Herpetological ReI,iell'. 2011, 42(2). 17()"177. © 2011 by Society for the Study ofAmphibians and Reptiles
Chesapeake Bay's Diamondback Terrapin (Malaclemy's terrapin terrapin) population ha made considerable recovery from the commercial exploitation that led to its near extirpation in the late 19'" and early 20'" centuries (Carr 1952; McCauley 1945). Nevertheless, it continues to face increasing threats from human population growth, degradation of the Bay, exposure to man-made hazards (Butler et al. 2006; Roosenburg 1991; Siegel and Gibbons 1995), and, during the period of this study (2003-2005), continued commercial harvest in the Maryland portion of the Bay. The state of Maryland maintained an active terrapin fishery that was closed only during the terrapin ne ting season (May-July), had no daily or ea onal catch limit, and exclu ively targeted breeding-age female, i.e., terrapins with plastron length ;,,15.2 cm. Although the fishery was perceived a small, with little market demand, the long season and 'no limit' regulation left the fishery vulnerable to over harvest. Concern over this vulnerability led us to examine the method of traditional winter harvest and characteristics of terrapins occupying hibernacula. Although generations of Chesapeake Bay watermen have pursued winter harvest of terrapins, fishing methods, locations of hibernacula, and the terrapins occupying them have not previously been documented.
About Terrapin Institute
The Terrapin Institute began in 1998 as a consortium of concerned citizens, scientists, resource managers, and educators dedicated to the understanding, persistence, and recovery of Diamondback Terrapins and other turtles through effective management, thorough research, and public outreach. We work to protect an abundance of adult turtle populations, preserve nesting and forage habitat, and improve recruitment. In return the terrapin has become the perfect metaphor for natural resource stewardship and public engagement; the face of estuarine restoration, and a gateway to the many wonders of our rich tidewater heritage.
170 ARTICLES
Herpetological ReI,iell'. 2011, 42(2). 17()"177.
© 2011 by ociely for the lUdy of Amphibians and Reptiles
Using Scrape Fishing to Document Terrapins
in Hibernacula in Chesapeake Bay
Chesapeake Bay's Diamondback Terrapin (Malaclemys
terrapin terrapin) population ha made considerable recov-
ery from the commercial exploitation that led to its near ex-
tirpation in the late 19'" and early 20'" centuries (Carr 1952;
McCauley 1945). Nevertheless, it continues to face increas-
ing threats from human population growth, degradation of
the Bay, exposure to man-made hazards (Butler et al. 2006;
Roo enburg 1991;
iegel and Gibbons 1995), and, during the
period of this
tudy (2003-2005), continued commercial har-
ve t in the Maryland portion of the Bay. The state of Mary-
land maintained an active terrapin fishery that was closed
only during the terrapin ne ting season (May-July), had no
daily or ea onal catch limit, and exclu ively targeted breed-
ing-age female, i.e., terrapins with plastron length ;,,15.2
cm. Although the fi hery was perceived a small, with little
market demand, the long season and 'no limit' regulation left
the fishery vulnerable to overharvest. Concern over this vul-
nerability led us to examine the method of traditional winter
harvest and characteri tics of terrapins occupying hibernac-
ula. Although generations of Chesapeake Baywatermen have
pursued winter harvest of terrapins, fishing methods, loca-
tions of hibernacula, and the terrapins occupying them have
not previously been documented.
Here we utilize watermen fishing skills and harvest meth-
ods to locate, ample, and de cribe terrapins occupying es-
tuarine bay hibernacula. Because the region has long been a
center for the commercial harve t of Blue Crabs (Callinectes
sapidus), we al 0 examine terrapin capture for potential ef-
fects of selective by-catch mortality of mall terrapins associ-
ated with decades of commercial crab pot use in the region
(Roo enburg et al. 1997; Warner 1977; Wood 1997).
MATERIALS AND METHODS
We sought collaboration with a former terrapin harvester
to learn fishing methods and to access hibernacula in the
Tangier ound region of Che apeake Bay (Fig. 1). Tangier
Sound is bordered by extensive brackish tidal marshes and
historically has harbored large terrapin populations (Mc-
Cauley 1945).
