring all
consumables used on this project. The cost of delivery was presumed to be equal in both groups and
was not factored into these figures. Clearly, if the project had paid a shop to fill cylinders, the cost of
labor and overhead would have been much higher. Shop charges are approximately $0.35⋅ft-3 for
trimix, $0.25⋅ft-3 for oxygen and $0.20⋅ft-3 for 70% nitrox. With shop prices for fills, the expense for
CCR is approximately half that of OC.
CCR is less expensive than OC per dive hour when making helium-based dives. CCR requires much
less in-field logistics for gas procurement, fill station time, storage, weight, etc. There is also less
danger from out-of-gas emergencies, increased diver comfort with more options when problems do
arise. There were no significant differences in efficiency or maintenance between the different
rebreather models used on this project. However, the initial cost of CCR is significantly higher than
96
OC (purchase price and specialized training for CCR technology), and divers can have longer inwater exposures at deeper depths thus with exposure to increased risk.
CCR has definite advantages as a long-term investment, but not necessarily for short term gains (i.e.,
training divers and purchasing gear for a specific project). However, anywhere that gas procurement
is an issue, any CCR disadvantages are outweighed.
We recommend a cost/benefit analysis when considering OC vs. CCR life support technology on a
research project. Clearly, for long term investment in UW research, CCR technology will provide
better life-support to the science diving community on deeper, longer duration, more arduous
exposures. Research requiring minimal noise intrusion, water column mixing, clandestine operations
(bubbles) or when gases are limited and/or a small footprint is required, CCR technology will become
more beneficial than OC. Availability of gases, functional CCR rigs, trained researchers, and
environmental considerations are all cost considerations that should be compared to the cost of OC
options.
Selection of the CCR technology, maintenance and repair options and training remain critical to the
success of such a venture. The WHS relied upon personal CCR units and the assistance of a training
and maintenance facility to insure reliable technology during the project. An institution will need to
either provide competent internal monitoring or work closely with external companies to insure CCR
reliability.
Rigging and Hull
The wreckage indicates that the rig of the Rouse Simmons was essentially as seen in historical
photographs and typical of great lakes schooners of the era. The wreck is remarkably intact and the
majority of the spars were found and identified. The fore lower mast was positively identified because
of the presence of the top, a small platform found in place of spreaders that are normally found on
fore and aft rigged masts. The top allows a secure place to work aloft and acts as a spreader and or
fairlead for topmast shrouds. A top is characteristic of square rigged masts as well as fore and aft
rigged masts that are fitted with a yard. Only the foremast of a schooner would be fitted with a yard.
In the Rouse Simmons case, the yard served to support the base of a triangular sail known as a Raffee
(or Raphe) topsail for sailing down wind (Underhill, 1946). This style of topsail is characteristic of
great lakes and coastal schooners but was not in common use elsewhere. One anomaly of the foremast
was that the futtock shrouds appeared more modern than those expected on a ship of the Rouse
Simmons age. The futtock shrouds are solid metal rods that form a connection between the wire rope
topmast shrouds above the top or crosstrees and the futtock band located on the lower mast just below
the cheeks. Those found on the wreck were of heavier gauge than would be expected and had
turnbuckles forged onto them. Normally, futtock shrouds have an eye at each end, the lower eye
connects to the futtock band, and the upper end runs through a groove in the spreader or top and
terminates in a larger eye or loop forged around a dead-eye. The futtock shroud/turnbuckle
arrangement does not appear to be original and is consistent with a report that the Rouse Simmons was
refitted in 1905 after being partially dismasted. Turnbuckles were not found elsewhere on the wreck
and were not in general use on wooden ships of the era. The more modern design we found is
consistent with a historical report that the Rouse Simmons was partially dismasted in 1905
(Milwaukee Sentinel, 22 Oct 1905) and subsequently refitted.
The main mast was identified because it is nearly the same diameter as the foremast but did not have
a top. Instead it was fitted with simple spreader bars which act as a fairlead for topmast shrouds and
some backstays. This is typical of fore and aft rigged masts that did not have yards because a secure
place to work aloft was not necessary. Historical records show that some great lakes schooners had
97
tops installed on all masts, and although it is common to find a top on the foremast but not the other
masts, it is highly unlikely that a top would be installed on the mainmast but not the foremast. Several
pieces of evidence allowed the identification of the main topmast. First, both the main lower mast and
the main topmast were located aft of their foremast counterparts. Second, a heavy loop of standing
rigging, probably a main lower shroud, is loosely wrapped over the topmast and extends under the
lower main mast. Third, it appears that a piece of standing rigging from the fore topmast (a topmast
shroud) is still connected to one of the futtock shrouds associated with the fore lower mast.
With an intact (and presumable watertight) outer hull at the time of the sinking, the question remains
as to why the Rouse Simmons sank that November day. It will never be known for certain, but it is
possible the missing deck planks played a role in the foundering. Given the intact nature of the hull
and the less dynamic environment due to the water depth, the deck planking was expected to be intact
as on other wooden vessels in similar Lake Michigan depths. The presence of salt channels on the
deck beams suggests the iron deck fasteners may have been compromised by corrosion, allowing the
deck planks to dislodge in the storm and allow water entry into the hull.
What happened to the Rouse Simmons?
The standing rigging of a ship is a system integrated with the masts and the hull. Under stress, rigging
failures can lead to dismasting and weaknesses in the wooden masts can lead to rigging failures.
Wood rot in the masts or corrosion of wire standing rigging are potential weaknesses of a wooden
ship of this type. However, it would be unlikely to find evidence of rot or corrosion this long after the
wreck. Otherwise, under stress, the failure points are the connections of the standing wire rigging to
the chain plates which was accomplished by hemp lanyards threaded through the dead-eyes. Other
than the bowsprit/jib boom, we found no evidence of standing rigging still attached to hull chain plate
dead eyes by hemp rope, nor did we find standing rigging positioned near hull chain plates. Instead,
all of the masts and standing rigging appeared to have been thrown forward of the hull in a relatively
small area (Figures 1 and 3).
The mizzen mast broke off above the deck and the upper portion was not found. We found evidence
that the main lower mast also broke off above the deck but if so, the base portion has been lost. It is
unclear whether the breaking of these masts occurred prior or during the sinking. The fore lower as
well as the fore and main topmasts are intact. Because the mounting hardware is still present and
undamaged, it appears that these masts fell out of their steps while the ship was sinking. The finding
of a broken bowsprit with an intact jib boom is understandable in that it broke where it passed the
stem of the ship. The outboard portion of the bowsprit is unlikely to have broken because it is doubled
with the jib boom. Because the jib-boom was not broken by the same forces that severed the bowsprit,
it is possible that the bowsprit broke before the wreck collided with the bottom of the lake. It is also
unusual that the port anchor chain runs atop, rather than below, the bowsprit.
Historic records suggest that the ship sank in the afternoon in temperatures slightly above freezing
during gale force winds coming out of the northwest. The position of the port side anchor chain
suggests the ship was facing northwest and dragging an anchor. The day of the sinking the winds had
come from the southeast and had changed to the northwest during the day, possibly creating large
choppy waves. There is no evidence that the ship had significant sail set at the time of the sinking.
One scenario consistent with our observations is that the ship foundered as a result of the gale force
winds and large choppy seas. The crew may have dropped the anchor in an effort to control the ship.
The ship foundered, rolling to port when the bow went under. Wind and water forces wrenched the
masts partially aft as evidenced by the distorted chain plates on the wreck. At some point, either on
the surface or as the ship sank, the fore and main masts were whipped forward and fell in the
98
relatively small area immediately forward of the wreck as reflected in our survey. As the ship sank,
air pressure grew in the stern of the hull, further tilting the bow downward. Air pressure trapped in the
stern of the wreck escaped through the transom, partially dislodging the transom and several outer
hull planks as evidenced by the transom damage (Figure 5).
References
Bozanic J. Mastering Rebreathers. Best Publishing: Flagstaff AZ, 2002; 548 pp.
Gurr K. Technical Diving From The Bottom Up (Phoenix Oceaneering 01202 624396), 2002; 250 pp.
Hirthe WM, Hirthe MK. Schooner Days in Door County. Voyager Press, Minneapolis, MN, 1986;
147 pp.
Historic image credit: Historic Collections of the Great Lakes, Bowling Green State University.
Pennington R. The Historic Christmas Tree Ship: A True Story of Faith, Hope and Love. Pathways
Press: Wast Bend, WI, 2004; 323 pp.
Underhill HA. Masting and Rigging the Clipper Ship and Ocean Carrer: With Authentic Plans,
Working Drawings and Details of the Nineteenth and Twentieth Century Sailing Ship. Brown, Son
and Ferguson, Ltd: Glasgow, UK, 1946 (reprinted 1949).
99
100
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Rebreather Fatality Investigation
Richard D. Vann, Ph.D., Neal W. Pollock, Ph.D., Petar J. Denoble, M.D., Sc.D.
Divers Alert Network, 6 Colony Place, Durham, NC 27705, USA
*corresponding author: rvann@dan.duke.edu
Abstract
Sixty-six representatives of rebreather manufacturers, training agencies, government agencies,
rebreather users, and DAN met in November 2006 to discuss objectives for rebreather fatality
investigations. DAN has collected information on 80 recreational diving rebreather deaths from
1998 through 2006. The annual number of rebreather fatalities appears to have tripled since
1998. The percentage of fatalities involving rebreathers among US and Canadian residents
increased from about 1 to 5% of the total number of diving fatalities captured from 1998
through 2004. Rebreather fatality investigations attempt to reduce future occurrences by
identifying causative factors, primarily focusing on three areas: medical, equipment, and
procedural. Medical investigation dwells on diver health and final cause of death. Equipment
investigation addresses potential hardware issues. Procedural problems appear to be more
common than equipment problems but are often difficult to identify. Witness reports and 'black
box' recordings of rebreather function could help untangle procedural and equipment issues.
Enhanced international training and cooperation will facilitate effective incident investigation
and, ultimately, the education of the diving community.
Introduction
This paper is a compilation of information from multiple sources: 1) a two-hour meeting before the
2006 Diving Equipment and Marketing Association (DEMA) Show in Orlando with 66
representatives from rebreather manufacturers, training agencies, government agencies, and
rebreather users; 2) comments on a draft of the meeting report; and 3) a preliminary review of
rebreather fatality data collected by the Divers Alert Network (DAN). The data are from recreational
diving with open-circuit fatality data presented for comparison. The term rebreather is used to
describe closed-circuit or semi-closed circuit mixed gas scuba. The paper attempts to give the reader a
sense of the issues and potential opportunities.
The Problem
Figure 1 summarizes 80 rebreather deaths collected by DAN America for 1998 through 2006. The
information was obtained from internet searches or sent to DAN by interested persons. This does not
necessarily include all rebreather fatalities in the world, just those about which DAN has been
informed. The total number of rebreather fatalities per year appears to have tripled since 1998. In
general, annual fatalities among non-US or Canadian residents have been greater than among US and
Canadian residents but both are increasing. The most likely explanation for the increase in rebreather
fatalities is that rebreathers have become more popular, and manufacturers are selling more. We do
not know this for sure, however. Sales data are not currently available.
101
18
Total
Number of Fatalities
16
Non-US/
Canadian
residents
14
12
10
8
6
US/Canadian
residents
4
2
0
1998
1999
2000
2001
2002
2003
2004
2005
2006
Figure 1: All known worldwide rebreather fatalities.
10%
100
90
80
70
60
50
40
30
20
10
0
Total
number
8%
6%
Percentage
of
Number of
Rebreather
4%
2%
Percent Rebreathers
Number of Fatalities
Figure 2 shows that the percentage of fatalities involving rebreathers among US and Canadian
residents has increased from about 1 to 5% of the total US/Canadian fatalities from 1998 to 2004.
(This is approximate as the total number of US and Canadian fatalities is not known for certain.)
Once again, this probably reflects the growth in rebreather sales, but no sales data are available.
0%
1998
1999
2000
2001
2002
2003
2004
2005
2006
Figure 2: All known fatalities and all known rebreather fatalities among
US and Canadian residents.
Risks Other than Death
Does the increase in rebreather deaths suggested by Figures 1 and 2 indicate a rebreather safety
problem? Rebreather meetings in 1994 and 1997 were not followed by action to make the root causes
102
of fatalities easier to identify. In the absence of such action, the present rise in deaths might lead to
the perception of excessive risk and the imposition of restrictive limitations. Only the rebreather
community (i.e., users, trainers, manufacturers, etc.) can change this perception.
The threats may be legislative, legal, or operational. In addition to local ordinances, one opinion at the
2006 DEMA rebreather meeting held that federal terrorism laws passed since 9/11 might be used as
pretexts to shut down rebreather diving, not because of terrorist threat, but because of public
perception. Another opinion was that private litigation by next of kin is a greater risk. One
manufacturer has reported that three dive boats have banned rebreathers in the Los Angeles basin, and
some dive shops would not fill pony bottles because they can be used in rebreathers.
There may or may not be a real safety problem, but if private individuals or the public perceive one
exists, the consequences will be the same. Multiple voices stated that now is the time to put the house
in order before public or private perception becomes reality. Tact, diplomacy, technical development,
training, community cooperation, public relations, and cooperation with law enforcement authorities
and legislative bodies/agencies will be required. Fatality investigations are unpleasant, and the best
outcome is the prevention of similar events. That is why investigations are important.
The community needs to self-regulate. An attorney who is involved in rebreather litigation and is also
a rebreather diver stated that local, state, and federal agencies simply do not know what to do. This is
a problem for both the industry and the families of rebreather divers who die.
Rebreather vs. Open-Circuit Fatalities: A Preliminary Look
Factual information from fatality investigations is needed to identify the most important points for
action and to avoid restrictions on rebreather use. To illustrate how this process might work, we
compared 964 open-circuit fatalities from 1992-2003 with 80 rebreather cases from 1998-2006. These
are preliminary data and no conclusions are warranted.
We used a simplified form of Root Cause Analysis shown as a four-event sequence in Figure 3
(Rooney and Vanden Heuvel, 2004). Event (a), called the trigger, was the earliest identifiable root
cause that transformed an unremarkable dive into an emergency. Event (b), called the disabling agent,
was a root cause identified immediately before the disabling injury. Event (c), the disabling injury,
caused death or rendered an incapacitated diver susceptible to drowning. Event (d) was the cause of
death specified by the medical examiner that might be the same as the disabling injury or might be
drowning secondary to the disabling injury. It was not unusual for one or more of the four events to
be unidentifiable. This was particularly true for rebreather fatalities that occurred with nonUS/Canadian residents which DAN America could not investigate.
(a) Trigger
(b) Disabling agent
(c) Disabling injury
(d) Cause of death
Figure 3: Root cause analysis as modified for diving fatalities.
103
Figure 4 illustrates triggers that were identified in 338 open-circuit and 30 rebreather cases.
Equipment trouble and buoyancy problems appeared more common for rebreathers than for opencircuit. The trigger category equipment trouble included both procedural problems and less common
equipment malfunctions that occurred during diving. Only three apparent equipment malfunctions
were identified: flooded display; oxygen supply failure; and an unspecified malfunction at 330 fsw
(100 msw) in a cave. There were 11 apparent procedural problems that reflected inappropriate
preparation (including maintenance) or equipment operation by the diver: oxygen valve not on;
electronics not turned on (two cases); gases not checked and displays not on; oxygen sensor
incorrectly installed, oxygen valve partly blocked, loose connections; pre-dive malfunction of oxygen
system, used emergency semi-closed mode during dive instead; gave display to student and relied on
automatic oxygen addition; gas leak in breathing loop and bad oxygen sensor; removed rebreather in
wreck to bypass an obstruction; gas supply valve set to an external rather than internal source; and
mouthpiece valve sticking but dived anyhow. There were seven buoyancy problems. Four appeared to
be rebreather-related and involved mouthpiece removal after ascent with failure to close the
mouthpiece followed by sinking. Three were not rebreather related and included: tangled in lift bag,
pulled to surface, fatal DCS; drysuit valve failure, blow-up, fatal AGE; and corroded drysuit valve,
blow-up from 300 fsw (91 msw), and fatal DCS.
45
40
338 Open-circuit cases
Percent of Total
35
30 Rebreather cases
30
25
20
15
10
5
0
Insufficient Gas
Entrapment
Equipment
Trouble
Rough Seas
Trauma
Buoyancy
Figure 4: Triggers in open-circuit and rebreather diving fatalities.
Figure 5 shows the disabling agents that were identified. Emergency ascent was important for both
open-circuit and rebreathers. Insufficient gas seemed more common with rebreathers which might
seem surprising until it is recalled that the diluent supply in a rebreather is small and can be quickly
exhausted by a leak or multiple upward and downward excursions. The largest difference between
open-circuit and rebreathers was for inappropriate gas which represented apparent hypoxia, oxygen
toxicity, or carbon monoxide poisoning. There were four cases of insufficient gas: direct ascent from
250 fsw (76 msw) with DCS and AGE; gas leak in breathing loop with rapid ascent and AGE; loss of
consciousness at depth; and open-circuit until 180 fsw (55 msw), buddy-breathed, and drowned.
There were five seizures suggesting oxygen toxicity: on switch to open-circuit; at 35-40 min; when
caught in a downdraft from 28 to 82 fsw (8.5 to 25 msw); and two with no obvious explanation.
Thirteen lost consciousness early in the dive suggesting hypoxia or a cardiac incident. Entrapment
with rebreathers was less common than with open-circuit and included: tangled in line while
attempting body recovery at 880 fsw (268 msw), drowned; tangled in lift bag at 150 fsw (46 msw)
decompression stop, pulled to surface, fatal DCS; tangled in line at 130 fsw (40 msw) in cave,
drowned; and tied self to coral during 20 fsw (6 msw) decompression stop, drowned.
104
50
Percent of Total
40
27 Rebreather cases
30
20
10
0
Emergency
Ascent
Entrapment
Insufficient
Gas
Buoyancy
Trauma
Inappropriate
Gas
Equipment
Trouble
Figure 5: Disabling agents in open- and rebreather diving fatalities.
Figure 6 suggests that drownings and cardiac incidents were less common with rebreathers than with
open-circuit while presumed hypoxia and oxygen toxicity appeared to be responsible for over half the
disabling injuries with rebreathers.
60
24 Rebreather cases
Percent of Total
50
40
30
20
10
0
Drowning
AGE
Cardiac
Trauma
DCS
LOC
Inappropriate
Gas
Figure 6: Disabling injuries in open- and rebreather diving fatalities.
In Figure 7, the main cause of death was drowning for both open-circuit and rebreathers. Bear in
mind, however, that the disabling injury was more relevant to the prevention of diving fatalities
because a disabled or incapacitated diver was incapable of self-rescue and would drown if a buddy or
other divers were not present to lend assistance.
105
100
90
Percent of Total
80
15 Rebreather cases
70
60
50
40
30
20
10
0
Drowning
AGE
Cardiac
Trauma
DCS
Figure 7: Causes of death in open-circuit and rebreather fatalities.
Other characteristics of the 80 rebreather fatalities included: eight semi-closed; four kits, homemade,
or modified; and 26 diving alone, were separated, or lost contact. Practically no information was
available on 14 cases. In one case, a diver was dragged by a large speared grouper from 140 to 190
fsw (43 to 58 msw) and lost consciousness on the bottom. A diver who separated from his buddy was
found at 268 fsw (82 msw) with a shark bite although it was unknown if this was pre- or postmortem.
The information presented above was too incomplete on which to base useful conclusions. The
purpose of presenting it was to show that it is conceptually possible to identify the main factors
associated with different adverse events in diving fatalities. For such an analysis to be successful,
however, better and more complete information must be collected. This will require cooperation by
the entire rebreather community – divers, operators, and manufacturers – and in addition, law
enforcement agencies and medical examiners. There is much work to be done. Some of the issues are
addressed below.
Investigation of Diving Fatalities in the UK
In the United Kingdom (UK), when a sudden death of unknown cause occurs or someone dies at
work (including death of a student in a diver training course), an investigation or inquest is conducted
by a Coroner (Sheriff in Scotland) with the assistance of a jury. (The terms 'coroner' and 'sheriff' have
different meanings in the US and UK. A Coroner in the UK, or Sheriff in Scotland, conducts judicial
inquiries which are heard in Court.) The coroner's objectives are to establish: (a) the identity of the
deceased; (b) when, where, and how death occurred; and (c) whether recommendations to prevent
similar accidents are needed. Standard verdicts include: (a) natural causes; (b) accident/misadventure;
(c) killed unlawfully; (d) open verdict (insufficient evidence to reach a conclusion); or (e) a narrative
verdict in which the jury's factual conclusions are summarized.
Diving fatality investigations attempt to identify causative factors in three areas: medical, equipment,
and procedural. Medical investigation focuses on diver health and on the cause of death. All coroners
in the US, or post-morticians in the UK, are qualified to assess cardiovascular disease, the most
common health problem in diving deaths, but barotrauma, cerebral gas embolism, and decompression
sickness require special training.
106
Drowning is the most frequently assigned cause of death when the real culprit may have been a
traceless injury or unwitnessed event. Diving injuries that can result in unconsciousness with no
detectable trace other than a drowned diver include: a) deep water blackout; b) hypoxia with semiclosed equipment in shallow water or closed-circuit equipment with the electronics off; c) CNS O2
toxicity from an oxygen control system malfunction or from breathing the wrong gas at too great a
depth; d) CO2 toxicity from over-breathing (breathing at a higher ventilatory rate than the unit can
support with reasonable breathing resistance); and e) CO2 toxicity from canister channeling or using a
canister beyond its functional limit.
In the UK, attendees at an investigation may include the investigating officer (a police office and/or a
health and safety inspector), a forensic scientist, a rebreather manufacturer's representative, and a next
of kin representative. An opinion was expressed that if the manufacturer attends, the next of kin
should be present, or represented, so that all parties have the same information. During an
investigation, everything should be photographed and use of a tape recorder with an open microphone
should be considered. Video recording is strongly recommended. All evidence and information
should be preserved in compliance with the local rules.
A strong opinion held that incident 'investigation' by Internet chat room was to be avoided. Internet
forums are useful for communicating knowledge, but people without knowledge who speculate about
a fatality under investigation waste time, money and, worse, upset the next of kin. Those who run
Internet forums were encouraged to discourage inappropriate speculation or perpetuating missinformation regarding fatalities under investigation. When factual information is available, it should
be transmitted to the investigating authority.
Some reviewers were equally strong in support of using the Internet to exchange information about
rebreather fatalities as is commonly done. This is clearly an important tool that might be further
developed.
Equipment Testing
Equipment testing is important because next of kin tend to focus on equipment error rather than
human error as the root cause. In 18 post-incident rebreather tests conducted by the US Navy
Experimental Diving Unit (NEDU) since 2002, equipment malfunctions were easy to detect but rare.
When they occurred, the most common equipment problems found at NEDU included misadjusted
regulators; broken, corroded or poorly maintained components; and contaminated gas. Human error
or procedural problems appear to be more likely than equipment problems but can be difficult to
identify. NEDU has only focused on equipment testing, not procedural issues. The UK system is
more general.
Compared to open-circuit scuba incidents, rebreather testing is time consuming and expensive, and in
the UK, investigating authorities are not inclined to request equipment testing unless absolutely
essential because a typical cost is ~£3,000-5,000 per day for 5-10 days. The next of kin or
manufacturer sometimes offers to contribute, but the UK investigating authority must agree if the
evidence is to be used in court.
Rebreather testing is not easy and sometimes not possible, particularly when the equipment is
damaged, or there are long shipping delays. Only a fraction of the units are received in suitable
condition for testing. For example, a unit that had been stored after flooding may need to have the
absorbent chipped out before investigation. Investigation seeks to determine if the diver's hardware
(or software) was functioning properly, but equipment can never be tested exactly as it was during the
107
fatal dive. The primary aim of equipment testing is to replicate the dive profile in an unmanned test
facility. Complicated cases may require 'reenactment.' A secondary aim pertains to the validation of
design standards. Ideally, life support equipment that is sold should be designed to meet accepted
standards. The European Union and US Navy have defined the principal standards although they are
not in complete agreement. Testing for compliance with standards may require replacement parts
from a certified source, usually the manufacturer.
Testing cannot determine if early breakthrough of a CO2 absorbent canister occurred because a
canister cannot be repacked exactly as it had been dived. Another difficulty in investigating CO2
canisters is flooding which commonly occurs if the diver becomes unconsciousness and drops the
mouthpiece. Flooding may corrode and destroy components, particularly if they are aluminum, but
even a flooded canister can provide some information. If the canister contents (even a slurry that has
migrated to other parts of the breathing loop) are sealed in an airtight container and sent to a test lab
such as Qinetiq or Molecular Products, it may be possible to estimate how much of the absorbent had
been used by measuring how much calcium hydroxide had been converted into calcium carbonate.
Capabilities desirable for a facility that conducts rebreather tests include: a) knowledge of police
investigative procedures to maintain proper chain of custody; b) an unmanned and manned test/dive
facility; c) experts in the equipment to be tested; d) ability to conduct and interpret gas analysis,
sometimes from minuscule amounts of remaining gas; e) ability to download and interpret dive
computer or rebreather black-box data; f) ability to simulate underwater breathing apparatus (UBA)human interactions; and g) ability to review medical examiner reports for consistency with known or
discovered facts.
Two centers in the UK test rebreathers – Qinetiq and HSE. In the US, NEDU has conducted
underwater breathing apparatus testing as a public service for at least 30 years, and what the Navy
learns helps to make Navy diving safer. Testing can include inspection, unmanned breathing machine
evaluations, and manned reenactments in a test pool. Unless litigation is involved, the knowledge is
sometimes made public. NEDU tests open-circuit, surface-supplied, and military or civilian
rebreathers, but the scope of NEDU ends at equipment evaluation. The Navy does not do complete
incident investigations. Investigative agencies that have that responsibility and have requested
equipment tests include the US Coast Guard, coroner/medical examiner offices, local fire
departments, local police departments, state police, Occupational Safety and Health Administration
(OSHA), National Institute of Occupational Safety and Health (NIOSH), and military commands. In
the US, the NEDU is probably the best site for equipment testing, and what they learn about civilian
equipment is of value to Navy diving, but cost, time, and reluctance to become involved in litigation
makes the Navy uncomfortable to serve in that capacity. If the rebreather community could help
alleviate these problems, the US might achieve as good a testing system as the UK. There was some
discussion after the DEMA meeting about establishing a fund that might help to support equipment
testing, but this effort was apparently unsuccessful in gaining traction.
The opinion was expressed that it is in the manufacturers' best interest to make equipment testing as
easy as possible. This might include providing an equipment manual, a manufacturer's representative
who knows the equipment, replacement parts, and assistance with digital download from equipment
instruments. There was disagreement as to whether having a manufacturer's representative would
compromise the objectivity of a test, but this is done in the UK with full recording of all proceedings.
108
Procedural (Human Error) Investigation
Procedural investigations examine a diver's behavior against training and manufacturer standards for
equipment use or for environmental stress management (e.g., ascent rate, oxygen and nitrogen
exposures, buddy system, decompression procedures, training qualifications, etc.). Witness reports
and 'black-box' recordings can help untangle procedural and equipment problems. The importance of
black-box data was demonstrated at the DEMA meeting for two rebreather fatalities in which there
were recordings of depth, time, O2 sensor readings, CO2 scrubber utilization, and O2 and diluent gas
supply pressures. This information made it possible to identify: a) the presence of hypoxia or
hyperoxia; b) emergency or rapid ascent; c) buoyancy problems; d) use of the open-circuit emergency
mode including gas consumption as a work rate indicator; and e) previous dive history and sensor
calibration to show the extent of training on a new unit before a fatal dive. The examples presented
were helpful for identifying procedural problems while demonstrating that the rebreathers performed
satisfactorily. There appeared to be general agreement at the meeting that black-box data recording
should be standard on all rebreathers. However, a black-box can be an expensive add-on for smaller
rebreather manufacturers. Perhaps this could be an optional component manufactured by a thirdparty.
Training
New rebreather divers have a wide range of experience, and ideally, training should reduce the risk of
injury or fatality for all. The challenges to training include diver complacency, getting divers to
behave as they were trained, and instilling discipline and proper attitude. However, it is so easy to
forget even after the best initial training. Forgetting or ignoring training might be less likely if, where
possible, the equipment could provide immediate feedback. For example, an alarm that activated if a
unit was dived without being turned on might help prevent hypoxic drowning. Another example is a
carbon dioxide sensor that would guard against channeling or diving without absorbent.
One opinion held that training by the agencies and the manufacturers' pre- and post-dive checklists
were both good. Evidence was available that some instructors had taught incorrect procedures such as
how to extend a canister beyond the manufacturers' specifications. A US instructor suggested that
boat captains and open-circuit divers might be educated about rebreather divers so they could
recognize problems and know how to respond in emergencies. The meeting was informed that British
Sub-Aqua Club (BSAC) has done this to make open-circuit divers competent buddies for rebreather
divers and to make dive marshals competent supervisors for rebreather divers. It was suggested that
BSAC might provide these training courses to US agencies, but some voiced the strong opinion that
divers should be responsible for themselves including choice of buddies.
Solo diving was particularly controversial. There was strong discussion about how often rebreather
fatalities occurred in divers without buddies. Solo diving was cited as involved in 80-95% of
rebreather diving deaths. This estimate was questioned, but an attorney reported that 80% of 40
rebreather fatalities he had investigated involved solo diving. A case was cited that recovery of a
hypoxic diver was possible because a buddy was present. However, the points were also made that
divers should be trained to look after themselves rather than be buddy-dependent and that "divers are
going to do what they want." No one argued these points, but the question remained as to how much
emphasis training agencies and diving operators should put on buddy diving.
109
Actions at a Dive Site after an Incident
In the event of a diving accident, death may not be immediate, and first aid and medical attention take
precedence over everything else. But where possible, all equipment should be preserved, no matter
how seemingly inconsequential. Everything has to be counted in until it can be positively ruled out.
Even a bathing suit proved significant in one investigation. Buoyancy management is particularly
important. Weights and the buoyancy compensator or life jacket must be inspected. Many fatality
victims were over-weighted or did not ditch their weights. Equipment should not be disassembled. If
equipment is to be shipped, it should be packed and protected in hard containers, not shipped in
cardboard boxes. Long delays in reaching a test facility may result in the loss of valuable information.
Witness names and contact information should be recorded on-site and on-site interviews conducted
when possible. (How this might be done remains to be determined.) Finding recreational divers after
they have left a dive site can be difficult and time consuming.
Every diving fatality (or potential fatality) should be treated as a homicide and the equipment
subjected to a chain of custody with custody cards signed each time it changes hands. If the chain of
custody has not been maintained properly, there is no way to be certain that damage may not have
occurred during handling after the dive or during shipping rather than during the dive. Equipment
should be locked up to prevent tampering.
