Influence of soil chemistry on metal

Science of the Total Environment 374 (2007) 367 – 378
www.elsevier.com/locate/scitotenv
Influence of soil chemistry on metal and bioessential element
concentrations in nymphal and adult periodical
cicadas (Magicicada spp.)
G.R. Robinson Jr. a,⁎, P.L. Sibrell b , C.J. Boughton b , L.H. Yang c
a
U.S. Geological Survey, 954 National Center, Reston, Virginia, 20192, USA
U.S. Geological Survey, 11649 Leetown Road, Kearneysville, West Virginia, 25430, USA
Section of Evolution and Ecology, University of California, One Shields Avenue, Davis, California, 95616, USA
b
c
Received 26 July 2006; received in revised form 7 December 2006; accepted 8 December 2006
Available online 25 January 2007
Abstract
Metal and bioessential element concentrations were measured in three species of 17-year periodical cicadas (Magicicada
spp.) to determine how cicada tissue chemistry is affected by soil chemistry, measure the bioavailability of metals from both
uncontaminated and lead-arsenate-pesticide contaminated soils, and assess the potential risks of observed metal contamination
for wildlife.
Periodical cicada nymphs feed on root xylem fluids for 13 or 17 years of underground development. The nymphs then
emerge synchronously at high densities, before leaving their nymphal keratin exoskeleton and molting into their adult form.
Cicadas are an important food source for birds and animals during emergence events, and influence nutrient cycles in woodland
ecosystems.
Nymphal exoskeletons and whole adult cicadas were sampled in Clarke and Frederick Counties, Virginia and Berkeley and
Jefferson Counties, West Virginia during the Brood X emergence in May and June, 2004. Elements, such as Al, Fe, and Pb, are
strongly enriched in the nymphal exoskeleton relative to the adult body; Cu and Zn are enriched in bodies. Concentrations of
Fe and Pb, when normalized to relatively inert soil constituents such as Al and Ce, are similar in both the molt exoskeleton and
their host soil, implying that passive assimilation through prolonged soil contact (adhesion or adsorption) might control these
metal concentrations. Normalized concentrations of bioessential elements, such as S, P, K, Ca, Mn, Cu, Zn, and Mo, and
chalcophile (sulfur-loving) elements, such as As, Se, and Au, indicate strong enrichment in cicada tissues relative to soil,
implying selective absorption and retention by xylem fluids, the cicada nymphs themselves, or both. Element enrichment
patterns in cicada tissues are similar to enrichment patterns observed in xylem fluids from tree roots. Chalcophile elements and
heavy metals accumulate in keratin-rich tissues and may bind to sulfhydryl groups. Metal concentrations in the nymphal
exoskeleton show a positive correlation with soil metal concentrations, with Au exhibiting particularly strong enrichment in
the exoskeleton relative to soil concentrations. Metal concentrations in adult bodies do not correlate with soil chemistry.
Bioessential elements S, Ca, Mn, Fe, and Zn differed by sex in adults, whereas Na, Mg, K, Ca, Mn, Fe, Zn, and As differed by
species. Body concentrations of Ca differed by site conditions (orchard or reference setting). The high Pb contents of orchard
soils contaminated by arsenical pesticide residues might inhibit Ca uptake by cicada nymphs. The adult cicadas contain
concentrations of metals similar to, or less than, other invertebrates, such as earthworms. There does not appear to be a dietary
⁎ Corresponding author. Tel.: +1 703 648 6113; fax: +1 703 648 6383.
E-mail address: grobinson@usgs.gov (G.R. Robinson).
0048-9697/$ - see front matter. Published by Elsevier B.V.
doi:10.1016/j.scitotenv.2006.12.031
368
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
threat to birds or other consumers of adult cicadas based on Maximum Tolerable Dietary Level (MTDL) Guidelines developed
for agricultural animals.
Published by Elsevier B.V.
Keywords: Periodical cicadas (Magicicada spp.); Biogeochemistry; Metals; Orchard; Soil; Keratin; Lead arsenate; Gold
1. Introduction
Periodical cicadas are widespread insect herbivores
throughout the deciduous forests of the eastern United
States. Geographically separate broods of periodical
cicadas emerge synchronously at high densities over
large areas from underground burrows after 13 to
17 years of nymphal development. During their long
underground development, periodical cicada nymphs
feed on xylem fluids from tree roots (Williams and
Simon, 1995; Day et al., 2002).
Periodical cicada emerge at such high densities that
they commonly satiate the foraging abilities of their
predators. They have been recognized as important
events in the geochemical cycles of forest ecosystems
(Wheeler et al., 1992; Yang, 2004).
This study examined two focus points of the
biogeochemical cycle of bioessential elements and
metals: 1) differences in source factors between natural
soils and soils contaminated with arsenical pesticide
residues, and 2) the bioavailability of bioessential
elements, metals, and metalloids from these soils
through xylem fluids and direct exposure to cicadas.
This study addresses four questions:
1. Do concentration levels of metals, metalloids, and
bioessential elements in cicada tissues vary systematically with changes in soil chemistry?
2. How does cicada tissue chemistry vary by sex or
species?
3. To what degree does cicada tissue chemistry reflect
homeostatic processes to regulate the minerals in the
Fig. 1. Area of emergence in May–June 2004 of brood X periodic cicada, showing study area in northern Virginia and northeastern West Virginia.
Figure modified from online map at College of Mount St. Joseph (2004).
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
369
body by adjusting physiological processes in response
to environmental conditions versus passive uptake of
constituents from soil by adhesion and/or adsorption?
4. Do concentrations of arsenic and heavy metals in
periodical cicadas emerging from contaminated
orchard sites represent a risk to birds and other
cicada consumers?
traveling more than 50 m in a single flight (Karban,
1981). They congregate at chorus centers, often close to
their emergence sites.
We test the hypothesis that the chemistry of body
tissues reflects the concentrations of constituents in the
soil zone due to the long residence time and feeding
behavior of cicada nymphs in soil. We characterize
cicada tissues by species, type and sex, and compare the
bioavailability of metals and metalloids with the
geochemical behavior of bioessential elements.