ampled hibernacula were chosen to be
G. MICHAEl HARAMIS'*
e-mail: mharamis@usgs.gov
PAULA P. F. HENRY'
e-mail: phenry@usgs.gov
DANIEL D. DAY'
e-mail: dday@usgs.gov
I U.S. GeologicalSurvey, Patuxent Wildlife Research Center,
do BARC-EAST Bldg 308, 10300 Baltimore Avenue
Beltsville, Maryland 20705, U5A
*Corresponding author
geographically separated and to represent a selection of sites
of former terrapin harve t. Four ite were chosen near large
marsh i lands ca. 8-12 km off hore (Smith I land,
outh
Marsh Island, and Bloodsworth Island) and two addition-
al sites were located near the mainland (Janes Island and
anticoke River; Fig. 1). Two ites located within tributaries
( anticoke River and Janes Island) and a third located within
interior outh Marsh Island were within zones where com-
mercial crab potting is prohibited by state regulation. The
remaining three sites ( mith [Iand- , mith Island-S, and
Bloodsworth Island) were within areas of heavy potting activ-
ity. We also selected St. Jerome Creek, a site of fonner harvest
on the western shore of Chesapeake Bay that differed in be-
ing located in a navigable tidal creek having forested hore-
line in lieu of exten ive bordering salt mar h.
To capture terrapins watermen adapted a dredge, or
scrape, used to capture molting blue crabs from shallow,
near-shore waters using locally built, shallow-draft ve sel
(also called crapes; Warner 1977). Being hand-made, terra-
pin scrape vary in
ize, but most are ba ed on a 1.5 m wide
blue-crab design. In 2003, we u ed a blue crab scrape that wa
modified by attaching a larger mesh catch bag and welding
12 short 7.6 em teeth every 13 cm along th scraping bar. The
larger mesh bag (8 cm) helped minimize fouling and the teeth
dug terrapins from bottom sediment. In 2004 and 2005, we
used a slightly larger scrape framed from heavier 2 cm diam-
eter steel stock. Its entrance measured 1.7 m x 0.4 m and its
overall length was 2.4 m (Fig. 2a). The noteworthy difference
of this scrape was the addition of numerous (30), longer (15
cm), and more clo ely spaced (6 cm) teeth on the scrape bar.
The scrapes were towed in a circular pattern behind a 12
m work boat and retrieved by hydraulic winch. The length of
the tow rope was critical to proper operation as it adjusted
the drag or "bite" of the scrape in bottom ediment. Capture
success wa measured by the number of terrapins capturedl
tow. Although there was no fIXed time limit for each tow, sev-
eral hundred tows averaged about 7 min in length, or a rate of
8.6 tows/h. While some commercial crapes were equipped
with
ide baffles to help gUide terrapins into the catch bag,
neither of our scrapes had baffles.
Captured terrapins were given a visible mark by drilling a
5 mm diameter hole in the lOt" right marginal cute through
which we attached a erially numbered monel fish tag (tag
nO.l005-3:
ational Band and Tag Co., P.O. Box 72430,
ew-
port, Kentucky 41072). Passive integrated tran ponders, or
PIT tags (model TX1400L: Biomark, Inc., 7615 West Riverside
Drive, Boise, Idaho 83714), were used as permanent mark.
Size dimorphi m and tail characteristics (tail length and anus
position in relation to edge of carapace) were the primary
H~rp~t%gical Rroi~ 42(2), 2011
ARTICLES 171
FG. 1. Che apeake Bay and vicinity howing Tangier ound and
the location of ite of winter ampling of Diamondback Terrapin,
2003-2005: I) Blood worth Island, 2) outh Mar h I land, 3) mith
I land- onh, 4) mith Jsland- outh, 5) jane I land, 6)
anticoke
RiI"er, and 7)
l. jerome Creek.
criteria u ed to sex terrapins (Carr 1952). Because only terra-
pin about 10 year of age (YOA) Or younger could be aged by
annular growth rings (as read from plastral pectoral scutes),
we eparated captures into two ba i age classe : those s10
and tho e >10 YOA. Body mas wa mea ured with a digi-
tal electronic balance to the nearest g and midline plastron
length (PL) wa measured with calipers to the neare t mm.
To di pel concerns about po
ible d trimental physiolog-
ical effect a ociated from removal of terrapin from hiber-
nacula, we conducted a field tudy to determine if hart-term
acute effect did occur. Thi que tion ha management r 1-
erance becau e we observed commercial terrapin harve ter
leaving capture on boat decks for up to everal hour before
orting their catch and returning cull
to the bay. In our ex-
periment, one et of 12 newly captured terrapin
(6 of each
ex) were maintained in bay water (5°C) while
imultane-
ou Iy exposing a econd set of 12 terrapins in a heated boat
cabin at ca. 24°C, a temperature ~ ell above the range of am-
bient air temperatures (lO-iSOC) experienced on a boat deck
on a mild winter day. Both et of terrapin were held for a
DC
DE
®
\,
TANGIER
/SOUND
I
o
I
SOKM
FIG. 2.