Each type of rebreather is somewhat different from the others and may require special handling upon
recovery to preserve information for investigation. Military divers carry checklists for each breathing
apparatus they use. Civilian rebreather manufacturers could develop checklists specifically for their
own equipment. The US Navy, Dive Lab, Qinetiq, HSE, and Closed Circuit Research have checklists
that might be used as models. DAN volunteered to post links to the manufacturers' websites on the
DAN website so that checklists can be accessed for download through a central location.
Constructive interactions with law enforcement are essential. Police dive teams or representatives of
an official investigative agency will usually not be on-site. Ideally, dive supervisors and boat captains
would be familiar with evidence-preservation procedures and could assist on-site. Should evidencepreservation procedures specifically for diving fatalities be developed in coordination with law
enforcement agencies? Should police dive teams be trained in the special requirements for rebreather
investigations? Currently, it is not unusual for law enforcement agencies to send equipment to local
dive shops that are unfamiliar with rebreathers. Coordination between divers and police dive teams is
important to avoid situations in which police divers will not accept assistance from 'non-professionals'
who have been involved in a recovery. Written instructions or training courses that are developed
might be disseminated to the law enforcement community through umbrella organizations such as the
International Association of Dive Rescue Specialists, Inc. It was clear that law enforcement
representatives should be involved in future meetings of this nature.
Conclusions
This is a developing discussion that requires significant community cooperation if useful action is to
result.
Reference
Rooney J, Vanden Heuvel L. Root cause analysis for beginners. Quality Progress. 2004; July: 45-53.
110
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Pressure Related Incidence Rates in Scientific Diving
Michael R. Dardeau1 and Christian M. McDonald2
1
2
Dauphin Island Sea Lab, Dauphin Island, AL 36528, USA; mdardeau@disl.org
Scripps Institute of Oceanography, La Jolla, CA 92093, USA; cmcdonald@ucsd.edu
Abstract
In 1982 the US Occupational Safety and Health Administration (OSHA) examined records of
scientific divers and, based on their finding of an exemplary safety record resulting from selfregulation, partially exempted scientific diving from the strict control placed on commercial
diving activities. A retrospective examination of recent scientific diving statistics and incident
reports was undertaken to determine if the original reasoning behind the partial exemption is
still valid. Available data going back to 1998 indicated a total injury incident rate per 100
workers comparable to the earlier OSHA study.
Introduction
Training of scientists to use SCUBA as a research tool began in the early 1950s at Scripps Institute of
Oceanography and spread to many other research programs in the ensuing decades (Hanauer, 2003).
The scope of scientific diving may range from simple observations to use of sophisticated technology
(Sommers, 1988; Hicks, 1997; Sayer, 2005) but most agree that perceived hazards may include cold
and arduous dives, task loading and time at depth limitations. Given the potential for workplace
injury, in 1977, the US Occupational Safety and Health Administration (OSHA) established
mandatory occupational safety and health standards for commercial diving operations (29 CFR Part
1910, Subpart T - Commercial Diving Regulations). Because scientific diving was not specifically
exempted from these regulations, OSHA was requested by academic organizations involved in diving
to consider an exemption, based on effective self-regulation, for diving performed for research
purposes (Butler, 1996). OSHA examined the safety records of science diving programs, conducted
hearings regarding the safety of scientific diving and concluded in 1982 that, because of an exemplary
safety record, a partial exemption was warranted (Federal Register, 1982). Scientific diving has since
grown to encompass more than 100,000 dives per year. To determine if the original reasoning behind
the partial exemption is still valid, a retrospective evaluation of scientific diving statistics from 1998
to 2005 was undertaken.
Methods
The by-laws of the American Academy of Underwater Sciences (AAUS) contain a provision
requiring organizational members to submit statistical summaries of the number of dives and divers,
the mode of diving and any incidents associated with scientific diving at intervals (typically annual)
specified by the Academy. Although incident types (hyperbaric, near drowning, etc.) are defined,
incident itself is not, leaving the diving safety officer (DSO) to determine whether or not to report.
For the past eight years, these statistics have been collected in an identical manner and stored to a
database. Review of the statistics for quality assurance purposes revealed that comparisons to the
OSHA review of scientific diving safety from 1965-1981 were possible.
111
To evaluate scientific diving safety, OSHA calculated an annual incidence rate for pressure related
injuries in their Final Ruling (Federal Register, 1982) using a formula that is still used by the Bureau
of Labor Statistics (BLS) today:
(N/EH) x 200,000 = incidence rate per 100 full-time workers where-N = number of injuries and illnesses (including deaths) or lost workdays
EH = total hours worked by all employees during calendar year
200,000 = base for 100 full-time equivalent workers (working 40 hours/week, 50 weeks/year)
The AAUS database was employed to calculate an incidence rate for pressure related injuries for each
year from 1998 to 2005.
Results
Summaries of the annual diving activity by AAUS members from 1998-2005 are shown in Table 1.
Number of scientific dives increased from a low of 69,520 in 1999 to a peak of 124,722 in 2005.
Number of incidents ranged from 3 to 17, averaging 8.4 per year. Pressure related incidents ranged
from 2 to 13, averaging 6 per year.
Table 1. Summary of annual diving activity by AAUS organizational members from 1998-2005.
Organizational
Members Reporting
# of dives
# of divers
Total # of incidents (#
pressure related)
1998
1999
2000
2001
2002
2003
2004
2005
55
71,042
2,997
5
(3)
57
69,520
2,816
11
(8)
61
77,368
2,728
9
(8)
70
90,644
3,200
8
(5)
70
100,989
4,015
17
(13)
80
108,702
3,770
9
(5)
89
123,103
3,967
3
(2)
90
124,722
3,984
5
(5)
Each of the incidents reported between 1998 and 2005 was included in Figure 1. The only criterion
for inclusion was that a diver or a DSO thought it necessary to file an incident report. About half the
total of 67 incidents reported were thought to be some form of decompression illness. Sinus/ear
barotrauma involved mostly minor incidents, some not even requiring first aid. Environmental trauma
included bites, stings, shocks and injuries from contact with rigs or rocks. The other category
included no injury, anxiety, shallow water blackout, foreign object in the ear, hypothermia and other
infrequent occurrences. Neither the environmental trauma nor the other category was included in the
calculation of the incidence rate.
A survey of 88 organizations within the scientific diving community contracted by OSHA estimated a
diving population of 5,441 with a total of 39 pressure related incidents in the period between 1965
and 1981 (Federal Register, 1982; Sharkey and McAniff, 1984). Making various assumptions about
the number of divers and the number of incidents per year, OSHA calculated incidence rates ranging
from 0.04-1.66 per hundred divers per year (Table 2).
112
Other
12%
Environmental
Trauma 16%
DCI
>50 %
Sinus/Ear
Barotrauma
19%
Figure 1. Types of incidents reported, n=67
(DCI, n=35; Sinus/Ear barotrauma, n=13;
Environmental trauma, n=11; Other, n=8).
Table 2. Incidence rates calculated by OSHA for scientific diving from 1965-1981 (Federal Register, 1982).
Assumptions
Calculation
All 39 incidents in one year
All incidents attributable to an early 1970s
estimate by NOAA of a diving population of 2,340 divers in
one year
All incidents averaged over 15 years using 2,340 divers
All incidents averaged over 15 years using 5,441 divers
(39/(5441x2000))x200,000
Incidence
Rate
0.7
(39/(2340x2000))x200,000
1.66
(2.6/(2340x2000))x200,000
(2.6/(5441x2000))x200,000
0.1
0.04
OSHA also noted that these values compared favorably with BLS rates from other industries.
Incidence rates for AAUS divers and BLS statistics for 1998-2005, as well as the baseline 1979
OSHA numbers, are shown in Table 3.
Table 3. Total injury incidence rates for AAUS diving and various Bureau of Labor Statistics
industry summaries. Rates from 1979 are from (Federal Register, 1982) and industry rates
from 1998-2005 are from www.bls.gov
AAUS Diving
Construction
Transportation
Ag, Forestry
and Fishing
Wholesale
Finance
Manufacturing
Mining
Service
1979
0.04-1.66
1998
0.07
1999
0.28
2000
0.29
2001
0.16
2002
0.32
2003
0.13
2004
0.05
2005
0.13
16.2
8.7
7.0
6.2
8.4
7.0
7.0
8.2
6.7
6.8
7.8
6.6
7.0
7.1
6.1
6.4
6.8
7.8
6.2
6.4
7.3
6.4
6.3
5.2
6.1
6.3
1.7
8.5
4.7
4.9
6.0
1.6
8.0
4.1
4.6
5.8
1.6
7.8
4.6
4.6
5.4
1.5
7.0
3.9
4.4
5.3
1.7
7.2
4.0
4.6
4.7
1.7
6.8
3.3
2.5
4.5
1.6
6.6
3.8
2.4
4.5
1.7
6.3
3.6
2.4
1.7
13.3
11.4
113
Discussion
The use of scientific diving by researchers between 1998 and 2005 is higher than levels of annual
activity in the seventies estimated by OSHA. The average number of incidents reported per year is
also higher than during that period. Incidence rates among AAUS institutions between 1998 and
2005, however, remained within the 0.04-1.6 range calculated by OSHA for the late seventies. Some
of the apparent increase in both divers and incidents is, unquestionably, a function of more and better
record keeping resulting from participation in AAUS. Increased emphasis on reporting symptoms as a
result of training also probably contributed to some of the incidents reported. Many of these incidents
did not meet the OSHA criteria for inclusion (injuries and illnesses that result in days away from
work, restricted work or transfer to another job, medical treatment beyond first aid and loss of
consciousness) but were included to avoid underestimating the incidence rate.
Despite the obvious increase over time in the numbers of AAUS individuals and organizations
reporting dives, many agencies and organizations conducting scientific diving under the OSHA
partial exemption do not report to AAUS. For example, NOAA conducted 208,459 scientific dives
between 1981 and 2004 (Dinsmore and Vitch, 2005) and the Alaska Department of Fish and Game
made over 10,933 dives between 1990 and 2000 (Pritchett, 2001). Clearly, the 766,090 dives reported
to AAUS between 1998 and 2005 were not the only scientific dives being performed during that
period.
The operational definition of scientific diving adopted by OSHA implies oversight of the diving
activity. The partial exemption specifies a diving safety manual detailing operational procedures and
a diving control board to approve and monitor diving projects (Butler, 1996). The medical, training
and operational standards specified by AAUS exceed those detailed by OSHA and include record
keeping requirements, diving under the supervision of a diving safety officer, and advanced training
in rescue, CPR, first aid and oxygen administration. They distinguish scientific diving from
recreational diving and are generally considered the standard of practice for scientific diving in the
U.S. (Lang, 2003). It is clear that self-regulation and oversight of diving activity within the scientific
community has effectively maintained the incidence rate at levels considered acceptable by OSHA
when granting the partial exemption for scientific diving operations.
Acknowledgments
Thanks to all the divers and DSOs dedicated to safe scientific diving who turned in over three
quarters of a million dive records and to AAUS for allowing the use of them.
References
Butler SS. Exclusions and exemptions from OSHA's commercial diving standard. In: Lang MA,
Baldwin CC, eds. Methods and Techniques of Underwater Research. Proceedings of the American
Academy of Sciences 16th Annual Symposium, 1996: 39-44.
Dinsmore DA, Vitch ML. NOAA's diving accident management program: a review of current
capabilities and plans for improvement. In: Godfrey JM, Shumway SE, eds. Diving for Science 2005.
Proceedings of the American Academy of Sciences 24th Annual Symposium, 2005: 157-169.
Federal Register. Vol. 47, No. 228. Friday, November 26, 1982. Rules and Regulations: 1046-1050.
114
Hanauer E. Scientific diving at Scripps. Oceanog. 2003; 16(3): 88-92.
Hicks RE. The legal scope of 'scientific diving' an analysis of the OSHA exemption. In: Maney EJ,
Ellis CH Jr, eds. Diving for Science…1997. Proceedings of the American Academy of Sciences 1997
Scientific Symposium, 1997: 87-100.
Lang MA. Scientific Diving. In: Brubakk AO, Neuman TS, eds. Bennett and Elliott's Physiology and
Medicine of Diving, 5th ed. London: W.B. Saunders Co, 2003: 56-63.
Pritchett M. Scientific diving injuries in Southeast Alaska. In: Jewett SC, ed. Coldwater Diving for
Science 2001. Proceedings of the American Academy of Sciences 21st Annual Symposium, 2001: 7577.
Sayer M. The international safety record for scientific diving. SPUMS J. 2005; 35(3): 117-119.
Sharkey PIS, McAniff JJ. Scientific diving fatalities, 1970 through 1984--a fifteen year review.
Oceans '84. Proceedings, MTS-IEEE Conference, Sept. 10-12, 1984. Washington, DC, 1984:
517-520.
Sommers LH. Scientific diver training and certification. In: Lang MA, ed. Advances in Underwater
Science…88. Proceedings of the American Academy of Sciences Eighth Annual Symposium, 1988:
165-174.
115
116
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
When Everything Goes Right: Implications for Scientific Diving
Safety Programs
Vallorie Hodges
Oregon Coast Aquarium, 2820 SE Ferry Slip Road, Newport, OR 97394
val.hodges@aquarium.org
Abstract
In April 2006, a serious injury incident occurred involving a diver at the Oregon Coast
Aquarium, resulting in a near-drowning, full cardiac arrest, subsequent rescue, emergency
medical treatment, rehabilitation and ultimately a full recovery for the 43 year old female
patient. The diver was conducting a routine dive when she suddenly became unconscious at a
depth of 26 ft. No medical history or contraindications for diving were evident, no equipment
malfunctions were discovered, and the causative features of this event are not clear, however
radiographic finding were consistent with pulmonary edema. The diver spent six days on a
ventilator and several weeks making a full recovery, while the incident was investigated and
analyzed. Virtually everything went right, yet puzzles remain. While this dive was not a
scientific dive, many lessons can be learned which have implications for scientific diving
programs.
Background
On Sunday, April 23rd, 2006, a serious injury incident occurred involving one of the volunteer divers
at the Oregon Coast Aquarium, resulting in a near-drowning, full cardiac arrest, subsequent rescue
and ultimately a full recovery for the female patient. For the purposes of this story we will call the 43
year old female patient Danielle, and her husband dive buddy will be known as Shane.
Shane had been a volunteer diver at the aquarium for more than three years, and his wife had joined
him in the diving program first as a dive tender, and finally as a diver (once she had fulfilled the
prerequisite training requirements). These requirements include an advanced level certification, 25
logged cold water dives, 50 logged dives total, first aid, CPR and a specialty aquarium diving course
taught by Aquarium staff. She had been diving in the program for about six months. Both divers had
all required documentation on file, and were volunteers in good standing. Shane, a dive master
trainee, was partnered with his wife as a buddy for this dive. Volunteers use all of their own dive
equipment for these dives except for cylinders, which are provided by the aquarium. Shane and
Danielle owned top end dive equipment, which was kept in factory authorized annual repair. Danielle
also had a valid medical on file, dated 6 months prior to the dive incident. There was no history of
diving contraindications, such as medical problems or medications. Following the incident it was
related that Danielle had suffered a pneumonia-type event six to eight weeks earlier, for which she
received treatment from her local medical practitioner (she was not hospitalized), however it was
described that she had recovered from that ailment and had been diving since with no apparent
problem.
The dive location was Halibut Flats, one of three large exhibits that make up Passages of the Deep,
the largest aquatic gallery at the Aquarium. All three exhibits are 26 ft deep and each has a section of
acrylic viewing tunnel that guides visitors progressively through the three exhibits; Orford Reef,
Halibut Flats and Open Sea. The dive platform is reached via a flight of stairs that leads to the topside
117
of the exhibits, which are all housed indoors with a high arched ceiling. The entry areas for all three
exhibits are in this area. Halibut Flats features a large flat, rubble substrate area with three rocky reef
promontories and a simulated ship wreck. There are large and small halibut, skates, ling cod, and a
variety of rockfish species in this exhibit. The top of the tunnel is at about eight feet of depth, and the
bottom of the tunnel at about 16 ft, leaving approximately eight feet of clearance from the floor to the
bottom of the tunnel. The only obstruction to continuous visibility is the floor of the tunnel, however
there is also a floor window that allows some view. Divers enter the water using a controlled seated
entry technique and deep water exit, though there is also an entry/exit platform on the far side of the
exhibit. The entire exhibit is free of railings, making it possible to enter/exit the water from any point,
360 degrees around. With large open areas and a simple reef to flat bottom interface, Halibut Flats is
one of the easiest exhibits to dive.
The Incident
Pre-Dive
Danielle and Shane left their home in Corvallis at about 0700 and drove to the Oregon Coast
Aquarium, where they met the other members of the dive team, unloaded dive gear and prepared to
dive. A pre-dive shift meeting was conducted, and divers divided themselves into buddy pairs to dive.
Shane and Danielle prepped and checked their dive gear, and readied themselves to dive in Halibut
Flats. The dive plan was to do a thorough animal health check and then return to the surface for
acrylic tools to clean the tunnel. Samuel was tending for the first series of dives. He conducted the pre
dive safety checks of both divers and their equipment and logged them into the water at
approximately 0920. Both divers were wearing neoprene drysuits and high quality diving equipment
that appeared to Samuel to be functioning correctly.
The Dive
Danielle and Shane descended uneventfully. After about 15 minutes they both surfaced to get acrylic
cleaning tools and began to descend. They both returned to the surface almost immediately and
Danielle asked for additional weight to make using the acrylic tools easier. She sat on the corner grate
exit area and the tender, Samuel, added two 1.5 pound weights to the tank valve, while Danielle joked
about her weight needs. The two then descended, and Shane described that Danielle was several feet
below him as they approached the tunnel. It appeared to Shane that Danielle was going under the
tunnel and he expected her to come up the other side. He arrived at the tunnel and placed the suction
cup on the acrylic, but since Danielle hadn't arrived he stopped and looked down to see where she
was. Through the floor window in the tunnel he could see her fins in a stationary position. He
immediately dropped to the bottom where he saw her unconscious on the bottom with the regulator
out of her mouth.
The Rescue
Shane grabbed Danielle and took her to the surface, where he yelled for help. He tried to remove her
weight belt, but the buckle slid around her waist. He inflated his BC and started to swim to the side.
In the process he kicked out of one of his fins. He was able to get her to the side, where he removed
her weight belt. Another volunteer diver who was helping topside (Tony) and the dive tender Samuel
arrived and lifted her out of the water with her cylinder and buoyancy compensator still in place.
They described Danielle as blue and unresponsive. They removed her equipment and Samuel
immediately began CPR. Tony went to the radio and called for assistance, specifically requesting
security to call for an ambulance. Tony then returned to Danielle where Samuel was doing CPR and
118
began to administer breaths. They described that Danielle was attempting to breathe. This was later
determined to be agonal breaths. They experienced significant difficulty maintaining an airway due to
foam and water coming out of Danielle's mouth. Scissors were requested, and though there were two
pair in first aid kits on the diving deck, neither pair was located. Staff retrieved a pair from a nearby
office and the drysuit was cut off of Danielle. CPR was not delayed. Samuel had started CPR
immediately without removing the drysuit.
Within about 30 to 40 seconds of Tony's radio call, three staff arrived. These staff also noticed foam
and water coming out of Danielle's mouth during compressions. Samuel verbally described the foam
as pinkish in color. Security immediately called 911 and responded to guide emergency personnel
through gates and to the scene. Staff moved Danielle to a larger and better lit area, and noticed that
her color had begun to improve. Danielle was cared for by staff until the first (fire department)
paramedic arrived, approximately three to four minutes later. The fire paramedic began CPR. The
AED showed asystole. The fire paramedic administered cardiac drugs and was able to obtain a
cardiac rhythm, but no pulse. CPR was continued. When the ambulance arrived additional drugs were
administered and paramedics were able to obtain an effective rhythm and pulse. CPR was stopped at
this point. Danielle was intubated on scene. The ambulance transferred Danielle to the local
Samaritan Hospital.
Dive Safety Officer (DSO) Hodges received the call at home within about five minutes of the incident
and arrived at the scene just as the ambulance was leaving. The police had also arrived and took
photos of the dive equipment, evaluated the cylinder valve (it was all the way open) and asked to have
the gear locked up until it could be evaluated and any additional information provided to them. It was
taken to the compressor room and locked in.
The ambulance transported Danielle to the hospital. Hodges stopped briefly at the scene, picked up
Tony and drove to the hospital while Tony briefed Hodges on the event as he understood it. Hodges
met with the ER doctor and immediately requested the involvement of DAN (Divers Alert Network)
as a resource for diving emergencies. Though the ER had no information or phone numbers for DAN
they did call once Hodges provided them with the information. Later Hodges learned that the ER
doctor spoke with the on-call DAN physician and used medical protocol provided by DAN Within
about 10 minutes of Danielle's arrival, her husband/buddy Shane arrived at the hospital and provided
valuable details of the dive, including that there had not been an uncontrolled ascent or any sign of a
diving malady prior to the sudden unconscious event at depth. For that reason DAN and the ER
physician decided there was no reason to transport Danielle to a recompression chamber, and that she
should be treated for near-drowning.
Post-Dive
Danielle was placed on a ventilator, transferred to ICU and was listed in stable/guarded condition.
Hodges requested one of the volunteer diver shift captains to stay with Shane while Hodges returned
to the aquarium. A brief staff meeting was held to communicate what was known about the event. A
quorum of the DCB (Dive Control Board) was present at that meeting and the decision was made to
have Hodges (with a witness) do a cursory inspection of the dive gear for any obvious problems,
count the amount of weight and contact the police with any additional information they needed.
Hodges conducted a cursory inspection with Tony present. Hodges later contacted the police officer
assigned to the case and relayed the information that there was 38 pounds of weight total and no
obvious problem could be detected upon cursory inspection of the gear. Everything appeared to be
functioning correctly. Hodges then returned to the hospital to stay with Shane. Over the next four
days Hodges spent most of the time with Shane at the hospital. Shane involved Hodges completely in
119
the case, allowing access to Danielle in the ICU room, and involving Hodges in discussions with the
attending physician.
The incident occurred on a Sunday, and on Monday a different doctor was assigned Danielle's case.
This physician called the hyperbaric chamber in Portland, Oregon and after discussing the case with
them decided to transfer Danielle to the chamber for recompression treatment. Shane immediately
notified Hodges of this development, concerned that DAN had initially consulted with the ER
physician and recommended against this treatment. Hodges immediately called DAN for consultation
and with some difficulty Hodges and Shane were able to convince the attending physician to discuss
the case with DAN At first the physician was not convinced and still insisted on the transfer and
hyperbaric treatment, however DAN worked to get all of the parties on the phone together (including
the hyperbaric facility in Portland) and it was finally agreed to continue Danielle's current protocol
and not transfer her for chamber treatment.
Despite numerous efforts over the next several days to reduce Danielle's sedation and take her off of
the ventilator, Danielle remained on the ventilator. By Wednesday (day 4), Danielle's condition began
to worsen slightly, and Shane related to Hodges that medical staff were concerned that Danielle could
be beginning to suffer from ARDS (Adult Respiratory Distress Syndrome). Newport's hospital is a
smaller community facility and does not have many of the resources that larger trauma hospitals have.
Shane began to attempt to negotiate a transfer of Danielle to the Good Samaritan Hospital in
Corvallis, Oregon where pulmonary specialists and additional medical technology was available to
treat her. An emergency transfer was made on Thursday (day 5), and by Friday evening (day 6)
doctors were able to remove Danielle from the ventilator and reduce her sedation enough that she
became conscious. She clearly recognized family members and was responsive. By Monday (day 9)
she was taken out of ICU and moved to the general hospital ward for recovery. By Wednesday, May
3rd (day 11) she was transferred to a nursing facility for continued recovery including physical and
speech therapy. She experienced some balance and speech impairment early in her recovery, however
those issues resolved with time. During this time, Hodges visited Danielle and found her to be tired
but improving rapidly, in good humor and having no memory of the event and only scattered memory
of the hour or so prior to the dive. Over the next several weeks Danielle continued to improve, until
she was able to return home, return to work and has been described as completely recovered.
The Investigation
Oversight of the Incident by the Dive Control Board
The Dive Control Board discussed the events and evaluated them based on the dive program policies
and procedures as provided by the Oregon Coast Aquarium Dive Manual. No violations to these
policies or procedures could be found. In fact, it was seen by the board that staff and volunteers
reacted in an exemplary manner, from the buddy system, the rescue, the egress to the administration
of CPR, caring for the injured diver, contacting security, calling 911, notifying the Dive Safety
Officer, and controlling the scene. Access to the facility for emergency personnel was handled
exceptionally well by security staff, and personnel were stationed at the Passages stairs to direct them
to the patient. All elements of the accident management plan appeared to have worked very well.
Hodges provided an overview to the DCB of possible causes of the unconscious event underwater, as
suggested by DAN, including cardiac disrythmia, stroke, a tight drysuit neck seal, pulmonary edema,
or accidental aspiration of water that might be caused by sudden choking, panic, or loss of regulator
(M. McCafferty, personal communication, April 24, 2006). Deriving the cause was difficult because
there were no witnesses to the moment of the unconscious event. There was a current dive medical on
120
file for the diver, signed by a physician October, 2005. The diver had not notified the aquarium of any
serious medical injuries, illnesses or admission to hospital since that medical was conducted. During
the event, the diver's husband did relate to Hodges that the diver had a bout with pneumonia 6-8
weeks earlier, had received medication for it and had dived since that illness resolved. She was not
hospitalized for that illness. The Aquarium's Dive Safety Manual required divers that have a major
illness or injury/hospitalization to advise the Dive Safety Officer and to obtain a new release for
diving from a physician. Danielle's pneumonia prior to the incident did not appear to meet those
criteria, so no policy was violated in this case.
The DCB took the following information gathering actions:
1. The regulator used by the diver was sent to NOAA (National Oceanic and Atmospheric
Administration) to be tested and inspected. The regulator was found to be operating properly.
2. Hodges obtained the software and cable to download the data in the diver's computer. The
diver was using a Cobra dive computer that recorded time, depth, rates of ascent, and air
consumption data throughout the dive. Hodges also obtained the buddy's dive computer data
(also a Cobra). This information was helpful to establish a timeline for the dive, the profile
and rate of ascent, and information regarding air consumption. This data appears to indicate
the diver was rescued almost immediately and that there was no increase in the injured diver's
air consumption toward the end of the dive, which might be found in a panic event. The
computer recorded an in-water temperature of 52°F, 2,120 psi starting cylinder pressure and
960 psi ending cylinder pressure.
3. Hodges provided both data sets to DAN for further analysis, however no further insight was
provided by the data (PJ Denoble, personal communication, May 03, 2006)
4. The remaining gear used by the diver was held in a lock up storage until it was determined it
was no longer needed. It was eventually returned to the diver.
5. The air in the dive cylinder used by the diver was tested (following appropriate chain of
custody and sample analysis procedures required by Dr. Ed Golla of TRI-Environmental Inc.)
The air was found to be within standards. The dive program air fill station is tested every 6
months, and had just been tested the previous month.
6. Hodges reviewed the First Aid kits with divers, including the location of scissors.
7. An incident review/debriefing was held at the Aquarium on Thursday, April 27 2006 and all
staff and volunteers, plus responding emergency personnel were invited. Several of these
emergency personnel did attend and provided valuable feedback and answers to questions
brought forward by staff and volunteers.
8. The Fire Department also offered to hold a critical incident debriefing session Thursday
evening at 1800 for those staff and volunteers immediately involved in the event. A number
of those involved did participate in that session. It was recommended the Dive Safety Officer
not attend, in order to provide a non-supervisory atmosphere. Feedback from those attending
indicated this was a very beneficial session.
9. Hodges met with each dive team to provide a verbal overview of the event, sharing the details
appropriate to dive safety (while not sharing medical information) and emphasizing the
policies and procedures critical to diver safety – including the buddy system and effective
rescues, plus reinforcing the location of scissors and other first aid equipment.
10. Hodges spoke to Bob Hicks, General Counsel for AAUS regarding the incident. Hodges and
Hicks discussed OSHA and the extent of jurisdiction in this event with a volunteer. Hicks
related that determining whether the diver in this case should be considered a volunteer (not
reportable to OSHA) or an unpaid employee (and therefore reportable to OSHA) could be
evaluated on the basis of considering several criteria. Hicks draws his information from a
recent Supreme Court ruling that said 'In determining whether a hired party is an
employee…, we consider the hiring party's right to control the manner and means by which
the product is accomplished' (US Supreme Court). Hicks related twelve factors relevant to the
121
definition of employee which may be used to assess the employment status of the individual,
including: the skill required, the source of instruments and tools, the location of the work, the
duration of the relationship, whether hiring party can assign additional projects to hired party,
extent of discretion over when and how long to work, the method of payment, who hires and
pays assistants, whether work is regular business of hiring party, whether hiring party is in
business, the provision of employee benefits, and the tax treatment of the hired party.
Hicks suggested that the aquarium seek the advice of local counsel and that if OSHA made
contact with the aquarium that our legal staff and executive personnel should be prepared to
describe this position effectively. In balance, the Aquarium was confident this diver should be
treated as a volunteer and not as an employee.
Hicks additionally recommended that the Aquarium develop a written incident response
detailing the procedure for incident response management. This incident was managed very
well and should serve as a model. The plan should include specific details of who is in charge
(Hicks recommended Hodges should be for diving incidents), who should be the
spokesperson for the facility, what information should be released and when, and what
information should remain confidential.
Hicks also applauded the intensive care/involvement and ongoing maintenance of the
Aquarium's relationship with the patient and the husband and suggested that contact be
maintained in the months to come. (R. Hicks, personal communication, May 3, 2006 and
March 9, 2007).
A Medical Explanation
After her recovery, Danielle sought to return to diving as a volunteer at the Aquarium. Hodges
explained that Danielle would need to obtain a new medical release for diving and recommended that
she seek the professional advice of a physician with expertise in hyperbaric medicine.
Danielle's case was handled by Dr. Gordon Anderson, MD of Oregon Medical Group in Eugene,
Oregon. Dr. Anderson is a UHMS Dive Medical Examiner and hyperbaric consultant. He examined
Danielle, reviewed copies of her medical records and chest x-rays taken immediately after the
incident and reviewed an extract of the DCB report of the incident (provided to him with Danielle's
written permission). Dr Anderson then consulted with Dr. Cianci MD and discussed the case. Dr.
Cianci has studied pulmonary edema in divers and was a co-author on a leading case report and
review, Pulmonary Edema Associated with Scuba Diving (Slade et al., 2001).