A sampling program for soils, cicada molt exoskeletons and live mature adults was conducted from May
to June 2004 at orchards, former orchards, and nonorchard reference sites in Clarke and Frederick
Counties, Virginia and Berkeley and Jefferson Counties,
West Virginia. Cicada and soil sampling sites in relation
to site conditions are shown in Fig. 2. Sites were mostly
located on carbonate bedrock overlain with thin clayrich soils. The orchard locations are based on aerial
photographs and topographic maps (showing orchard
areas) prepared using information from the time period
of extensive use of arsenical pesticides between the
1920s and 1960s. A digital dataset identifying orchard
areas where arsenical pesticides were likely used in the
study area is available from Reed et al. (2006).
Cicada molt exoskeleton samples were collected at
emergence sites and adults were collected at emergence
sites and chorus centers. Soils and cicadas were sampled
at approximately 5 to 6 individual collection points
within a 1 km2 area at each site. The sex and species of
adults was determined. Cicada samples were frozen
until analysis, then dried at 60 °C. Dried adult insects
were macerated and analyzed individually; molt exoskeletons were composited into one sample for each
individual collection point.
Representative soil samples (B horizon) were
collected to depths of 15 to 25 cm (depending on
depth of regolith) by soil auger and as grab samples. The
soil sample intervals overlap much of the observed
depth range of 5th instar cicada nymphs, particularly in
orchard soils (Maier, 1980). Soil samples were air dried,
homogenized, sieved to b 200 micrometers (80 mesh),
and pulverized before analysis.
Element concentrations in cicada tissue and soil
samples were determined by Inductively Coupled
Plasma Mass Spectrometry (ICP-MS) following sample
digestion in aqua regia using the Ultratrace-2 method
performed by Activation Laboratories, Ancaster,
Ontario, Canada. Mercury was measured using coldvapor analysis techniques using the Code 1G method
performed by Activation Laboratories, Ancaster,
Ontario, Canada. Duplicate samples were submitted at
a rate of 5% of total samples. Data were accepted if the
relative standard deviation was b 15% at five times the
2. Background
2.1. Periodical cicadas
Brood X 17-year periodical cicadas (Magicicada
spp.) emerged during May-June, 2004, in the eastern
United States (Fig. 1). Cicada densities of ≤ 30,000 per
hectare to N 3.5 million per hectare are known to occur
during emergence in orchard and forest habitats
(Williams and Simon, 1995). Other cicada broods
emerge in other areas in other years.
Cicadas feed on root xylem fluids underground for 13
or 17 years before emerging to molt into their adult form.
Cicada nymphs pass through five nymphal instars during
their belowground development. Instars are insect
developmental periods between successive exoskeleton
moltings. The duration of these instars is variable in
periodical cicadas, but the fifth nymphal stage typically
lasts 5 years or more (Marlatt, 1907; White and Lloyd,
1975), during which time the exoskeleton is not shed until
emergence. The feeding depth of cicada nymphs seems to
depend largely on the depth of available plant roots. Maier
(1980) observed that 5th instar nymphs fed at shallower
depths in orchard soils compared to forest soils, and that
most 5th instar nymphs were found between 7.6 and
37.5 cm depth. Cicada nymphs have also been observed to
feed at depths of more than 60 cm (Williams and Simon,
1995). However, the soils developed over the carbonate
bedrock of the study area are typically thin, clay-rich and
rocky; tree roots, and likely cicada nymph feeding sites,
tend to be localized at shallow depths.
Nymphs emerge at dusk over a period of a few days
during their 13th or 17th year, depending on the species.
Adult cicadas typically shed their nymphal exoskeleton
within a few meters of their emergence site the next
morning. Adult cicadas have limited dispersal, rarely
3. Materials and methods
3.1. Study area
370
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
limit of detection for the duplicate samples. Analytical
methods, quality assurance and control methods, and
laboratory accreditation information are described in
more detail in Actlabs (2006).
3.2. Statistical methods
Normality tests were done on the geochemical data
for soils, cicada bodies and nymphal exoskeletons. Most
constituents analyzed for this report were neither
normally nor lognormally distributed and a few sample
measurements were below detection limits for some
elements. Concentrations less than detection limits were
reported as one-half the detection limit to calculate
sample means. We employed nonparametric methods,
which do not require assumptions about the distributions of the data, to compare sample group means.
The data for soil site chemistry were grouped by landuse status into orchard and former orchard sites or
reference sites. The data for cicada body chemistry were
grouped by land-use status, species type, and sex. The
cicada exoskeleton samples were not classified according
to species or sex and were grouped by land-use status.
Tests for differences in chemical distributions in soils
by site status and in cicada tissues by site status, sex, and
species classifications were performed using Mann–
Whitney (binary group) and Kruskal–Wallis (multiple
group) tests. The null hypothesis of these tests is that
there is no significant difference between the means of
the ranks of the concentrations of a chemical constituent
among the sample groups (Sokal and Rohlf, 1969;
Conover and Iman, 1981; Helsel and Hirsch, 2002).
Rejection of the null hypothesis at the 95-percent
confidence level (alpha = 0.05) was considered to be
evidence supporting the alternative hypothesis, that is,
the existence of a significant difference in chemistry
between groups. The general linear model (GLM)
procedure (SAS Institute Inc., 1999) was used to do
the non-parametric tests because the number of samples
was not the same for all of the category groups (Helsel
and Hirsch, 2002). Multiple comparison tests, based on
Kruskal–Wallis rank sums and comparisons of means of
ranks (Tukey) tests, were used to identify which groups
differed from others (Hollander and Wolfe, 1973).
Tukey's significant-difference test (Sokal and Rohlf,
1969; Stoline, 1981; Helsel and Hirsch, 2002) was used
Fig. 2. Location of soil and cicada sample sites in Clarke and Frederick Counties, Virginia and Berkeley and Jefferson Counties, West Virginia relative
to orchard areas that likely used arsenical pesticides. Orchard areas are from Reed et al. (2006). At each sample site, approximately six individual
sample collections were made for each sample type in a 1 km2 area.