) A crap ,or dredge, u ed to harve t DiamondbackTerrapin
in winter in he apeake Bay. De ign i ba ed on a crape u ed to har-
ve t blue crab but with heavier con truction, addition of teeth on
th
rape bar, and a larger me h catch bag ( ee text for detail ). B)
retri ved crape resting on the boat gunwale howing a good catch
of terrapin. The crape did not injure terrapin and a many a 24
were captured in a ingle tow. C) Two female caught] 8 March 2005
from t. jerome Creek demon trate size and age in the diamondback
terrapin. The female on the left (body rna
of 1922 g, PL of 202 mm)
wa known to be 20 years of age by previou marking on the nearby
Patuxent River (w. Roosenburg, per. comm.). The female on the
right i of unknown age and wa the large t terrapin caught during
our tudy (3022 g, PL of228 mm). Thi
pecimen approximate the
maximum size for the specie within the he apeake Bay and likely
within the pecie ' range.
H~rp~/ologicalRftj,roJ 42(2), 2011
172 ARTICLES
TABLE 1. Numbers and sex ratios of Diamondback Terrapin captured
from
ix hibernacula in Tangier Sound and the
t. Jerome Creek ite
on the western hore of Chesapeake Bay, winter 2003-2005.
ite are
Ii ted by rank in cumulative ex ratio given a proportion female (pf).
ite
o. captures
ex ratio (pf)
t. Jerome Creek
33
0.970 A"
mith IsJand-
74
0.784 B
Jane Island
222
0.734 B
mith Island-
377
0.719 B
outh Mar h I land
141
0.709 B
Bloodsworth I land
160
0.588 C
anticoke River
168
0.345 D
Totals
1175b
0.660
•Value sharing the arne letter are not different: individual 2-tailed
z-test of proportions (n =0.05).
bTotals exclude between-year recapture ( =22).
25
C
20
.2003
~
0200'
':
02005
..
15
co
~
~
,.
co
..
~
~
FIG. 3. A composite hi togram of year- and ite-specific terrapin cap-
rure rate (mean ± E capture ftow) using a modified crab scrape in
2003 and a lightly larger crape in winter 2004-2005. Capture rate
had marked year and site effects (P < 0.001, individual te t using Kru-
kal-Walli one-way ANOVA on rank ). Mean within site
haring the
I arne letter do not differ: Dunn's multiple comparison test (n =0.05).
period of 2 h. Following the holding period, all terrapins were
returned to the location of capture and placed in a 1.8 m x 2.4
m x 1.8 m wire holding cage made of 2.5 cm mesh galvanized
chicken wire. The cage at on the bottom and extended above
tide level. After 3 weeks, the cage was retrieved and urvival
and outward condition of the terrapin asse ed.
We used SigmaStat 3.1 (Sy tat oftware, Inc, Point Rich-
mond, California 94804) to evaluate assumption of nor-
mality and homogeneous variance and to conduct analyses
of variance. When nece sary, we either transformed data to
meet normality and equal variance or applied the nonpara-
metric Kruskal-Walli
OVA test on rank. Contrasts were
tested using Tukey'
test (parametric) or Dunn's method
(non-parametric).
ex ratios and proportion of females <10
YOA were tested by 2-tailed z-tests of binomial proportions
( okal and Rohlf 1995). We generated a simple Lincoln-Pe-
tersen e tirnate ba ed on site-specific between-year recap-
tures to estimate hibernaculum population size (Seber 1973).
This estimate is made under the as umption of population
closure and equal catchability. We te ted for two potential
effects of selective by-catch mortality of maller terrapins
in crab pots as noted by Dorcas et al. (2007) and Wolak et
al. (2010): larger size of both sexes based on PL and reduced
presence of younger female terrapins. Data were combined
across year within pot and no pot sample sites.
RESULTS
Hibernacu.la.-Hibernacula were located in semi-pro-
tected estuarine bays in near-shore shoal waters adjacent to
exten ive salt marsh. Sites had bottoms of moderately soft
mud and were deep enough (1.5-3.5 m) to offer little risk
of dewatering even under unusually low storm tides. Most
bottoms were relatively clean of shell and organic debri ; all
were expo ed to good tidal circulation but limited wave en-
ergy because of reduced wind fetch. Because terrapins were
often buried in soft mud, continued scraping at such sites
often increased capture succe s. On firm bottoms terrapins
were only partially buried or not at all. Terrapins were not
injured by capture and only rarely were they marked by the
metal teeth or frame of the scrape (Fig. 2b).