It is Dr. Anderson's opinion that Danielle suffered from idiopathic pulmonary edema which caused
her to become unconscious (G. Anderson, personal communication, July 7, 2006). The Physiology
and Medicine of Diving (Bennett and Elliot, 1993) describes immersion pulmonary edema as
'probably cardiogenic, caused by a combination of factors, including increased cardiac afterload due
to water immersion, high inspiratory breathing resistance and exaggerated cardiac aferload due to
cold-induced peripheral vasoconstriction.' Dr. Anderson explained that in some individuals, the
pressure of immersion causes 'increased preloading', or increased right atrial pressure and increased
pulmonary artery pressure which can result in disruption of the membrane of the lung. He noted that
it can typically be identified using radiological and physical findings, including listening to the lungs
for symptomatic 'crackles' and is often accompanied by pink frothy sputum. Witnesses did observe
pink frothy sputum, and Dr. Anderson reviewed Danielle's chest x-ray and noted that it showed a
diffuse interstitial alveolar pattern, consistent with pulmonary edema. Bennett and Elliott (1993) also
list symptoms of dyspnea, cough, haemoptysis or expectoration of frothy sputum, and describe that
122
these symptoms 'often occur during descent or while at the deepest depth' as contrasted with
cardiorespiratory decompression illness, which typically presents symptoms late in ascent or post
dive.
Dr. Anderson stated that the etiology of immersion pulmonary edema is still not known, and has
created something of a puzzle for diving physicians. He went on to say that currently there are no
tests to predict this malady, but that some apparently healthy people get pulmonary edema while
swimming, snorkeling or diving and we don't know why. The only preventive measure is to avoid
swimming, snorkeling or diving once pulmonary edema has presented.
Dr. Anderson advised Danielle that she should not dive again and would not provide her with a
medical clearance to dive. The patient and her husband agree that this is a good idea.
Pulmonary edema during scuba diving and swimming was first reported by Wilmhurst et al. (1981),
who reported episodes in 11 individuals. Since that time, pulmonary edema has been associated with
cold-water immersion and there have been further reports and reviews of cases, however the
mechanisms responsible are still unclear. Pons et al. (1995) looked at the prevalence of pulmonary
edema in healthy persons during scuba diving and swimming and concluded that it is extremely rare
in healthy individuals. In a case that presented to the Auckland, New Zealand hospital, Grindlay and
Mitchell (1999) discussed the highly effective treatment of continuous positive airway pressure
ventilation and provided guidelines for its use in diving related accidents. They also discussed the
occurrence of pulmonary edema as an isolated phenomenon (rather than a presenting feature of DCI
and barotrama) and listed the potential causes as 'salt water aspiration syndrome or near drowning;
and cold, immersion and exercise-induced pulmonary oedema.' More recently, Slate et al. (2001)
reported on eight scuba divers who developed acute pulmonary edema, all occurring in cold water.
This report described pulmonary edema in scuba divers as multifactorial and highlighted the
importance for physicians to be aware of 'this potential, likely underreported, problem in scuba
divers'.
Discussion
The impact of a serious or fatal incident on a dive program can be devastating, from a human
perspective as well as financially, legally, and operationally. In this incident, virtually everything that
could go wrong went right; from having appropriate training, documentation and protocols in place,
to all of the protocol being followed to the letter; from the proximity of a paramedic capable of
providing life saving drugs, to the availability and responsiveness of the DAN. From a buddy who
kept his cool, to a husband who was open, inclusive and never rattled, blaming or judgmental. In fact,
the only issues that appeared were problems associated with medical insurance caps, which were only
discovered post incident. These proved to not be a major issue in this case because the injured diver
held major medical and DAN diving accident insurance.
To discuss the implications of a diving incident, one must look at what went right, and what went
wrong, or what might have gone better. A primary concern and desire is the repeatability of what
went right. By evaluating a list of what went right, we were able to prioritize and reinforce our
training and procedures with the goal of repeating these positive outcomes, should it ever become
necessary. The list of what went wrong also contributed to focusing training efforts and identifying
program needs such as improved insurance coverage.
123
What went right:
• The diver met prerequisites for inclusion in the Aquarium dive program
• The diver completed all required training (per Aquarium policies)
• The Aquarium had all required documentation on file for the diver
• The diver used high quality equipment that was maintained within manufacturer
specifications, required annual maintenance and the Aquarium had proof of this on file.
• The Aquarium conducted all required air fill station tests and had documentation posted
• The divers conducted a pre-dive briefing
• The divers assembled their gear correctly
• The diver's equipment was checked prior to the dive (pre dive safety check) by the divers and
the tender
• The divers practiced proper buddy procedures
• The buddy responded promptly when he lost sight of his buddy
• The buddy conducted an effective rescue
• The buddy immediately called for topside assistance
• The tender and another diver responded immediately and appropriately to render aid and
remove the injured diver.
• One topside responder immediately began CPR while one went to call for help.
• The responder called correctly on the radio for help
• Security immediately called 911 without delay and responded to open gates and guide
emergency personnel.
• The responders provided adequate CPR prior to removing the drysuit.
• The injured diver's color began to improve.
• A paramedic less than one mile away heard the call and responded.
• The responding paramedic was capable of administering cardiac drugs to restore a cardiac
rhythym.
• The Dive Safety Officer was promptly contacted.
• The DSO was able to provide the emergency room doctor with contact information for DAN.
• DAN was able to provide medical protocol for near drowning and help evaluate the nature of
the diver's injury-recommending no hyperbaric treatment.
• DAN was able to liaise with subsequent physicians and prevent the transfer and hyperbaric
treatment of the injured diver.
What could have gone better:
• The rescuing diver lost a fin.
• The rescuing diver would ideally have taken the diver to the corner platform entry area for
easier egress, but the loss of a fin made that a poor option.
• The rescuing diver had difficulty removing the weight belt on first attempt.
• The topside responders might have removed gear before removing diver from the water to
minimize risk of injury to themselves.
• The topside responders were initially unsure of the injured diver's airway, mistaking agonal
breaths for signs of breathing.
• When responders attempted to provide rescue breaths they experienced significant difficulty
with the airway due to water and foam.
• The responders had to move the injured diver to a larger work area with better lighting.
• The responders and other staff did not find the two pair of scissors at the dive location and
had to retrieve a pair from another location.
• After the event it was discovered that medical insurance coverage was capped at a maximum
of $10,000 (the injured diver fortunately held major medical insurance and DAN insurance).
124
•
•
The lack of locally available diving medical expertise and advanced medical equipment to aid
in treating near drowning were problematic in this case, but this is not unusual. It is likely a
problem for many rural/remote dive operations around the globe. This case was further
complicated by the hospital's lack of knowledge of DAN and confidence in DAN's medical
expertise. The action of the husband, the DSO and the staff at DAN to create a dialogue
between the parties was important and provides a lesson for Dive Officers everywhere.
The impact upon personnel that are involved in these types of events should not be under
appreciated. A policy for dealing with the stress of such events can be valuable, such as
providing a debriefing and proactively providing access to counseling and professional care.
A stress incident debriefing was provided in this case, but could have been done earlier. Some
personnel that were involved in this event described feelings of isolation and anxiety, as well
as unpleasant visualizations and other stress related effects.
In summary, virtually everything in the management of this incident went right, demonstrating the
value of training and having emergency protocols in place. It also highlights some potential pitfalls
for dive programs, including the need to consider the resources and expertise of local medical centers,
knowledge of DAN and how they can help, medical insurance coverage limitations, and the need to
be prepared to handle stress responses. While this event did not result in any problems with the
media, it should be mentioned that most dive programs should be prepared to handle media responses.
The role of the DSO and the DCB should be carefully considered. Many DCB members have little
direct experience handling diving accidents and investigations and errors can have grave implications.
One of the tasks of the DSO and DCB is to juggle the priorities of all of the parties involved,
including the injured diver and their family, other dive team members, staff and personnel involved,
and the organization itself (risk management). An often overlooked area is that stress and trauma are
experienced by those involved in managing, not only the incident, but the subsequent events. The
organization should be ready and have a plan in place to address such issues.
The Puzzle Revealed
This incident provided more than a question of what caused the diver to become unconscious. A
salient point for Dive Safety Officers and all dive buddies is the notion that sudden unexpected
medical events can and do occur during dives. Another question worth looking at is whether the
relationship of the two divers (husband/dive master and wife/less experienced diver) played a
significant role in the effectiveness of the buddy system. Would other divers have responded so
quickly to a buddy that they had lost visual contact with?
It has been our experience that many of our divers become very comfortable with the diving
environment and their diving buddies. All divers are aware that the other divers in the program have
received proper training, medical screening and are generally moderately to highly qualified divers.
They come to trust their team members to perform to a certain standard. What could go wrong? So
over time, these divers become complacent about pre-dive checks, the buddy system and proficiency
in rescue/egress techniques. Dive Officers must be prepared to monitor, correct and reinforce these
behaviors.
Dress for the Crash, Not For the Ride
There are some questions we will never have answers for. It is not necessary to have all of the pieces
to get a pretty clear picture of what we need to know. More important is what we do with the
information that we do have. It is clear that the real puzzle here is how to continue to assure critical
safety behaviors among divers over time.
125
References
Bennett PB, Elliott DH, eds. The Physiology and Medicine of Diving. London, UK: WB Saunders
Co, 1993.
Grindlay J, Mitchell SJ. Isolated pulmonary oedema associated with SCUBA diving. Emerg Med.
1999; 11: 272-276.
Pons M, Blickenstorfer D, Oechslin E, Hold G, Greminger P, Franzeck UK, Russi EW. Pulmonary
oedema in healthy persons during scuba-diving and swimming. Eur Respir J. 1995; 8(5): 762-767.
Slade J, Hattori T, Ray C, Bove A, Cianci P. Pulmonary edema associated with scuba diving: case
reports and review. Chest. 2001; 120(5): 1686-1694.
Wilmhurst PT, Nuri M, Crowther A, Betts JC, Webb-Peploe MM. Forearm vascular responses in
subjects who develop recurrent pulmonary edema when scuba-diving; a new syndrome. Br Heart J.
1981; 45: 349.
126
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Behavior and Sound Production by Longspine Squirrelfish Holocentrus
rufus During Playback of Predator and Conspecific Sounds
Joseph J. Luczkovich 1,2 and Mark Keusenkothen3
1
Department of Biology, East Carolina University, Greenville, NC 27858, USA
Institute for Coastal and Marine Resources, East Carolina University, Greenville, NC 27858, USA
3
Coastal Resources Management Doctoral Program, East Carolina University, Greenville, NC 27858,
USA
luczkovichJ@ecu.edu; keusenkothenm@ecu.edu
2
Abstract
Fishes and marine mammals make sounds and listen for predators and conspecifics, i.e., they
communicate underwater using sound. Longspine squirrelfish Holocentrus rufus are nocturnal
reef fishes living in the Caribbean that commonly produce low-frequency sounds at dawn and
dusk. In order to determine the reactions of longspine squirrelfish to sounds made by their
conspecifics and by their potential predators, we performed experiments in which we played the
grunting sounds of longspine squirrelfish and the echolocation and signature whistle sounds of
bottlenose dolphin Tursiops truncatus through an underwater speaker on the reef at the Institute
for Marine Studies, Calabash Caye, Turneffe Atoll, Belize. At the surface, a portable laptop
computer was programmed to playback a series of sounds in the following sequence: preplayback period with no sounds (8 min in experiment 1, 5 min in experiment 2), 700-Hz tone
(10 min), longspine squirrelfish grunts (10 min), bottlenose dolphin echolocations (2 min),
bottlenose dolphin signature whistle (2 min), and a post-playback without sounds (1 min in
experiment 1, 2 min in experiment 2). Both the squirrelfish and dolphin sounds were recorded in
the coral reef areas surrounding the study site. We also played a 700 Hz tone as a control. We
monitored the sound production from free-ranging longspine squirrelfish in the area near the
speakers where we played the sounds, and recorded behavioral responses of the fish using a
digital video camera in an underwater housing with integrated hydrophone. To minimize
acoustic disturbance, we used closed-circuit rebreathers (Inspiration) during the experimental
playbacks. We compared the amount of time the squirrelfish stayed in view, how long the fish
performed visual displays (fin erections), and how often they vocalized during each of the
playback treatments. We found that the amount of time longspine squirrelfish remained in view
did not significantly differ among treatments. Likewise, the duration of visual displays did not
significantly differ among treatments. However, the fish appeared to perform fewer
vocalizations during the playback of bottlenose dolphin sounds relative to vocalizations made
during other playbacks. Longspine squirrelfish may be listening for the hunting sounds made by
predators and responding to those sounds by performing fewer vocalizations.
Introduction
Fish use sound to communicate in a variety of ecological interactions such as courtship and mating,
aggressive encounters, and to signal alarm (e.g., Winn et al., 1964; Myrberg, 1981; Crawford et al.,
1997; Mckibben and Bass, 1998). An example of such a sound-producing fish is the longspine
squirrelfish Holocentrus rufus, which inhabits shallow coral reefs throughout the Caribbean Sea.
They produce low frequency vocalizations at a dominant frequency of 75-600 Hz (Moulton, 1958;
Fish and Mowbray, 1970; Carlson and Bass, 2000). Vocalizations are produced by the contraction of
muscles that vibrate ribs surrounding the swim bladder (Winn and Marshall, 1963). These lowfrequency sounds are produced during both day and night, but appear to increase after sunset
127
(Luczkovich, 2002) with peaks at dawn and dusk related to territorial behavior; during the day the
fish are inactive hiding in reef crevices and at night they venture away from their territories in search
of food (Winn et al., 1964). During the transition between day and night, they interact vocally at their
territory boundaries.
Longspine squirrelfish behavior
The following summary of behavior and sound production of longspine squirrelfish is based on the
description by Winn et al. (1964), except where noted. Longspine squirrelfish typically produce either
single grunts or a series of grunts (staccatos). The fish are territorial and often maintain territories
adjacent to other longspine squirrelfish territories. During the day, squirrelfish hover near reef
crevices and at night, they are more active, but do not venture far from daytime territory (Collette and
Talbot, 1972). Longspine squirrelfish display a variety of behaviors and vocalizations when their
territorial boundaries are violated. Single grunts are often issued by resident fish when territories are
violated by conspecifics. The resident fish may also erect its fins and dash at the encroaching fish.
Should the resident and intruder meet, both individuals may shudder so that the whole body vibrates.
Sometimes only the resident fish shudders. In the case of an abrupt meeting, the resident fish might
infrequently initiate a staccato call. Longspine squirrelfish may also initate shudders, grunts, or
staccatos when fishes of other species, such as the bluestriped grunt, violate territorial boundaries.
Interestingly, Winn et al. (1964) found that longspine squirrelfish sounded longer staccatos when
moray eels were introduced into their territories relative to other non-predatory fish such as mullets
and grunts.
Marine Mammals and Predator-Prey interactions
Marine mammals also use sound in predator-prey interactions. For instance, bottlenose dolphins,
change their direction of travel and orient toward playbacks of sound-producing (soniferous) fishes,
suggesting that they listen passively for low-frequency prey sounds (Barros and Odell, 1990; Barros
and Wells, 1998; Gannon and Waples, 2004; Gannon et al., 2005). Bottlenose dolphins also produce
high-frequency sound (active sonar) for echolocation (Au, 1993). Bottlenose dolphins may also
employ low frequency 'pops' when locating prey (Remage-Heley et al., 2006). Recent studies have
suggested that some species of fish are capable of hearing and reacting to these sounds used by
dolphins while hunting. Mann et al. (1998) conducted an experiment in which bottlenose dolphin
echolocation clicks were played to captive shad Alosa sapidissima. The researchers used classical
conditioning of heart rate and muscle activity to determine whether or not their experimental subjects
could detect echolocation clicks. Their results suggest that shad can detect clicks with a peak
frequency of 80 kHz. Mann et al. (1998) propose that shad evolved this ability to hear ultrasonic
sounds in response to selective pressure brought about from predation by marine mammals. RemageHealey et al. (2006) played bottlenose dolphin sounds to gulf toadfish Opsanus beta in their natural
environment. The males of the species use vocalizations during courtship. The researchers found that
these vocalizations were suppressed by as much as 50% when low frequency dolphin 'pops' were
played within the hearing range of the toadfish. Additionally, the scientists took blood samples from
the toadfish immediately following playback of the dolphin sounds. These samples were analyzed and
found to have significantly elevated levels of the stress hormone cortisol. Finally, Luczkovich et al.
(2000) recorded chorusing silver perch Bairdiella chrysoura in North Carolina waters, and found that,
when bottlenose dolphin whistles also occurred, the intensity of perch chorusing dropped nearly eight
fold. The perch resumed chorusing following the cessation of dolphin whistling. This significant drop
in chorusing intensity was also observed when the researchers performed experiments involving the
playback of recordings of dolphin whistles (3500-6000 Hz) to naturally occurring populations of
chorusing silver perch (Luczkovich et al., 2000). We believe that bottlenose dolphins are predators of
the longspine squirrelfish, based on the fact that they normally consume fishes, and elsewhere it has
been shown that up to 75% of the diet of this marine mammal is composed of soniferous fish species
(Barros and Odell, 1990; Barros and Wells, 1998; Gannon and Waples, 2004).
128
Study Objectives
The purpose of our study was to determine the reactions of longspine squirrelfish to sounds of their
conspecifics and their potential predators (bottlenose dolphin) on a coral reef in Belize. We observed
the behavior of these fishes during the crepuscular period in a natural setting in order to determine
what types of sounds they would respond to with naturally occurring background sound levels
(sounds of other fishes, snapping shrimp, waves, etc.), but without any diver's bubbles and associated
sounds. We examined the responses of free-ranging longspine squirrelfish to a variety of acoustic
stimuli including vocalizations made by conspecifics and sounds produced by bottlenose dolphins
recorded in nearby locations. We determined if these sounds would alter the amount of time
longspine squirrelfish stayed out in the open, the duration of visual displays such as fin erection, and
vocalization frequency of occurrence. We hypothesized that the amount of time in view, the display
time, and the vocalization rate would be higher during the playback of the longspine squirrelfish
vocalizations, and lower during the playback of the bottlenose dolphin sounds relative to the controls.
Methods
Playback Studies
Playback studies were conducted using an underwater speaker mounted on the bottom in 8 m of water
on the Calabash Caye at dusk (1700-1900 local time in June 2005) (Figure 1). Sounds were played
through a Clark AC339 underwater speaker. A portable car audio amplifier (Sony XPlode 200 watt)
was used to drive the underwater speaker. Sounds were played back on a Panasonic Toughbook CF29 using a pre-programmed Windows Media play list. Longspine squirrelfish were videotaped by
divers using Inspiration rebreathers to allow the recording of fish sounds while minimizing the
bubbles and sounds associated with SCUBA diving. Fish were videotaped and sound recorded using a
Light and Motion Mako housing and Sony PC100 camera. The housing had an integrated
hydrophone. Behavior was monitored during a series of playback treatments: 1) pre-playback period
with no sounds (8 min in experiment 1, 5 min in experiment 2); 2) pure 700-Hz tone (10 min) as a
positive control; 3) longspine squirrelfish grunts (10 minutes); 4) bottlenose dolphin echolocation (2
min); 5) signature whistle (2 min); and 6) a post-playback period with no sounds. Sounds were played
back in sequence as noted above in both experimental trials; no attempt was made to randomize
playback sequence because this was a preliminary study. Both the longspine squirrelfish and
bottlenose dolphin sounds were previously recorded in the coral reef areas on Turneffe Atoll and
stored in wav files (16 bit, 44.1 kHz).
Video Analysis
Tapes were reviewed and scored for the following responses. Dorsal fin erections (displays) were
counted and timed during each of the various playback treatments. Also, the amount of time
individual fish remained within view during each playback treatment was determined. Finally, we
used Ulead Video Studio (8.0) software (www.ulead.com) to make wave files of the video
soundtracks. These files were then made into spectrograms using Raven Lite 1.0 Bioacoustics
Software (Cornell Lab of Ornithology, 159 Sapsucker Woods Rd, Ithaca, NY, 14850) so that sounds
could be visualized and counted, and the frequency of occurrence of longspine squirrelfish
vocalizations was determined for each playback treatment. It should be noted that low-light
conditions during the dolphin playback made the analysis of longspine squirrelfish displays and time
in view more difficult and potentially less accurate. Also, background noise in the dolphin whistle
playback prevented analysis of longspine squirrelfish vocalizations.
129
Calabash Caye
IMS
Turneffe Atoll
Figure 1. Landsat image of the coastal area of Belize, with inset of the experimental area for playback
studies, near University of Belize Institute of Marine Studies (IMS), Calabash Caye, Turneffe Atoll, Belize.
Statistical Analysis
Although these were preliminary studies, we attempted a statistical test in spite of the low power to
reject our hypotheses. The small sample size reported here (n=2) necessitated a non-parametric
testing procedure and we report probability values for informational value, as it would be difficult in
any circumstance to obtain a significant test at such low statistical power. We used a nonparametric
Kruskal-Wallis (KW) test for comparison of median response variables (Systat Version 11.0, Systat
Software, San Jose, CA, USA). We feel that plots of the response variables under variable treatments
are very informative for our preliminary experiments. We plotted response variables with a box plot
(also in Systat 11.0). Examination of the box plots allows the reader a reliable and quick assessment
of the fish's responses for patterns that may be replicated under greater statistical power.
Results
Longspine squirrelfish did not appear to spend less time in view during the conspecific and dolphin
sound playbacks, and they did not exhibit different durations of visual display. The amount of time
individual fish remained in view during each playback treatment did not vary in a statistically
significant way (KW test p=0.898, Figure 2). Likewise, no statistically significant differences were
found among treatments in display duration (KW test p=0.610, Figure 3). No clear patterns were
observable among median values shown in the box plots, although the range of observed values
declined greatly during the longspine squirrelfish and dolphin sound playbacks and did not return
during the post-playback period.
130
Figure 2. Box plot of the time in view
(s) of individual longspine squirrelfish
(LSSF) during various playback
treatments. which shows the median as
a horizontal line, a box showing the
range in which 50% of the values fall
(the mid-range, defined by the first and
third quartile range boundaries), a
vertical line above and below the box
which shows the range of the observed
values within a range equal to 1.5* the
mid-range above and below the first and
third quartile range boundaries. Values
falling outside this range are indicated
with asterisks (<3* the mid-range above
or below the box limit) or circles (>3*
mid-range above or below the box
limit) and are considered far outside
values.
400
200
100
0
l
tro
n
co
e
ton
in
ck
SF
lph
ba
LS
y
o
d
la
stp
o
p
Playback Treatment
40
Time Displaying (s)
Time in View (s)
300
30
20
10
0
n
co
tro
l
ton
e
k
SF
hin
ac
p
l
b
LS
ay
do
tpl
s
po
Playback Treatment
Figure 3. Box plot of the time spent displaying by longspine
squirrelfish (LSSF) during various playback treatments. See Figure 2
for explanation of box plot symbols.
131
Vocalizations did not appear to vary under the pure tone and longspine squirrelfish playback
treatments; however vocalizations became less frequent during dolphin echolocation playback (Figure
4). Experimental Trial 1 had high levels of natural grunting by longspine squirrelfish and there were
several longspine squirrelfish in the experimental area. This resulted in a high variation in the rate of
grunting between experimental trials. Nonetheless, vocalizations that had been recorded at median
rate of 3.15 and 4.25 per minute during the control and 700-Hz tone periods dropped to 2.45 per
minute during the longspine squirrelfish playbacks and 0.25 per minute during the dolphin
echolocation playbacks. In addition, the range of values decreased greatly during these two playback
treatments. Although no vocalizations occurred during the dolphin echolocation playback treatment in
Experimental Trial 1, a single 'staccato' sound was recorded during that treatment in Experimental
Trial 2. During the post-playback period, the vocalization rates increased again to a median of 3.15
per minute. This suggests that the vocalization rate was not declining as a simple function of time, as
might be expected if it were declining as ambient light levels declined at dusk. Indeed, the opposite
pattern would be expected to occur, as vocalizations should have increased as darkness increased. It
appears that the dolphin echolocation sounds decreased the vocalization rate of longspine squirrelfish.
No.of Grunts/min
10
5
0
-5
e
o
ts
ck
nd
ch
on
un
ba
e
ou
r
T
y
S
G
z
in
la
No
SF
lph
0H
stp
o
0
S
o
7
L
D
P
TREATMENT
Figure 4. Box plot of the number of 'grunts' per minute by
longspine squirrelfish (LSSF) recorded during each of the
playback treatments. Although there was a non-significant
statistical result, (KW test p=0.320), the statistical power was low
(n=2) and there was a clear pattern of reduction in sound
production during the dolphin playbacks. See Figure 2 for box
plot explanation.
132
Discussion
Our hypothesis that the amount of time in view, the display time, and the vocalization are higher
during the playback of the longspine squirrelfish vocalizations was not supported by our preliminary
observations. However, vocalization rate was lower during the playback of the bottlenose dolphin
echolocation sounds relative to the controls.
The apparent suppression of longspine squirrelfish vocalizations during playback of bottlenose
dolphin sounds is similar to the observed behavior of other species of fish such as Gulf toadfish
(Gannon et al., 2006) and silver perch (Luczkovich et al., 2001). Both of these species suppressed
vocalizations in apparent response to sounds made by bottlenose dolphin. Similarly, Mann et al.
(1999) observed that shad seemed to be able to detect dolphin echolocation.
One squirrelfish staccato call occurred during one of the two dolphin echolocation playbacks. Winn et
al. (1964) surmised that longspine squirrelfish use staccato calls as signals of alarm, and it is possible
that our recorded stacatto was an alarm call to other nearby fishes. We noted another possible
Holocentrid alarm call in an on-line video recording of cooperative hunting by grouper and moray
eels in the Red Sea (Bshary et al., 2006; Luczkovich and Keusenkothen, 2006).
The behavior we noted during playback differs somewhat from the results of Winn et al. (1964).
Winn found that longspine squirrelfish, in an aquarium study, initially sought cover during playback
of three types of fish sounds (longspine squirrelfish staccato, toadfish Opsanus tau, and big eye scad,
Selar crumenophthalmus). Some fish emerged from cover after a minute or so. Our experimental
subjects did not appear to seek cover during the various treatments. However, the Winn et al. (1964)
experiments were performed in tanks, and it is possible that the behavior of the fish was influenced by
their close proximity to the underwater sound source in an enclosed space.
Thus, it appears that fish inhabiting coral reefs are acoustically aware of their potential predators in
the coral reef environment. We conclude that there is sufficient indication of 'acoustic avoidance'
behavior in squirrelfish to dolphin echolocation calls that further verification in a controlled replicated
study with randomized playback treatments is warranted. We hope to perform these experiments in
the near future.
References
Au WWL. The Sonar of Dolphins. New York, NY: Springer, 1993.
Barros NB, Odell DK. Food habits of bottlenose dolphins in the Southeastern United States. In:
Leatherwood S, Reeves RR, eds. The Bottlenose Dolphin. San Diego, CA: Academic Press, 1990:
309-328.
Barros NB, Wells RS. Prey and feeding patterns of resident bottlenose dolphins (Tursiops truncatus)
in Sarasota Bay, Florida. J Mammal. 1998; 79: 1045-1059.
Bshary R, Hohner A, Ait-el-Djoudi K, Fricke H. Interspecific communicative and coordinated
hunting between groupers and giant Moray Eels in the Red Sea. PLoS Biol. 2006; 4(12): e431.
Collette BB, Talbot FH. Activity patterns of coral reef fishes with emphasis on nocturnal-diurnal
changeover. In: T. J. Bright and S. A. Earle (Eds.), Results of the Tektite Program: Ecology of Coral
Reef Fishes. Bulletin of the Natural History Museum of Los Angeles County, 1972; 14: 98-111.
133
Carlson BA, Bass AH. Sonic/Vocal motor pathways in squirrelfish (Teleostei, Holocentridae). Brain
Behav Evolut. 2000; 56: 14-28.
Crawford JD, Cook AP, Heberlein AS. Bioacoustic behavior of African Fishes (Mormyridae):
potential cues for species and individual recognition in Pollimyrus. J Acoust Soc America. 1997; 102:
1200-1212.
Fish MP, Mowbray WH. Sounds of Western North Atlantic Fishes. Johns Hopkins Press, 1970.
Gannon DP, Waples DM. Diets of coastal bottlenose dolphins from the U.S. mid-Atlantic coast differ
by habitat. Mar Mammal Sci. 2004; 20(3): 527-545.
Gannon DP, Barros NP, Nowacek DP, Read AJ, Waples DM, Wells RS. Prey location by bottlenose
dolphins, Tursiops truncates: an experimental test of the passive listening hypothesis. Animal Behav.
2005, 69: 709-720.
Luczkovich JJ, Daniel HJ III, Hutchinson M, Jenkins T, Johnson SE, Pullinger RC, Sprague MW.
Sounds of sex and death in the sea: bottlenose dolphin whistles suppress mating choruses of silver
perch. Bioacoustics. 2000; 10: 323-334.
Luczkovich JJ, Stewart C, Sprague MW. Nocturnal sound production by longspine squirrelfish
(Holocentrus rufus) with notes on sound production by fishes on the Turneffe Atoll coral reef in
Belize. J Acoust Soc Am. 2000; 112: 2202.
Luczkovich J, Keusenkothen M. Sounds recorded during the cooperative hunt may be prey fish alarm
call. Response to: Bshary R, Hohner A, Ait-el-Djoudi K, Fricke H. Interspecific communicative and
coordinated hunting between groupers and giant Moray Eels in the Red Sea. PLoS Biol. 2006; 4(12):
e431 http://biology.plosjournals.org/perlserv/?request=readresponse&doi=10.1371/journal.pbio.0040431#r1400
McKibben JR, Bass AH. Behavioral assessment of acoustic parameters relevant to signal recognition
and preference in a vocal fish. J Acoust Soc Am. 1998; 104: 3520-3533.
Mann DA, Lu Z, Hastings MC, Popper AN. Detection of ultrasonic tones and simulated dolphin
echolocation clicks by a teleost fish, the American shad (Alosa sapidissima). J Acoust Soc Am. 1998;
104: 562-568.
Moulton JM. The acoustical behavior of some fishes in the Bimini area. Biol Bull. 1958; 114: 357-374.
Myrberg AA. Sound communication and interception in fishes. In: Reese S, Lighter F, eds. Hearing
and Sound Communication in Fishes. New York, NY: Springer, 1981: 395-426.
Remage-Healey L, Nowacek DP, Bass AH. Dolphin sounds suppress calling and elevate stress
hormone levels in a prey species, the Gulf Toadfish. J Experim Biol. 2006; 209: 4444-4451.
Winn HE, Marshall JA. Sound-producing organ of the squirrelfish Holocentrus rufus. Physiol Zool.
1963; 36: 34-44.
Winn HE, Marshall JA, Hazlett B. Diel activities, and stimuli that elicit Sound production and
reactions to sounds in the longspine squirrelfish. Copeia. 1964: 413-425.
134
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Deploying Benthic Chambers to Measure Sediment Oxygen Demand in
Long Island Sound
Prentiss H. Balcom1, Jeff M. Godfrey1*, Diane C. Bennett1, Gary A. Grenier1, Christopher G.
