Butler University Butler University
Digital Commons @ Butler University Digital Commons @ Butler University
Scholarship and Professional Work - LAS College of Liberal Arts & Sciences
2005
Descriptive ecology of a turtle assemblage in an urban landscape Descriptive ecology of a turtle assemblage in an urban landscape
Conner C. A
B A. Douthitt
Travis J. Ryan
Butler University
, tryan@butler.edu
Follow this and additional works at: https://digitalcommons.butler.edu/facsch_papers
Part of the Animal Sciences Commons, and the Biology Commons
Recommended Citation Recommended Citation
Conner, C. A.*, B. A. Douthitt*, and T. J. Ryan. 2005. Descriptive ecology of a turtle assemblage in an urban
landscape. American Midland Naturalist 153:426-435.
This Article is brought to you for free and open access by the College of Liberal Arts & Sciences at Digital
Commons @ Butler University. It has been accepted for inclusion in Scholarship and Professional Work - LAS by an
authorized administrator of Digital Commons @ Butler University. For more information, please contact
digitalscholarship@butler.edu.
Am. Mid!. Nat.
1.53:428-435
Notes aiid Discussion
Descriptive Ecology
of a
Turtle Assemblage
in an
Urban Landscape
cr.—We studied luixle populations inhabiting a canal and a lake (both mau-tnadc)
within a hoavi!y disturbed, urban setting. Six aquatic and semi-aquaiic turtle .species were
collected in both liabitats: spiny softshell turtle (Afiohne
apinifera).
painted turtle (Chrysimys
pkta),
coninmn snapping tnil!e {Uielydra strfuniliua), common map turtle {Grnptpmys
gungrtiphira).
common musk turiJe (Sttmolhentf odimilus) and rod-caied siidcr
{Trachemys
srripla).
VVIiik'
('-.
gpograpliiai
WAS
the must common species in the canal habit:it, 7! sniplfi
WAS
most (ommon in ihe lake habitat. We describe patterns of sexual size tlinioiphism and sex
rdtio.s for the three most abundant species (fi.
gengmphica,
T.
scripta
and S.
udoralus).
We
discuss our data in light of problems facing turtle assemblages in urban settings.
iNTRODlKmON
Habitat conversion and degradation is generally recognized as the most pervasive and important of
the six major threats to biodiversity (other tJireats being invasive species, enviionmenlal poltutioii,
disease/parasitism, unsustainable use and global climate change; Gibbons ft
uL.
2t)00). The major effect
of habitat conversion is the outright loss of critical habitats for essential life functions, including feeding
{Vickery
el
«/., 2001), courting and nesting (Heckert
et
ai. 2(>03) and hibernation (Ball, 2002). Habitat
conversion as the result of increasing urbanization, in particular, affect.s a wide aiTay of
organisms,
from
large carnivores (Reilty
et
al..
2003) tn buticrflies (Collingc
et al..
2003) to plants (Frtinsisco-Ortgca
et
al.,
20((0) in terrestrial situations and from salaniatiders (Willsoti and Dorcas, 2003) to fish (Paul and Meyer.
20111) to algae (Fore and Gmle, 2002) in aquatic environments.
Tiuile populations have Ijecn significantly impacted by human activity, development and
urbanization. Negative effects include (ragincntation of genetic structure (Rubin el riL, 2001).
demographic effects (Garber and Burger, 1995; Lindsay and Dorcas, 20tll) and direct mortality
{e.f(.,
through collision witli autotnobiles, Gibbs and Sliriver, 2002). Nonetlieless. some turtle species may be
very resilient in the face of human activity and continue to exist In highly modified habitats when other
i\itdlife is extirpated (Mitchell, 1988). Data on the specific impacts of human activity on turtle
l>opulations and assemblages, and how these effects may be amelionited, provide essential components
ro sound consen-ation practices in human-<iominateii landscapes. The pnr|)osf of the present study is lo
understand the basic ecology of
a
turtle assemblage living within au urban landscape. These descriptive
population and community ecology data can then sei"ve as a bas<'line For more thorough investigations
of the effects of urbanization.
MATEWALS
AND
METHODS
The Central Canal is a man-made riverine habitat created in the 1830s in Indianapolis, Indiana, the
12"'
largest city in tin; USA (2000 census population 791,90tVf residents). The remnant of
a
much laiger
uncompleted canal sysietn, the Cetural Cunal originates from the White River and Mows south through
commercial, residential and recreational areas for 11.2 km. At least a dozen roads cross the cajial.
including fom' major thoroughfares and one iniriNtate highuay. At the scjuthem terminus, the canal
enters a water treatment facility operated by the Indianapolis Water Company (IWt;). The canal trans-
ports approxitnately 70% of the city's annual water use; water
level,
flow rate, submergent and emergent
aquatic vegetation are all controlled in part by the IWC. The canal varies from 15 to 25 m wide and is
usually less than 2 m at its deepest points. Shorelines aie practically non-existent in most places, with
banks
1—2
m high on either side. P'ragmented woodloLs border portions of the canal and fallen trees and
snags serve ;is basking sites; however, many of these basking sites are removed on a regular basis.