Capture success.-During February and early March,
2003-2005, we captured 1175 terrapins from seven hibernac-
ula (Table 1). Capture rate had a marked site and year effect
(P < 0.001: individual tests, Kruskal-Wallis one-way A OVA
on ranks; Fig. 3). Capture rate for
ite
in Tangier Sound
averaged 4.18 ± 1.02 SE captures/tow ( = 18), or about 36
terrapins/h. Capture rate generally increased in 2004 and
2005 with the u e of the heavier scrape with longer, more
closely paced teeth (Fig. 3). The exceptions were the Smith
lsland- site where harvest activity was suspected to have
depleted the population in 2004 and 2005, and the St. Jerome
site where capture uccess was low in all years (mean = 3.1
terrapin /h). St. Jerome Creek wa the deepe t site ampled
at 3.5 m and during three winters produced the fewest (32
female and one male), but on average the largest terrapins.
The largest, a 3022 g female with PL of238 mm, was consider-
ably larger than a known-age 20-year-old terrapin captured
at the same site (Fig. 2c) and marked previously on the Patux-
ent River (w. Roo enburg, Ohio University, pers. comm.).
Mass of26 females averaged 2151 ± 56 g and PL ranged from
191 to 228 mm (mean 205 ± 1.6 mm). At our South Marsh site,
heavy growth of eelgrass (Zostera marina) precluded capture
of terrapins in 2003. However, in the absence of vegetation
in 2004, capture success increased sharply, and in 2005, the
site was our mo t productive hibernaculum, averaging19 ter-
rapins/tow, or 160 terrapin /h (Fig. 3).
Sex ratio.- ex ratio across all site and years favored fe-
males: 0.66 proportion female (pf), or 1.94 females/male ( =
1175; Table 1). Cumulative yearly sex ratio declined from 0.730
Ht'rpt'/ological Rt«IJirw 42(2 ), 2011
------
- ~ -
-------
---
ARTICLES 173
TABLE 2. Mean mass of male and female terrapin captured at six site within
Tangier ound and
t. Jerome Creek on the we tern hore of Che apeake Bay,
winter 2003-2005. Includes only unique adult capture (exclude between-year
recapture) of female >10 year of age and male >9 year of age. For the
ix
Tangier ound ite
there wa no rank correlation between ite- pecific mean
mass of males and female (r =0.54, P> 0.05, 4 dO.
• One-way
OVA found marked
ite effect for female ma s (Fs.sI< =52.2, P <
0.001). Mean within columns sharing the arne letter do not differ: Tukey' te t
(et = 0.05).
b Insufficient sample size.
c One-way
OVA on In transformed data found a marked site effect for male
mass (F5.J63 =16.0, P < 0.00l). Mean within column haring the ame letter do
not differ: Tukey's te t (et = 0.05).
180
10 YO""
n-8
170
160
150
140
y = O.000787xUS51
R'=O.995
Plallron Length mm
130
120
5 YO""
n-8
Mean mas
Female
Male
26
2150.6 ± 56.4 At
_b
49
1589.2 ± 34.1 B
16
455.2 ± 5.03 N
236
1386.7 ± 17.3 C
104
407.3 ± 5.03
81
1324.8 ± 34.5 0
63
436.5 ± 7.91 AB
72
1285.8 ± 34.0 CO
32
390.9 ± 13.03 CD
III
1248.1 ± 23.2 0
53
358.3 ± 7.14 0
46
1181.7 ± 30.9 0
101
412.3 ± 5.39 BC
621
1393.5 ± 22.6
369
407.3 ± 3.19
but only 22 terrapins, 21 females (2.6% of female captures),
and a ingle male (0.25% of male captures) were recaptured
during winter ampling. All 22 recaptures were recorded at
the ite of original capture. Recaptures were sufficient to e ti-
mate hibernaculum population size only at the mith Island-
N site, where sampling was increased a part of an ongoing
population study. Based on 13 female recaptures in 2005,
FIG. 4. Growth curve for 5- to 10-year-old female terrapin captured
during winter ampling in the Tangier Sound region of Che apeake
Bay. Age was determined by counting growth annuli on pia tral pec-
toral cute. Points are means; bracket are ± E. mall ample size
in 5-year-old re ulted from general ab ence of the e and younger
( mailer) terrapin in e tuarine bays in winter. mall ample ize of
10-year-old terrapins re ulted from hell wear that ob cured growth
annuli and limited the number that could be identified.