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
371
Table 1
Summary of geochemical statistics for sample media, characterized by sample groups
Sample group
P
S
K
Na
Ca
Ce
Mg
Mn
Fe
Cu
Zn
Al
As
Pb
Au
mg/kg dry weight
Hg
ug/kg
Soil; 72 samples: 45 reference, 27 orchard
Mean
Total
360 230 1140 90 1590 65.9 2100 769 19550 18.3 43.6 14220 15.2
57.7
Median
Total
340 200 1070 80 1350 62.5 1460 690 18240 14.3 36.9 13270
8.1
22.5
Minimum Total
150
90
500 50 110 17.2 560
76 7210
7.0 15.9 6730
2.4
12.3
Maximum Total
780 480 2150 140 4370 125.8 9070 2110 38720 155.5 131.9 27640 263.1 1150.0
Mean
Reference 360 220 1150 90 1680 71.6 2000 893 20650 14.4 36.5 15220
6.9
21.6
Std. Error Reference
20
10
60
3 150
3.8 250
78 1490
0.7
1.9
750
0.5
1.0
Mean
Orchard
370 230 1130 90 1430 56.4 2250 563 16510 24.7 55.5 12560 28.9 117.9
Std. Error Orchard
30
20
90
4 170
3.6 360
67 1300
5.3
5.2
640
9.5
41.3
Exoskeleton; 55 samples: 36 reference, 19 orchard
Mean
Total
880 1130 1980 60
Median
Total
790 1080 1810 60
Minimum Total
400 750
720 30
Maximum Total
2030 2120 4040 100
Mean
Reference 890 1090 2020 60
Std. error Reference
60
40
130
3
Mean
Orchard
840 1220 1910 60
Std. error Orchard
90
70
190
3
3140
2790
1650
8010
3290
250
2850
220
Adult body; 127 samples: 71 reference, 56 orchard;
Mean
Total
6810 4880 7110 270
Median
Total
6770 4790 7370 270
Minimum Total
1750 3590 3010 150
Maximum Total
8720 6540 10670 460
Mean
Reference 6870 4910 7110 270
Std. error Reference
90
80
140 10
Mean
Orchard
6740 4840 7120 260
Std. error Orchard
130
80
170 10
Mean
Female
6690 5110 7220 270
Std. error Female
110
70
90 10
Mean
Male
6980 4550 6960 280
Std. error Male
90
50
220 10
74 female, 53 male
1300
1.3 1420
1280
1.2 1450
460
0.7 580
2500
3.2 2000
1370
1.3 1430
50
0.0
30
1210
1.3 1410
40
0.1
30
1480
1.2 1440
40
0.0
20
1050
1.4 1400
40
0.1
50
11.7 800 468 3690
10.5 730 420 3280
2.6 410 228 1190
33.2 1740 1140 13420
11.4 740 493 3470
0.8
30
30
280
12.2 900 423 4100
1.5
80
47
620
211
198
32
534
226
13
193
13
268
11
132
8
16.4
12.3
8.9
44.7
12.4
0.4
23.9
2.5
133.8
122.4
25.8
526.3
108.0
9.7
182.5
22.5
2280
1970
610
5830
2250
180
2340
250
6.3
4.8
1.6
32.0
4.3
0.2
10.2
1.8
210 62.4 165.7
220 56.6 161.1
b100 19.1 52.9
480 185.1 395.6
200 64.8 165.0
10
3.7
6.2
210 59.5 166.7
10
3.6
7.4
240 63.4 190.1
10
3.2
5.9
170 61.2 131.7
20
4.4
5.1
b100
b100
b100
b100
b100
2.5
2.6
0.1
4.9
2.6
0.1
2.5
0.1
2.5
0.1
2.6
0.1
b100
b100
b100
10.6 55.9
b0.2 46.7
b0.2 20.3
394.9 203.8
6.4 50.3
3.0
2.0
17.4 76.6
14.8 12.1
14.5 610.6 37.0
4.8 501.8 14.1
1.7
29.2
7.6
132.8 2730.0 690.0
4.1 501.1 22.3
0.3
55.5
5.6
34.2 817.9 64.8
9.6 185.6 36.3
0.4
0.3
0.0
1.9
0.4
0.0
0.4
0.0
0.4
0.0
0.3
0.0
4.0
b0.2
b0.2
103.5
4.8
1.7
3.0
1.3
4.6
1.7
3.1
1.1
15.3
13.6
6.9
45.6
18.6
6.9
14.3
1.0
16.2
3.3
14.6
1.5
to discriminate which category group or groups of data
differed when the analysis of variance rejected the null
hypothesis. Rank–order correlation (Spearmans rho)
was used to determine if the concentration levels of
elements in the paired cicada tissue and soil samples
were associated.
heavy metal concentrations (Tables 1 and 2), consistent
with the presence of Pb-arsenate pesticides residues at
these sites (Peryea, 1998). With the exception of As and
Pb values, the chemistry of B-horizon soils was similar
to other untilled soils in the eastern United States
(Shacklette and Boerngen, 1984).