Cooper2, David R. Cohen1, Dennis A. Arbige1, and William F. Fitzgerald1
1
Marine Sciences, University of Connecticut, 1080 Shennecossett Rd., Groton, CT, 06340, USA
Ocean Technology Foundation, 1080 Shennecossett Rd., Groton, CT, 06340, USA
*corresponding author: jeff.godfrey@uconn.edu
2
Abstract
In order to address the need for high-quality sediment oxygen demand (SOD) measurements,
diver deployed benthic flux chambers equipped with sondes were used to assess SOD (aerobic
uptake) and the sediment-water fluxes of constituents in regions of Long Island Sound (LIS)
characterized by periods of summer hypoxia/anoxia. Procedures were developed for deploying
chambers from small vessels, and our sampling techniques employed equipment custom
designed at our laboratory. SOD measurements in western LIS during both winter (850 µmol O2
m-2 h-1) and summer hypoxic conditions (870 µmol O2 m-2 h-1) showed good agreement over
small spatial scales. Although the benthic oxygen utilization constant (k) was 6-7 times higher
in the summer, low ambient DO (hypoxia) in western LIS bottom waters at this time accounted
for reduced SOD rates.
Introduction
The goal of the Long Island Sound (LIS) Integrated Coastal Observing System, LISICOS, is the
development of a sustained capability to observe the LIS ecosystem, and to understand and predict its
response to natural and anthropogenic changes (LISICOS, 2007). Efforts are specifically focused on
combining real-time data (monitoring buoys) with a numerical model to predict dissolved oxygen
(DO) concentrations and the occurrence of summer hypoxia. Our primary objective was to determine
benthic oxygen demand, a key unknown in water quality models of hypoxia in LIS, at LISICOS
sampling stations (Figure 1). Benthic chambers were used to assess sediment oxygen demand (SOD;
aerobic uptake) in regions of LIS characterized by periods of summer hypoxia/anoxia. The apparatus
is oceanographically robust and user friendly and, therefore, does not require a specialist for its
deployment. Aerobic and anaerobic bacterially mediated sedimentary processes (i.e., diagenesis) and
infaunal activities, especially in coastal regions, have a major role in the cycling of oxygen, carbon,
nutrients, essential and toxic trace metals, and organics (Warnken et al., 2000, 2001; Mason et al.,
2006). Fluxes of these biologically active constituents into and out of the sediment provide essential
process information needed for understanding the eutrophication dynamics in the water column and
sediments.
Benthic in situ flux chambers have had wide application (Tengberg et al., 1995, 2004, 2005) and are
integral to the essential tasks of this project. Based on a design provided by Dr. Gary Gill and
colleagues (Warnken et al., 2000; 2001) we have built an innovative apparatus equipped with an
oxygen electrode/data logger for time sequenced measurements of oxygen. The design also allows
manual sampling of waters from the flux chamber by syringe. Development of the chambers required
creative designing, engineering, and numerous field and laboratory tests. The benthic chambers yield
high quality time series information on sediment-water exchanges.
135
Methods
Benthic in situ flux chambers were developed and tested during this study. Chambers were deployed
at LISICOS sampling stations in western Long Island Sound (WLIS; LISICOS, 2007) to measure
spatial and temporal variations in sediment/benthic oxygen demand (SOD) during March and August
2005. Deployments were done at the FB1 buoy (40.994°N, 73.678°W; 13.7 m. depth) and the WLIS
buoy (40.950°N, 73.567°W; 18.3 m. depth; Figure 1).
Figure 1. LISICOS monitoring buoy sites in LIS. The five primary sampling
sites for LISICOS process studies are enclosed by the square (figure adapted
from LISICOS, 2007).
The devices are dual chambered (Figure 2), offering a variety of flexible sampling protocols (e.g.,
replication, short spatial variability tests, time series water sampling). Each chamber has an internal
volume of 13L and covers a 0.1 m2 area of surface sediment. The chambers were built at the UCONN
Marine Science and Technology Center (MSTC) using sheet polycarbonate (Modern Plastics), and
only teflon and polycarbonate components (McMaster-Carr) came in contact with the water in the
chambers. Dissolved oxygen was measured at 15 minute intervals using Yellow Springs Instruments
(YSI) 600 series sondes, and other standard hydrographic data (temperature, salinity, depth) were also
measured by these instruments. An Ikellite submersible battery powers a stirring motor (Hankscraft;
25 rpm) inside a watertight polycarbonate housing on each chamber. The motor rotates a floating stir
bar inside the chamber, keeping oxygen and other constituents uniformly distributed throughout the
chambers (Figure 2).
Benthic chambers were manually deployed by davit from the UCONN MSTC vessel R/V Challenger
(25 ft. hull). During deployment, the chambers were attached to a subsurface float located 3 meters
from the bottom of a deployment shot-line (Figure 3). This prevented the chambers from sitting on
the bottom during deployment, and allowed for diver placement on undisturbed sediment. Prior to
deployment, the legs are extended to add stability once the chambers are on the sediment (Figure 2).
The side ports are open during deployment on the shot-line to assist with removal of trapped air, and
closed once the chambers are in position. Due to heavy commercial and recreational boat traffic in the
deployment area, the chambers were not left attached to the deployment shot-line. Gear attached to
136
surface marker buoys is often dragged off by boats after being snagged, or removed intentionally by
user groups that feel the gear will impede their use of LIS. To prevent this, a tether (2m) and weight
(4.0 kg) was attached to the shot-line weight (23-45 kg) before deployment. After the divers had
placed the benthic chamber in undisturbed sediment, the tether weight was detached from the shotline weight and attached to the chambers (Figure 3). The two weights sat next to each other, providing
a continuous guideline from the surface to the chambers. This aided the divers in locating the
chambers in the zero to three foot visibility water without the risk of the equipment being removed if
the shot-line was snagged. Sonic tags (pingers) were also used to allow divers to locate the chambers
with an underwater acoustic receiver if the surface marker buoy was removed. To recover the
chambers at the end of the deployment, the divers attached the chambers to the subsurface buoy and
the tether to the shot-line weight. The entire apparatus could then be hoisted aboard the boat.
Benthic Chambers
stainless
steel cage
Chamber (bottom view)
manually
extendable legs
floating stir bar
electrodes
sampling tubing
Apparatus 0.9 x 0.5 x
0.6 m
0.35 m
manually operated side ports
Figure 2. Benthic chambers used for sediment oxygen demand measurements. The stainless steel cage is an aid to
divers during deployment and protects the sondes once deployed. The two-chamber configuration measures 0.9 x
0.5 x 0.6 m (L x W x H).
The power supply for externally-powered sondes was modified using submersible 5-pin SeaCon
fittings (Brantner & Associates; Figure 4). A similar battery-powered calibration cable was also used
in the field prior to deployment. These power/calibration cables were designed by David Cohen for
use on the LISICOS monitoring buoys, and wiring was done by UCONN MSTC. This system allows
the user to avoid buying and modifying YSI field cables, and sondes can be changed out underwater
when using SeaCon fittings.
Divers sampled water from chambers at deployment and retrieval using syringes (60 ml
polypropylene luer-lock or 1L acrylic/polycarbonate; Figure 5). Sample water used for Winkler DO
titrations was added without disturbance to 60 ml BOD (biological oxygen demand) bottles. Water
was sampled through either polypropylene luer-lock tubing or large inner diameter silicone tubing
(Figure 2). Winkler titration (APHA, 1995; Strickland and Parsons, 1968) comparisons to sonde DO
readings were made in the laboratory following calibration, and with samples collected from the
chambers in the field. In August 2005, water samples were collected as a test of the procedures and
techniques for the determination of biogeochemically important chemical constituents. The suite of
measurements included dissolved organic carbon, nutrients, total mercury, and monomethylmercury
(results not presented here).
137
Surface buoy
Subsurface buoy
Tether
weight
Chambers during deployment
4 kg
Shot-line
weight
Chambers after deployment
Figure 3. Schematic of benthic chamber deployment using a shot-line, and position of the
tether weight and chambers on sediment relative to the shot-line weight after deployment.
submersible
battery
Submersible power supply
Calibration cable
5-pin connectors
battery
dummy plugs
computer
sonde
5-pin connectors
Figure 4. Externally-powered sonde shown with submersible 5-pin power cable and 5-pin battery-powered
calibration cable.
138
Figure 5. 1L syringe used to sample water from the chambers for Winkler
titrations and biogeochemically important chemical constituents.
An oxygen utilization constant (k) is estimated from a linear regression of the log of the scaled DO
observations (ct/c0) and time (ct and c0 are the DO content of the chamber at time t and initial
deployment), and was used to estimate SOD (kc0; James O'Donnell, personal communication). Since
the exponential decline is substantial (Figure 6), the initial phase (four hours) of the oxygen record
that was not influenced by start-up transients was chosen to estimate these rates.
14
WLIS buoy - March 2005
Dissolved Oxygen (mg/L)
11
10
9
8
7
6
5
4
3
2.0
1.5
1.0
L
L
H
L
H
12
:
00
:
00
00
00
:
06
:
00
:
00
00
00
:
00
:
12
:
00
:
00
0.0
0:
00
0:
00
H
12
:0
L
0:
00
0:
00
H
00
:0
0:
00
L
FB1 buoy - August 2005
0.5
sunrise
06
:0
sunset
18
:0
H
0:
00
L
0:
00
0:
00
H
sunrise
12
:0
12
:0
0:
00
0
sunset
18
:0
1
06
:0
2
00
:0
Dissolved Oxygen (mg/L)
12
14.0
12.0
10.0
8.0
6.0
4.0
18
:
13
Time Sampled
Time Sampled
Figure 6. Benthic chamber dissolved oxygen measurements (YSI sondes) from March and August
2005. Each plot shows one of the duplicate chambers deployed at each site. Approximate times of high
(H) and low (L) tides are indicated.
The calculated SOD rates from duplicate chambers at the WLIS buoy in March 2005 (Figure 6) were
834 and 871 µmol O2 m-2 h-1, indicating good agreement (4% relative difference;
[difference/average]*100) over small spatial scales, and k ranged from 0.018 to 0.022 h-1. The
deployment of duplicate chambers at the FB1 buoy in August 2005 (Figure 6) produced SOD
estimates of 928 µmol O2 m-2 h-1 (k = 0.148 h-1) and 824 µmol O2 m-2 h-1 (k = 0.120 h-1), indicating
slightly greater variability (12% relative difference) on a small spatial scale. Although the SOD
estimates are similar in March (2°C) and August (19°C), k values were 6-7 times higher in August.
The low ambient DO (hypoxia) in the bottom waters of WLIS in August (22% oxygen saturation)
139
accounted for reduced SOD rates, even at elevated values of k. This suggested that enhanced SOD
fluxes would be evident in the early summer and fall when aeration processes produce substantially
higher benthic water column DO in WLIS.
In the laboratory, the difference between sonde oxygen measurements and Winkler determined DO
levels was small (0-4% relative difference) in March and August 2005. Winkler samples collected
from benthic chambers deployed in March 2005 (average ambient DO 400 µM or 12.8 mg L-1)
differed from the sonde reading by 7%. Oxygen measurements agreed within 1-4% in August 2005
(average ambient DO 56 µM or 1.8 mg L-1).
The instrument deployment and recovery method was effective. A significant amount of time was
saved by having a continuous guideline to the chambers, and the divers required the use of an
underwater acoustic receiver to find an instrument only once. The polycarbonate construction proved
to be robust enough for deployment by shot-line from a small boat. The threaded connection between
sonde and chamber failed once during a deployment, and the problem was resolved by removing the
bails (lifting wire) from the sondes as they proved to be a snag point. The initial results show that the
dual benthic in situ flux chambers yield high quality time series information on sediment-water
exchanges and allow for manual sampling of waters from the chambers. Our procedures for deploying
and recovering the chambers were reliable, and became easier in subsequent deployments (2006)
from a larger, more stable platform, the UCONN MSTC vessel R/V Lowell Weicker (36 ft. hull, Aframe).
Additional surveys were conducted in 2006 and have addressed the seasonal component to SOD rates
and estimating k. Concurrent deployments (duplicate chambers) at several of the LISICOS sampling
stations have allowed us to examine spatial variability in SOD over a larger area. Although
successful, the current approach where the chambers are deployed and sampled entirely by divers, is
labor intensive and ultimately limits the spatial and temporal coverage. Therefore, longer term plans
include adapting the apparatus to a benthic lander and development of an automated version that will
provide timed water sampling. These modifications may allow benthic chambers to be deployed from
a boat without diving.
Acknowledgments
Many thanks to Peter Boardman, Turner Cabaniss, Bob Dziomba, Adam Houk, Kay Howard-Strobel,
Marco Liebig, Nathalie Morata, Dr. James O'Donnell, and Brennan Phillips (UCONN Marine
Sciences) and Dr. Gary Gill (Battelle Marine Sciences Laboratory, WA). This research was partially
funded by NOAA through grants to LISICOS.
References
APHA (American Public Health Association, American Water Works Association, and Water
Environment Federation). Standard Methods for the Examination of Water and Wastewater, 19th ed.
American Public Health Association, Washington, DC, 1995.
LISICOS (Long Island Sound Integrated Coastal Observing System). University of Connecticut,
Department of Marine Sciences, 2007. http://uconn.lisicos.edu.
Mason RP, Kim E-H, Cornwell J, Heyes D. An examination of the factors influencing the flux of
mercury, methylmercury and other constituents from estuarine sediment. Mar Chem. 2006; 102: 96110.
140
Strickland JDH, Parsons TR. In: Stevenson JC, ed. A Practical Handbook of Seawater Analysis.
Bulletin 167, Fisheries Research Board of Canada: Halifax, NS, 1968.
Tengberg A, de Bovee F, Hall P, Berelson W, Chadwick D, Ciceri G, Crassous P, Devol A, Emerson
S, Gage J, Glud R, Graziottin F, Gundersen J, Hammond D, Helder W, Hinga K, Holby O, Jahnke R,
Khripounoff A, Lieberman S, Nuppenau V, Pfannkuche O, Reimers C, Rowe G, Sahami A, Sayles F,
Schurter M, Smallman D, Wehrli B, de Wilde P. Benthic chamber and profiling landers in
oceanography—a review of design, technical solutions and functioning. Prog. Oceanogr. 1995; 35:
253-292.
Tengberg A, Stahl H, Gust G, Muller V, Arning U, Andersson H, Hall POJ. Intercalibration of
benthic flux chambers I. Accuracy of flux measurements and influence of chamber hydrodynamics.
Prog Oceanogr. 2004; 60: 1-28.
Tengberg A, Hall POJ, Andersson U, Linden B, Styrenius O, Boland G, de Bovee F, Carlsson B,
Ceradini S, Devol A, Duineveld G, Friemann J-U, Glud RN, Khripounoff A, Leather J, Linke P,
Lund-Hansen L, Rowe G, Santschi P, de Wilde P, Witte U. Intercalibration of benthic flux chambers
II. Hydrodynamic characterization and flux comparisons of 14 different designs. Mar Chem. 2005;
94: 147-173.
Warnken KW, Gill GA, Santschi PH, Griffin LL. Benthic exchange of nutrients in Galveston Bay,
Texas. Estuaries. 2000; 23(5): 647-661.
Warnken KW, Gill GA, Griffin LL, Santschi PH. Sediment-water exchange of Mn, Fe, Ni and Zn in
Galveston Bay, Texas. Mar Chem. 2001; 73: 215-231.
141
142
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
An Evolution of Scientific Mixed Gas Diving Procedures at the
National Park Service Submerged Resources Center
Jeffrey E. Bozanic
Island Caves Research Center, P.O. Box 3448, Huntington Beach, CA 92605-3448
JBozanic@HQonline.net
Abstract
The Submerged Resources Center (SRC) of the National Park Service has been conducting an
assessment of sites of cultural significance in the Lake Mead Recreational Area since 2002.
Many of the sites were flooded when Hoover Dam construction was completed and the reservoir
filled, such as the aggregate plant and associated structures. Others have been deposited after
Lake Mead filled, including a B-29 lost in 1948. Because many sites of interest are located at
depths exceeding 150 ffw (45 msw), work initially began with ROV surveys. Later activity
involved mixed gas training and in situ work with divers using heliox. Equipment profiles and
techniques have evolved as subsequent phases of the project were fielded. Initially, open-circuit
equipment was used, then mixed teams using closed-circuit rebreathers (CCRs) with the opencircuit divers, and most recently teams solely using CCRs. In addition to the SRC divers,
participants from NOAA and NGOs have been involved. This paper will examine the
background of the project, diving standards utilized for the heliox operations, and the
procedures and safety concerns involved with teams using multiple modes of diving with mixed
gas on these research operations.
Introduction
The Submerged Resources Center (SRC) of the National Park Service (NPS) was originally
established in 1980 as the Submerged Cultural Resources Unit. In 2000, their mission was expanded
and their name changed. Their current mission is to provide direct support to superintendents and
partners responsible for stewardship of submerged resources, and it enhances and facilitates public
appreciation, access, understanding, and preservation of those resources.
In the past, SRC personnel have been involved in such projects as the original inventory of the ships
sunk in Bikini Atoll as part of the 1946 atomic bomb tests (Delgado et al., 1991), assessment of the
U.S.S. Arizona (Lenihan et al., 1989), investigation and raising of the Confederate submarine H.L.
Hunley (Murphy et al., 1998; Conlin and Russell 2006), documentation of the shipwrecks of Isle
Royale (Lenihan, 1987), cataloguing the submerged cultural resources of Fort Jefferson (Murphy
1993), and examination and evaluation of the submerged artifacts of Nan Madol, Pohnpei (Lenihan,
2002). The vast majority of their diving work has been conducted at depths considered to be within
reasonable air diving limits (150 fsw [45 msw]), however, many working dives have been made to
deeper depths during the course of field operations. For example, dives in excess of 150 fsw (45 msw)
were conducted in Isle Royale, Bikini, various reservoirs, and during examination of CSS Alabama
sunk off Cherbourg France (Lenihan, 2002).
In 2002 they were asked by management of the Lake Mead National Recreational Area (LAME) to
assist in the inventory and assessment of their submerged cultural resources. Lake Mead was formed
when the Colorado River was impounded by the construction of Hoover Dam in 1935. There are three
categories of submerged resources: preexisting structures unrelated to dam construction inundated by
143
rising lake water, structures related to dam construction but subsequently flooded, and vessels sunk in
the lake since flooding.
Lake Mead, at highest water mark, is as much as 589 ft (180 m) deep. The surface of the lake
fluctuates depending on annual recharge from precipitation in the Rocky Mountains and other parts of
the water basin and the amount of water released from the dam. In 2002, at the time SRC was
requested to provide assistance, some sites of particular interest were located in as much as 240 ffw
(80 mfw). No unit in the NPS, including SRC, had the capability to deploy working divers to these
depths. Accordingly, SRC Chief Larry Murphy developed a training strategy to develop this
capability to continue their mission of assessment, monitoring, preservation, and enforcement
assistance to depths beyond 200 fsw (61 msw) in the safest manner possible, both in terms of project
dive safety and minimization of longitudinal physiological consequences for career occupational
scientific divers. This paper will cover the training, equipment, and procedures used to achieve this
objective.
Training
Five NPS divers were to be involved in the project. Three modes of diving were considered for the
deep assessment work: surface supply, open-circuit SCUBA (OC), and closed-circuit rebreathers
(CCRs). All divers had extensive (more than ten years) occupational experience in the use of
commercial-type surface supplied equipment as well as open-circuit SCUBA. Surface supplied gas
was rejected as an option because of the logistical requirements necessary and the restrictions to
mobility required for archeological documentation of fragile sites like the B-29. Use of CCRs was
similarly dismissed from immediate consideration because none of the divers had mixed gas or
rebreather training, suitable equipment was unavailable within the NPS, and project deadlines did not
allow sufficient time for this option to be researched and developed.
The anticipated working depths mandated that a helium-based breathing mixture be utilized.
Replacing some of the nitrogen and possibly oxygen in air reduces inert gas narcosis at depth, as well
as reducing the potential for CNS oxygen toxicity and physiological complications from long-term
deep air exposures. Two options available were the use of heliox (a mixture of helium and oxygen) or
trimix (a mixture containing nitrogen, helium, and oxygen). Use of either type of gas mixture offer
positive safety aspects, but entails additional risk to the divers, including in part increased potential
for hypoxia, hyperoxia, hypothermia, decompression sickness, high pressure nervous syndrome,
isobaric counterdiffusion, and other risks.
In general, the lower the nitrogen fraction in breathing gas, the lower the level of inert gas narcosis
experienced by the diver. Because there is no commercially available open-circuit heliox training
available, the decision was made to complete a training course based on trimix, which is standardized.
The training program was designed to utilize trimix, as per the certifying agency training standards,
allowing the divers to reduce their equivalent narcotic depth (END) to 100 ffw (30 mfw ) (i.e., the
same level of narcosis they would experience as though they were breathing air at that depth).
In addition, since trimix was not available commercially at either the training location in Santa Rosa,
NM or at Lake Mead, training included instruction in the production of both trimix and the nitrox
mixtures used for decompression.
Training of the five divers took place over a sixteen-day period in 2003. Curriculum of the program
was patterned after IANTD Technical Diver, Trimix Diver, and Trimix Gas Blending standards
(Anon, 2003a). Additional components were added to prepare the divers for the type of work they
144
would be doing at depth in Lake Mead. Each diver did a minimum of 12 training dives. They were
also responsible for mixing all of the gases they each utilized during those dives. Some of the course
objectives included training in use of double cylinders, use of stage cylinders, gas switching,
decompression procedures, surface air consumption (SAC) rate determination, gas management
planning, emergency procedures, and breathing gas selection.
Equipment used by each of the divers included the following: back-mounted double 120 ft3 (19 L)
cylinders with a dual valve non-isolation manifold, two 80 ft3 (11 L) stage cylinders containing
multiple decompression gases, 6.0 ft3 (1.0 L) air drysuit inflation cylinder, regulators for all of the
preceding, lift bag and reel, canister light, various slates and dive tables, Hydrospace Engineering
Explorer (HSE) dive computer, multiple cutting devices, drysuit, watch, mask and fins.
Training decompression was planned using the IANTD tables (Mount, 1998), which are based on a
Buhlmann algorithm. Tables were backed up by the HSE computers. The HSE computer has nine
different models programmed into the units. Computational formulas (CF) 0-2 are based on the
RGBM model developed by Wienke. CF3 to CF9 are based on the Buhlmann algorithm, and are
progressively more conservative. Use of the RGBM model was rejected due to the SRC dive team's
concerns with validation and total operational experience. CF3 was utilized for all of the dives. In
addition, a deep stop (halfway between maximum depth and the first required decompression stop) of
two minutes was added to the required decompression, with the maximum bottom time reduced by
that time.
In all cases, computers allowed for a shorter decompression time and quicker ascent time than did the
tables, as would be expected because the training dives were not true square profiles. However, had
there been an emergency mandating a minimal ascent time, the computer parameters for required
decompression would have been followed.
Progressively deeper training dives were made at Rock Lake, to a maximum depth of 190 ffw (58
mfw). Altitude corrections were applied to the tables due to the 7,000 foot (2,133 m) elevation of the
highway between the dive site and nearest medical facilities in Albuquerque (dive site elevation was
5,000 ft (1,524 m). This yielded an equivalent physiological depth of 230 fsw (70 msw). Bottom
times were as much as 27 minutes, resulting in average decompression times of 75 minutes.
Nitrox was generally used as the bottom gas for dives 100 ffw (30 mfw) and shallower. On deeper
dives, various Trimix blends were used, ranging from Tx28/32 to Tx18/40, etc (Tx18/40 denotes a
mixture containing 18% oxygen, 40% helium, and 42% nitrogen). Decompression was conducted
using EAN32 (a nitrox mixture containing 32% oxygen) for decompression stops in the 110 ffw (33
mfw) to 31 ffw (10 mfw), and oxygen from 30 ffw (9 mfw) to the surface. (Note: this exceeds AAUS
and NOAA recommendations.) This added an additional margin of safety relating to decompression
sickness, as the tables called for the use of EAN35 and EAN75 respectively.
Breathing gas was blended on site using a continuous blend methodology. A mixing stick capable of
accepting oxygen, helium, and air simultaneously was utilized in conjunction with a Rix oil-free
compressor. All final gas mixtures were analyzed for oxygen and helium fractions during blending
and each mix was analyzed using independent Oxycheq analyzers prior to each dive.
As the instructor, I preferred the safety and logistical advantages afforded by mixed gas CCRs, so
during all training dives I used a PRISM Topaz rebreather. Tx10/50 was utilized as diluent for all of
the deep dives (>100 ffw [30 mfw]). I carried sufficient bailout gas to get either myself or any one of
the NPS divers to the surface in the event of an emergency.
145
During the class, the dive team conducted extensive research into comparisons of various
decompression algorithms, Trimix fractions and heliox use. Upon course completion, three days of
concentrated deep diving physiology and dive accident prevention and management were conducted
by Dudley Crosson of Delta P, Inc., a company that provides diving consultation to commercial
companies, military and NASA.
Operational Diving Procedures
B-29 Plane Wreck
The initial objective of LAME management was the archeological assessment of a B-29 that crashed
into the lake during a scientific mission in 1948. At the time of the two field deployments in 2003 and
2005, the lake level was about 1,140 ft (347 m), yielding a maximum diving depth of 185 ffw (56
mfw).
During fieldwork SRC divers decided to use heliox instead of trimix as the bottom gas, based on their
research. There were several reasons for this change. First, trimix decompression models are not as
well validated as heliox models. Heliox has been in use by the US Navy (Anon, 1999), commercial
diving corporations, and DCIEM (Anon, 1995) for decades, and there has been extensive scientific
experimentation with this mixture. Heliox tables had varying amounts of controlled validation testing,
were readily available, and had significant operational history. In contrast, the IANTD Trimix tables,
released in 1993, had, in the opinion of the dive team, insufficiently rigorous scientific validation and
limited operational usage.
Second, all of the divers felt varying degrees of narcosis and apprehension during the deeper training
dives in Rock Lake. This was a cold, dark environment, very similar to that expected in Lake Mead.
Based on the training experience, elimination of all nitrogen from the bottom mix would facilitate the
research work and accident management and also provide the best risk minimization approach. To
minimize field logistics and to ensure consistency of heliox mix, commercial blended heliox with an
oxygen fraction of 0.20 (HeO220) would be used for OC. Gas was boosted to final filling pressures
using an electric Masterline oxygen-clean booster.
This change was not without downsides. The cost-per-dive-per-diver increased from about $25 to
about $100. The increased helium fraction increased decompression time and also induced a
potentially greater heat loss in the divers. In SRC's view, the risk minimization offered by heliox was
worth the increased cost.
This project required development of a project-specific Safe Practice Manual that included a Project
Dive Emergency Plan. Some key elements of the diving protocol may be relevant to others. Bottom
mix was normoxic heliox (HeO220) premix on OC or heliox diluent containing 5-15% oxygen with a
set point of 1.3 atm when using a constant PO2 heliox in rebreathers. The only approved dive tables
for this project were the DCIEM tables and 'straight' Buhlmann tables, and the HSE Explorer or VR3
computer using the straight Buhlmann algorithm calculated for EAN32 decompression. However,
during operations surface-supplied 100% O2 from 30 ffw (9 mfw) to surface was required, and all
stops shallower than 20 ffw (6 mfw) were conducted at 20 ffw (6 mfw). Rebreather divers adjusted
PO2 accordingly.
Fieldwork was conducted for 14 days in 2003, 10 days in 2005, and two deployments of 16 days and
seven days in 2006. During that time diving operations were conducted in three-day segments, with
the fourth day being an enforced break from diving. This was done to eliminate concerns with long
146
half-time compartment loading and to provide an extended break for pulmonary oxygen toxicity
exposure. Divers were limited to one deep dive per day.
Initial site assessment was done using side scan sonar and ROVs. ROVs were utilized nearly every
day throughout the project to augment the data collected by the divers. They were also used to
penetrate and record images and water quality in the plane's interior, which was inaccessible to the
divers (access would have had significantly detrimental effects to the B-29).
All diving operations were conducted from a 40-ft (12-m) NPS maintenance barge, the Tamarisk. The
barge was motored out daily and secured over the site in a four-point mooring. All dive equipment,
emergency gear like AED and first aid kit, surface supplied oxygen, and other gear were kept on the
barge. The barge was moved in from the site every night because of weather uncertainty (wind and
squalls could blow up at virtually any time).
Conditions at depth were extremely dark. In the summer months in which the project was run in 2003,
the upper 50 ft (15 m) of the water column was a very dense plankton assemblage. Below this layer,
ambient light was extremely limited, being inadequate to read gauges, for example. For this reason,
all divers carried 10-watt HID canister lights, as well as one to three backup lights per person. In
addition, a custom built frame holding four 1,200-watt HMI lights was deployed at a depth of about
130 ffw (39 mfw). These lights were powered by three generators located on the surface. The light
provided by these HMIs enabled the divers to work most of the time without using their individual
lights, and greatly facilitated video and photographic documentation.
Water temperature at depth was about 52°F (11°C). Air temperatures ranged as high as 105°F (40°C).
This was of concern because of dehydration issues, especially with the suited standby diver.
Conscious efforts were taken to keep all of the divers adequately hydrated, including providing
drinking water in Camelback bags at the decompression stages.
A triangular shaped decompression stage was constructed from PVC pipe for the project. Each side of
the triangle was eight feet (2.5 m) long, sufficient for two divers to comfortably decompress on each
side. The assembly consisted of two levels. The decompression stage was deployed prior to
commencing operations every day, with the two stages at 30 ffw (9 mfw) and 20 ffw (6 mfw)
respectively. The stage was suspended by a single line which splint about 10 ft (3 m) below the
surface to each corner of the shallower decompression stage. One corner of the deeper stage was
secured to the descent/ascent line used by the divers.
The descent line was moored to a 200 lb (90 kg) concrete block deployed for that purpose. It was
located about six feet (2 m) from the starboard wing. Attached to block was an 80 ft3 cylinder filled
with heliox bottom mix. A cylinder containing EAN32 was tied to the line at 110 ffw (33 mfw), the
location where the team members would switch from their heliox bottom mix to a nitrox
decompression mix. These cylinders were left in their respective locations for the duration of the
project and were for emergency use only. Another cylinder of heliox was staged on the bottom at the
trailing edge of the port wing near the fuselage. All cylinders were wrapped with photoreflective tape
and left with their valves closed. The first dive team of the day checked the submersible pressure
gauges of each cylinder at the beginning of their dive to ensure that the cylinders were full.
Bottom times ranged from 20 to 45 minutes. Various decompression models were compared,
including USN, DCIEM, and HSE Explorer. Because the USN and DCIEM tables utilized surface
oxygen decompression (involving egress of the diver from the water and pressurization in a chamber
for surface decompression, software derived HSE tables were followed for the dives. Again, a straight
Buhlmann-based algorithm was used. Total decompression times ranged to approximately two hours.