II
1400
1200
1000
.. 800
:
600
400
200
0
100
110
Site
t. Jerome Creek
mith I land-
Smith 1 land-N
Bloodsworth I land
South Marsh Island
Jane 1 land
anticoke River
Combined
pf in 2003 to 0.654 in 2004 and 0.641 pf in 2005 (P
< 0.05: individual 2003 vs 2004, and 2003 vs 2005
z-tests of proportions). Discounting po sible year
effects, the decline in sex ratio was likely a result
of use of the larger crape with more numerous
and closely spaced teeth that increased capture
efficiency of males.· ot only were generally more
males captured, but the minimum size wa low-
ered in ome cases by as much as 15 mm. For site
within Tangier ound, year- and site-specific sex
ratio varied from a low of 0.32 pf to a high of 0.91
pf. The anticoke River, a site of known frequent
terrapin harvest, was the only site where male
capture outnumbered females (Table 1).
Estimate offemale recruitment.-Using growth
annuli to age terrapins, we found females s10YOA
to compri e 19.2% (145) of764 captures. Site- and
year-specific percentages varied from 0 to 39%.
We estimated annual recruitment by a suming
100% annual survival in the 6-to-9 YOA classes
and averaging the percent of these year classes
to total annual female captures. Terrapins <6
YOA were not included because only 9 were cap-
tured (8 five-year-old and 1 four-year-old); and
lO-year-old were excluded because advanced
shell wear obscured growth annuli limiting the
number that could be identified ( =8). The 128 terrapins in
the 6-to-9 year-old age classes represented 10.4%, 18.3% and
21.6 % of female captures during the respective three winter
and yielded estimates of annual recruitment of 2.6%, 4.6%,
and 5.4% (mean = 4.2 ± 0.83).
To eliminate potential error a ociated with reading
growth annuli, we additionally estimated recruitment ba ed
olely on size metrics. We used mean ize
tatistics from 44
known 9-year-old female, i.e., mean ma s (1043 g). PL (165
mm). and mass-to-pla tron ratio (6.340 g/nun), to separate
female s9 YOA from aU older female . The re ulting mean
e timates of percent annual recruitment were about 50%
higher than those based on growth annuli: 6.0 ± 0.29% based
on PL, 6.5 ± 0.35% based on mass, and 6.7 ± 0.16% based on
mass-to-PL ratio. Although we recorded 19.2% of females s10
YOA, only 29 3-to-8-year-old males (7.9% of male captures)
were so captured (X2 =22.8, P < 0.001, 1 dO. We concluded
that too few male in the younger age c1as es were captured
to provide a meaningful estimate of recruitment.
Growth rate.-Females in the 5-to-1O-year-old age classes
exhibited marked growth as mean mass increased by a factor
of 3 (from 397 to 1184 g) and mean PL increa ed 50% (from
115.9 to 174.0 mm; Fig. 4). Based on the Maryland harvest
regulation of a minimum PL of 152 mm, mo t female terra-
pins would reach harvestable size at between 7 and 8 YOA.
Too few males were captured to adequately interpret growth
in the 5-to-9-year-old age classes.
Recaptures.-We recaptured 102 terrapins (8.7% of cap-
tures) by variou means up to four years after initial capture,
Herpetological Rrview 42(2), 2011
174 ARTICLES
TABLE 3. Mean plastron length (PL) for male and female terrapins and the proportion of young female
(pf) sl 0 years of age (YOA) captured at
ix
ites within Tangier ound, winter 2003-2005. Ranking in relation to zone with and without crab potting i shown. PL means include only
unique capture (excludes bet\veen-year recapture) of female >10 YOA and male >9 YOA.
Mean plastron length (mm) ± E
Crab potting
Site
Male
Rank
Female
Rank
pfslOYOA
zone
(
)
(
)
Pots
120.9 ± 1.16 A"
186.8 ± 1.33 N
0.157 AB<
(16)
(49)
Pots
Bloodsworth Island
118.3 ± 0.73 A
2
176.3 ± 1.49 BC
3
0.133 B
(63)
(8l)
Pot
mith I land-
116.2 ± 0.59 AB
3
179.2 ± 0.72 B
2
0.135 B
(104)
(236)
No pot
anticoke River
115.2 ± 0.57 B
4
167.8 ± 1.40 0
6
0.300A
(101)
(46)
o pots
outh Marsh Island
113.7 ± 1.24 BC
5
174.5 ± 1.43 C
4
0.309 A
(32)
(72)
o pot
Jane I land
111.5 ± 0.84 C
6
173.9 ± 0.98 C
5
0.186AB
(53)
(Ill)
Combined
115.6 ± 0.33
178.2 ± 0.53
0.192
(369)
(595)
(764)
Rank
4
6
5
2
3
• Kru kal-Walli one-way A OVA te t on rank found a marked site effect for pia tron length (1-1 = 48.0, P < O.OOl, 5 df). Means within column sharing the same
leiter do not differ: Dunn's lest (n = 0.05).
b One-way A OVA found marked site effect
for plastron length (Fs.SIIO = l8.3, P < 0.001). Means within column sharing the ame leiter do not differ: Tukey's te t
(u = 0.05).
a mark-recapture estimate of 1210 female populated the
site in winter 2004 (normal 95% C.L: 684-1737; Seber 1973).