4. Results and discussion
4.2. Cicada tissue chemistry
4.1. Soil chemistry
Average adult cicada body concentrations for As, Cu,
Hg, Pb, and Zn were 3, 62, 0.015, 0.4, and 166 mg/kg
(dry weight), respectively (Table 1). The adult cicadas
contained concentrations of metals similar to other
invertebrates, such as earthworms (Saxe et al., 2001),
consistent with the results of Clark (1992). The average
chemistry of adult cicada bodies from orchard sites was
compared to the Maximum Tolerable Dietary Level
(MTDL) Guidelines (Table 3) developed for small
Soil concentrations at the collection sites varied over
one order of magnitude for bioessential elements and
many metals and over two orders of magnitude for As,
Au, and Pb (Table 1). Approximately one third of the
soil sites were from orchard sites where arsenical
pesticides were used. Relative to non-orchard sites,
orchard site soils contain elevated Pb, As and other
372
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
Table 2
Summary of attained significance levels (p-values) for Mann–Whitney and Kruskal–Wallis tests on element concentrations on sample groups defined
by sample criteria
Sample Soil
Exoskeleton Cicada body
criteria site
site
Site
Sex
element
p-value
Ce
0.558 0.908
Na
0.557 0.497
Mg
0.693 0.032
P
0.912 0.633
S
0.912 0.090
K
0.613 0.524
Ca
0.370 0.391
Sr
0.884 0.313
Ba
0.446 0.972
Mn
0.013 0.025
Fe
0.083 0.507
Pb
0.0001 0.0001
Cu
0.018 0.0001
Zn
0.0007 0.0005
As
0.0001 0.0001
Hg
0.128 0.026
n
72
55
(Sample numbers by criteria)
Comments
Species
0.544
0.224
0.637
0.159
0.674
0.0004
0.528
0.794
0.017
0.458
0.057
0.150
0.676
0.0001
0.205
0.577
0.406
0.010
0.014
0.0001
0.006
0.030
0.0001
0.040
0.648
0.319
0.732
0.086
0.0001
0.004
0.795
0.001
0.001
0.092
0.016
0.809
0.161
0.398
0.097
0.969
0.0001
0.0002
0.307
0.548
0.0003
0.969
0.843
0.232
127
127
127
p-values in bold are different at 95% confidence level
M. cassini N M. septendecim, M. septendecula
Orchard N reference; M. cassini N M. septendecim or M. septendecula (split by sex)
Female N male
M. cassini N M. septendecim or M. septendecula (split by sex)
Reference N orchard; female N male; M. cassini N M. septendecim, M. septendecula
Reference N orchard; female N male; M. cassini N M. septendecim
Reference N orchard; female N male; M. cassini N M. septendecim
Female N male; M. cassini N M. septendecim
Orchard N reference; female N male
Orchard N reference
Orchard N reference; female N male; M. cassini N M. septendecim, M. septendecula
Orchard N reference; M. cassini N M. septendecula
Orchard N reference
Soil (orchard:reference, 27:45); exoskeleton (orchard:reference, 19:36)
Cicada body site (orchard:reference, 56:71) (female:male, 74:53)
Cicada body species (M. cassini:M. septendecim:M. septendecula, 74:46:6)
The p-values show the probability that the observed differences are due to chance rather than the factor tested; p-values significant at alpha = 0.05
(95% confidence) are shown in bold. The number of samples evaluated in each sample group is denoted by “n”. The comments column identifies the
direction of the differences within sample groups by sample criteria and identifies the number of samples tested by sample criteria.
barnyard animals and poultry (NRC, 1980, Table 1) to
evaluate if the heavy metal content of cicadas pose a
dietary threat for birds and other cicada consumers. In
all cases, the average cicada body concentrations were
less than half of the MTDL guidelines, implying no
significant dietary risk to consumers, even if adult
cicadas were consumed at twice normal dietary
requirements during the emergence event (Clark, 1992).
Periodical cicada emergence events coincide with, and
provide abundant food for, the nestling–rearing activities
of many birds. These nestlings might represent a special
population sensitive to heavy metal and toxic element
loading. Based on our data and on feeding rates calculated
from previous field and lab studies, Red-Winged
Blackbird (Agelaius phoeniceus) nestlings feeding on
cicadas may consume between 0.08–0.19 mg As, 0.01–
0.03 mg Pb, 1.8–4.6 mg Cu, and 0.42–1.10 ng Hg from
adult cicadas in the first 10–12 days before fledging
(Strehl and White, 1986; Yasukawa and Searcy, 1995;
Pope et al., 2001; Cartwright et al., 1998; Fiala and
Congdon, 1983, A. Clark and L. Yang, personal
communication). The xenobiotic metals, such as Pb and
Hg, may accumulate and concentrate in specific target
organs of the nestlings, such as the liver and kidney
(Hunter and Johnson, 1982). However, the concentration
levels of As, Pb, Cu, and Hg in adult cicadas from orchard
sites contaminated with arsenical pesticide residues are
comparable to the concentrations of these elements in
Table 3
Chemistry of adult cicada bodies and pre-molt nymph relative to maximum tolerable dietary level limits defined for small animals
Site
Mn
Fe
Cu
Zn
Al
As
Pb
Hg
59
56
200
167
168
500
b100
260
200
2
3
50
0.4
4.2
30
0.01
0.02
2.00
mg/kg dry weight
Adult cicada (mean)
Nymph (estimated mean)
MTDL† (NRC, 1980)
Orchard
Orchard
193
218
400
210
642
500
† Maximum Tolerable Dietary Level, minimum of poultry–rabbit levels.
The estimate of mean cicada nymph chemistry is based on a mass-weighted average of nymphal exoskeleton and adult body means.
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
other food sources, from uncontaminated sites, that are
likely to be consumed by nestlings, such as earthworms
(Saxe et al., 2001). It is unlikely that these metal loadings
exceed the detoxification capacity or homeostatic regulatory capability of the nestlings, although the evaluation
of biomagnification of heavy metals in natural ecosystem
foodchains is a controversial issue (Laskowski, 1991; van
Straalen and Ernst, 1991).
Average elemental concentrations in cicada tissues
differ by tissue type (Table 1; adult body vs exoskeleton).
Bioessential elements, such as Na, K, P, S, and Mg, were
concentrated in adult body tissues. Heavy metal
concentrations, with the exception of Cu, were low in
body tissues. Al, Fe, Ce, Pb, and Au were concentrated in
the nymphal exoskeleton. Ca, Mn, As, and Hg show
approximately similar distributions.
Many nocturnal predators, such as mice and shrews,
consume the vulnerable cicada nymphs before they shed
their nymphal exoskeleton. Using a mass-weighted
average of the mean exoskeleton and adult body
chemistry as an estimate of nymph chemistry, we estimate
that the mean Al and Fe values for cicada nymph slightly
exceed MTDL guidelines, whereas other element concentrations were below the MTDL guidelines (Table 3).
The cicada populations sampled in this study appear
to constitute a safe diet for birds and small mammals and
probably have lower average contamination levels than
do other invertebrates that may be consumed in this
geographic area. Consistent with the results of Clark
(1992), adult cicada bodies have higher concentrations
of Cu than host soil and other invertebrates, such as
earthworms. Periodical cicadas from areas with histories
of Cu contamination may need to be analyzed before the
safety of these cicadas can be evaluated. However, Cu is
an essential trace element in animal nutrition and
efficient homeostatic regulation of Cu is observed in
field populations of small mammals in areas of metal
contamination (Hunter and Johnson, 1982; Laskowski,
1991).