147
Open-circuit divers utilized the same decompression gas mixtures previously listed. Again, oxygen
decompression began at 30 ffw (9 mfw), yielding a PO2 of 1.9 atm. While this exposure is greater
than that recommended by the recreational sport diving agencies, AAUS or NOAA, SRC divers have
been using this approach for decompression purposes since the inception of the original dive team in
1975 without incident. Members of SRC pioneered the use of pure oxygen for decompression without
altering times for scientific diving (Murphy, 1978; Lenihan, 2002). SRC experience determined that
the benefits of increased inert gas washout (maximum alveolar tissue pressure gradient) outweighed
CNS oxygen toxicity issues. To alleviate concerns with CNS toxicity, oxygen was breathed for 20minute intervals, followed by five minute air breaks. The air break times were not counted towards
the required decompression obligations, again adding a margin of safety. The last decompression stop
was at 20 ffw (6 mfw), followed upon completion by a two minute ascent to the surface.
Oxygen was supplied from the surface. Four 200 ft3 (38 L) cylinders were manifolded together to
supply the gas. Only three were online at any given time, with one kept in reserve to cover supply
failures. Six second stages on individual hoses were attached at the deeper stage by the safety diver
prior to commencing research diving operations. Two emergency supply cylinders of oxygen were
placed at both decompression levels, as well as two air cylinders for air breaks. Each cylinder had its
own regulator, with two second stages each. Water in Camelback bags was also placed at each level,
allowing the divers to stay hydrated during the decompression stops.
A topside dive supervisor maintained logs of all entries and exits, as well as overseeing pre-dive
safety equipment emplacement.
Again, for these dives I utilized a CCR, using HeO210 in the diluent cylinder. In 2003, I used a
PRISM Topaz. The factory supplied 19 ft3 (3 L) cylinders were replaced with 42 ft3 (7 L) cylinders,
allowing additional available onboard bottom gas for open-circuit bailout. In 2005, a stock Evolution
CCR was used, and in 2006 a stock Dive Rite Optima with 27 ft3 (4.5 L) cylinders was used. Drysuit
inflation gas was supplied by an off-board cylinder containing either air or nitrox.
Aggregate Plant
During the early 2006 field season, various structures associated with the aggregate plant used to
construct Hoover Dam were investigated and documented. The aggregate sorting and washing facility
was a huge industrial complex designed to provide clean aggregate of differing sizes which, when
mixed with cement, would form concrete with specific structural properties for use in different
sections of Hoover Dam. In all, more than 118 million cubic feet of concrete were poured to create
the dam. Following completion of the dam, most of the facility was removed, but significant portions
remained and were subsequently submerged by rising lake levels.
Dives in 2006 were conducted from February to March and in December, eliminating the problems
associated with summer heat and dehydration of the divers. Shallow visibility was also better due to
reduction of plankton density in the winter months.
This phase of the project used mixed modes of diving. After the 2005 field season, SRC divers had
decided to utilize CCRs for their diving. In addition to the units used and evaluated during prior field
seasons, the SRC also considered the USN Mark 16, Ourobouros, Megalodon, Oceanic and
Inspiration CCRs. The team selected the AP Evolution as the best currently available unit for their
needs, and accordingly two divers spent two weeks training on the unit in December 2005. Further
experience was gained during field work in Biscayne National Park after the completion of training,
and while completing geological and hydrological studies in Montezuma Castle National Monument
in January 2006.
148
Training for the use of helium-based diluents was accomplished at Lake Mead prior to beginning
research activities. SRC divers had in excess of 50 hours each using the Evolutions prior to beginning
this advanced training. Maximum depth of the training dives was 210 ffw (64 mfw).
Two additional divers from NOAA joined the SRC team during this phase of the project. In all, there
were three CCR divers and three O/C divers.
While no particular effort was made to match divers using the same mode of diving, dive planning
considerations generally led to this result. In those dives where mixed CCR and OC modes were
utilized, the CCR divers always carried sufficient OC bailout gas to aid their buddies in the event of a
gas supply failure emergency.
Prior to being paired with CCR divers, OC divers were provided a basic overview of CCR theory and
operation. They were also instructed in CCR failure modes, and how to handle CCR emergencies and
CCR specific rescue procedures. This basic level of education permitted them to monitor CCR
partners, and would have allowed them to provide emergency assistance had it been required.
The shallowest site was the water clarifier, originally built to remove sediment from the water used to
clean the aggregate prior to use in the dam concrete. The clarifier ranged from 3-55 ffw (1-17 mfw) in
depth in early 2006. By December 2006, portions of the clarifier were above the surface due to
reservoir lowering. The second site surveyed was a section of railroad tracks used to transport
materials to and from the aggregate plant. These were 110-115 ffw (33-35 mfw) deep.
The bulk of the aggregate plant included railroad hoppers, piles of aggregate, and various equipment
associated with the aggregate plant. This third site contained at least four areas in which divers
entered structures in which direct access to the surface was not possible. During these penetrations
(maximum linear distance of about 70 ft (21 m) at a maximum depth of 155 ffw (47 mfw) guidelines
were used as navigational aids. Guidelines were also used when swimming from one structure to the
next to ensure that divers would be able to return to the upline and the decompression trapeze at the
end of the dive.
In general, the same decompression protocols established during the 2003 and 2005 B-29 portion of
the project were utilized. However, because part of the team was in residing Boulder City instead of
on a houseboat at lake level, an altitude adjustment of 3,000 ft (900 m) was incorporated into the
decompression model used to generate the decompression tables. Because NOAA divers were using
VR3 computers, tables were generated using both the VR3 and HSE models, with the more
conservative figures used for any given dive.
Catalina PBY-5A Wreck
The most recent diving activities took place during December 2006. The lake level at this time was
~1,127 ft (344 m). The primary focus during this field deployment was documentation of a Catalina
PBY-5A flying boat which crashed and burned during a water landing in 1949 (Anon, 1949), sinking
soon after hitting the water. Maximum depth reached during this phase was 191 ffw (58 mfw).
The primary difference during this deployment was the fielding of a complete CCR dive team. Plans
called for using three rebreather divers, two diving at any given time, and one situated on the surface
as a standby diver. The goal was to minimize dive and support equipment, and simplify gas supply
logistics.
149
Unfortunately, one of the divers was unable to dive, due to a pre-existing problem with alternobaric
vertigo. To accommodate this issue, plans were modified to utilize an OC diver as the standby diver
for all dives. Since he was not diving daily, his gas cylinders only needed to be filled one time at the
beginning of the project. Therefore, total gas requirements for the project increased by a negligible
amount.
Heliox10 was used as the diluent for these dives. Divers carried two 46 ft3 (8 L) bailout cylinders, one
containing HeO220 for use on the bottom, and a second of oxygen for decompression. For suit
inflation divers carried 6 ft3 (1 L) cylinders of air.
Again, general decompression procedures remained the same as previous field deployments. Both
VR3 and HSE computers were used to back up the Vision electronics on the Evolutions during this
phase, with Buhlmann-based HSE tables cut to determine decompression obligations. Divers were
again residing in Boulder City, so the same altitude adjustment was used. The dives were done using
a set point of 1.3 atm. Since decompression was done on the CCRs instead of OC gas, PO2s were
increased to 1.4 atm for the decompression, and no air breaks were utilized. To incorporate a further
margin of safety, the tables cut for the dives were based on a set point of 1.1 atm for both the dive as
well as the decompression. OC nitrox, surface supplied oxygen and OC air was available to the divers
in the event of a CCR failure.
Emergency Procedures
A variety of plans were defined in advance of actual dive operations to cope with any foreseeable
emergencies that might occur. These were defined in the SRC Safe Practices Manual, which was
revised prior to every new field deployment (Murphy and Seymour, 2003-2006).
Key components of this plan were communication and transport management. Central dispatch was
notified immediately prior to initiation of diving activities, and again when divers had exited the
water. The nearest operational chamber was identified and contacted prior to the initiation of all
projects and transport protocols established in the event of need. During the time NOAA personnel
were diving, a portable recompression chamber and operator were on site, and all personnel were
briefed on its use prior to onset of diving operations.
A standby diver was suited and ready to deploy within five minutes whenever deep diving operations
were underway. A dive supervisor was always present on the surface to oversee operations and
maintain dive time and depth information.
General emergency plans included caching cylinders of bottom gas at two locations on the bottom,
typically at the base of the down line and at a site near where the divers would be working on the
structure. Emergency decompression gases were cached on the upline at or below the depth of
planned use (EANx). At the decompression trapeze, a redundant cylinder of oxygen was clipped off
with the valve turned off in the event of catastrophic failure of the surface supplied equipment, and
extra air cylinders were staged for air break usage. In addition, one cylinder each of bottom gas mix,
EAN34, oxygen, and air were available on the surface rigged and ready to deploy via the standby
diver if needed.
All divers carried the decompression schedules on slates on every dive. They each also carried
contingency tables for worst case scenarios, in case they exceeded planned maximum depth by 10 ffw
(3.0 mfw) or time by 10 and 15 minutes.
150
Divers carried a separate slate with a list of predetermined emergency scenarios (equipment failure,
out of bottom gas, out of EANx, out of oxygen, DCS, lost buddy, lost upline, etc). This allowed the
diver to merely circle the appropriate problem in the event of difficulties and send it to the surface via
lift bag. At the surface it could be retrieved, and the standby diver deployed knowing what to expect
and carrying any necessary equipment or supplies. Each lift bag was clearly labeled with the diver's
name so surface personnel would know who was experiencing difficulties.
Lines were run on the bottom to/from the down line, insuring navigational capability back to the line
for ascent. Whenever structures were penetrated, guidelines back to the entrance were utilized.
During the 2006 operation, divers also tagged the upline with a waterproof strobe light to facilitate
reacquisition in the event of light failure or siltout.
Rebreathers were prepared and maintained after the dives by following detailed checklists. The
checklists used were developed in greater detail than those provided by the various manufacturers.
Any problems with CCRs were resolved prior to diving.
OC divers were trained on what responses were necessary should a CCR dive partner experience
difficulties while underwater. In this manner they were prepared to provide effective assistance if
necessary. CCR divers carried OC gas appropriate for whatever depth they were at for both self
rescue as well as buddy assistance. Decompression schedules for OC bailout from depth with a worst
case depth and time profile were carried as a backup to the closed-circuit decompression schedules
should a bailout at depth become necessary.
Rebreather divers were required to use fresh carbon dioxide absorbent fills for any dive deeper than
150 ffw (46 mfw). Emergency procedures for recovery from different rebreather problems or failures
were practiced or reviewed before each field deployment. Bailout procedures were similarly
practiced.
Discussion
During the three years that deep diving operations have been used to support research by the SRC,
118 helium-based deep dives have been conducted without incident (2003, 48 dives; 2005: 16 dives;
2006: 54 dives). Both open- and closed-circuit dive modes worked well and these projects offer the
opportunity to compare the logistics, costs and efficiencies among OC, mixed and rebreather field
operations.
The SRC protocols have evolved through time, with their Safe Practices Manual reflecting this
evolution. In general, the more active divers have switched from an open-circuit to a closed-circuit
mode. This has resulted in more cost-effective diving operations, as gas costs have been enormously
reduced (by as much as 95% for consumables, including gas, batteries, and absorbent cost).
Rebreather use has also resulted in an overall increase in efficiency. The time needed to prepare the
CCRs and maintain them after the dives, including filling cylinders, has been less than half that
needed to fill the larger double cylinders necessary for OC operations. This has freed more time for
research and support activities, and it has also reduced diver fatigue and stress.
The SRC divers made several modifications to their Evolutions to facilitate maintenance chores.
Primary among these was changing the manner in which the counterlungs attach to the unit. The
standard lungs were cut off the mounting harness, and Fastex clips utilized to attach them. With this
system, the lungs and breathing hoses may be removed from the unit and easily carried inside to a
151
comfortable location for disinfecting. They also had special quick-release mounting brackets designed
so that accessories like video camera battery packs, small bailout cylinders and drysuit inflations
systems can be quickly and easily attached and removed from the system.
Finally, rebreather use has reduced shipping and logistical costs and time expenditures. A fully
supported CCR team can be deployed with the equipment, including gas supplies, in a large pick-up
truck, while a similar open-circuit effort would require a fully laden panel truck. Conceivably, a flyaway capability could be reasonably easily established, so long as the appropriate gases are available
at the research destination.
These benefits have some significant costs associated with them. The initial capital outlay of about
$10,000 per diver, plus the associated training cost and time commitment (about four weeks for
helium-based diluent use) is a serious barrier to entry for this mode of diving. Skill maintenance costs
are also high, as CCR skills must be practiced to maintain currency. These issues reduced the
available pool of fully qualified deep SRC divers from six persons to two during the time frame of
this project.
Therefore, it would seem if deep diving operations are to be sporadic in nature or of limited scope,
then an open-circuit approach is probably the most cost effective. If such operations are to be more
extensive, and if the personnel turnover is low, then a closed-circuit approach may be justifiable.
With the use of CCRs, there has been a move away from using oxygen at 30 ffw (9 mfw) depths,
reducing the planned oxygen exposure experienced by the divers from 1.9 atm to 1.4 atm. There has
also been a shift in tools used to plan dive exposures from published tables to software developed
tables, and from the HSE Explorer dive computer to the VR3 computer.
A variety of agencies have been involved in the SRC diving activities at Lake Mead. Some of these
have fielded divers, other have participated as observers, surface research personnel, support
personnel or collaborators. These have included, in part: other branches of the National Park Service,
NOAA, Parks Canada, Lone Wolf Documentary Group, History Channel, the Outside Las Vegas
Foundation, Forever Resorts, the National Park Service's Student Conservation Association,
University of Nevada Las Vegas, Discovery Channel Canada, Our World Underwater Scholarship
Program, PBS, Archaeology Magazine and Clark County Coroner's Office.
In some cases interagency differences in diving protocols have initiated modifications to the dive
operation procedures. For example, NOAA's standards for mixed gas diving require an on site
recompression chamber, and the use of different decompression planning tools than those used by
SRC. In these cases, both agencies' standards were examined, and the most conservative of each was
implemented for that portion of the project. Interagency involvement was seen as beneficial, as it
raised new issues for consideration, fostered discussion of previously established policies, and
generally resulted in a more robust diving operation. A beneficial discussion with the NOAA Dive
Officer and SRC was initiated and is ongoing.
Diving efforts on this project have resulted in a wide variety of end products. These include, in part:
active criminal investigative work leading to the recovery of objects illegally removed from some of
the sites, survey work, maps, interpretative brochures and movie presentations, emplacement of
mooring blocks and sub-surface navigational guidelines for use by the general public as sites are
opened for visitation, water chemistry analysis, photographic and videographic libraries, resource
assessments and baseline data collection and other site documentation (Chenoweth et al., 2004; Anon
2004, 2006).
152
Finally, these projects provide a model in which active research divers may utilize both open-circuit
and closed-circuit modes of helium-based diving operations that may be exported to other agencies
and projects.
Acknowledgments
I would like to thank Brett Seymour for the use of his photographic images for the 2007 symposium
presentation. I would also like to thank Larry Murphy, Dave Conlin and Brett Seymour for their
reviews of the draft of this paper. All of these individuals are employees of the Submerged Resources
Center, National Park Service, Santa Fe, NM.
References
Anon. DCIEM Diving Manual. Universal Dive Techtronics, Inc. Toronto, ON, Canada, 1995: Table
7.
Anon. U.S. Navy Diving Manual. Revision 4, Volume 3, 1999.
Anon. IAND/IANTD, Inc. Standards. IANTD, Miami Shores, FL, 2003a: 56-57, 62-63, 67.
Anon. A Factory Underwater. National Park Service, Lake Mead National Recreational Area, 2006; 1
pp.
Chenoweth B, Pepito R, Warshefski G, Conlin D. Lake Mead's Cold War Legacy: The Overton B-29.
National Park Service. Santa Fe, NM: Submerged Resources Center, 2004; 6 pp.
Conlin DL, Russell MMA. Archaeology of a Naval Battlefield: HL Hunley and USS Housatonic. Int
J Naut Archaeol, 2006: 35(1): 20-40.
Delgado JP, Lenihan DJ, Murphy LE. The Archeology of the Atomic Bomb: A Submerged Cultural
Resources Assessment of the Sunken Fleet of Operation Crossroads at Bikini and Kwajalein Atoll
Lagoons. Submerged Cultural Resources Center Professional Paper No. 37. Santa Fe, NM: SCRU,
National Park Service, 1991; 208 pp.
Lenihan DJ, ed. Submerged Cultural Resources Study: Isle Royale National Park Service. Submerged
Cultural Resources Center Professional Paper No. 8. Santa Fe, NM: SCRU, National Park Service,
1987.
Lenihan DJ, Delgado JP, Dickinson B, Cummins G, Henderson S, Martinez DA, Murphy LE.
Submerged Cultural Resources Study: USS Arizona Memorial and Pearl Harbor National Historic
Landmark. Submerged Cultural Resources Center Professional Paper No. 23. Santa Fe, NM: SCRU,
National Park Service, 1989; 192 pp.
Lenihan DJ. Submerged: Adventures of America's Most Elite Underwater Archeology Team. New
York, NY: Newmarket Press, 2002: 44-53, 233-240.
Mount T. Technical Diver Encyclopedia. Miami Shores, FL: IANTD, 1998: 244-270.
153
Murphy LE. 8So19: Physiological, Methodological, and Technological Aspects of the Excavation at
Warm Mineral Springs, Florida. In: Arnold JB III, ed. Beneath the Waters of Time: The Proceedings
of the Ninth Conference on Underwater Archaeology. Austin, TX: Texas Antiquities Committee
Publication No. 6, 1978: 123-128.
Murphy LE, ed. Submerged Cultural Resources Assessment: Dry Tortugas National Park. Submerged
Resources Center Professional Papers No. 13. Santa Fe, NM: US National Park Service, 1993.
Murphy LE, Lenihan DJ, Amer CF, Russell MA, Neyland RS, Wills R, Harris S, Askins A, Smith TJ,
Shope SM. H.L. Hunley Site Assessment. Submerged Cultural Resources Unit Professional Paper
No. 62. Santa Fe, NM: SCRU, National Park Service, 1998; 198 pp.
Murphy LE, Seymour BT. Submerged Resources Center Safe Practices Manual: Lake Mead NRA.
Santa Fe, NM: SRC, National Park Service, 2003, 2004, 2005, 2006.
154
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Diving at Extreme Altitude: Dive Planning and Execution During
the 2006 High Lakes Science Expedition
Robert Morris2, Randy Berthold1*, and Nathalie Cabrol 1,2
1
NASA Ames Research Center, Moffett Field, CA 94035, USA; randall.w.berthold@nasa.gov
SETI Institute Carl Sagan Center, 515 N. Whisman Road, Mountain View, CA 94043, USA
2
Abstract
The NASA Ames Diving Safety Office supported the successful diving operations of the 2006
High Lakes Project, HLP, in the 19,400 ft (5,913 m) crater lake of the Licancabur volcano in
Bolivia. The HLP explores the limits of life in some of the highest lakes in the world located in
the Andes at elevations up to 20,290'. The unique extreme diving environment required the
development of new sets of diving standards, techniques, and technical diving protocols, early
literature surveys having established the limitation of existing recreational, commercial, and
military guidelines. Here we document the process of developing and executing a dive plan that
enabled the dive team to achieve its scientific objectives while adhering to a high standard of
safety. Key elements of the dive plan included use of closed-circuit oxygen rebreathers to
mitigate combined altitude- and decompression-related risks, derivation of new dive tables, a
novel diver suspension system, and definition of protocols for a broad spectrum of dive-related
activities and contingencies. The elements of the dive plan were tested individually and in
combination, to the extent possible. A description of the two successful dives is given along
with a summary of lessons learned and recommendations for future expeditions.
Introduction
In November of 2006 the High Lakes Project (HLP) mounted an expedition to the Andes mountains
of southern Bolivia that included diving operations in the 19,400 ft (5,913 m) crater lake of the
volcano Licancabur. The team's objectives were to retrieve biological and sediment samples and to
document the unique underwater habitat. Since the remote location and extreme altitude of the dive
site posed unique safety challenges, a highly specialized dive plan was developed in order to meet
those challenges and maintain a high standard of safety.
Background
Funded by the NASA Astrobiology Institute (NAI), the HLP has made annual expeditions to the
region since 2002. Its purpose is to explore the limits of life in some of the highest lakes in the world
located at altitudes up to 20,290'. The project focuses on characterizing lake habitats, the organisms
that populate them, and the short and long-term effects on these organisms of the high ultraviolet
(UV) radiation found only at these altitudes and latitudes. Understanding how life has adapted to cope
with extreme environmental stresses will give scientists insight into both the evolution of life on early
Earth and the possibilities for life on other planets. Most specifically, the lakes in this region are of
unparalleled interest as Earth-analogs to the ancient lakes thought to have existed on early Mars, c.a.
3.5 billion years ago. The high UV flux (up to 216% that at sea level [Cabrol et al., 2007]), dramatic
temperature swings, aridity, low oxygen, low pressure, and volcanic geology combine to give us what
is perhaps the closest approximation on Earth to conditions on early Mars. Learning about the limits
of life, its survival strategies, and its bio- and geosignatures is a critical step in preparing the next
155
generation of NASA planetary surface missions which will be searching for evidence of past or
present life on Mars. Since conditions on Mars are thought to have changed rapidly, the HLP also
seeks to understand the effects of rapid climate change on life and habitats and to identify adaptations
that have aided survival in the face of such change.
So far, the HLP has investigated a dozen lakes at altitudes above 14,240 ft (4,340 m). The highest of
these is located in the crater of the volcano Licancabur at 19,400 ft (5,913 m). Based on a shore
sampling from 2005 (the lake was covered by 18 ft [5.5 m] of ice) it was determined that 12% of the
microscopic organisms isolated are currently not classified at the phylum level, and 70% are not
classified at the level of genus [Cabrol et al., 2007]. Free diving efforts in 2003 and 2004 yielded
some samples from the lake bottom, but divers tired quickly and were unable to obtain sufficient
samples given the diversity of the ecosystem they discovered there.
The diversity and uniqueness of the ecosystem encountered in the summit lake provided justification
for further diving efforts. It was felt that divers in the water could much more effectively identify and
sample interesting locations than could be done by remote sampling from the surface.
Diving Objectives
The diving objectives were: 1) to sample the diversity of organisms and sediments in the lake and 2)
to obtain high-quality underwater video and image documentation of a) the diversity and extent of the
lake's ecosystem (habitat and life), b) the sampling sites in order to provide context, c) the sampling
methodology, and d) of the team's underwater activity in general. Required supporting data for each
sample consisted of its GPS latitude and longitude, its depth, the time at which it was taken, and a
contextual image or video of the sample site prior to collection. The final sample set was to include
samples of each visually discernible type of bottom material and samples collected at depth intervals
of 50 cm over the full depth range of the lake.
Diving Environment
The physical parameters of the dive site are listed in Table 1. Licancabur's summit crater is a rocky
amphitheater roughly a quarter of a mile across with slopes of 25-35 degrees running down to a small
(300 ft x 200 ft) lake at its center. Since there is precious little flat area inside the crater, camp was
placed on the more spacious North shoulder of the volcano at 19,300 ft (5,883 m). Each of the four
work days at the lake required the team to hike up over the 19,620 ft (5,980 m) rim into the crater and
then back out and down at the end of the day. Weather inside the crater is usually sunny and calm
from morning through midday with winds increasing after about 1400. Although precipitation,
already scarce in the region (≤100 mm⋅y-1), is rare in November, bad weather can develop rapidly.
The lake itself is exceptionally clear with visibility in excess of 40 ft (12 m) with the bottom
undisturbed. The bottom material consists of rocks, sand, and a soupy muck that reduces visibility to
near zero once disturbed. Ice cover has varied each year ranging from none in 2003 and 2006 to 18 ft
(5.5 m) covering the entire lake in 2005. The dive staging area was situated along the northern
shoreline where there was sufficient flat area for a small tent and diving equipment. Figure 1 shows
the staging area and lake bottom.
Dive Planning
Dive planning was driven jointly by science objectives and the need to mitigate risk. In accordance
with American Academy of Underwater Sciences standards the final plan was required to pass the
156
scrutiny of a Diving Control Board, which included personnel from five NASA offices (medical,
safety, diving safety, management, and science). Early consultations with technical diving experts
identified closed-circuit oxygen rebreathers as the safest option for shallow diving at extreme
altitudes. This was the starting point from which the rest of the plan evolved through an iterative
process of testing, revision, and review. This section describes the elements of the final plan along
with the rationale for important decisions along the way. In its final form, the plan called for two
dives, each to be carried out by three divers. The purpose of the first dive was to perform exploratory
sampling of bottom materials and to obtain video- and photo-documentation of the underwater
environment. The purpose of the second dive was to obtain bottom samples at 50 cm depth intervals.
Table 1. Summary of the diving
environment
altitude
19,400 ft
surface pressure
0.48 atm
air temperature
~5°C at midday
wind speed
<20 mph in crater
water
temperature
4 C at surface
2°C at 2 m depth
lake dimensions
300 × 200 ft.
max. depth
16 ft
pressure at
max. depth
0.91 atm
visibility
1-40 ft
bottom
rocks, sediments
organic; muck
ice cover
none in 2006
shoreline
talus, scree, sand
lake access
easy from North,
South, and West
shores
Figure 1. Clockwise from upper left: location of the Licancabur
volcano, Licancabur as seen from Juriquès, a neighboring peak,
the underwater environment of the summit lake (photo Clayton
Woosley), and the dive staging
Special Challenges of Diving at Extreme Altitude
Absence of Safety Guidelines for Extreme Altitudes
Early surveys of the available literature revealed the absence of tested dive tables giving safe
decompression and ascent rate limits above 14,000 ft (4,267 m) [Heine et al., 2004]. Although the
NASA Diving Safety Office extrapolated its own tables from US Navy tables, these were untested
and were therefore treated with caution. Following the standard practice of rounding up to the nearest
1,000 ft (the daily high point of the team was 19,620 ft [5,980 m]) a maximum safe ascent rate of 14
ft⋅min-1 (4.3 m⋅min-1) was calculated for altitude of 20,000 ft (6,096 m).
157
Maintaining Safe Ascent Rates
The calculated safe ascent rate of 14 ft/min (4.3 m/min) is extremely slow and it was determined
through pool trials that most divers would be unable to reliably maintain it using fins and buoyancy
control devices. To address this problem along with others, a novel diver suspension system was
developed and tested, enabling divers to precisely control their depth and maintain safe ascent rates.
Combined Risk of Diving- and Altitude-related Illnesses
How does extreme altitude affect the incidence of diving illnesses? How does diving affect the
incidence of altitude illnesses? The absence of statistics makes these questions difficult to answer.
Because of the uncertainties, steps were taken to eliminate mechanisms of injury and otherwise
mitigate risks wherever possible. Perhaps the most significant such step was the decision to dive on
closed-circuit oxygen, which prevented the loading of tissues with inert gases while simultaneously
mitigating risks stemming from hypoxia at high altitudes.
While supplemental oxygen can prevent and/or relieve the symptoms of acute mountain sickness
(AMS), symptoms may then be exacerbated by its removal [Houston, 1980]. Concerns about diving
on pure oxygen followed by a period of oxygen starvation upon exiting the water led to the planning
of a transition period following each dive, during which divers rested on medical oxygen while being
closely monitored for signs and symptoms of illness.
Hypoxia
Almost by definition, people become hypoxic at extreme altitudes. Although they may experience no
illness, individuals at high altitudes show diminished physical and mental performance. Since diving
can be physically demanding and requires alertness for safety, the effects of hypoxia pose a serious
danger that increases with altitude. The HLP dive team addressed this danger both by using 100%
oxygen for diving and by carefully scripting and rehearsing all diving-related activities.
Health and Safety: Concerns, Prevention, and Response
Health and safety planning consisted of anticipating potential problems and defining preventative
measures, in-the-field responses, and a plan for evacuation to advanced medical facilities. Baseline
prevention consisted of rigorous medical screening prior to departure and daily health monitoring in
the field for early detection of developing problems before they could become emergencies. Daily
medical statistics included blood oxygen saturation, blood pressure, heart and respiratory rates,
weight, and responses to subjective questions.
Altitude Illnesses
Altitude illnesses comprising AMS, High Altitude Pulmonary Edema (HAPE), and High Altitude
Cerebral Edema (HACE) are caused primarily by hypoxia, though the physiology behind individual
susceptibility is not well-understood [Wilkerson, 1992]. HAPE and HACE, both life-threatening
conditions, can be thought of as advanced forms of AMS. The best prevention is proper
acclimatization and the only effective treatment for advanced illness is immediate descent or
equivalent re-pressurization.
Primary preventative measures included a rigorous acclimatization regimen and close monitoring of
all team members for signs and symptoms. Secondary measures included avoidance of overexertion
by climbing slowly, maintaining a high-carbohydrate diet while on the mountain, and adequate
hydration. The team developed an acclimatization program similar to those implemented by other
mountaineering expeditions, including a multi-stage ascent over 10 days, modified diet, and
medication with acetazolamide.
158
The previously mentioned concerns regarding re-exposure to the oxygen-depleted atmosphere
following a dive on 100% oxygen were addressed by introducing an oxygen transition period into the
dive plan. Divers exiting the water would transition from their rebreathers onto medical oxygen,
which could then be regulated to slowly reintroduce them to the hypoxic conditions at the summit.
During this period divers would also be closely monitored (visual/verbal, blood oxygen saturation,
and vitals) for any signs of deteriorating health.
Individuals experiencing mild to moderate symptoms of AMS would make an accompanied descent
to base camp, using supplemental oxygen if needed. In the event of advanced altitude illness, the sick
individual would either walk assisted or be transported by litter down to the base of the mountain at
14,800 ft (4,511 m) and then evacuated by vehicle from there. In the event of unforeseen delays in
transporting a seriously injured party, the team carried a Gamov bag which could be pressurized to
roughly the equivalent of a 7,000 ft (2,134 m) descent and then carried down.
Accidents Due to Hypoxia
At altitudes approaching 20,000 ft (6,096 m), some degree of hypoxia is unavoidable. The
accompanying decline in physical and mental performance increases the risk of accidents resulting
from fatigue, inattention, and poor judgment. Prevention consisted of highly detailed written
procedures covering every aspect of diving operations, well-rehearsed and familiar to everyone on the
team. In theory, divers breathing pure oxygen could not be hypoxic and were therefore at no risk of
its effects for the duration of the dive.
Diving Illnesses and Injuries
When diving on closed-circuit oxygen, nitrogen and other inert gases are purged from the breathing
loop save in small amounts. Loading of tissues with inert gases is therefore not possible, eliminating
one of the principal mechanisms behind DCS. An ambient absolute pressure of less than 1 atm at the
deepest point in the lake (16 ft [4.9 m]) meant that the risk of Central Nervous System (CNS) oxygen
toxicity was also small. The risk of pulmonary oxygen toxicity, which occurs with long-term
exposure at relatively low pressures, was mitigated by limiting dive duration to 45 min or less.