Males were excluded from the estimate becau e no male
were recaptured. However, ba ed on the 2004 ample ex ra-
tio of 0.744 pf, the estimate increases to 1613 when males are
included. Based on capture rates, populations at two oUler
hibernacula, Janes Island in 2004 and 2005 and SOUUl Mar h
Island in 2005 (Fig. 3), likely exceeded uli estimate.
ize charaeteristics.-Off- hore sample site near Blood-
sworth, Smith, and oum Marsh Islands, had lightly heavier
females than the near- hore ite ofJanes I land and ule an-
ticoke River (Table 2). The largest females were at the St. Je-
rome site and me mallest at me Nanticoke River, a site offre-
quent harvest.
imilarly, the mean rna
of adult males (all >9
YOA) was highe t for off-shore Smim Island-S/Blood worth
Island sites and lowest for near-shore Jane I land (Table 2).
As e entially all male are immune to commercial harve t
by virtue of their mall size,
anticoke River males ranked
comparatively higher in site-specific mean mas (Ulird) ver-
u
the females captured there that ranked sixth (la t). For
the
ix Tangier ound ample sites, site-specific mean mass
of males and female was not correlated ( pearman rank r =
0.54, P > 0.05, 4 df). Consistent with the known dimorphism
in the species, the mean mas of female terrapins (1276 g,
=767) was about three time greater than that for males (401
g,
= 399). Only 4.6% of females (35 of 767) fell below a PL
measurement of the large t male (138 mm).
Effects ofcrab pot bycatch.-We tested for the effect of se-
lective crab pot mortality on female size and found female PL
to be greater in pot versus no-pot zone (one-way ANOVA:
FI.593 =47.7, P < 0.001; mean PL, pot zon =179.6 ± 0.61,
366; mean PL, no-pot zone = 172.9 ± 0.73,
= 229). Because
the male PL distribution was found to be non-normal, we
applied ule Kruskal-Wallis ANOVA test on ranks and found
a larger median value in pot versus no-pot zones (H = 25.6,
P < 0.001, 1 df; pot zone median PL = 117,
= 183; no-pot
zone median PL = 114,
= 186). Site-speciiic tests showed
a marked pattern of larger PL means at
ites with crab pot-
ting that produced a consi tent alignment of ranks to pot
and no-pot zone (Table 3). We also found females to have
a higher proportion of young s10 YOA in no-pot versus pot
zone (one-way A OVA: FI.I3= 5.2, P < 0.04; mean proportion
no-pot zone = 0.26 ± 0.13,
= 8; mean proportion pot zone
= 0.11 ± 0.11,
= 7). Individual ite-specific tests produced a
H~rptfologi{QlReviroJ -12(2 ). 2011
imilar pattern of a higher proportion of female at site
in
no-pot zones with a consi tent alignment of rank to pot and
no-pot ite (Table 3).
Testing for acute effects.--Dn initial placement of the 24
terrapin
in the holding cage, three individual were active
enough to wim to the top of the water column. Thi
indi-
cated that their metaboli m had increa ed enough to be-
come active wimmer, but none were ob erved to breathe
air at the urface. The swimming activity cea ed shortly after
placement ugge ting that any increase in body temperature
and metabolism was quickly rever d by return to cold wa-
ter. All other terrapins were Ie s activ and simply sank to the
bottom of the cage. Following the three-week holding period
in February 2005, all terrapin were alive and observed to
behave normally, i.e., were luggi h but active; there wa no
ign of morbidity.
DISCU SlO
Our discovery that upwards of a thou and or more terra-
pin can be concentrated at heavily populated hibernacuJa
under core the vt1lnerability of the pecie to winter scrape
harve t. Our capture of 160 terrapin /h at outh Marsh I land
demon trate how, in certain in tance ,hundred of terra-
pin can be removed from a local area in a matter of hours,
a majority of which would be harvest -size females. Becau e
terrapin population trait
include low recruitment, delayed
maturity, long life, and limited di per al (Gibbon et al. 2001;
Harden et al. 2007; Tucker et at. 2001), high urvival of long-
lived adult
i critical to ustaining population (Mitro 2003).