4.3. Soil and cicada chemistry in relation to site and
sample characteristics
Mann–Whitney (binary group) and Kruskal–Wallis
(multiple group) tests were used to identify differences
in chemical distributions in soils by site status and in
cicada tissues by site status, sex, and species classifications (Table 2). Cicada exuviae and soils collected from
orchard site soils contain elevated Pb, As and other
heavy metal concentrations (Cu, Zn) relative to nonorchard (reference) sites (Table 2), consistent with the
presence of lead-arsenate and other metal-based pesti-
373
cide residues at these sites (Peryea, 1998). The elevated
Mg and Hg contents of cicada exuviae from orchard
sites relative to reference sites may reflect interaction
effects with other metals in the contaminated soils. The
group differences in Mn concentrations between soil
(and exoskeleton) by site classification likely reflect
natural variation, not contamination, because Mn
compounds are not known to have been used as
agricultural chemicals.
Except for Ca and Sr concentrations, the chemistry
of adult cicada bodies collected from orchard sites did
not differ from non-orchard sites (Table 2). The lower
concentrations of Ca and Sr in cicada bodies from
orchard sites cannot be explained by differences in Ca
and Sr in the orchard site soils. Calcium and Sr have
similar chemical properties. The high Pb contents of
orchard soils contaminated by arsenical-pesticide
residues might limit Ca uptake and retention by cicada
nymph. Lead has been shown to inhibit nutrient
assimilation and water uptake in plant roots (Malecka
et al, 2001; Burzynski, 1988) and to disrupt Ca-ion
channel functions regulated by Ca-channel proteins in
animals (Anetor et al, 2005). The inhibition of Ca
uptake and retention may occur either at the soil–root
tip–xylem fluid interface (where Pb concentrations are
highest) or at the cicada nymph–xylem fluid interface.
The bioessential elements S, Ca, Mn, Fe, and Zn
differed in cicada bodies by sex; all elemental concentrations were greater in female bodies. These sex-based
differences may influence the dietary resources available
to cicada predators, as female cicadas are preferentially
consumed by some predators (Steward et al., 1988;
Karban, 1983). Both Red-Winged Blackbirds (Agelaius
phoeniceus) and House Sparrows (Passer domesticus)
preferentially forage on female cicadas, perhaps due to
their higher nutrient content or lack of a disturbance
squawk behavior (Steward et al., 1988; Karban, 1983).
Concentrations of Na, Mg, K, Ca, Mn, Fe, Zn, and As
differed in cicada bodies by species, and these
differences persist when the data are split by sex. For
all elements, concentrations in M. cassini are greater
than in M. septendecim or M. septendecula species
(Table 2). The observation of higher elemental concentrations in M. cassini cicadas might reflect different
habitat preferences between the different cicada species.
M. cassini prefers wetter floodplain sites compared to
the drier upland habitats of M. septendecim and M.
septendecula (Dybas and Davis, 1962). These wetter
habitats might explain the observed species differences
in elemental composition either through increased
concentration of some elements in these soils, or due
to increased absorption and assimilation rates from
374
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
Table 4
Summary of Spearmans rho rank-correlation coefficients for element concentrations between paired soil and tissue samples
Sample
Ce
Exoskeleton
0.658
Exoskeleton; Ce-normalized
data
Male body
(0.038)
Female body
(0.298)
Na
K
P
(0.196) (− 0.280) (0.265)
0.606 0.351
0.396
S
Ca
(− 0.028) 0.385
0.499
0.535
(0.124) (− 0.099) (− 0.070) (0.022)
(0.110) (0.271) (0.216) (0.078)
Mg
Fe
Cu
Pb
As
Au
0.581
0.786
0.647
0.848
0.473
0.662
0.754
0.907
0.669
0.636
0.747
0.638
(0.108) (−0.060) (0.140) (0.132) (0.094) (−0.005) –
0.444 (0.142) 0.487 (− 0.039) (0.290) (−0.371) –
Significant correlations, the probability that the observed correlations are due to the relation tested rather than to chance, at alpha = 0.05 (95%
confidence level), are shown without brackets. Ce-normalized data correlation coefficients for cicada bodies are similar to the raw data correlations;
most are not significant at 95% confidence.
wetter soils. An alternate hypothesis is that since M.
cassini is substantially smaller than M. septendecim
(Karban, 1983), this species has a higher proportional
ratio of body mass to exoskeleton that might lead to a
concentration of elements in body tissues. However, M.
septendecula is almost the same size as M. cassini, but
does not have the same elevation of chemical constituents, favoring the habitat hypothesis.
4.4. Soil–cicada tissue correlation
Metal concentrations in nymphal exoskeletons tended
to vary in relation to soil chemistry as indicated by the
consistency in elemental group differences for exoskeleton and soil chemistry as classified by site (Table 2) and
by the rank-correlation (Spearmans rho) coefficients
between paired soil and tissue samples (Table 4).
Spearman correlation coefficients, relating concentrations
of elements such as Ce, Ca, Mg, Fe, Cu, Pb, and As in
exoskeleton to soil, were consistently positive and low to
moderate (0.385 b r b 0.754) in value (Table 4). For soil
and exoskeleton chemical data normalized relative to Ce,
a relatively inert soil constituent, the Spearman correlations increased in value and are significant for more
elements (Ce-normalized data in Table 4). These results
imply that assimilation through prolonged soil contact
may influence the concentrations of these elements in the
nymphal exoskeleton.
In contrast, the cicada body-soil correlation coefficients for these elements were low (− 0.371 b r b 0.487),
inconsistent by sex category, and generally not significant at the 95% confidence level. Ce-normalized
Spearman correlation coefficients for cicada bodies
relative to soil are not shown in Table 4, but were similar
to the raw data correlations in being inconsistent by sex
category and generally not significant at the 95%
confidence level. These data indicate that biological
regulation processes influenced cicada body chemistry,
consistent with the geochemical features of other soil
invertebrates, such as earthworms (Saxe et al., 2001).