The main concern among diving illnesses was arterial gas embolism (AGE) due to rapid ascent. As
previously mentioned, maintaining the calculated safe ascent rate of 14 ft⋅min-1 (4.3 m⋅min-1) was
problematic. This and other problems led to the development of a diver suspension system that both
enabled precise depth control and enabled the diver to return to depth quickly even in the event of
accidental release of a weight belt. Extensive pool practice with these systems habituated divers to
ascending at a rate well within the safe limit.
Using rebreathers introduced the risks associated with failure or improper use of equipment, including
hypoxia, hypercarbia, hypocarbia, and chemical injury. It was decided that, with sufficient training,
careful maintenance, careful testing of oxygen sources, and careful handling and storage of CO2
absorbent, these risks were small in comparison to the margin of safety gained.
Divers were to work closely together in groups of two or three. The level of coordination their
activities required would make it obvious if one of the divers became incapacitated or was
experiencing serious symptoms. In such an event, able divers would commence a rehearsed rescue
procedure to bring the injured diver to the surface and return to the staging area. Once there, the dive
master would begin treatment and would decide whether evacuation was necessary.
In the event of an AGE, the patient would be administered supplemental oxygen and evacuated
immediately. Likewise, field treatment for hypoxia and hypercarbia consisted of supplemental
oxygen. A diver with hypocarbia would be calmed and encouraged to breathe slowly. Chemical
159
injuries would be treated by having the diver rinse his/her mouth with water and then drink water to
dilute stomach contents. Squeezes, including those caused by middle ear oxygen absorption would be
treated with a combination of decongestants and valsalva.
Exposure and Hypothermia
Hypothermia was a risk for all personnel on the mountain, but diver's exposure to near-freezing water
followed by exposure to freezing air and potentially high winds on the surface made the risk
particularly serious for them. Moreover, the tendency for people to become dehydrated and
hypoglycemic at high altitude added to the concern since both conditions increase a person's
susceptibility. Risk was mitigated by using dry suits and sufficiently warm undergarments, limiting
dive times to prevent overexposure, adequate food and water, and post-dive rewarming in a tent with
bottles of warm water.
Emergencies Requiring Prompt Advanced Medical Care
The remote location of the dive site limited our access to advanced medical care. It also meant that
self-rescue was the only option for the team in the event of an emergency. It was estimated that
transporting an injured party down the mountain and then to the nearest hospital and hyperbaric
facility would require a minimum of eight hours. Transportation to the hospital was complicated by
the need to cross the border from Bolivia into Chile. This minimum time estimate assumed excellent
logistical support from below, good communications with base camp, and a litter or backboard on
hand at the site of the injury. Had any of these assumptions proved false, the actual evacuation time
would likely have been significantly longer.
Equipment
Each diver was equipped with a rebreather unit, dry suit, undergarment, neoprene hood, gloves, mask,
fins, and a suspension system comprising a buoy, tether, and weight belt. Additional equipment
pertaining to sampling and documentation was distributed to divers according to their roles and
included mesh bags containing sample bottles, a digital still camera, a video camera, a GPS unit, and
a tape measure. Safety equipment at the staging area included a throw line, medical oxygen, a medical
kit, a Gamov bag, and a tent and stove. A complete check list of all diving-related equipment was
created and used prior to equipment portage and again upon arrival at the dive site.
Aqualung CODE Rebreather
The CODE (Compact Oxygen Diving Equipment) rebreather is an extremely simple, compact, light,
and durable closed-circuit oxygen rebreather unit used mainly for military applications. Oxygen is
injected into the counter lung by a demand valve and carbon dioxide is removed from the system by a
canister containing a soda lime absorbent. Prior to use, the breathing loop must be purged of gases
other than oxygen and carbon dioxide. Closed-circuit oxygen rebreathers have a distinguished history
in mountaineering and were used with great success by Hillary's 1953 Everest ascent team (Hunt,
1954).
Accurate knowledge of one's breathing gas is fundamental to safe use of any closed-circuit system.
To this end, a small certified bottle of oxygen was obtained in advance from a facility in Santiago,
Chile. The certified bottle was used to calibrate a hand-held gas analyzer with which larger volumes
of gas obtained near the dive site were tested.
Although the CODE was a major asset to diving safety, there were some disadvantages in using it.
Since the system is completely closed there are no bubbles, making divers difficult to locate from the
surface or in poor visibility. Also, the oxygen bottles contained only a small volume of gas, 0.6
L/2030 psi, so a separate air cylinder would be required in order to use a buoyancy compensator (BC)
160
for depth control. These issues and others unrelated to the CODE were addressed by the diver
suspension systems described below.
Table 2. Summary of rationale for CODE rebreathers
Advantages
• mitigates high-altitude hypoxia
• avoids loading of bodily tissues with inert gases
• oxygen is indicated for any diving or altitude illnesses
• compact and light-weight
Disadvantages
• not conducive to BC usage (insufficient gas for BC inflation)
• no bubbles – lost divers may be difficult to find
• cannot surface and talk without re-purging the rigs
• consequences of failure or improper use of equipment
Diver Suspension System
The difficulty of maintaining safe ascent rates and the need to hover above the soft lake bottom while
working motivated the development of a novel Diver Suspension System (DSS), which allowed the
diver to hang suspended at any depth (Figure 2). The DSS consisted of a buoy, adjustable tether, and
modified weight belt connected together with locking aluminum carabiners. Buoys were constructed
using inner tubes and nylon chord, while tethers consisted of 20 ft (6 m) off-the-shelf cam-buckle
straps. A taught line running from hip to hip on the weight belt allowed the diver to adjust his roll
angle from left to right. Divers were weighted to be about 10 lbs heavy, requiring between 40 and 60
lbs of lead per diver. Having weights evenly distributed around the diver's waist made it easy to
maintain the desired roll angle.
In addition to enabling precise depth control, the DSS addressed several other safety concerns and
technical challenges. The overhead buoy served as a marker for tracking divers from the shore, a
platform for GPS tracking of sampling locations, a reference for depth measurements, an attachment
point for equipment, and an emergency float. The configuration of weight belt and tether served to
mitigate the risk of rapid ascent and AGE following accidental release of a weight belt. Although
releasing the belt would detach the diver from the DSS, the belt and tether would remain connected
and would still provide a viable up/down line. In the event of an accidental release, the diver could
grab the tether to stop an uncontrolled ascent, or use it to quickly return to depth afterward. These
advantages were felt to strongly outweigh the disadvantages listed in Table 3.
Table 3. Summary of rationale for diver suspension systems
Advantages
• precise control of depth and ascent rate
• ability to hover over work site with no up/down drift
• easy tracking of divers from surface
• mitigated danger of losing a weight belt
• enabled GPS tracking of sampling locations
• provided a place to clip and keep equipment
• provided a reference for depth measurements
• buoy could serve as an emergency float
• materials are cheap and easy to obtain almost anywhere
Disadvantages
• some added complexity over a BC and/or fixed ascent line
• risk of tangling
• inability to dive under ice
• transportation of extra weights to dive site
161
Using the DSS required some training for efficiency, but was otherwise straight forward. Downward
depth adjustments could be made by pushing the button on the cam buckle and allowing the tether to
extend. This could be done easily with one hand while maintaining a downward-facing orientation
and working with the other hand. Upward adjustments were made by pulling on the tail of the tether
and were facilitated by the 2:1 mechanical advantage inherent in the configuration (Figure 2).
Because of the relative ease of moving downward, dive profiles were planned from shallow to deep
water. At the end of a dive, divers would slowly ascend and would release their weight belts when
they reached the surface. The entire system, with the belt hanging just underneath the buoy, could
then be either towed back to shore or left for retrieval.
Figure 2. Diver suspension system (photos Clayton Woosley and Christian Tambley). All photo credits
The High-Lakes Project: NAI/SETI CSC/NASA Ames Research Center
Diving Safety Limits
Quantifiable diving limits for the HLP dive team included ascent rates (Table 4), oxygen diving tables
(Table 5), and limits for safe usage of the rebreathers. Ascent rates were extrapolated from existing
navy tables using the formula
Altitude Rate = Sea Level Rate × exp(-0.0381 × altitude/1000ft).
Table 5 gives oxygen exposure limits for single depth dives. Although these limits prove safe for the
vast majority of divers, certain individuals succumb to oxygen toxicity even when maintaining them
[U.S. Department of the Navy, 2005]. Because of the remote setting and extreme altitude the HLP
planned very conservative dive profiles, limiting dive times to 45 minutes or less.
Safe use of rebreathers required aviation-grade oxygen (99.5% O2) in order to avoid build up of
impurities in the counter lung leading to hypoxia. The brand of CO2 absorbent used by the HLP was
selected mainly for its specified storage temperature range (-30°C to 50°C). Absorbent materials
exposed to temperatures outside of the recommended range should not be used. The lifetime of a
162
single canister was estimated to be roughly 60-90 minutes depending on work load and diving
environment.
Sampling Technique
The dive team consisted of a lead diver responsible for collecting samples from the lake bottom, a
sampling assistant whose job was to photograph both the sample sites and (numbered) sample bottles
just prior to collection, and a third diver using a video camera to document the dive and the
environment. A GPS unit set in tracking mode was placed in a waterproof container and clipped to
the lead diver's buoy prior to diving. By synchronizing the internal clocks of the GPS and the
assistant's digital camera, one could later determine where the lead diver was when a particular photo
was taken (digital image files contain a time stamp). The latitude and longitude of each sample site
was obtained using this technique.
Table 4. Ascent rate limits as a function of
altitude.
Altitude
(ft)
Pressure
(atm)
Ascent Rate
(ffw/min)
0
1.000
30.9
1,000
0.963
29.7
2,000
0.927
28.6
3,000
0.892
27.6
4,000
0.859
26.5
5,000
0.827
25.5
6,000
0.796
24.6
7,000
0.766
23.7
8,000
0.737
22.8
9,000
0.710
21.9
10,000
0.683
21.1
11,000
0.658
20.3
12,000
0.633
19.6
13,000
0.609
18.8
14,000
0.587
18.1
15,000
0.565
17.4
16,000
0.544
16.8
17,000
0.523
16.2
18,000
0.504
15.6
19,000
0.485
15.0
20,000
0.467
14.4
Table 5. Single depth oxygen exposure limits
per 24 hour period.
Depth
Maximum Oxygen Time
25 fsw
240 min
30 fsw
80 min
35 fsw
25 min
40 fsw
15 min
50 fsw
5 min
Determining the depth at which a given sample was taken
was done differently for the first and second dives. The
approximate depths of sample sites from the first dive
were recovered by correlating the GPS position with a
bathymetric map of the lake obtained via sonar prior to
diving. On the second dive, a pre-marked tape measure
was attached to the overhead buoy and was used to
sample at precise depth intervals.
To obtain a sample, the lead diver would indicate the
desired location by pointing and removing an empty
sample bottle from his/her bag. After the sampling
assistant had photographed the number on the top of the
bottle and the undisturbed site, the lead diver would turn
the bottle upside down and carefully remove the top,
preventing the air inside from escaping. The bottle was
then scooped down into the bottom material and turned
upright, effectively sucking the material into the bottle.
The bottle was then recapped and returned to the bag.
163
Procedures
Highly detailed written procedures and checklists were developed as the dive plan evolved. These
documents served as scripts during training and then again as references in the field to prevent errors
and oversights in the hypoxic high-altitude environment. Table 6 illustrates the level of detail that was
characteristic of procedural planning. Written procedures covered dive preparation, activities in the
water, post-dive activities, emergencies and other contingencies, and some aspects of operations
management.
Training
All dive team members underwent 32 hours of training on the CODE rebreather under a USIA diving
instructor. All team members were certified in medical oxygen administration.
Extensive pool training was an important factor in both dive plan development and diver safety. The
pool served as a test bed during development of procedures, techniques and equipment. Integrated
tests were done in the form of simulations of both planned dives.
Table 6. An example of the level of procedural detail typical of dive planning
Dive Preparation
1. Prepare tent with bags/blankets, warm clothing, and thermos
2. Prepare diver suspension systems
(a) assemble buoy (inner tube, cord, and carabiner)
(b) attach knotted yellow line with a carabiner
(c) set up cam strap with locking carabiner and attach to buoy
(d) set up weight belt (marked weights, front carabiner, hip-to-hip hang line)
3. Pre-dive CODE units (refer to CODE pre-dive checklist)
4. Prepare medical O2 cylinders at entry/exit point
5. Prepare cameras, GPS, sonar, and sample bottles (10 for each dive)
(a) record times of camera and GPS clocks and/or image the GPS clock with the still camera
(b) set GPS to tracking mode
(c) seal GPS unit in a dry bag or dry goods container
6. Attach GPS, sonar, cameras and sampling bottles to buoys and place buoys at entry/exit point
7. Diver's don dry suits, CODEs, weight belts, gloves, hoods, and masks
8. Dive Master dons dry suit
9. Dive Master performs final gear checks:
(a) layering (vest, CODE, weight belt)
(b) check for good vacuum and good seal at canister insertion
(c) straps (have diver rotate)
(d) mouthpiece (orientation, hose attachment, zip tie, head band)
10. Dive Master turns on O2, checks, and records pressure
11. Dive Master walks divers through purge protocol, ensuring proper execution
12. Divers breathe O2 for two minutes before entering water
13. Dive Master checks each diver's alertness and either gives or denies permission to enter water
Dive Execution
The only major setback suffered by the expedition occurred one day before departure for South
America when a key member of the team was medically disqualified from both diving and climbing
above base camp. The approved dive plan included contingency plans for such an event and called for
164
a redistribution of roles and responsibilities on the mountain. This in turn required additional training
and pool simulations to familiarize team members with the changes.
The team was fortunate to find access to pool facilities in Chile where training was completed. Three
20-minute dive simulations were performed using the same equipment and procedures used for the
summit dives. Changes to the original dive plan included reassignment of roles and responsibilities
and assignment of an assistant to each diver to help him/her get suited up and in the water.
A full evacuation plan was established in the field prior to deployment up the mountain. The plan was
based upon experience from previous non-diving expeditions to the area, advice from local points of
contact including the Bolivian National Park Service (SERNAP) and the contracted in-country
logistics manager. The plan included the identification of local and regional medical services. With
the help of the Divers Alert Network (DAN), provisions were made for specialty medical care for
divers, including chamber identification and access. General transportation logistics were arranged,
including provisions for all-wheel drive vehicles that could accommodate a stretcher. A companion
communications plan was framed utilizing anticipated local resources in the region.
The evacuation plan was revised and improved to account for unanticipated complications involving
access to vehicles and nonexistent services and capabilities, e.g. access to advanced medical care
24/7, and communication equipment. Unanticipated were the problems with crossing international
borders by certain types of licensed vehicles. The nearest medical facility was a local clinic that
operated on an irregular business hours basis. However, a more substantial hospital 60 miles (100 km)
away would have been needed to treat any serious or off-hours medical emergencies. Chamber access
turned out to be via a local mining company, which operated on a limited schedule. Contact was
never established with the local diving medical officer referenced by DAN. The communications plan
integrated the use of identified local radio, which proved to be unreliable. Our satellite phone proved
to be the only reliable means of communications.
Equipment was assembled at base camp and prepared for portage to the dive site. The majority of the
load was distributed over eight hired porters who made daily trips to the summit, pre-staging much of
the equipment before the team's ascent. The remainder, including CO2 absorbent, cameras, and other
sensitive electronic equipment, was carried by expedition team members on their ascent to the
summit. Absorbent was carried by diving team members who ensured that it was never exposed to
temperatures below the specified threshold of -30°C.
In accordance with the acclimatization plan, the team ascended from base camp at 14,200 ft (4,328 m)
to a mid camp at 17,700 ft (5,395 m) on November 16. On November 17 they ascended to the summit
camp at 19,300 ft (5,883 m). November 18 was spent setting up equipment and doing science in the
crater. The first and second dives were carried out on November 19 and 20, respectively. The team
descended to base camp on November 21.
Dive Day I
A morning medical assessment indicated that all divers were fit to dive. The team arrived in the crater
at 0900. After staging equipment at the entry point, divers suited up and entered the water at
approximately 1030. In the first few minutes of the dive it became obvious that one of the divers was
too buoyant. The divers returned to shore, made the necessary adjustment, and re-entered the water.
Sampling and documentation proceeded as planned, although the team wasn't able to cover much
ground before succumbing to cold. The maximum depth was approximately 2 m and the total bottom
time was 31 min. Average oxygen consumption during the dive was 58 bar from a 0.6 L bottle.
165
Dive Day II
The team arrived in the crater at 0900. Divers entered the water at approximately 1100. At the
beginning of the dive the lead diver experienced some vertigo and mild nausea due to unequal
vestibular stimulation by frigid water after rolling on his side. Samples were taken at depth intervals
of 0.5 m starting at 1.5 m and continuing to the deepest point reached at 4.5 m. Total bottom time was
25 min. Diver's surfaced near the center of the lake, detached from the DSSs and towed them back to
shore. All divers were fatigued and cold and two experienced mild headaches during the transition to
medical oxygen. Divers recouped in the tent for roughly one hour under the supervision of the Dive
Master. Average medical statistics for the oxygen transition and recuperation period are shown in
Table 7. Average oxygen consumption per unit of dive time was similar to day one (actual data was
not logged due to divemaster tasking overload, a lesson learned) and average post-dive medical
oxygen consumption was roughly two D-size cylinders per diver.
Table 7. Average post-dive medical statistics for the three divers in the aftermath of dive II. Time
begins when all divers are out of dry suits, in the tent, and on medical O2. Average blood pressure (BP)
is given as average systolic pressure over average diastolic pressure. Respiration rates were determined
manually, BP and heart rate with a digital wrist BP cuff, and blood oxygen saturation with a fingertip
pulse oximeter. Note the dramatic drop in oxygen saturation following the removal of medical oxygen.
Time
(min)
Blood Oxygen
(%)
Heart Rate
(bpm)
Blood Pressure
(mm Hg)
Resp. Rate
(bpm)
0
96
101
120/89
17
18
Off Medical O2
20
80
99
137/86
19
33
80
96
124/83
19
53
78
104
127/86
19
Conclusions
The scientific research requirements to manually obtain samples at this unique high-altitude site
required the development of unique diving equipment, training and the extrapolation of existing
tables. A number of lessons were learned through the experiences of the High Lakes Project, which
led to the following recommendations for projects considering similar constraints and environments.
Diving teams which must operate in such extreme environments require specialized training in
disciplines outside of normal diver training, e.g., high altitude acclimatization techniques, advanced
medical care, general survival techniques for long term field site occupation and the development of
an extensive set of plans and procedures.
All the elements of the dive plan (equipment, procedures, and techniques) must be tested in as high a
fidelity simulation as possible. Contingency operations must be planned for, options defined, tested
and exercised for each element. The plans must consider the remoteness, time to obtain or access
advanced medical care and provide for self-sufficiency as much as possible. Assumptions must not be
made for local support availability. Communications plans must be exercised in advance. Special
consideration must be made to assure that facilities necessary for emergency operations are contacted
in advance and all aspects of availability are confirmed: hours of operation, staff availability,
limitations of services, transport time, etc.
166
The success of any expedition in a remote location depends on the team's training and cross training.
The expedition must have sufficient depth to absorb the loss of a member and personnel must be
ready to perform functions outside of their normal skill set. This is especially important when
considering the multiple stresses experienced by team members diving in an extreme environment.
Even the most well trained team, armed with adequate planning must consider the variability of
individual responses to the environment. Thus far there are no models that can predict how an
individual will respond physically or psychologically to extreme environments. This is especially true
at 20,000 ft (6,096 m).
Finally, no expedition should rely solely only on its internal knowledge base. A review of all the
elements, science requirements as they drive data gathering, training, planning, and logistics, by
external experts can provide mission-critical insight.
Acknowledgments
The HLP project is funded by the NASA Astrobiology Institute/SETI CSC lead team. Additional
funding, equipment and technical support which made the 2006 expedition possible were provided by
Aqualung, Specialized bicycles, and CHEP Chile. We also want to express our gratitude to:
Dominique Sumian, Eric Heid, and Ariadne Delon Scott. Special thanks go to the local institutions
supporting the deployment of HLP in the field: SERNAP (Bolivian National Parks) and the
Universidad Catolica del Norte in Antofagasta, Chile
References
Cabrol NA, Minkley EG Jr, Yu Y, Grin EA, Woosley C, Morris RL. Unraveling Life's Diversity in
Earth's Highest Volcanic Lake, Bioastronomy Conference, Puerto Rico, 2007.
Heine J, Bookspan J, Oliver P. NAUI Master SCUBA Diver, National Association of Underwater
Instructors, 2004.
Houston CS. Going High: The Story of Man and Altitude. New York, NY: American Alpine Club,
1980: 101.
Hunt J. The Conquest of Everest, Appendix V: Oxygen Equipment. New York, NY: EP Dutton and
Co., 1954.
US Department of the Navy, U.S. Navy Dive Manual, Vol. 4, Rev. 5, Washington DC: Naval Sea
Systems Command, 2005.
Wilkerson J, ed. Medicine for Mountaineering, 4th edition. Seattle, WA: The Mountaineers, 1992.
167
168
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Long Term Monitoring of a Deep-water Coral Reef: Effects
of Bottom Trawling
John K. Reed1, Christopher C. Koenig2, Andrew N. Shepard3, R. Grant Gilmore, Jr.4
1
Harbor Branch Oceanographic Institution, 5600 US 1 North, Fort Pierce, FL 34946, USA;
jreed@hboi.edu
2
Department of Biological Science, Florida State University, Tallahassee, FL 32306, USA;
koenig@bio.fsu.edu
3
NOAA Undersea Research Center, University of North Carolina at Wilmington, 5600 Marvin Moss
Lane, Wilmington, NC 28409, USA; sheparda@uncw.edu
4
R. Grant Gilmore, Jr., Estuarine, Coastal and Ocean Science, Inc., 5920 1st St. SW, Vero Beach, FL
32968, USA; rggilmorej@aol.com
Abstract
The deep-water Oculina coral reef ecosystem is unique and exists solely off the east coast of
central Florida. Oculina varicosa forms azooxanthellate colonies up to 2 m in diameter which
coalesce into dense thickets on 20-m tall mounds that are thousands of years old. Recently
restored videotapes that were made in the 1970s with the Johnson-Sea-Link submersibles show
large breeding aggregations of grouper associated with the coral habitat. Historical photographic
surveys provide evidence of the status and health of the reefs prior to heavy fishing and trawling
activities of the 1980s and 1990s. Recent quantitative analyses by point count of photographic
images reveal drastic loss of live coral cover between 1975 and present. Submersible and ROV
surveys conducted from 2001 to 2006 suggest that much of the Oculina habitat has been
reduced to rubble by bottom trawling which unfortunately is a trend for deep-water reefs
worldwide. In 1984, the Oculina reefs were the first deep-water coral reefs in the world to be
designated a marine protected area (MPA). Unfortunately, the northern two-thirds of the reef
system remained unprotected and was legally open to bottom trawling until the year 2000 when
the boundaries were expanded to 1029 km2 (300 nm2) from the original 315 km2 (92 nm2).
However, portions of the original reserve are still healthy and signs indicate improving grouper
populations. In 2006, a high resolution multibeam map was completed which details the
hundreds of pinnacles and ridges making up the reef system. Many new reef features were
discovered both inside and outside the designated MPA.
Introduction
Deep-water coral reefs in the United States are formed principally by two framework constructing
species of scleractinian corals, Oculina varicosa and Lophelia pertusa. Deep-water Oculina reefs
grow at depths of 60-100 m and are only known off central eastern Florida, whereas Lophelia reefs in
this region occur at depths of 490-870 m from North Carolina to Florida but are also known
worldwide. Both types of reef form bioherms that are high-relief mounds of unconsolidated sediment
and coral debris and are capped with thickets of the azooxanthellate coral. Both provide essential
habitat for diverse communities of fish and invertebrates (Reed, 2002a,b; Koenig et al., 2005; Reed et
al., 2005). Deep-water coral reefs have recently gained considerable attention as fisheries and oil/gas
production expand to deeper habitats. Unfortunately many deep-water reefs have been severely
impacted by bottom trawling but few have been mapped and little is known regarding the ecology of
these diverse and fragile ecosystems.
169
The majority of deep-water Oculina habitat is known in Florida between 27º30'N (Fort Pierce) and
28º30'N (Cape Canaveral) in a zone 2-6 km wide and paralleling the 80ºW meridian along the shelf
edge break (Figure 1; Avent et al., 1977; Reed, 1980). The coral forms azooxanthellate colonies, 1-2
m in diameter and 1-2 m tall, that coalesce into thicket-like habitats and high-relief bioherms that are
similar in structure to deep-water Lophelia coral reefs (Reed, 2002a,b; Reed et. al., 2005).
The Oculina reefs were the first deep-water coral reefs in the world to be designated as a marine
protected area. The Oculina Habitat Area of Particular Concern (OHAPC) was enacted in 1984 by the
South Atlantic Fishery Management Council (SAFMC) which banned bottom trawling, bottom
longlines, and anchoring in a 315 km2 (92 nm2) area (NOAA, 1982). Historical accounts from
Johnson-Sea-Link submersible dives in the 1970s described large populations of economically
important reef fish including spawning aggregations of grouper associated with the coral habitat
(Gilmore and Jones, 1992; Koenig et al., 2000, 2005; Reed et al., 2005). By the early 1990s, the grouper
and snapper spawning aggregations had been decimated from commercial and recreational fishing
(Koenig et al., 2000, 2005). This stimulated the SAFMC to ban hook-and-line bottom fishing in 1994
to test the effectiveness of a fishery reserve. Unfortunately, the northern two thirds of the Oculina reef
system remained unprotected outside the boundaries of the OHAPC and was legally open to
mechanically destructive activities such as bottom trawling. Bottom trawling within Oculina
ecosystem was primarily for rock shrimp and brown shrimp and this was the primary cause of major
habitat destruction. In 2000, the Oculina HAPC boundaries were expanded to 1029 km2 (300 nm2)
which banned bottom trawling, and the boundaries of the original OHAPC, termed the Oculina
Experimental Closed Area (OECA), continued the additional moratorium on bottom fishing (NOAA,
2000). In addition, the SAFMC has mandated that the rock shrimp industry implement a vessel
monitoring system (VMS) for the fishery to aid in enforcement of the closed Oculina HAPC areas.
Historical photographic records from submersible surveys of the eastern Florida shelf between 1975
and 1977 provide evidence of the status of the deep-water Oculina reefs prior to heavy fishing and
shrimp trawling activities of the 1980s and 1990s (Avent et al., 1977; Avent and Stanton, 1979; Reed,
1980). The Oculina reefs were first discovered during these submersible transects when over 50,000
35-mm photographs were taken. Twenty-five years later, in 2001, portions of these original
photographic transects were resurveyed. In this report two sites are compared, one within and one
outside the boundaries of the original OHAPC. The northern site remained open to bottom trawling
during this time period. Quantitative analyses by point count of video and photographic images from
1970s compared to 2001 allows a direct comparison of changes in percent cover of live Oculina coral
during this 25 year time period. This study has resulted in the restoration of these rare and invaluable
photographic and video images, and will provide marine managers and scientists a quantitative
assessment of the health of these reefs in the 1970s, prior to intense trawling, compared to the same
sites today (Figure 2).
Methods
Historical (1975-1983) Submersible and ROV Surveys
From 1975 to 1977, Dr. Robert M. Avent conducted benthic surveys of the east Florida shelf and
slope, consisting of 12 east-west photographic transects with Harbor Branch Foundation's (Harbor
Branch Oceanographic Institution, HBOI) Johnson-Sea-Link (JSL) I and II submersibles and CORD
Remotely Operated Vehicle (ROV) from Cape Canaveral to Palm Beach (Figure 1; Avent et al.,
1977; Avent and Stanton, 1979; Reed, 1980). The transects were spaced approximately 19 km apart,
extended from depths of 30 m to 300 m, and consisted of 55 submersible dives covering 298 km. It
was during this survey that the live deep-water Oculina banks were first observed and described in
170
detail (Avent et al., 1977; Reed, 1980). Photographs were taken randomly every 1-2 minutes with a 35mm Edgerton-type still camera and flash system (Benthos model 372, Benthos Inc.) that was mounted
29° from vertical providing average frame coverage of 6.3 m2 at 3.0 m height and 2.5 m mid-frame
width. Viewing angle in water was 54° wide and 42° fore-aft. Area was estimated from camera height
and in-water viewing angle of the lens and was verified by flying the sub over a 10-m grid on the bottom.
Photographs were originally analyzed using a microfilm reader with a grid overlay for describing
habitat type, dominant fish and benthic invertebrates. Over one hundred 30-m rolls of 35-mm
Ecktachrome film (or black and white) were used and archived at HBOI. Each photograph has a date
and time stamp in the corner. A software program was used to interpolate the navigation coordinates
throughout each submersible dive. Navigation used LORAN-A which in this region had an accuracy
of ±150-300 m. Coordinates and depth were determined for each photographic image. During each
submersible dive, meticulous written transcripts were compiled by Dr. Avent and research assistants
which described habitat, fish and benthic communities, and physical factors. Ship and submersible
logs recorded coordinates, depth, and other data throughout each dive. Hydrographic and navigation
information were entered into a computer database (unfortunately, the computer tapes are obsolete,
but hard copies and photos are archived at HBOI).
Figure 1. Map of the deep-water Oculina Habitat Area of Particular
Concern (OHAPC) with historical submersible transect lines and
Oculina reef study sites.
171
Figure 2. A. South slope of healthy deep-water Oculina reef (Jeff's Reef, 80 m); B. Peak of trawled Oculina reef.
Some of the original videotapes (3/4 inch and 1/2 inch open reel) and 30-m rolls of 35-mm
Ektachrome film were recently restored and archived. Unfortunately videotapes of that age are prone
to hydrolysis problems. The restoration process stabilizes the polymers of the tape coating. The
restored tapes were recently archived onto Beta SP videotapes and copied to digital video disk
(DVD). Nearly 3,000 35-mm photographs that were over Oculina coral habitat were recently
digitized with a Nikon LS-2000 Coolscan scanner, enhanced in Photoshop®, saved as high resolution
TIFF files (300 dpi, 3.75 mB file), and copied to DVD. An Access database was compiled for the
metadata and list of images for the photographic and video archives. Each photographic image
filename includes JSL dive number and time (e.g., JSL I 0233 T 09-55-42.jpg = JSL I, dive number
233, photo time = 9:55:42 am).