.Approximately ft.5 km of the canal (7ft%) is bordered by a greeniv'ay (the Central Canal Towpath)
maintained by IndyParks, the (^ity of Indianapolis Deparunent of Parks and Recreation. Most of our
field work f(»r this sttidy in the cuna! was in this 8..5 km section. In this section, the canal is never more
than 1 km from the White River and is as close a.s 2.'V-4O m at several points.
428
2005
NOTES AND DISCUSSION 429
In addition to the canal, we also studied the turtle community inhabiting a man-made lake owned by
the Indianapolis Museum of .^rt {IMA Lake). The 14.7 ha lake is situated in close proximity to both the
canal (Iti5 m) and the W\mv River {30 m). Relictual woodlots surround about 75% of the lake's
shoreline and the lake is frequently used by recreational fisherman.
We captured 1044 individual turtles a total of 14119 times between April-October 2002 (<0.5% of the
captures were made during a preliminary ir,^pping period in September-October 2001), Most captures
were made through the use of aquatic hoop traps {7tj..1 cm diameter hoops. 30 X 30 cm coated nylon
mesh with a funned at one end and a closed bag at the other) although occasional capnires
{<1
%) were
made by hand or with a dip net. Wliile no trapping method is without .species-.specific biases {see
Gibbons, 1990a), the use of aquatic traps for turtle population and community studies ha.s gained wide
acceptance (Bodie
et
ni. 2000; Smith and Iverson, 2002; Burv' and Crfrmano, 2003) when limitations are
properly acknowledged
(see
Results and Discussion, below). In the canal we deployed 6-20 traps spaced
approximately
10(1
m apart; spacing was considerably greater in the lake (>25() ml where trappiiig was
limited to September-Otober 2002. We baited traps with sardines and/or chicken livers {refreshed
every 4-5 d). checked traps daily and changed trap locations weekly in order to maximiw- coverage of
the canal. Traps were submerged save for ihe top ,'>-20 cm. For each turde we recorded mid-line
carapace length (CL to the nearest mm) using calipers, mass (to the nearest g) using a benchtop
electronic balance, species, sex, location of capture and any notable damage. Each turtle was given an
individual mark by notching the marginal scutes in a unique pattern to allow for future identification
(C'«igle. 1939). Turtles were processed and returned to the point of capiure within 24 h.
RESULTS
AND
DiscusstON
COMMUNtTy
COMPOSITON
The same assemblage of
six
species
w-as
captured in both the canal and lake habitats: spiny suftshell
turtle
(Apolnne
spinifera).
painted turtJe
{Onysemys
pula). common snapping turtle
{Chelydm
serpentina),
<ommon map turtle
(Graptemys
geogiapbira).
common musk turtle
(Stemotherus odoratus)
and red-eared
.slider
{Trarhrmys
srripla).
However, relative abtindances of tJiese species dilfered significantlv in the two
habitats (Fig. 1). For the purposes of a quantitative comparison, we used data collected only during
September-October 2002, the titne period when both habitats were sampled contemporaneously (Table
I).
Three species, A. spinifem. C. fnrtn and C. snpentiiuu collectively constiuited <20% of tlie total
captures in either habitat. In the Central Canal, C
geuginphiia
was most abundant with T. scripta
represenling <I0% of all captures, whereas in IM.^ l.ake
T.
Am/*/^
alone accounted for more (j5% of the
captures and
G.
gengrapkirn
represented <5%.
Differences in composition may be due in pan tn the unequal distiibution of aquatic mollusks
between tliese habitaLs. Freshwater mollusks are the primar>' prey for
Grnptemysgeographica
(Gordon and
MacCulloch, 1980; Vogt, 1981; White and Moll, 1992). The Central Ganal supports several species of
freshwater snails (eg'., members of the genera Pipw/ncCT-fl,
Goniohasis-and Vivapanis)
and are found at high
densities locally; similar sampling efforts in IMA Lake have failed to detect the presence of aquatic
snails. The lake, thus, represents sui>optimal habitat for
G.
gengrapfuca
due to a lack of preferred food.
In contrast, 7^
srripta
is more omnivorous and opportunistic in feeding (Ernst et al., 1994) and is the
most abundant species in the lake. It may be that the lack of
G.
gengraphicu
in IMA Ijke due to the
absence of suitable food has allowed the
Treirhnnys scriptn
population lo grow tnore successftilly than in
the habitat where G.
gpographirrj
is abundant. The extent to which these two species interact
competitively, however, is unclear and the differences in distribution may simply reflect microhabitat
preferences
(icc
Ernst etai, 1994). The relative consistency of the rest of the species abundances speaks
to the overall similarity of the two sites despite the differences inherent in lentic and lotic habitats.
The rarit>- of
Chri.semys
piita in Ixith habitats is unexpected, as it is one i>f the most common and
abundant species throughout its lange and particularly in the Midwest (Anderson
et
ai, 2002; Bury and
(iennano, 2003) and Southeast (Litidsayanci Dorcas. 2001). Moreover, it can be vei-y abtmdant in urban
habitats; Mitchell {1988) estimated more than 500 individuals in a C/«r/« population inhabiting a small
creek {6 m wide) and two small associated beaver ponds in urban Richmond, Virginia. We speculate that
the Wliite River is tlie source of the turtle populations inhabiting the Central Canal and IMA Lake.