4.5. Soil–exoskeleton comparison
The soils contained approximately 1 to 12% of Albearing clay minerals, by weight. These clay minerals
are relatively insoluble in soil fluids and soil concentrations of Al were high relative to the concentration of Al
in soil pore waters and xylem fluids (Smith and Shortle,
2001). The high concentration of Al in cicada
exoskeleton, in contrast to the low concentration of Al
in the adult cicada body, was likely the result of 1) direct
exposure to insoluble clay minerals in soil involving
adhesion and adsorption processes, or 2) efficient
sequestration of Al derived from xylem fluids by the
nymph into exoskeleton.
The role of passive assimilation (adhesion–adsorption) resulting from cicada nymph exposure to soil
relative to other biological regulation processes was
assessed by normalizing the cicada tissue chemistry to
relatively insoluble, but abundant, soil constituents such
as Al and Ce (the precision for Ce is better than Al for
our dataset). For cicada exoskeletons, Al:Ce:Pb ratios
were similar to soil ratios and the data follow a 1:1 trend
line over two orders of magnitude (Fig. 3A). Concentrations of Pb, Al, and Ce in cicada exoskeleton appear
to be largely controlled by soil adsorption or adhesion.
Normalized plots for bioessential elements (Ca, P,
and Na) indicate correlation with soil chemistry and
demonstrate consistent fractionation into exoskeleton
tissue (Fig. 3B). These bioessential elements are
expected to fractionate into biological tissues. Sulfur
and As also preferentially fractionate into exoskeleton
(data not shown in graph), but their fractionation
appears to be biologically regulated as well as varying
as a function of soil chemistry (resulting in weak Cenormalized Spearman correlations, Table 4).
Similar plots for chalcophile elements (Cu and Au
are shown examples, Fig. 3C and D) exhibit consistent
and strong fractionation into exoskeleton tissue from
soil. The strong enrichment in Au (fractionation factor
of approximately 50) is surprising for a reported bio-
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
375
Fig. 3. Comparison of cicada nymphal exoskeleton geochemistry with soil geochemistry. Element ratios are normalized by Ce or Al concentrations.
Ce and Al are considered to be inert and insoluble soil constituents. Element ratios falling along the 1:1 line indicate tissue concentration ratios
equivalent to those in host soil. Element ratios falling above the 1:1 line indicate biogeochemical fractionation into the exoskeleton tissue relative to
concentrations of inert constituents Ce and Al.
nonessential element, but is consistent with some other
data reported for keratin samples, such as Desert
Tortoise scute (keratin fraction of shell) (Haxel et al.,
2000) and other biota (Reith, 2003).
4.6. Cicada tissue–soil comparison
The general pattern of biogeochemical fractionation
into cicada tissues is shown in the plot of normalized
cicada tissue chemistry relative to soil chemistry
(Fig. 4).The fractionation factors, shown on the logtransformed Y axis, are cicada tissue chemistry,
normalized by Ce concentrations, divided by Ce-normalized soil concentrations. Symbols occur at the
median fractionation factor, bounded by the 10–90
percentile envelopes, for the sample population groups.
A fractionation factor of 1 indicates a pattern consistent
with passive assimilation resulting from soil contact.
Fraction factors N 1 indicate preferential incorporation of
the element from soil conditions into biological tissues.
Normalized concentrations of Al, Pb and the
transition metals Fe, Co, and Ni in both exoskeleton
(Fig. 4a) and cicada body tissues (Fig. 4b) indicate
fractionation ratios of approximately 1, consistent with
their concentrations being controlled by passive assimilation through dermal exposure. Bioessential (Ca, P,
Na) and chalcophile (Cu, Zn, Mo, Au) elements show
strong enrichment in cicada tissues relative to soil and
the fractionation patterns are similar for both nymphal
exoskeleton and adult body tissues (although the
376
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
fractionation values are generally higher for the body
tissues). The strongest elemental enrichments for both
tissue types are for S and for the chalcophile elements,
consistent with the hypothesis that chalcophile elements
may be bound to sulfhydryl groups in organic tissues.
Divergence of the mean fractionation values for male
versus female bodies in Fig. 4b indicate preferential
enrichment in female bodies of the elements: S, Ca, Cr,
Mn, and Mo, consistent with the differences shown in
Table 2.
Cicadas feed on xylem fluids from plant roots, and
xylem sap is one component of the linkage between the
soil environment and cicada physiology. We could not
identify the host tree for the individual adult cicadas
collected in our study and we did not sample tree root
sap for comparison with soil chemistry. Although the
chemistry of xylem sap in tree roots varies in relation to
soil element availability and geochemical discrimination
in root uptake by plant species, the ionic chemistry of
xylem sap appears to be largely regulated by the plant
(physiologic homeostatic control) in response to conditions in their soil environment (Smith and Shortle,
2001). Xylem fluid from various host plants for
periodical cicada show similarities in ionic composition
(Cheung and Marshall, 1973). We use some limited, but
available, data on tree root sap chemistry, in relation to
soil data, to illustrate what we believe is likely to be a
general pattern of homeostatic control on xylem
chemistry relative to soil chemistry (Fig. 4b). The tree
root sap (xylem) enrichment values shown in Fig. 4,
derived from data for red spruce reported by Smith and
Shortle (2001), are not from a host tree for periodical
cicada in the study area. However, this data is probably
representative of the likely variation in element
composition of typical root xylem fluids relative to
host soil chemistry. The pattern of geochemical
enrichment in cicada tissues relative to soil conditions
is similar to those observed in these reference xylem
fluids from tree roots. The cicada body composition
appears to reflect xylem composition (Fig. 4b) much
more than soil composition (Table 4), whereas the
nymphal exoskeleton reflects soil (Table 4) better than
xylem fluid (Fig. 4).
4.7. Gold in cicada tissues
The observed correlation between Au concentrations
in paired soil and nymphal exoskeleton samples
(Table 4, Fig. 3D) and the strong enrichment in Au
concentrations in the nymphal exoskeleton relative to
host soil concentrations (Table 1) implies the presence
of mechanisms to dissolve and transport Au in the soil
Fig. 4. Cicada tissue:soil geochemical ratios normalized relative to Ce. Data are shown for cicada nymphal exoskeleton, adult female bodies, and adult
male bodies. Symbols are plotted at the median value of the geochemical ratio for each sample population group, bounded by the 90th and 10th
percentile value for the geochemical ratio. Element ratio values of 1 indicate a biogeochemical fractionation similar to cerium, inferred to be an inert
soil component. Element ratios N1 indicate preferential incorporation of the element from soil conditions into biological tissues. Tree xylem root sap
data is from Smith and Shortle (2001).