Tethered, mixed-gas dives (lockout) were also made with the Johnson-Sea-Link submersibles from
1976 to 1983 on the deep-water Oculina banks for studies on biodiversity of animals associated with
the coral, coral growth rates, and geology (Reed, 2002a). The scientist-lockout diver used a KirbyMorgan band mask attached to a 30-m umbilical hose which supplied the gas mix (10% oxygen/ 90%
helium) and voice communications from the submersible. During ascent of the submersible after each
dive, decompression stops with oxygen and air breaks were initiated at a depth 58 m in 3-m
increments and took place in the submersible's aft, two-man dive chamber. Once the submersible was
on deck the R/V Johnson, it was mated to a double-lock decompression chamber (152-cm diameter)
for the remaining decompression period (3-4.5 hours).
Recent (2001-2006) Submersible and ROV Surveys
Recent data (2001-2006) were collected using the human-occupied Clelia submersible (HBOI) and
the NOAA NURC ROV (SuperPhantom). Submersible dives that were in proximity to the historical
transects were selected for comparison of reef habitat over a 25 year period. Color videotapes (MiniDV digital) were recorded with a pan and tilt videocamera which provided a 72.5° x 57.6° field of
view (Sony DX2 3000A, three chip 2.6 mm CCD, with Canon J8X6B KRS lens, 6-48 mm zoom, and
0.3 m minimum focus). The downward-looking camera was equipped with four parallel lasers (17.5
cm apart along the edge of the diamond shape) providing scale for quadrat size. Ship navigation was
determined with differential global positioning system (Magnavox MX 200 GPS) which is accurate to
±5 m, and submersible tracks were plotted with the Integrated Mission Profiler (Florida Atlantic
University) which links to the ship's GPS. Still images were derived from randomly selected video
frame grabs for quantitative point count analyses.
172
Quantitative Analyses of Photo Transects
The historical photographic transects were recently analyzed and compared with the recent dives on
the Oculina reefs to determine changes of percent cover of living coral over time (Reed et al., in
review). For this paper, only the northern-most (trawled) study site (Cape Canaveral) is compared to
the southern-most (untrawled) study site (Jeff's Reef). Each photographic image was overlaid with
100 randomly distributed points to determine percent cover for each habitat type (live coral, standing
dead coral, coral rubble) using CPCe software (Kohler and Gill, 2006). Coral colony diameters were
also calculated with the CPCe program.
Northern Trawled Site (Cape Canaveral)
In 1976, the Cape Canaveral photographic transect consisted of six Johnson-Sea-Link (JSL)
submersible dives covering a total E-W distance of 42.0 km from a depth of 30 m to 305 m at the
latitude of ~28°29'N. The depth zone of distribution for living Oculina was 70-87 m over a distance
of 2.4 km. One major, high-relief Oculina bioherm was encountered.
In 2001, fathometer transects were made at the Cape Canaveral Oculina bioherm site previously
mapped in 1976. Although there were some discrepancies in the coordinates from LORAN A used in
1976 versus GPS in 2001, there were no other mounds or features in the region. In this case, the
LORAN converted coordinates were 460 m SE from the GPS coordinates. Multibeam maps generated
from surveys made in 2002 (Reed et al., 2005), further verified that this was indeed the same reef
(GPS at peak- 28°29.802'N, 80°01.268'W). This site had remained open to bottom trawling until
2000, when the OHAPC was expanded up to Cape Canaveral and incorporated this reef.
Cape Canaveral Reef (1976)
The Cape Canaveral Oculina reef was discovered during a submersible dive in 1976 (JSL II-063).
Bottom visibility was 5-10 m, current 21 cm⋅s-1 (30 cm s-1 at top of reef), and temperature 20°C. This
bioherm was described from written transcripts and submersible logs as a coral mound with 30-45°
slopes, 18 m relief, and ~0.35 km wide at the base. Maximum depth during the transect was 87 m at
the east base and 70-72 m at the top which was a SE-NW oriented ridge. The transect followed the
east flank, the top, and west flank. No exposed rock was observed on the slopes or crest of the reef
but appeared to be entirely covered with living and dead Oculina coral and sediment. Thus it is a
bioherm and not a rock mound or lithoherm. Colonies of live Oculina were ~1 m tall on the flanks
and the observers estimated coverage of 25%. Coral colonies along the peak were 45-60 cm tall.
Some colonies appeared to have been severed into two or three pieces, possibly by an anchor or cable.
A 6-m long, 5-cm diameter cable was found on the bottom near the reef. Dominant fish associated
with the reef were snowy grouper (Epinephelus niveatus), greater amberjack (Seriola dumerili), and
smaller reef fish including bank butterflyfish (Prognathodes aya), blue angelfish (Holacanthus
bermudensis), and various damsels and wrasses. Continuing the transect west of the reef were a series
of smaller reefs of moderate to low relief (3-5 m to flat pavement). Some had spurs of live coral, 1.0
m tall and 2.5 m long. The bottom undulated with exposed bedrock and a thin veneer of sediment,
coral rubble, and live coral mostly 45-60 cm diameter. At a depth of 70 m to the end of the transect at
65 m, the bottom was sandy and no Oculina was found.
173
Cape Canaveral Reef (2001)
The submersible dive (Clelia 616) in 2001 consisted of six transects on the 18-m tall bioherm. After
25 years since the first submersible dive, we found a reef that had been reduced to coral rubble. The
peaks and flanks (N, S, E, and W) were covered in thick layers of unconsolidated dead coral rubble
consisting of pieces ~2-10 cm in length. Some of the dead coral rubble on the upper south and east
slopes and peak were somewhat consolidated rubble and well encrusted with demosponges and
possibly blue-green algae. The flanks and peak appeared to be nearly 100% cover of dead coral based
on visual observations. Apparent trawl tracks were observed as deep, straight grooves (~30 cm deep,
60 cm wide) cut into the coral rubble on the upper slopes and may be the result of doors from bottom
trawls that are known to work in the area. A few 1.0 m colonies of standing dead coral were found at
the west base. The only living Oculina coral observed was at the southeastern base from 82-85 m
where a few 15-40 cm live colonies were observed apparently unattached on sand. There were also
some 30-60 cm standing dead coral with live tips. The only fish were amberjack and few small reef
fish.
Southern Non-trawled Site (Jeff's Reef)
The original photographic transect (1976-1977) off Fort Pierce consisted of four submersible dives
from 30-261 m over 23.3 km; however, no live Oculina coral or coral rubble were encountered on
this transect. During the same time period, a massive live Oculina bioherm was discovered just 4 km
north of the transect line and is the southern-most living Oculina bioherm known.
This site (named Jeff's Reef after the JSL pilot, Jeff Prentice) was used for various experiments and
studies over the past 25 years and so the exact location of the reef was certain as navigation
equipment evolved from LORAN A to GPS. This reef was near the southern end of the original
boundaries of the OHAPC that was designated in 1984 and so remained protected from bottom
trawling during the entire period. However, it was open to bottom hook and line fishing until 1994
when that was also banned. Bottom fishing at these depths and within the Florida Current (Gulf
Stream), requires the use of heavy weights which can certainly crush the fragile coral but not to the
extent of a trawl net.
Jeff's Reef (1977)
An extensive photographic transect was made by the PI in 1977 on this Oculina reef (JSL II-164).
This 18-m tall bioherm was ~300 m wide and consisted of three E-W oriented ridges (80 m maximum
depth). Described in Reed 1980, the mound appeared to be entirely coral and sediment and a true
bioherm. The dive was divided into five photographic transects on the flanks and peak of the reef.
The east, west, south slopes and peak ridges were covered with massive live Oculina coral, 90-150
cm tall. The steep south slope (45º) and south faces of the ridges were covered with dense coral,
forming nearly continuous rows of coral bushes. The 30º north slope had more coral rubble, less live
coral, and generally smaller colonies of live coral. On this dive numerous snapper and grouper were
noted along with amberjacks, 3.0 m shark and 2.4 m tall giant ocean sunfish (Mola mola). On
subsequent dives dense spawning aggregations consisting of hundreds of scamp and gag grouper were
described in the early 1980s (Gilmore and Jones, 1992).
Average growth rate of Oculina varicosa at a depth of 80 m is 16 mm⋅y-1 (Reed, 1981). At this rate a
large 1.5 m colony may be nearly a century old. During a lockout dive from the submersible, the PI
using a steel rod was able to probe the flank of Jeff's Reef to a depth of 4.0 m without hitting bedrock.
A 6-cm diameter sediment core was also taken during the dive. The core consisted of dead coral branch
fragments and mud sediment; a piece of Oculina branch within the core had a radiocarbon age of 480±70
y B.P. (Hoskin et al., 1987). Using the radiocarbon date yields an estimate of 980 years for the
development of this Oculina bank. Considering that the limestone base of these Oculina reefs would
174
have been exposed ~18,000 years ago during the low water stand (–80 m) at the height of the Wisconsin
glacial period, these deep-water Oculina reefs maybe thousands of years old.
Bottom temperatures averaged 16.2°C and ranged from 7.4 to 26.7°C at the 80-m Jeff's Reef site
during a long-term survey (Reed, 1981). Upwelling of bottom water from the Florida Straits produces
episodic intrusions of cold water throughout the year at the shelf edge in this region which causes
temperatures to drop below 10oC (Smith, 1981; Reed, 1983). During these upwelling periods, levels
of nutrients and chlorophyll increase nearly an order of magnitude: nitrates increased from <2 uM
during non-upwelling to 9-18 uM during upwelling; phosphate from <0.25 to 0.5-2.0 uM; and
chlorophyll-a from <1 to 1-9 mg⋅m-3 (Reed, 1983). Salinity on the deep reef averages 36.0. The clear,
warm water of the northerly flowing Florida Current in the region of the Oculina reefs typically only
extends down to a depth of 50-60 m. Seldom does this water mass extend to the bottom and the reefs
are often inundated with a turbid, bottom nepheloid layer. Bottom currents averaged 8.6 cm⋅s-1 but
may exceed 50 cm⋅s-1 (1 kn); currents of 50-100 cm⋅s-1 due to the Florida Current may affect the peaks
of the higher Oculina pinnacles and may be strong enough to break the coral branch tips (Reed, 1981;
Hoskin et al., 1983). Long-term light measurements with Lambda quantum meters recorded an
average of 0.33% transmittance of surface light which usually does not support macroalgae on the
deep-water Oculina reefs or zooxanthellae within the coral (Reed, 1981).
Jeff's Reef (2001)
In 2001, ten video transects were randomly laid out on Jeff's Reef (Clelia 606). Since the 1970s, the
reef looks relatively healthy compared to the northern (recently protected) sites. There is no evidence
of trawl damage, but long lines and fishing lines are present on the reef. However, the fish
populations remain impacted from 20 years of overfishing. Population densities for the dominant fish
species correlated highly with habitat type (Koenig et al., 2000; Koenig et al., 2005). Gag
(Mycteroperca microlepis) and scamp (M. phenax) grouper and juvenile speckled hind (Epinephelus
drummondhayi) are predominately associated with the intact coral habitat. Although the fish surveys
in 2001 and 2003 were not directly comparable to previous surveys, there was a noted increase in
grouper numbers and size since the fishing moratorium in 1994. There was also an increase in the
abundance of male gag and scamp grouper since the 1995 survey, suggesting the possible
reoccurrence of spawning aggregations of both species. Still, very few commercial reef fish (snapper
and grouper) were observed in 2003, even after a 10 year moratorium on bottom fishing. The most
common larger grouper observed in 2003 were red (Epinephelus morio), scamp, gag, and snowy
grouper (E. niveatus). In the 1970s and 1980s, black sea bass (Centropristis striata) were abundant,
and large (50-100 kg) Warsaw grouper (Epinephelus nigritus) were common (G. Gilmore, pers.
observations; Reed, 2002a). After 20 years, black sea bass and juvenile speckled hind were observed
for the first time on the reefs during surveys in 2005 but the large Warsaw grouper were still absent.
Quantitative Changes in 25 Years
Trawled Site (Cape Canaveral)
In 1976, based on quantitative analyses of the photographic transects, live coral coverage ranged from
7.0-35.1% ( x = 19.2%), 31.4% standing dead coral, and 17% coral rubble (Figure 3). Maximum coral
density for individual photographic quadrats ranged from 32.0-73.2% ( x = 44.4%) and maximum
coral colony diameters were 1.4-1.7 m.
By 2001, quantitative analyses revealed living Oculina coral cover ranging from 0-0.9% ( x = 0.2%).
Mean standing dead coral cover was 1.68% and coral rubble 80.5%. Over a period of 25 years, nearly
100% of the live coral had been lost apparently due to trawling damage. Mean live coral cover was
175
reduced from 19.2% to 0.2% (p<0.0001; Figure 4), but concurrently, coral rubble cover had increased
by 64%, from 17 to 80.5% (p<0.0001).
Non-Trawled Site (Jeff's Reef)
Quantitative photo analyses from 1977 transects showed a mean range of live coral of 30.6-47.7%
( x = 39.3%), 25.8% standing dead coral, and 25.3% coral rubble. Maximum coral density from
individual quadrats ranged from 46.3-67.4%, and maximum coral diameter was 1.75-2.01 m although
in many cases the corals appeared to grow together into continuous hedges (which exceeded the width
of the photographs, ~2.5 m) and were difficult to determine individual colonies.
In 1996, the PI revisited the reef for the first time in over a decade (JSL II-2800). Video transects
were made with similar methodology and generally in similar locations as the 1977 dive. Five
transects were selected for quantitative analyses. Mean live coral cover ranged from 8.2-15.0% ( x =
10.3%), standing dead coral 60.2%, and coral rubble 25.4%. Maximum coral cover ranged from 1236%.
In 2001, the mean live coral cover (7.2-18.8%, x=13.4%) was similar to the 1996 results but both had
drastically reduced from 1977. Mean standing dead coral had decreased to 34%, and coral rubble was
43.4%. Maximum coral density ranged from 20.4-55.0%. There was a significant difference in
percent cover of live Oculina coral between 1977 and 1996 and between 1977 and 2001 (p<0.0001).
Live coral cover decreased 30% between 1977 ( x =39.3%) and 1996 (10.3%), but increased slightly
by 2001 (13.4%).
Figure 3. Mean percent live Oculina coral at reef sites in 1977
and 2001 (bars= range of transect means). CC= Cape Canaveral
Reef (trawled site), JF= Jeff's Reef (protected site).
Figure 4. Percent loss of live Oculina coral in a 25 year period
from 1977 to 2001. CC= Cape Canaveral (trawled site), JF= Jeff's
Reef (protected site)
176
Conclusions
It is apparent that the protected reef did not have the devastating destruction from trawling, however it
did show a loss in percent coral coverage over the 25 year time period. Although this could be due in
part to a variety of factors including global warming, disease, or pollution like nearby shallow water
reefs in the Florida Keys; damage from bottom hook and line fishing is certainly a factor. To date
there is no evidence regarding the effects of global warming on deep reefs, nor is their any record of
coral disease in deep reef corals. The 1996 survey was completed just after the moratorium on bottom
hook and line fishing was placed in effect in 1994. Since 1996, however, the reef has shown a
positive increase in live coral cover. This could be the result of fewer impacts of fishing weights and
line breaking the coral. Fish populations have yet to recover from overfishing in the 1980s and 1990s
but populations are showing signs of recovery. Speckled hind, which may be designated in the near
future as a threatened species, and gag and scamp grouper which are predominately associated with
the intact coral habitat, are showing up in greater numbers since the 1994 moratorium on bottom
fishing.
During ROV surveys in 2002 and 2003 it was apparent that some rock-shrimp trawling and bottom
hook and line fishing continue illegally within the Oculina HAPC. These observations included
sightings of trawlers during the surveys; evidence of fishing lines and bottom longlines wrapped
around coral colonies and remnants of new bottom trawl nets; artificial reef modules damaged by
heavy gear; and apparent trawl tracks in the rubble noted near the damaged modules. Since 2000
several shrimp trawlers have been caught trawling illegally within the OHAPC, and as recently as
February 2007, one trawler was caught within the OHAPC with 4500 kg of rock shrimp on board and
allegedly had lost his 1300 kg trawl net and doors somewhere in the sanctuary.
An artificial reef restoration project has placed over 100 1-m diameter, hollow, concrete domes (Reef
Balls, Reef Ball Foundation, www.reefball.org) in rubble areas of the OHAPC in 2000 and 2001
(Koenig et al., 2005). Scamp and snowy grouper were observed to associate with the Reef Ball
clusters very soon after placement. Several modules that were deployed in 1998 were also revisited in
2003 where numerous small colonies (1-10 cm) of Oculina were photographed growing on the blocks
providing evidence of settlement by planula larvae (Brooke and Young, 2003, Reed et al., 2005).
The expansion of the OHAPC in 2000 now covers the majority of the reef system, but hundreds of
Oculina bioherms that were discovered in 2003 during multibeam sonar surveys still remain
unprotected and vulnerable to bottom trawling. In 2003, the OECA (the original OHAPC boundaries)
closure to bottom fishing was extended indefinitely to protect the overfished deep-water species of
grouper and snapper. In addition, recent implementation of a vessel monitoring system (VMS) for the
rock shrimp industry by the SAFMC has already proved to aid in enforcement of the closed Oculina
HAPC areas.
Additional measures that could help protect these deep-water reefs would be surface monitor buoys
with acoustic and video recorders which could relay via satellite real-time data on boat traffic and
illegal trawlers. These could also be used by scientists studying the fish population dynamics and
could include arrays of thermographs, current meters, and other equipment to help understand this
remote yet valuable resource.
Acknowledgments
This manuscript is dedicated to the memory of Robert Avent who first described these magnificent
deep-water Oculina reefs using the Johnson-Sea-Link submersibles. Historical surveys were funded
177
solely by Harbor Branch Oceanographic Institution (HBOI) and recent research has been funded by
NOAA National Marine Fisheries, NOAA Office of Ocean Exploration, NOAA Undersea Research
Center at the University of North Carolina at Wilmington, and NOAA Ocean Service. Andy Shepard
(Director NOAA NURC) was instrumental in the funding and recent multibeam mapping of the
Oculina reefs, and Chris Koenig (Florida State University) conducted the statistical analyses of the
quantitative transects, fish surveys and the Reef Ball project. Grant Gilmore (Estuarine, Coastal and
Ocean Science, Inc.) conducted all the historical fish surveys. Kathy Scanlon, USGS, drafted the map
(Figure 1). We thank the crews of HBOI's R/V Johnson, R/V Sea Diver, R/V Seward Johnson I and
II, Johnson-Sea-Link I and II and Clelia submersibles, and NOAA NURC's SuperPhantom ROV for
their exceptional operational support. Funding for this project was provided in part by the 2005 Mia J.
Tegner Memorial Research Grant in Marine Environmental History and Historical Marine Ecology.
The Banbury Foundation and the Robertson Coral Reef Research and Conservation Program at HBOI
are gratefully acknowledged for their support. This is contribution #1652 from Harbor Branch
Oceanographic Institution.
References
Avent RM, Stanton FG. Observations from research submersible of megafaunal distribution on the
continental margin off central eastern Florida, Harbor Branch Foundation Tech. Rep. 25, Harbor Branch
Oceanographic Institution, Ft. Pierce, FL, 1979; 40 pp.
Avent RM, King ME, Gore RH. Topographic and faunal studies of shelf-edge prominences off the
central eastern Florida coast. Int Rev Ges Hydrobiol. 1977; 62: 185-208.
Brooke S, Young CM. Reproductive ecology of a deep-water scleractinian coral, Oculina varicosa, from
the southeast Florida shelf. Continental Shelf Res. 2003; 23: 847-858.
Gilmore RG, Jones RS. Color variation and associated behavior in the epinepheline groupers,
Mycteroperca microlepis (Goode and Bean) and M. phenax Jordan and Swain. Bull Mar Sci. 1992; 51:
83-103.
Hoskin CM, Geier JC, Reed JK. Sediment produced from abrasion of the branching stony coral Oculina
varicosa. J Sed Petrol. 1983; 53: 779-786.
Hoskin CM, Reed JK, Mook DH. Sediments from a living shelf-edge reef and adjacent area off central
eastern Florida. In: Maurrasse F, ed. Symposium on South Florida Geology, Miami Geological Society
Memoirs. 1987; 3: 42-57.
Koenig CC, Coleman FC, Grimes C, Fitzhugh G, Scanlon K, Gledhill C, Grace M. Protection of fish
spawning habitat for the conservation of warm-temperate reef-fish fisheries of shelf-edge reefs of
Florida. Bull Mar Sci. 2000; 66: 593-616.
Koenig CC, Shepard AN, Reed JK, Coleman FC, Brooke SD, Brusher J, Scanlon KM. Habitat and fish
populations in the deep-sea Oculina coral ecosystem of the western Atlantic. Amer Fish Soc Symp. 2005;
41: 795-805.
Kohler KE, Gill SM. Coral point count with Excel extensions (CPCe): a visual basic program for the
determination of coral and substrate cover using random point count methodology. Computers and
Geosciences. 2006; 32: 1259-1269.
178
National Oceanic and Atmospheric Administration. Fishery Management Plan for Coral and Coral Reefs
of the Gulf of Mexico and South Atlantic. Gulf of Mexico and South Atlantic Fishery Management
Councils, Tampa, Florida, 1982; 342 pp.
National Oceanic and Atmospheric Administration. Final Rule, Amendment 4 to the Fishery
Management Plan for Coral, Coral Reefs, and Live/Hard Bottom Habitats of the South Atlantic Regions
(Coral FMP). Federal Register, Vol. 65, No. 115, June 14, 2000.
Reed JK. Distribution and structure of deep-water Oculina varicosa coral reefs off central eastern
Florida. Bull Mar Sci. 1980; 30(3): 667-677.
Reed JK. Nearshore and shelf-edge Oculina coral reefs: The effects of upwelling on coral growth and on
the associated faunal communities. In: Reaka M, ed. The Ecology of Deep and Shallow Coral Reefs,
Symposia Series for Undersea Research, Vol. 1, NOAA, 1983: 119-124.
Reed JK. Deep-water Oculina coral reefs of Florida: biology, impacts, and management. Hydrobiologia.
2002a; 471: 43-55.
Reed JK. Comparison of deep-water coral reefs and lithoherms off southeastern U.S.A. Hydrobiologia.
2002b; 471: 57-69.
Reed JK, Koenig C, Shepard A. Impacts of bottom trawling on a deep-water Oculina coral ecosystem
off Florida. Bull Mar Sci. 2007; 8: 481-496.
Reed JK, Shepard A, Koenig C, Scanlon K, Gilmore G. Mapping, habitat characterization, and fish
surveys of the deep-water Oculina coral reef Marine Protected Area: a review of historical and current
research. In: Freiwald RA, Roberts JM, eds. Cold-water Corals and Ecosystems: Springer-Verlag: Berlin
Heidelberg, 2005: 443-465.
Reed JK, Weaver D, Pomponi SA. Habitat and fauna of deep-water Lophelia pertusa coral reefs off the
Southeastern USA: Blake Plateau, Straits of Florida, and Gulf of Mexico. Bull Mar Sci. 2006; 78(2):
343-375.
Smith NP. An investigation of seasonal upwelling along the Atlantic coast of Florida. Proc 12th
Internat. Liege Colloque Ocean Hydrodynamics, 1981: 79-98.
179
180
Pollock
NW, Godfrey
eds.American
Diving for Science
2007.Of Underwater Sciences
Diving For ScienceIn:
2007
Proceedings
OfJM,
The
Academy
Measuring Structural Complexity on Coral Reefs
Anders Knudby*, Ellsworth LeDrew
Department of Geography, University of Waterloo, 200 University Avenue West, Waterloo, N2L
1W6, Canada.
*corresponding author: knudby@gmail.com
Abstract
Structural complexity on coral reefs has been shown to positively influence several measures of
biodiversity, and is thus an important ecological variable. The dominant field method used by
reef scientists to measure structural complexity is the chain-and-tape method, which produces a
measure of rugosity calculated as the ratio of contour–following vs. straight distance between
two points on the reef. Expanding on this method, we developed simple and easy-to-use tools to
measure rugosity at four spatial scales for a range of typical coral reef structures, and also used
the data to calculate fractal dimensions at three intermediate scales. We show that measures of
structural complexity change unpredictably across spatial scales, and illustrate that typical coral
reef structures are too complex for any single measure to function as a comprehensive index of
structure over a range of scales. This illustrates that considerations of spatial scale are important
when measuring structural complexity, and that the smallest scale obtainable with current
remote sensing technology and methods is not directly related to the scale used in most studies
of fish ecology. We also illustrate that the fractal dimension measure is more closely related to
human intuitive perception of structural complexity than is rugosity, though we are unable to
test its value in fish ecology with our current dataset. Future research will relate fish body size
to the scale of reef structural complexity, and develop remote sensing-based methods to map
structural complexity over large spatial extents.
Introduction
The influence of structural complexity on the biodiversity of fish assemblages has received much
attention in the fish ecology literature (McCormick 1994; Szmant 1997). Most studies have concluded
that high structural complexity is spatially correlated with overall fish species richness (Chabanet et
al., 1997), or the abundance of specific trophic guilds (Hixon and Beets 1993) or mobility guilds
(Friedlander and Parrish 1998). Other studies have found that the fish biodiversity variables are
temporally correlated with structural complexity, declining as reef frameworks disintegrate after mass
bleaching events (Jones et al., 2004; Wilson et al., 2006) or experimental manipulation (Syms and
Jones 2000). However, the in-situ quantification of structural complexity has received surprisingly
little attention, and there is neither consensus on the best measure of structural complexity, nor on the
most relevant spatial scale at which to measure or assess it. Table 1 illustrates approaches taken to
quantify structural complexity in studies of coral reefs and other near-shore environments, and the
different scales at which the complexity has been measured. The variety of methods seen in the table
exists despite the acknowledged importance of structural complexity for coral reef biodiversity and
ecology.
Few studies have investigated what measure of structural complexity best predicts a positive effect on
fish biodiversity (McCormick, 1994; Gratwicke and Speight, 2005), and none have determined the
best spatial scale at which to measure it. In this study, we investigated the influence of spatial scale on
measures of structural complexity. We did this by measuring rugosity, the most commonly used
181
measure of structural complexity, at four spatial scales for six characteristic substrate types on coral
reefs.
Table 1: A review of methods for assessment of structural complexity of coral reefs and other nearshore environments.
Source
Environment
Organism(s)
Chain length (m) Link length (cm)
Friedlander and Parrish, 1998 Coral reefs
Fishes
3
Chabanet et al., 1997
Coral reefs
Fishes
Coverage of branching corals
1.3
McCormick, 1994
Coral reefs
Fishes
3
Hixon and Beets 1993
Coral reefs
Fishes
Holes of varying size and number
Ormond et al., 1996
Coral reefs
10 and Unknown
Pomacentrids
Geomorphologic zones, microhabitats
Luckhurst and Luckhurst, 1978 Coral reefs
Territorial fishes
3
1.5
Garpe and Ohman, 2003
Coral reefs
Unknown
Unknown
McClanahan, 1999
Patch reefs
Fishes
Echinometra
10
Nylon line
Chapman and Kramer, 1999
Coral reefs
Fishes
14
?
Kostylev et al., 2005
Rocky shore
Macrofauna
0.15
0.1
Gratwicke and Speight, 2005
Artificial reefs
Fishes
1-3.64
Unknown
Syms and Jones, 2000
Coral reefs
Fishes
Vertical relief
McClanahan, 1988
Coral reefs
Sea urchins
10
Kohn, 1968
Multiple types
Gastropods
No quantification
Unknown
Lara and Gonzalez, 1998
Coral Reefs
Fishes
6-point scale (visual estimation)
Sleeman et al., 2005
Model coral reef
None
18
10
Brock et al., 2004
Patch reefs
None
Variable, ~60
80
Rogers et al., 1991
Coral reefs
None
10 (transect)
1.3
Bainbridge and Reichelt, 1988 Coral reefs
None
Vertical relief
Methods
To measure rugosity at four spatial scales we constructed four simple chains, each with a different
link length. Chains with the two longest link lengths (82 and 12 cm respectively) links were
constructed of cut pieces of PVC pipe pulled over a rope, with knots separating each link. Chains with
shorter link lengths (3.1 and 0.2 cm) were ready-made metal chains. All chains were five meters long,
and negatively buoyant.
We chose our smallest scale (0.2 cm) as similar to the smallest of those reported in the literature to
ensure coverage of the scales that fish ecologists have found to correlate with fish biodiversity
variables (see Table 1). We chose our largest spatial scale (82 cm) to simulate the best available
resolution of current large-coverage remote sensing instruments (Brock et al., 2004) to assess the
possibility of measuring structural complexity remotely. We also chose two intermediate scales (3.1
and 12 cm) that are closer to the body sizes of most reef fish to create some continuity between the
smallest and largest scales.
In April and May of 2006, on Ngederrak forereef in Palau, we measured rugosities for six distinct
substrate types: sand, rubble, coarse branching coral, tabulate coral, foliose coral, and fine branching
Millepora coral. These substrate types were chosen because they were distinct, easily identifiable, and
182
could be found in patches five metres across. A typical example of each substrate is illustrated in
Figure 1.
Figure 1: The six substrate types measured in this study. Top row from left: rubble, coarse branching coral,
fine branching Millepora coral. Bottom row from left: tabulate coral, foliose coral, sand.
Each measurement consisted of draping the chain as closely as possible in a straight line over the
substrate, then measuring the straight-line distance between the two end points with a tape measure.
When end points were located in depressions, the straight-line distance was estimated by holding the
tape measure over the opening of the depression. For each substrate type, using each of the four
chains, this measurement was repeated at least seven times over different areas. The mean and
standard deviation of rugosity could then be calculated for each substrate, at each spatial scale. Figure
2 illustrates the four chains, and the tape measure, draped over a rubble substrate. Rugosity is
calculated by dividing the known length of the chain (five meters) by the measured distance between
the end points (Risk, 1972).
Using the mean rugosities we calculated the fractal dimensions (D) of the six substrates, at three
intermediate spatial scales. The fractal dimension is a measure of the change in rugosity with
changing scale of measurement, and can be calculated as D = 1 – S, where S is the slope of rugosity
values on a log-log plot with rugosity on the y-axis and spatial scale on the x-axis (Mandelbrot,
1977).
Figure 2: Four chains draped over a rubble substrate.
183
Results
Figure 3 presents the mean rugosities of each substrate type, at each scale. As expected, each
substrate type shows variation in rugosity with variation in the scale of measurement (one-way
ANOVA, p<0.001). T-tests were used to confirm that sand substrates had the lowest rugosity values
at all scales (p<0.001) and that coarse branching coral substrates had the highest rugosity values at the
0.2 cm and 3.1 cm scales (p=0.007 and p<0.001, respectively). At the largest scale, tabulate coral
substrates had the highest mean rugosity, though not significantly higher than that of foliose coral
substrates (p=0.142).
At each spatial scale, the six substrate types demonstrated significant differences in rugosity (one-way
ANOVA, p<0.001), though not all substrate types are distinguishable at all scales (e.g., tabulate and
fine branching Millepora coral has similar rugosity when measured at the 3.1 cm scale).
