ANTHONY D. BARNOSKY MARC A. CARRASCO

1.0
Results from the Northern Rocky Mountains
Species Richness
2.0
Atlantic d18O
1.0
Under our MIOMAP Project, we are compiling all mammal records of Miocene age in
the U.S. A Geographic Information System (ARC/INFO) and other statistical techniques
are used to determine how closely species turnover patterns correlate on a region-byregion basis with the timing of the mid-Miocene Climatic Optimum and with tectonic
activity in the northern Rocky Mountains and Great Basin.
The maps to the right indicate areas from which fossil
data have been compiled and entered in the MIOMAP
database. Only the northern Rocky Mountain data
(Idaho, Montana, Wyoming) are analyzed in this poster.
Most of the areas indicated by white dots contain
multiple collecting localities. Numbers of specimens
range from tens to thousands per locality. A full
explanation of the MIOMAP project is available at
http://ib.berkeley.edu/labs/barnosky.
HE1
AR4
AR4
AR3
AR3
O
L
I
G
O
C
E
N
E
C.
80
20
60
40
20
0
Northern Rockies
Species Richness
25
10
1
Number of Localities Collected
9.
20
20
Split Rock (He 2&3)
15
10
5
0
-.25
.25
.75
1.25
log NISP
1.75
2.25
16
12
E.
Mid/Upper Cabbage Patch
and Peterson Creek (Ar 2)
Barstovian 2
Localities
8
4
0
-.25
ALPHA DIVERSITY (Graphs D, E, F)
Species diversity at discrete localities
may have increased coincident with
the Miocene Climatic Optimum.
The He2&3 localities contain the
most species at sample sizes of 30
to 200 specimens. However, these
point estimates are potentially
subject to time and community
biases.
.25
.75
1.25
1.75
2.25
log NISP
F.
40
35
30
Split Rock (He 2&3)
Barstovian 2
Localities
25
20
15
10
5
0
-.5
0
.5
1
BETA DIVERSITY (Graphs G, H, I)
Bootstrap analyses that accumulate
species with successive samples
suggest higher species richness in
He2&3 and Ba2 (Graph G);
however, these results may simply
be an artifact of differing numbers
of individual specimens (NISP).
Bootstrap analyses that control for
numbers of specimens per sample
and assume either equal (Graph H)
or best estimate time spans
(Graph I) for each collecting area
suggest a decrease in species
richness across the warming event.
However, these analyses may be
biased if different habitats were
sampled.
1.5
2
2.5
3
3.5
Hemingfordian
(21 locs.)
es
M
n'
s
0
1
7 9 11 13 15 17 19 21 23 25
Number of Samples
3 5
H.
60
50
40
ge
a
bb
L.
1(
Ca
30
20
10
)
ch
t
a
P
)
ock
R
t
li
Ar
(Sp
3
&
r) He2
lte
o
C
2
Ba
10
I.
60
50
40
10
20
30
Square Root of NISP
40
30
20
Antelopes /
Oreodonts x 10
10
0
16
12
8
4
(Oreodonts /
Horses) x 10
0
Begin
Warming
)1
a
es
M
u
b
ep
(H
's
rn
Within the northern Rockies coincident with the late-early Miocene Climatic
Optimum:
Ma
.5 Ma
ock) 1
plit R
3 (S
He2&
Ar2 (Peterson Creek) 3.5 Ma
10
0
40
Ba2 (Colter) 2 Ma
0
10
Conclusions
2
20
Pocket Mice /
Gophers
p
He
2(
Ba
30
20
Rhinos /
Horses
Pocket Mice /
Gophers
Post-Warming
r
bu
Ba
0
Pre-Warming
a)
n's
2(
Antelopes /
Oreodonts
s
Me
0
0
log NISP
Arikareean
(236 locs.)
Ar2 (Peterson Creek) 16.8
10
Camels /
Oreodonts
The relative abundance of species within
higher taxa (left graph) changes dramatically
at the warming event. Pocket mice replace
gophers. Antelopes, camels, and horses
become more dominant over rhinos and
oreodonts.
The available data suggest shifts in the
relative abundance of individuals within
species (right graph) accompanied the
warming event. Numbers of identifiable
specimens (NISP) overestimate actual
numbers of individuals, but nevertheless
provide a reliable gauge of relative
abundance. Unfortunately, requisite data
for the critical time intervals Ar3, Ar4, and
He1 do not yet exist.
20
h)
Mid/Upper Cabbage Patch
and Peterson Creek (Ar 2)
30
tc
25
RICHNESS = Species / Interval Length
40
Pa
40
)
a
abb
ock
C
R
.
17.7
(L
lit
)
1
p
h
r
c
A
Pat
3 (S
e
&
g
2
a
He
abb
C
.
(M.U
2
r
A
e
30
a)
14
20
50
9
94.
ag
10
30
60
6
(Oreodonts+
Camels+
Antelopes)
/ Horses
.5
) 23
h
c
at
ge P
.
1 Cab
(L
. C bag
ab e P
ba
a
ge tch
Pa ) 2.
2
tc
M
h)
a
2
M
a
0
G.
10
Beginning of
Warming Event
Ar
AR1
30
D.