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
environment and concentrate Au in keratin-rich tissues
at the soil interface. Gold in soils is likely transported as
organometallic or chelated compounds (Boyle et al.,
1975). Root exudates of cyanogenic plants and some
deciduous trees are known to dissolve Au, which can be
transported across the root interface and accumulate in
plant tissues (Lakin et al, 1974; Shacklette et al., 1970).
The Au content of xylem fluid has been observed to be
significantly higher than either the aqueous extracts of
the plant material or of the soil itself (Kyuregyan and
Burnutyan, 1972; Krendelev and Pogrevniak, 1997;
Boyle, 1979). In addition, the cicada nymph voids most
of the ingested xylem fluid, and this copious dilute urine
contains amino acids and other organic compounds that
might influence Au accumulation (Cheung and Marshall,1973). The nymph and adjacent soil is exposed to
this urine bath for several years. It seems possible that
this urine could be similar to xylem fluid in Au
concentration. The extreme enrichment of Au in
exoskeleton tissues might reflect the solubilization and
mobility of Au in the soil environment by organic acids,
thiosulfate, and cyanogenic metabolites exuded by tree
roots (Reith, 2003; Neumann and Romheld, 2001) or
excreted by the cicada nymph from metabolized xylem
fluids, coupled with precipitation from these soil fluids
into keratin (c.f. Ishikawa and Suyama, 1998).
5. Conclusions
Concentrations of potentially toxic elements in adult
cicada bodies are low and probably do not pose a
significant risk to birds and other consumers that
preferentially feed on cicadas during emergence years.
These findings support the conclusions of previous
studies investigating metal accumulation in cicadas
(Clark, 1992). Periodical cicadas can represent a
substantial flux of nutrients (Yang, 2004, 2006) and
other chemicals from soil environments into aboveground food webs during a substantial resource pulse.
These data suggest that Fe, Al, Pb, and As are
concentrated in the nymphal exoskeleton, which is
shed within hours of emergence. Metal concentrations,
except for Cu, are low in adult cicada bodies. The
shedding of the nymphal skin may reduce the risk of
metal exposure to predators of adult cicadas, but
nocturnal consumers of the cicada nymphs, such as
mice and shrews, may experience greater exposure to
accumulated metals.
Normalized concentrations of bioessential and chalcophile elements are enriched in cicada tissues relative
to soil. These metals may bind to sulfhydryl groups in
organic tissue. The patterns of element enrichment
377
relative to soil concentrations are similar between
nymphal exoskeleton and adult body tissues. The
enrichment patterns in the tissues are similar to
enrichment patterns reported in the literature for tree
root sap, implying that cicada nymphs might be able to
minimize their internal energy requirements for homeostatic control by utilizing the homeostatic processes
exerted by their host tree biology. Although soil
chemistry does not appear to strongly influence cicada
body chemistry, our data support the hypothesis that
high Pb soils might inhibit Ca (and Sr) uptake and
retention by cicada nymphs.
The strongest fractionation observed is for Au, and
might reflect the solubilization and mobility of Au in the
soil environment by organic acids, thiosulfate, and
cyanogenic metabolites exuded by tree roots and
excreted by cicada nymph. Soluble Au has a strong
affinity for keratin (c.f. Ishikawa and Suyama, 1998)
and Au concentrations measured in nymphal exoskeleton are very high relative to soil conditions.
Acknowledgements
The authors thank Larry Gough and Helen Folger, U.
S. Geological Survey, and an anonymous reviewer for
their comments and suggestions to improve the
manuscript. The authors also thank Peter Larkins and
Brad Reed who prepared figures and graphics.
References
Actlabs, Method description. http://www.actlabs.com/methods_usa.
htm. July 2006.
Anetor JI, Akingbola TS, Adeniyi FA, Taylor GO. Decreased total and
ionized calcium levels and haematological indices in occupational
lead exposure as evidence of the endocrine disruptive effect of
lead. Indian J Occup Environ Med 2005;9:15–21.
Boyle RW. The geochemistry of gold and its deposits, together with a
chapter on geochemical prospecting for the element. Geol Surv
Can Bull 1979;280: 584 pp.
Boyle RW, Alexander WM, Aslin GEM. Some observations on the
solubility of gold. Geological Surv. Canada Paper, vol. 75. 1975.
24 pp.
Burzynski M. The uptake and accumulation of phosphorus and nitrates
and the activity of nitrate reductase in cucumber seedlings. Acta
Soc Bot Pol 1988;57:349–59.
Cartwright MA, Beauchamp DA, Bryant MD. Quantifying cutthroat
trout (Oncorhynchus Clarki) predation on sockeye salmon
(Oncorhynchus Nerka) fry using a bioenergetics approach. Can J
Fish Aquat Sci 1998;55:1285–95.
Cheung WWK, Marshall AT. Water and ion regulation in cicadas in
relation to xylem feeding. J Insect Physiol 1973;19:1801–16.
Clark Jr DR. Organochlorines and heavy metals in 17-year cicadas
pose no apparent dietary threat to birds. Environ Monit Assess
1992;20:47–54.
378
G.R. Robinson Jr. et al. / Science of the Total Environment 374 (2007) 367–378
College of Mount St. Joseph. 2004 17-year periodical cicada
emergence map. http://www.msj.edu/cicada/. May 2004.
Conover WJ, Iman RL. Rank transformations as a bridge between
parametric and nonparametric statistics. Am Stat 1981;35:124–9.
Day E, Pfeiffer D, Lewis E. Periodical Cicada. Virginia Tech
Publication number 444–276; 2002. http://www.ext.vt.edu/pubs/
entomology/444–276/444–276.html. May.
Dybas HS, Davis DD. A population census of seventeen-year
periodical cicadas. Ecology 1962;43:432–44.
Fiala KL, Congdon JD. Energetic consequences of sexual size dimorphism
in nestling red-winged blackbirds. Ecology 1983;64:642–7.