As expected, branching coral substrates have high rugosities at the smaller spatial scales, at which the
complexity of the branching network plays a role. At the larger spatial scales, coral colonies with
tabulate form have high rugosity, due to the overall shape of the colonies.
Figure 3: Mean rugosity for the six substrate types - at the four scales of measurement.
Figure 4 presents the mean fractal dimension values of each substrate type, at each scale. T-tests
confirmed that the two branching coral substrates had the highest fractal dimension at the smallest
scale (p<0.001), coarse branching coral had the highest fractal dimension at the intermediate scale
(p=0.006), and non-significantly at the largest scale (p=0.157). Sand substrates had the lowest fractal
dimension at any scale (p<0.001). As for rugosity values, the six substrate types demonstrated
significant differences in fractal dimension (one-way ANOVA, p<0.001), though not all substrate
types are distinguishable at all scales
184
For individual substrate types, the fractal dimension varies significantly between scales. Fine
branching Millepora coral and coarse branching coral have their highest fractal dimensions at the
intermediate scale, and tabulate coral has its highest fractal dimension at the largest scale. Sand is the
only substrate type with a near-constant fractal dimension. The highly varying fractal dimensions of
all substrate types other than sand stand in contrast to the near-constant fractal dimensions reported
for individual coral polyps (Basillais, 1997) and for larger reefscapes (Reichelt and Bradbury, 1984;
Purkis et al., 2005).
Figure 4: Mean fractal dimension for the six substrate types, at the three scales of measurement.
Discussion
For each substrate type, the spatial scale at which rugosity is measured significantly influences the
resulting values of both rugosity and fractal dimension (one-way ANOVA, p=0.0014). The structural
complexity of a substrate relative to other substrates also changes with scale, and with measure.
Several conclusions can be drawn:
1. Structural complexity at the smallest scale, a scale similar to that used in the fish ecology literature,
cannot be directly inferred by measurements of structural complexity at a larger scale. Structural
complexity, as measured by either rugosity or fractal dimension, does not vary with scale in a
straight-forward or predictable manner, but is highly dependent on substrate type. Information
about structural complexity at one scale therefore does not enable predictions about structural
complexity at another scale unless the substrate type, and hence the scaling relationship, is
known.
185
2. Fish and other organisms respond to the structural complexity of their habitat at a scale similar to
that of their own body size (Gratwicke and Speight, 2005). The spatial scale at which rugosity is
measured is therefore important for studies of fish ecology. More careful considerations when
selecting the appropriate spatial scale to measure structural complexity is likely to yield stronger
relationships between fish biodiversity and structural complexity variables. The scale of
measurement should preferably match the typical body size of the organisms whose habitat is
being investigated.
3. Because of the scale-dependency of the two measures of structural complexity, rugosity
measurements based on the best available remote sensing technology (equal to the largest scale
measured in this study) cannot be related directly to rugosity measured in situ at the smaller
scales, as typically measured in the literature.
4. For each substrate type, the measure of fractal dimension is more similar to an intuitive perception
of structural complexity than rugosity. The branching coral types, arguably having the most
complex structure at the smallest scale, has the highest fractal dimension of any measured
substrate at the smallest scale. At the intermediate scale, only the coarse branching coral has the
highest fractal dimension, because at this scale the small spaces between individual fine branches
are not registered, whereas the larger spaces between the coarse branches are. At the largest scale,
coarse branching coral still has the highest fractal dimension as a result of the 'valleys' between
individual colonies, but does not differ significantly from tabulate and foliose coral substrates
which also have high fractal dimensions at this scale. Fine branching Millepora coral colonies
often form continuous and relatively flat surfaces, and therefore have a low fractal dimension at
the largest scale. Measuring and calculating fractal dimensions thus give a picture of structural
complexity at several spatial scales that is easily interpretable.
5. It is well documented that reef fishes have specific substrate preferences (Myers, 1999). These
preferences are likely to influence the normally positive relationship between rugosity and the
diversity of a fish assemblage (see Table 1). However, given the scale-based variation in the
relative rugosity of the observed substrate types, the scale of measurement will influence the
nature and strength of this relationship, and therefore the accuracy with which the diversity of fish
assemblages can be predicted from the rugosity measurements. Given that efforts are being made
to operationalize such predictions using airborne lidar measurements of rugosity at several scales
(Kuffner et al., 2007), further insight into the influence of spatial scale is needed.
6. Given the variation between substrate types in relative values of both structural complexity
measures, no one measure at one scale can adequately describe the structural complexity of a
substrate. However, measuring rugosity at multiple scales, as required for the calculation of the
fractal dimension, is time-consuming. It requires more time spent under water than standard
rugosity measurements, which can already be time-consuming and cumbersome. Handling
multiple five-metre chains of varying buoyancy on a coral reef, without severely damaging the
substrates being measured, requires patience, especially when working in a current of surf
environment. Efficient measurement of structural complexity on a coral reef will remain
dependent on the objective of the study, and particularly on the relevant spatial scale.
Future Research
Future research will include collection of data on fish biodiversity variables, and establish guidelines
for choosing the best spatial scale at which to measure structural complexity in fish ecology studies. It
will also explore remote sensing methods to obtain information on coral reef structural complexity at
the relevant spatial scales, with a view to ultimately replace in-situ measurements.
186
References
Bainbridge SJ, Reichelt RE. An assessment of ground truth methods for coral reef remote sensing
data. International Coral Reef Symposium, Townsville, Australia, AIMS. 1988; 2: 439-444.
Basillais E. Coral surfaces and fractal dimensions: a new method. Comptes Rendus De L'Academie
Des Sciences Serie Iii-Sciences De La Vie-Life Sciences. 1997; 320(8): 653-657.
Brock JC, Wright CW, Clayton TD, Nayegandhi A. LIDAR optical rugosity of coral reefs in
Biscayne National Park, Florida. Coral Reefs. 2004; 23(1): 48-59.
Chabanet P, Ralambondrainy H, Amanieu M, Faure G, Galzin R. Relationships between coral reef
substrata and fish. Coral Reefs. 1997; 16(2): 93-102.
Chapman MR, Kramer DL. Gradients in coral reef fish density and size across the Barbados Marine
Reserve boundary: effects of reserve protection and habitat characteristics. Marine Ecology-Progress
Series, 1999; 181: 81-96.
Friedlander AM, Parrish JD. Habitat characteristics affecting fish assemblages on a Hawaiian coral
reef. J Experim Mar Biol Ecol. 1998; 224(1): 1-30.
Garpe KC, Ohman MC. Coral and fish distribution patterns in Mafia Island Marine Park, Tanzania:
fish-habitat interactions. Hydrobiologia. 2003; 498(1-3): 191-211.
Gratwicke B, Speight MR. Effects of habitat complexity on Caribbean marine fish assemblages.
Marine Ecology-Progress Series. 2005; 292: 301-310.
Hixon MA, Beets JP. Predation, prey refuges, and the structure of coral-reef fish assemblages. Ecol
Monog. 1993; 63(1): 77-101.
Jones GP, McCormick MI, Srinivasan M, Eagle JV. Coral decline threatens fish biodiversity in
marine reserves. Proceedings of the National Academy of Sciences of the US. 2004; 101(21): 82518253.
Kohn AJ. Microhabitats abundance and food of Conus on atoll reefs in Maldive and Chagos islands.
Ecol. 1968; 49(6): 1046-1062.
Kostylev VE, Erlandsson J, Ming MY, Williams GA. The relative importance of habitat complexity
and surface area in assessing biodiversity: Fractal application on rocky shores. Ecol Complex 2005;
2(3): 272-286.
Kuffner IB, Brock JC, Grober-Dunsmore R, Bonito VE, Hickey TD, Wright CW. Relationships
between reef fish communities and remotely sensed rugosity measurements in Biscayne National
Park, Florida, USA. Environ Biol Fishes. 2007; 78(1): 71-82.
Lara EN, Gonzalez EA. The relationship between reef fish community structure and environmental
variables in the southern Mexican Caribbean. J Fish Biol. 1998; 53: 209-221.
Luckhurst BE, Luckhurst K. Analysis of influence of substrate variables on coral-reef fish
communities. Mar Biol. 1978; 49(4): 317-323.
187
Mandelbrot BB. Fractals. San Francisco, WH Freeman and Company, 1977.
McClanahan TR. Coexistence in a sea-urchin guild and its implications to coral-reef diversity and
degradation. Oecologia. 1988; 77(2): 210-218.
McClanahan TR. Predation and the control of the sea urchin Echinometra viridis and fleshy algae in
the patch reefs of Glovers Reef, Belize. Ecosystems. 1999; 2(6): 511-523.
McCormick MI. Comparison of field methods for measuring surface-topography and their
associations with a tropical reef fish assemblage. Marine Ecology-Progress Series. 1994; 112(1-2):
87-96.
Myers RF. Micronesian Reef Fishes. Davie, FL: Coral Graphics, 1999.
Ormond RFG, Roberts JM, Jan RQ. Behavioural differences in microhabitat use by damselfishes
(Pomacentridae): implications for reef fish biodiversity. J Experim Mar Biol Ecol. 1996; 202(1): 8595.
Purkis SJ, Riegl BM, Andrefouet S. Remote sensing of geomorphology and facies patterns on a
modern carbonate ramp (Arabian Gulf, Dubai, UAE). J Sediment Res. 2005; 75(5): 861-876.
Reichelt RE, Bradbury RH. Spatial patterns in coral-reef benthos - multiscale analysis of sites from 3
oceans. Marine Ecology-Progress Series. 1984; 17(3): 251-257.
Risk MJ. Fish diversity on a coral reef in the Virgin Islands. Atoll Res Bull. 1972; 153: 1-6.
Rogers CS, McLain LN, Tobias CR. Effects of hurricane Hugo (1989) on a coral-reef in St-John,
USVI. Marine Ecology-Progress Series. 1991; 78(2): 189-199.
Sleeman JC, Boggs GS, Radford BC, Kendrick GA. Using agent-based models to aid reef restoration:
enhancing coral cover and topographic complexity through the spatial arrangement of coral
transplants. Restor Ecol. 2005; 13(4): 685-694.
Syms C, Jones GP. Disturbance, habitat structure, and the dynamics of a coral-reef fish community.
Ecology. 2000; 81(10): 2714-2729.
Szmant AM. Nutrient effects on coral reefs: A hypothesis on the importance of topographic and
trophic complexity to reef nutrient dynamics, Panama. 1997; 2: 1527-1532.
Wilson SK, Graham NAJ, Pratchett MS, Jones GP, Polunin NVC. Multiple disturbances and the
global degradation of coral reefs: are reef fishes at risk or resilient? Global Change Biol. 2006;
12(11): 2220-2234.
188
In: Pollock
NW, Godfrey
Diving forAcademy
Science 2007.
Proceedings
Of JM,
Theeds.
American
Reef Status Protocol (RSP): A Prognostic Reef Survey Methodology,
Rapid Yet Comprehensive
Hannah L Markham*, and Nicola K Browne
Society for Environmental Exploration, 50-52 Rivington Street, London EC2A 3QP
*corresponding author: hannah.markham@gmail.com
Abstract
A comprehensive biological assessment of the marine environment to determine reef health has
traditionally relied on access to extensive economic and material resources. Resources typically
include expensive survey equipment, a multidisciplinary team of specialist scientists and an
adequate time period over which a representative data set can be collated. The Reef Status
Protocol (RSP) survey technique provides a rapid yet comprehensive and prognostic data set
with minimal resource use from which management recommendations can be made with a high
degree of confidence. A number of sites within a target area can be surveyed to a high standard
over a short time period thus reducing the required levels of both economic and material
resources. The floral and faunal composition of a site is assessed using standard biological
parameters and noted in parallel with the environmental variables of the site. Low manpower is
required for the technique, and surveyors can be of a varying skills base (highly skilled
specialists to non-specialist volunteers), thus increasing the efficiency of ecological data
collection.
Introduction
There are a number of underwater data collection techniques that can be applied in the assessment of
biological diversity and resource use (English et al., 1997; Hill and Wilkinson, 2004). Evaluation of
the most appropriate method will depend on the objective of the study usually dependant on
characterising the reef, maximising sampling efficiency and/or incorporating quantitative data
collection for a number of biological parameters. The extent of information required and time frames
involved will ultimately decide which method is to be implemented (Montebon, 1992). Surveying
new sites rapidly and efficiently using a ubiquitously implemented technique has in the past required
high levels of resource use and a large task force of highly trained personnel (Pattengill-Semmens and
Semmens, 2003). As a result such factors have prevented smaller resource-limited organisations
contributing to marine management strategies and thus aiding in the establishment of protected areas.
However, said organisations are often characterised by high levels of man power, and hence have the
capability of increasing sampling efficiency and geographical coverage. Thus, they should not be
overlooked as a large source of data for conservation initiatives.
Information required to fulfil biodiversity action plans is commonly labour intensive, but
scientifically undemanding (Foster-Smith and Evans, 2003). An example to highlight the sheer
implausibility of scientists to conduct these surveys was calculated by Goffredo et al. (2004) who
determined that it would have taken a professional researcher 20 years at a cost of more than $1
million to collect the same amount of data about seahorse distribution that their self-funded
volunteers had collected over a period of just three years. Hence, by far the most valuable resource of
not-for-profit organisations is the high manpower available to conduct the underwater census.
189
Self-funded survey teams, such as Frontier/Society for Environmental Exploration have their own
distinct niche within the scientific community. These organisations have the ability to move into areas
of reef that have yet to be studied in detail as they are not constrained by restrictions that commonly
govern the locations of study, such as private investment and time limitations. This factor combined
with the ability to collect large quantities of data over relatively short time periods can tag regions,
and thus pave the way for teams of highly skilled specialists in the identification of biodiversity
hotspots and the implementation management plans.
Reef Status Protocol (RSP) was developed by Frontier-Madagascar in 2005 in order to facilitate the
biological assessment of Diego Suarez Bay, Northern Madagascar (Figure 1). The methodology was
required to implement a rapid assessment with basic monitoring components for implementation in a
wide range of sub-aqua habitats, conducting reconnaissance as to the structure and locations of the
reef communities. A comprehensive output to assess the biological status of the bay's reef ecosystems
included a broad range of both physical and biological parameters to allow valid conclusions to be
drawn. Furthermore, the technique would rely heavily on non-specialist volunteers, and hence needed
to allow for a varying skills base without compromising data accuracy. Attention was also given to
the ability of the technique to assess marine resource distribution on both a spatial and temporal scale.
Figure 1. Map of Diego-Suarez Bay, northern Madagascar (12°10'S 049°15'E)
Methodology
A Reef Status Protocol can be performed snorkelling or using SCUBA depending upon the equipment
available, the habitat to be surveyed and the visibility of the water. Transect sites were chosen using
stratified random sampling method and were located 1-2 km apart along the shore line. Stratified
190
random samples may represent one of the most suitable sampling methods as it compromises between
the need to avoid bias due to a lack of randomization and the need to take only a limited number of
samples due to time constraints (Lessios, 1996). In addition, integrating a level of stratification may
provide additional information on factors affecting organism distributions.
Three individuals are required for the survey including both a physical and biological surveyor
alongside a boat marshal. The surveyors enter the water 150 m from the shore line or at a maximum
depth of 15 m. The physical surveyor takes a compass bearing to land prior to either descent of the
dive or before commencing a snorkel survey. The surveyors swim along the bearing towards the
shore at a rate of 10 m⋅min-1 for 15 minutes (covering 150 m) or until it is too shallow to continue
(Figure 2).
Figure 2. Diagrammatic representation of the Reef Status Protocol (RSP).
Boat Marshal
The boat marshal's primary responsibility is to ensure the safety of the in-water surveyors by
monitoring their progress. Additional duties include: establishing the start and end GPS coordinates
of the survey; obtaining surface water samples for nutrient testing; measuring turbidity using a secchi
disk; noting weather conditions (wind strength, direction and cloud cover); and recording sea surface
activities such as fishing and boat movements.
Physical Surveyor
The physical surveyor takes a bearing perpendicular to the shore on the surface, prior to descending.
On reaching the bed, the surveyor is responsible for obtaining a sediment and water sample, noting
temperature and estimating horizontal visibility. They should position themselves half a body length
behind the biological surveyor during data collection to allow the biological surveyor a clear 180°
view and minimise disturbance to fish populations, however it should be noted that the physical
surveyor controls both the pace and direction of the underwater swim.
Biological Surveyor
The biological surveyor is responsible for recording the substratum type on a minute basis and marine
taxa, as indicated in Table 1, sighted during the survey. In order to conduct a tiered monitoring
system, this methodology allows surveyors to implement the biological survey at a number of discrete
191
levels. The practical application of each level of the biological survey is shown in Table 1 and should
be conducted according to the competency of the surveyor. It must be noted, however that with
multiple surveyors conducting surveys at varied application levels the data will only be analysed at
the lowest level. In association with the biological identification of fauna, the abundance categories
vary according to the taxa to be noted. Categories (good, medium, poor or absent) are used to
describe substratum cover and the extent of macrophyte cover. Invertebrates are noted only for their
presence or absence and individual fish should be tallied.
Table 1. Practical application of each of the levels of the biological survey
Level
Advanced
Intermediate
Octocoral
Genera
Lifeform
Scleractinian Coral
Genera
Lifeform
Fish
Species
Family
Elementary
Identification of Presence / Absence at all levels
Invertebrate
Species
Phyla / Family
Algae
Species
Colour and
Lifeform
Data is recorded on underwater slates and transferred to RSP survey data forms. Statistical analysis
following data input will permit the generation of habitat maps and allow further spatial and temporal
evaluation for specific sites within a region.
Data Validation
There have been a number of studies of public based surveys that validate the use of volunteers and
highlight the contribution that non-specialist based organisations have made to marine conservation
(Chou, 1994; Mumby et al., 1995; Hodgson, 1999; Roxburgh, 2000). However, questions are still
raised as to the reliability of data collected by non-specialist volunteers and whether its use in
accurately projecting an ecological profile of a study site is valid (Darwell and Dulvy, 1996). As a
result, research organisations using non-specialist volunteers for data collection inevitably have to
self-analyse and regulate their volunteers to ensure competency. They must also ensure that the
supporting methodologies are both within the capabilities of the volunteer and do not allow
opportunity for error in collection.
Frontier-Madagascar maintains the accuracy of their data by thoroughly training and assessing
volunteers prior to survey commencement during an intensive marine species identification course
lasting between 14 and 18 days. Table 2 indicates the minimum level of knowledge they are required
to possess prior to survey dives. However, as the individual gains more experience in the water they
are expected to continue their private study in order to become more specialist within a specific area
and conduct more comprehensive assessments within other parts of the survey programme.
Table 2. An indication of the minimum level of knowledge expected for each data set prior to survey dive
Data Set
Fish
Coral
Invertebrates
Algae
Sea grass
Phyllum
Class
6
18
Family
30
Genus
Species
40
37
27
8
Training is conducted both in the class room and during point out snorkels/dives, and is completed
with standardised tests. Tests are carried out both in situ and using computer images with a 95% pass
rate required for fish, invertebrate and marine plant identification and an 80% pass rate for corals. In
192
addition volunteers are trained in survey methodologies, broken down into component parts to
include swim pace, compass work and dry-runs in order to ensure diver competency in the water.
RSP has a number of discrete levels within its survey technique to allow volunteers to record data to
the level of their capability. If data collection is undertaken by a number of individuals with a varied
skills base, data can only be evaluated at the lowest level of data collection, i.e., to genus level as
opposed to species. This does in effect invalidate the complex data sets collected until enough
replicates at the higher level can be attained for the site. Thus, when undertaking a survey programme
the use of RSP should be closely monitored according to the skills base of the team in order to allow
an even distribution of higher skilled surveys across the site. This will allow for a more
comprehensive and directed analysis.
Data Analysis
The RSP's standard parameters allow the technique to be used to determine the biological value of a
site in the context of its chemical and physical properties. The accumulated data sets can be analysed
at a number of different discrete levels according to the surveyor competency. This can result in the
generation of simple GIS habitat maps, analysis of spatial relationships between and within reefs and
realise broad patterns in faunal distribution. However, data sets containing species level information
may be further subjected to rigorous statistical analysis between and within sites in order to evaluate
the spatial interactions between certain fish species, corals genera, invertebrate fauna and their
favoured physical, chemical and environmental variables. Furthermore, RSP is useful in the
preliminary examination of temporal changes to the floral or faunal composition within sites and thus
has the potential to determine both reef fish and invertebrate community dynamics, including
monitoring the recruitment of new species and range extensions of individual species within the zones
surveyed. Moreover, it is a protocol that can be used to assess coral reef health at local and regional
levels, thus facilitating the recognition of biologically diverse sites which are then tagged for further
investigation and a more detailed assessment in the future by specialist scientists.
Discussion
Broad scale monitoring techniques are particularly useful for habitat mapping, site selection and
covering large expanses of water (Hill and Wilkinson, 2004). The two most commonly used
methodologies include the Rover Diver Technique (RDT) and the Manta Tow (Table 3). The focus of
the RDT is on the compilation of fish species lists where as the latter is used to assess broad scale
changes in benthic communities.
The RSP represents a new technique that incorporates the assessment of both variables together with
algal cover, invertebrate composition and physical variables. Species lists for coral, fish, algae and
invertebrates together with quantitative data on the latter three parameters will provide an ecological
profile of a site. Although, these species lists are not conclusive, they are comparable across the
region being assessed and thus can be used to identify sites with higher levels of biodiversity.
Furthermore, data sets created provide an overall assessment of the benthic habitat based on scaled
categories thus aiding in the generation of habitat maps. Organisations based on non-specialist
volunteers will be able to generate large amounts of data over comparably short time periods, and
hence provide relevant information representing the first level of site evaluation. Such data will
provide the assessment on the availability and quality of the reef allowing decisions to be made on the
necessity to investigate further using a highly skilled specialist investigative teams, as such
investigations will inevitably increase resources required.
193
Table 3. Comparison of common broad scale monitoring techniques to the RSP technique
1
Resource use
2
Ease of
implementation
3
Descriptive
parameters
4
Quantitative
parameters
5
Skills base
Reef Status Protocol
- 3 surveyors
- 15 min. swim
- Snorkel or SCUBA
- Use of non-specialist
Volunteers
- Recommend min of 1
week training period
- Some degree of inwater/SCUBA
experience required
- Benthic cover
- Site Biological Profile
Roving Diver Technique
- 1 surveyor
- Varies
- Snorkel or SCUBA
- Experienced surveyor
Recommended
- Training periods will
vary significantly.
- Some degree of inwater/SCUBA
experience required
- Fish, coral,
invertebrate
and algae species lists
- Species abundance
- Visibility, temperature,
salinity,
meteriological
conditions
- Data forms takes into
consideration varying
skills base
- Fish species and
abundance using log
categories.
Manta Tow
- 2 surveyors
- Varies
- Snorkel
- Use of non-specialist
volunteers
- Less than 1 week
training period
- Minimum in-water
experience required
- Benthic cover
- Site Biological
Profile
- Experienced surveyors
recommended.
- Experience level of
surveyor not taken
into account
Furthermore, it is characterised by minimal resource use, and hence is suited to situations where rapid
marine surveys are required, skills base is low and costs must be kept to a minimum. The technique
can be adapted depending on habitat to be surveyed and resources e.g., transects from the shore using
snorkeling equipment will reduce equipment demands and thus costs, but are just as effective in
assessment of close shore environments. The specificity of data collected will depend on surveyor
experience. However, data collection forms take into account such considerations allowing for the
surveyor to complete forms to their knowledge level.
The RSP technique has allowed Frontier Madagascar to achieve its initial goal in the assessment of
Diego Suarez Bay. Prior to the organization's arrival, no underwater visual census had been
conducted in the region due to a lack of resource availability. Hence, the health status of reefs inside
the bay was unknown. The RSP technique provided a large yet comprehensive data set from which
management recommendations can be made with a high degree of confidence. Complication of
species lists and overall assessment of the reef health identified reefs of high biological diversity, thus
highlighting regions for further quantitative assessment. The technique relied on the use of nonspecialist volunteers who completed an initial training period within a week. Hence, data sets were
compiled relatively quickly. The RSP technique has proved to be valuable for an organisation based
on non-specialist volunteers, but would also suit organisations with minimal resource availability
particularly those involved in resource management training, potentially aiding in the creation of a
long term sustainable monitoring programme.
194
References
Chou LM. Living coastal resources of Southeast Asia: management through continuing education by
institutions of higher education. Aquatic Conservation: Freshwater and Marine Ecosystems. 1994; 4:
179-184.
English S, Wilkinson CR, Baker V. Survey manual for tropical marine resources. Australian Institute
of Marine Science: Townsville, Australia, 1997; 390 pp.
Darwall WRT, Dulvy NK. An evaluation of the suitability of non-specialist volunteer researchers for
coral reef fish surveys. Mafia Island, Tanzania – A case study. Biological Conservation. 1996; 78:
223-231.
Foster-Smith J, Evans SM. The value of marine ecological data collected by volunteers. Biolog
Conserv. 2003; 113: 199-213.
Goffredo S, Piccinetti C, Zaccanti F.Volunteers in marine conservation monitoring. A study of the
distribution of seahorses carried out in collaboration with recreational scuba divers. Conserv Pract.
2004; 18: 1492-1503.
Hill J, Wilkinson CR. Methods for Ecological Monitoring of Coral Reefs. Australian Institute of
Marine Science, Townsville, Australia, 2004; 117 pp.
Hodgson G. A global assessment of human effects on coral reefs. Marine Pollut Bull. 1999; 38: 345355.
Lessios H, Harilaos A. Methods for quantifying abundance of marine organisms. Smithsonian
Scientific Diving website, Methods and Techniques, 2004: 149-157.
Montebon A. Use of the Line Intercept Technique to Determine Trends in Benthic Cover.
Proceedings of the Seventh International Coral Reef Symposium, Guam, 1992; 1: 151-155.
Mumby PA, Harborne AR, Raines RP, Ridley JM. A critical appraisal of data derived from Coral Cay
Conservation volunteers. Bull Marine Sci. 1995; 56: 737-751.
Pattengill-Semmens CV, Semmens BX. Conservation and management applications of the reef
volunteer fish monitoring program. Environment Monitor Assessment. 2003; 81: 43-50.
Roxburgh T. Checking Reef Check in Tanzania. Reef Encounter. 2000; 27: 21.
195
AAUS 2007 Symposium Schedule – University of Miami
Friday March 9th
0935 – 1000
Temporal Dynamics and Changes from Historical Levels in Reef Fish
Assemblages in Biscayne National Park (Todd Kelison)
1005 – 1025
Coral Restoration in the Florida Keys Using Colonies Derived from
Aquacultured Fragments (Ilze Berzins)
1030 – 1055
Spatial and Temporal Variability of Groundfish Populations in Proposed
Marine (J. Henry Valz)
1115 – 1200
Exploration of the Woodville Karst Plain
(Jarrod Jablonski)
1300 – 1325
Baseline Survey Protocol (BSP) (Hannah L Markham)
1330 – 1355
AAUS Diving Officer and Scientific Diver Certifications
(Michael A. Lang)
1400 – 1425
Test and Evaluation of Two, Commercial Off-The-Shelf, Multi-Gas Dive
Computers for Providing Accurate Depth Measurements and Acceptable
Mixed Gas and Air Decompression Schedules (J. Morgan Wells)
1430 – 1455
Physical Fitness of Scientific Diving: Standards and Shortcomings
(Alison C. Ma)
1515 – 1540
Scientific Diving Safety: Integrating Institutional, Team and Individual
Responsibility (Neal W. Pollock)
15:45 – 16:10
Using SCUBA and Snorkeling Methods to Obtain Model Parameters for an
Ecopath Network Model for Calabash Caye, Belize, Central America
(Rebecca A. Deehr)
1615 – 1640
Title TBA (David L. Jones)
1645 – 1710
Video IPOD Instructional Design Consideratons for Dive Training and
Underwater Subject Matter (Mike Dermody)
Saturday March 10th
Location: Auditorium
Concurrent Session A
0920 – 0945
Closed-circuit Rebreathers in the Forensic Study of the Rouse Simmons
Shipwreck (Gregg Stanton)
0950 – 1015
Evolving Strategies for Rebreather Fatality Investigations (Neal W. Pollock)
196
1020 – 1045
Comparing Potential Differences In The Assessment Of Fish Populations
And Assemblages Using Open-Circuit Versus Closed-Circuit Modes Of
Diving: A Study In Silence (Derek Smith)
1100 – 1125
When Things Go Wrong: A Look at Scientific Diving Incident Reports
(Michael R. Dardeau)
1130 – 1155
When Everything Goes Right (Vallorie Hodges)
Saturday March 10th
Concurrent Session B
Location: Seminar Room Science and Administration Building Rm 120
0920 – 0945
Underwater Methods Enhance Study of Shelter Competition Between Native
and Invasive Species of Crayfish (Karl Mueller)
0950 – 1015
Behavior and Sound Production by Longspine Squirrelfish During Playback
of Predator and Conspecific Sounds (Joseph J. Luczkovich)
1020 – 1045
Deploying Benthic Chambers to Measure Sediment Oxygen Demand in Long
Island Sound (Prentiss Balcom)
1100 – 1125
The Influence of Sedimentation on Southeast Florida Coral Reef Community
Composition (Melissia Phillips)
Saturday March 10th
Location: Auditorium
Afternoon Session
1300 – 1325
An Evolution of Scientific Trimix Diving Procedures
at the Submerged Resources Center, National Park Service
(Jeffery E. Bozanic)
1330 – 1355
Quantifying the In Situ Survivorship of Recently Settled Coral Spat
(Wade Cooper)
1400 – 1425
Coral Disease and Bleaching Relationships in South Florida in 2005
(Marilyn Brandt)
1430 – 1455
Conservation Efforts of Threatened Rockfish Species at the Oregon Coast
Aquarium (Kevin Clifford)
1515 – 1540
The Effects of Size-Selective Fishing Pressure on the Mating System,
Population Structure, and Sex-Change Dynamics of the California
Sheephead, Semicossyphus Pulcher (Lynne Wetmore)
1545 – 1610
High Altitude Diving Operations (Randall Berthold)
1615 – 1640
Long Term Monitoring of a Deepwater Coral Reef: Effects of Bottom
Trawling (John K. Reed)
197
198
199
200

diving for science 2007 - Rubicon Research Repository

Transcription

Similar documents

Snorkel feature of the month - February 2013

November - Bay Area Divers

48x96 poster template

May 2015 FCDA Surface Interval. - Fairfield County Diving Association

art snorkelling

9 marine life: Reef life includes elkhorn, brain and many other hard

the aqua tooter - Aqua Tutus Diving Club

3_Adventure Playground.indd

rubiconhaulers.com 102" Wide Body Toy Haulers

the aqua tooter - Aqua Tutus Diving Club

Dissemination events diary and results