100
7
AR2
14
bu
r
AR1
15
HE2/3
HE1
AR2
Length of Time Interval (Ma)
.U
25
BA1
4
ep
20
BA2
3
2
(M
HE2/3
1
(H
BA1
0
2
15
18
0
2
BA2
BA3
20
so
Ar n Cre
e
2
(M k)
.U
.C
ab
b
BA3
M
I
O
C
E
N
E
10
Ar
This map exemplifies output from climate
SUMMER SOIL MOISTURE
models that assume continued growth of
world greenhouse gas emissions (source:
Climate Change Impacts on the United
States, U.S. Global Change Research
Program). All models predict resultant
global warming will change local climatic
parameters to different extents. Our study
area (white line) represents a region that
shows a more or less uniform climate
response given modern geographic
http://www.gcrio.org/NationalAssessment/overpdf/overview.html
constraints. While climatic patterns in the
Miocene were clearly different than these, we assume that the study area also represented
a uniform climatic zone then. Our lab currently is working on ways to more accurately
delineate Miocene climate patterns in the western U.S.
CL1
Species Richness
Species Richness
We test these suppositions by examining effects of the late-early Miocene Climatic
Optimum warming event on faunas of the northern U.S. Rocky Mountains. The
warming event began 18.5 Ma, peaked around 17 Ma, and lasted until about 14 Ma.
Thus it is of a temporal scale that characteristically demonstrates little correlation
between global climate change and faunal turnover patterns. However, the geographic
scale is small enough to prevent mixing of substantially different climatic zones as seen
in the climate prediction model below.
CL1
Difference in Species Richness
(Lower minus Higher Interval)
40
(Pe
ter
We suspect that studies that focus on too large a scale blur the effect of climate change
on mammalian communities, because a given global change elicits differing local
climatic responses. We also suspect that species richness by itself can be an ambiguous
gauge of change in mammalian communities, in part because it is so difficult to reliably
determine in paleontological samples.
CL2
0.3
Study Goals and Methods
CL2
20 40 60 80
Relative Abundance
60
Ba
As the scales resolve to local sites over decades, centuries, or millennia, however, some
species clearly appear and disappear. For example, faunal effects that have been attributed
to Pleistocene climate changes in numerous studies include changes in: (1) species
composition of communities, (2) biogeographic ranges of individual species, (3) relative
abundance of species and individuals, (4) phenotype/genotype, and (5) extinction rates.
10
HP1
Ar2
Because species richness correlates with the
duration of the time interval spanned (shown
in Graph B), analyses in which time intervals
are standardized (Graph A) are frequently used.
A comparison of the Atlantic d18O curve
during the Miocene to a time-standardized
species richness curve for the northern Rocky
Mountains suggests an increase in richness
corresponds with the warming event
(highlighted in yellow). However, as shown
in Graph C, species richness also correlates
with the number of localities sampled, which
is a proxy for the number of individual
specimens collected. Other analytical methods
(Graphs D through I) attempt to account for
such potential biases but provide conflicting
results.
Age in Millions of Years
Atlantic d18O Curve
HP1
(Col
ter) 7
5
Ba2
15
80
Estimated Species Richness
25
HP2/3
Estimated Species
Richness/ Time Interval
35
Species Richness
45
Species Richness
55
0
B.
5
5
HP2/3
65
Age Ma
Ar
1
Ar
2
Ar
3
Ar
4
He
He 1
2&
3
Ba
1
Ba
2
Ba
3
1.5
Species Richness
A.
2.0
Age
Estimated Species Richness
Ma
Species Richness
3.0
Ratios of Numbers of Species
in Compared Groups
Number of Species
-60 -40 -20
2
0
B CA
Ba
100
carrasco@socrates.berkeley.edu
(510) 642-5318
3
-1.0
University of California
Berkeley, CA 94720
0.2
0.1
0
2&
barnosky@socrates.berkeley.edu
(510) 643-6275
0.5
0.4
0.3
He
Time Period
of Study
Atlantic d18O
Curve
MARC A. CARRASCO
The survival index provides a crude estimate
of faunal turnover and is computed as SI=
(Survivors/Species Pool), where 'Survivors' is
the number of species surviving across an
interval boundary, and 'Species Pool' is the
number of species found in the lower interval.
Plotted as a function of the difference between
species richness in adjacent stratigraphic
intervals, anomalous SI values appear
coincident with the onset of the warming event.
A and B designate the points for the transition
between He1 and He2&3 for the entire dataset
and for the 'mountains only' dataset,
respectively. C denotes the Ar1-Ar2 transition
for the localities near the Wyoming-Nebraska
border only.
Ar
2
ANTHONY D. BARNOSKY
-2.0
Atlantic d18O
Number of Species
200
MUSEUM
of
MUSEUM
of
and
PALEONTOLOGY
PALEONTOLOGY
Faunal Turnover
Survival Index (SI)
Studies that focus on large geographic (continental) and long temporal scales (millions
of years) tend to demonstrate little correlation between climate change and mammalian
species turnover rates or diversity as shown below in a graph that compares the North
American mammal species richness through time to the Atlantic d18O curve (sources of
data: Alroy, 2000, Geology 28:1024; and Miller et al., 1987, Paleoceanography2:3).
EFFECTS OF CLIMATIC AND TECTONIC
PERTURBATIONS ON MAMMALIAN COMMUNITIES
Ratios of Numbers of Specimens
(NISP) in Compared Groups
Introduction
20
30
Square Root of NISP
40
• Alpha diversity most likely increased
• Beta diversity may have changed
• Species compositions changed dramatically
• Relative abundance of species within higher taxa shifted
• Relative abundance of individuals within species shifted
Future Work
Barstovian
(105 locs.)
Tectonic uplift as well as climate change characterized the study area during
the Miocene Climatic Optimum. As the MIOMAP database becomes more
complete, we will test hypotheses designed to tease apart the relative effects
of climate change and tectonism on mammal faunas. Currently we are
analyzing data we have just finished compiling for the northern Great Plains
and Pacific Northwest.
Funded by