Haxel GB, Knight RJ, Berry KH, Homer BL, Chaffe MA. Trace
element systematics of desert tortoise shell. Proceedings 25th
Desert Tortoise Council Symposium, April 21–24, 2000, Las Vegas,
Nevada. http://www.deserttortoise.org/abstract/abstracts2000/
2000abs15.html. June 2006.
Helsel DR, Hirsch RM. Statistical methods in Water Resources. US
Geol Surv techniques of water resources investigations, book 4
(chapter A3); 2002. http://water.usgs.gov/pubs/twri/twri4a3/. May
2005.
Hollander M, Wolfe DA. Nonparametric statistical inference. New
York: John Wiley & Sons, New York; 1973.
Hunter BA, Johnson MS. Food chain relationships of copper and cadmium
in contaminated grassland ecosystems. Okios 1982;38:108–17.
Ishikawa S, Suyama K. Recovery and refining of Au by gold–cyanide
ion biosorption using animal fibrous proteins. Appl Biochem
Biotechnol 1998;70–72:719–28.
Karban R. Flight and dispersal of periodical cicadas. Oecologia
1981;49:385–90.
Karban R. Sexual selection, body size, and sex-related mortality in the
cicada Magicicada cassini. Am Midl Nat 1983;109:324–30.
Krendelev FP, Pogrevniak IF. Possibility of searches for ore
manifestations of gold and zinc according to their contents in
birch sap. Akad Sci USSR Dokl 1997;234:184–7.
Kyuregyan EA, Burnutyan RA. Gold in the juice of plants and a
method for its detection. Akad Nauk Arm SSR, Izv Nauki Zemie
1972;25:83–5 (Chem Abstr 81: 130520c).
Lakin HW, Curtin CG, Hubert AE. Geochemistry of gold in the
weathering cycle. US Geol Surv Bull 1974;1330, 80 pp.
Laskowski R. Are the top carnivores endangered by heavy metal
biomagnifications? Okios 1991;63:387–90.
Maier CT. A mole's-eye view of seventeen-year periodical cicada
nymphs, Magicicada settendecim. Ann Entomol Soc Am
1980;73:147–52.
Malecka A, Jarmuszkiewicz W, Tomaszewska B. Antioxidative
defense to lead stress in subcellular compartments of pea root
cells. Acta Biochim Pol 2001;48:687–98.
Marlatt CL. The periodical cicada. USDA Bur Entomol Bull 1907;71:
181 pp.
National Research Council (NRC). Mineral tolerance of domestic
animals. Nat Acad Press, Washington D.C., 1980: 577 pp.
Neumann G, Romheld V. The release of root exudates as affected by the
plant physiological status. In: Pinton R, Varanini Z, Nannipieri Z,
editors. The rhizosphere: biochemistry and organic substances at
the soil–plant interface. New York: Marcel Dekker; 2001. 424 pp.
Peryea FJ. Historical use of lead arsenate insecticides, resulting soil
contamination and implications for soil remediation. Proc. 16th
World Congress of Soil Science. Montpellier, France; 1998.
Science Reg. No. 274. Symposium 25.
Pope K, Brown ML, Duffy WG, Michaletz PH. A caloric-based
evaluation of diet indices for largemouth bass. Environ Biology
Fisches 2001;61:329–39.
Reed BW, Larkins P, Robinson Jr GR. Digital data set of orchards
where arsenical pesticides were likely used in Clarke and Frederick
Counties, Virginia and Berkeley and Jefferson Counties, West
Virginia. U.S. Geol. Surv. Open-File Rept. 2006–1330; 2006.
http://pubs.usgs.gov/of/2006/1330. October.
Reith F. Evidence for a microbially mediated biogeochemical cycle of
gold — a literature review. In: Roach IC, editor. Advances in
Regolith. CRC LEME; 2003. p. 336–41.
SAS Institute Inc. SAS OnlineDoc®. Ver. 8. Cary, NC: SAS Institute
Inc; 1999.
Saxe JK, Impellitteri CA, Peijnenburg WJ, Allen HE. Novel model
describing trace metal concentrations in the earthworm, Eisenia
Andrei. Environ Sci Technol 2001;35:4522–9.
Shacklette HT, Boerngen JG. Element concentrations in soil and other
surficial materials of the conterminous United States. U.S. Geol.
Surv. Prof. Paper 1270; 1984.
Shacklette HT, Lakin HW, Hubert AE, Curtin GC. Absorption of gold
by plants. US Geol Surv Bull 1970;1314b: 23 pp.
Smith KT, Shortle WC. Conservation of element concentration in
xylem sap of red spruce. Trees – Struct. Funct 2001;15:148–53.
Sokal RR, Rohlf FJ. Biometry. San Francisco, Calif: W.H. Freeman;
1969.
Steward VB, Smith KG, Stephen FM. Red-winged blackbird predation
on periodical cicadas. Oecologia 1988;76:348–52.
Stoline MR. The status of multiple comparisons — simultaneous
inference of all pairwise comparisons in one-way ANOVA designs.
Am Stat 1981;35:134–41.
Strehl CE, White J. Effects of superabundant food on breeding success and
behavior of the red-winged blackbird. Oecologia 1986;70:178–86.
van Straalen NM, Ernst WHO. Metal biomagnifications may endanger
species in critical pathways. Oikos 1991;62:255–6.
Wheeler GL, Williams KS, Smith KG. Role of periodical cicadas
(Homoptera:Cicadidae:Magicicada) in forest nutrient cycles. For
Ecol Manag 1992;51:339–46.
White J, Lloyd M. Growth rates of 17- and 13-year periodical cicadas.
Am Midl Nat 1975;94:127–43.
Williams KS, Simon C. The ecology, behavior, and evolution of
periodical cicadas. Annu Rev Entomol 1995;40:269–95.
Yang LH. Periodical cicadas as resource pulses in North American
forests. Science 2004;306:1565–7.
Yang LH. Interactions between a detrital resource pulse and a
detritivore community. Oecologia 2006;147:22-532.
Yasukawa K., Searcy W.A. Red-winged blackbird (Agelaius phoeniceus). In: Poole AF, Gill F, editors. The birds of North America,
No. 184. The academy of natural sciences, Philadelphia, and the
American Ornithologists’ Union, Washington, D.C., 1995.