It follows that harvest removal of a large portion of breeding-
age female would be deva tating to local population. Ter-
rapins, like other long-lived turtles, have no compensatory
mean to replace uch 10 se (Brook et at. 1991; Congdon et
al. 1993; Heppe1l1998) and recovery would be predicted to be
e pecially protracted.
As e tuarine bay hibernacula have never been studied
previou Jy, no comparative data exi t on numbers, size,
and sex ratio. Moreover, we hav no knowledge of how our
capture characteristics have been altered by the confound-
ing influence of two principal anthropogenic effects: 1) the
direct 10
of females to commercial harvest and 2) seJectiv
mortality of mall terrapin as bycatch in crab pot (Roosen-
burg et at. 1997; Roosenburg 2004). Becau e our data how
that heavy terrapin harve t would quickly devastate the adult
female portion of the population, the prevalence of a female
bia ed ex ratio in Tangier ound i
trong evidence of mini-
mal harve t activity in the region in the recent past. Terrapin
mortality in recreational (not commercial) crab pots primar-
ilyaffect maller terrapin (male and young females) that
occupy habitat near hore where they are at ri k to recre-
ational crab pots (Roosenburg et at. 1997; Roo enburg 2004).
Terrapin expo ure and ub equent mortality in commercial
crab pot in Tangier ound therefore seem mediated by vir-
tual re triction of commercial pots to offshore u e. Indeed,
Roosenburg (2004) sugge ted that Maryland' deep-water
1
ARTICLES 175
re triction on commercial crab potting likely has averted the
decimating los e to Bay terrapin populations such a ha e
been reported in Florida ( eigel 1993), outh Carolina (Dor-
ca et al. 2007; Gibbon et al. 2001; Hoyle and Gibbons 2000;
Tucker et al. 2001), and more recently, Georgia (Gro e et al.
2009).
onethele
,decade of crab pot expo ure eem ap-
parent as our re ult
indicate an effect of increased
ize of
both exe and reduced number of young females. The e
result
are consi tent with demographic effect attributed
to elective crab pot mortality in tidal creek of the Kiawah
River, outh Carolina (Dorcas et aI. 2007), and at the Good-
win Island at the mouth of the York River, Virginia (Wolak et
al. 2010). Although other interpretations are possible, a non-
normal PL distribution for male, in contrast to a normal
distribution for female, lend
upport to an active mortality
proces affecting males. Additionally, and most significantly,
we found the proportion of young female in the no-pot zone
to be 2.4 time that in the pot zone. It follows that this ap-
parent 10 s of female to crab pot, and the projected loss in
recruitment it represents, would have the greatest long-term
effect on terrapin productivity in Tangier ound.
Our winter
ampled ex ratio
and the ex- and age-
related vulnerability of terrapin
to crab pot al 0 may be
driven by a fundamental di tribution proce
related to ize,
i.e., larger terrapin dominating open-water e tuarine bay
and mailer terrapins seeking protected areas in tidal creek
and interior alt mar h. Roo enburg et al. (1999) provide evi-
dence of larger adult female moving farther and pending
more time offshore while mailer male and juvenile remain
in near-shore shallow water. Finding from our ummer cap-
ture at mith I land upport thi notion as our interior alt
mar h bait-trap captures yielded maller terrapin of more
ven sex ratio (0.58 pf) and open-water Bay-shore fyke net
captures produced primariJy large females (0.87 pf; P. Hen-
ry, U G , pers. comm.). That terrapins overwinter in inte-
rior tidal creeks and creek bank at a ha been documented
(Yearick et at. 1981). Our amples show that few female <6
YOA (with mean PL of 134 mm and mean mass of 560 g) and
few males <6 YOA (mean PL of 105 mm and mean mass of300
g) occupy estuarine bays in winter.
The comparatively pristine nature of Tangier ound, in
contra t to highly developed areas of the Bay, likely provide
re ilience to terrapin population ,a evidenced by our e -
timated female recruitment rate of 4-7%. Thi recruitment
Ie eJ eem favorable and et a bench mark for future win-
ter ampling, but its relation to population statu i unclear
without knowledge of other vital population
tati tic, e -
pecially an e timate of population growth rate (Mitro 2003).
Further demographic tudy i needed to better establi h the
tatu ofTangier ound terrapin population.
Although terrapins were gen rally abundant in Tangier
ound, the paucity of number in
t. Jerome Creek, a former
harve t site, is enigmatic. The los of terrapins at this for-
merly productive site may be related to harve t or perhap
a broader population decline do ument d during long-term
Herpetological Review 42(2),2011
,
176 ARTICLES
study of terrapin on the nearby Patuxent River (w. Roosen-
burg, Ohio University, pers. comm.).