430
THE AMERICAN MIDLAND NATURALIST
IS 3(2)
DCental Canal (n=873)
IIMA Lake (n=171)
0
0.2
0.4
0.6
Proportion
of
captures
F[(..
I,—Tbe proporiion of
six
aquatic atid st-ini-aquaLic luitle species captured in a man-made caiial
(Oniral Canal) aiitt lake (IMA Lake} within an urban landscape in Indianapolis. Indiana, USA. The
ttual number of individuals collected between is indicaicd in the figure legend (see text for details).
Abbreviations
are as follows: \!i
—
Apohne
spinifera.
Cp -
Chrysemys
picta,
Cs =
{Chelydra
serpmtina), Gg =
Graptemys
^ographica.
So -
Stenuitherus
odoratus, Ts =
Tracliemys
scripta
2005
NOTES
AND
DisajssioN
431
1.—Number
and proportion of six aquatic and semi-aqtiaiic turtle species collected in
September-October 2002. B«-cause tbe sample size was low for Apolonf \pinifrra,
Chelydra
serpmtina and
CJirysems
picta in botli habitats, the goodnessof-Ht-tesi was conducted usinR only the ibree more
CO nun on species
Ap<»lo»f
spiritjna
Chehdra
snprntuia
Chrysrmys
pic I a
Crraplfmys
grogt'aphifa
Sttmothenis odtiraliu
T^arhrmys
sciipta
TOTAL
Otiiral
(
nrnnbrr
1
18
5
50
43
6
126
.nal
proportion
0.032
0.143
0.040
0.397
0.341
0.048
niimhci
3
4
4
7
32
102
152
IM.A hik<-
priipi-niiiiii
0.020
0.026
0.026
0.046
0.211
0.671
tkHsdnessol-fit-tesi:
%-
= 115.8. df = 2, P < O.OOOl
Althougb C. picta is frequently abundant in punds and lake.*, it is notably less toinmon in rivers (Emst ft
al., 1994) and. tliiw, the current low density may reflect a histonral low density in this region, It
Ls
worib
noting also that tlicst- populations may sufTcr from very low recniitnieni rales, as all 19 indi\iduak we
captured in Ixjlh habitats were mature adults, whereas we have collected or observed batchlings for ibe
other live .sjiecics.
SEXUAL
SIZF.
DIMORPHISM
We found significant sexual si/e dimoqihism for lhe three tnost frequently captured species, with
females significantly larger iban ibe males in Iwth CI, and mass in each case (Table 2). Sexual si/e
dimorphism, with females lai-ger tban males, is tbe norm for most emydid turtles (Emst ft
aL.
1994) and
has been dotumentcd in populations of
Grfiplemys grographica
and Trarhnnys scriptn throughout their
ranges
[e.g.,
Ciigle, 11)50; Vogt, 19K(); Ciibbons and Lovicb. 1990). as well as in (entral Indiana (Minton,
2001).
Sexual size dimorpbi.sm, however, is less cotnmon in kinosterid turtles, paniculai'ly in SUmollieriu
Tible 2.—Sex ratios and bod>' sizes of C
grngraphica,
S.
odorauts.
and 7^
scriptn,
the major species of tbe
fk-ntral Canal (ranal) and lM.-\ Lake (lake) turtle assemblages. For each population we report tbe
number, and mean (and St.) cuitpiice length {V.i.) and mass of eacb sex; we tised Chi-square goodness-
of-fit tesLs tu detect skewed sex nuios and one-way analysis of variance on log-traiisformed CL and mass
to detect sexual size dimorphism
Nuiiiljer
(;.
grtigrapbiea
(canal)
fctiiale ni;ili"
Ufi 15!)
X" =
0.554
P =
0.457
S. «Mi
tein;.l<.-
122
i
aulf (ranal)
malt'
170
' - 7.89
= 0.005
T. SI
45
)
/
•nptri (t.iiinll
k- mak-
.t6
'=0.317
y: »Tipi,i (lake)
It-nuilf aiiili-
5(1 4H
X'
=
0.041
/' =
0.840
CL (mm)
Mass
(g)
168,7
(4..')4)
P<
800.6
(51.9)
f^ l.MU
P<
102.7
(1.03)
=
209.7
0.0001
137.3
(.S,45)
=
235.2
O.(K)O1
107.1
(0.81)
P<
212.4
(4.34)
P<
98.6
(0.82)
0 = 46.9
0.0001
157.7
(3.36)
„
=
76.1
O.O(K)1
I7.'i.l
(6.98)
P =
988.0
(130.0)
f'' 1,7«
P =
]
51.0
(•1.90)
;
= 5.04
0.028
542,0
(46,6)
,
=
9.82
0.002
178.2
(6.36)
P =
972.8
(78.0)
;^<
148.6
(4.61)
s
=
n.o
0.001
522.3
(44.4)
fi = 14.4
0.0001