CONCLUSIO S
The nature of adult female terrapins to aggregate in hi-
bernacula often in high densities and in easily accessible
estuarine bays has made them exceptionally vulnerable to
commercial winter harve t. Such harvest, if unregulated, can
be devastating to local terrapin population. Although no ex-
act history can be reconstructed, winter scrape fishing may
have played a part in the near extirpation of the species in
Che apeake Bay in the early 20'h century (Carr 1952; Ernst
and Lovich 2009; McCauley 1945). In 2007, a well-organized
conservation movement successfully lobbied the Maryland
legislature to permanently close the terrapin fishery. Com-
mercial harvest of diamondback terrapins is now prohibited
throughout Chesapeake Bay.
ow that the commercial terrapin fishery is closed,
adopting scrape fi hing as a winter sampling method could
be of particular value in assessing the past effects of har-
vest and recovery of terrapin population. Scrape fishing of
hibernacula offers unique access to a large portion of the
adult female segment of the population, including female
approaching breeding age. Moreover, scrape sampling offers
a novel mean to advance scientific study of terrapins in win-
ter, an aspect of the biology of the species that has been little
addre sed. In thi light, we recommend further work be con-
ducted to better understand physiological effects of removal
from hibernacula and effects on long-term survival. We be-
lieve winter scrape sampling could be an important element
of terrapin population and scientific study in Chesapeake
Bay and perhaps elsewhere throughout the north temperate
range of the species.
Acknoru/edgments.-This study could not have been conducted
without the boating kills and pecial knowledge of terrapin hiber-
nacula and harve t methods provided by Smith I land re ident D.
Mar hall. Hi willingne
to hare his intjmate knowledge of terra-
pins has made this a very pecial experience. Additional thank go to
W. Roosenburg for a sisting with marking techniques and freely shar-
ing hi
pecial knowledge of terrapin from hi exten ive research on
the Patuxent Rjver. dditional thanks are extended to C. Driscoll and
K. Brittingham for field as i tance and
. Beyer, M. Erwin, E. Grant
and W. Link for review of an early draft of the manu cript. Animal
welfare protocols u ed herein were detailed in a peer-reviewed tudy
plan and approved by the U G Patuxent Wildlife Re earch Center's
Animal Care and Use Committee.
LITERATU RE CiTED
BROOKS, R. J., G. P. BROWN, ADD. A. GALBRAITH. 1991. Effects of a sud-
den increase in natural mortality of adults on a population of
the common napping turtle (Chelydra erpentina). Can. J. Zool.
69:1314-1320.
BUTLER, J. A., G. L. HEINRICH, AI DR.
. EIGEL. 2006. Third workshop on
the ecology, status, and conservation of diamondback terrapin
(Ma/aclemys terrapin): result
and recommendations. Chelon.
Con erv. BioI. 5:331-334.
CARR, A. 1952. Handbook ofTurtle . Cornell University Press, Ithaca,
ew York. 542 pp.
CO GDON, J. D., A. E. Du HAM, A DR.
. VA LOBEN ELS. 1993. Delayed
exual maturity and demographic of Blanding' turtle
(Emydoi-
dea b/alldingit1: implication for con ervation and management
of long-lived organism . Con erv. BioI. 7:826-833.
DORCA, M. E., J. D. WILLSON, AND j. W. GIBBON. 2007. Crab trapping
causes population decUne and demographic change in diamond-
back terrapin over two decade. BioI. Con erv. 137:334-340.
ERNST, C. H., ADJ. E. LoVl H. 2009. Turtle of the United States and
Canada. 2nd ed. John Hopkin University Press, Baltimore, Mary-
land. 827 pp.
GIBBO S, J. w., J. E. LoVlCII, A. D. TUCKER,
. FITZSIMMONS,
DJ. L.
GREEI E. 2001. Demographic and ecological factors affecting con-
ervation and management of the diamondback terrapin (Mala-
c/emys terrapin) in outh Carolina. Chelon. Conserv. BioI. 4:66-74.
GROSSE, A. M., J. D. VAN DIlK, K. L. HOLCOMB, ADJ. C. MAERZ. 2009. Dia-
mondback terrapin mortality in crab pots in a Georgia tidal mar h.
helon. Con erv. BioI. 8:98-100.
HARDEN, L. A.,
. A. DILUZIO, J. W. GIBBON', AND M. E. DORCAS. 2007. pa-
tial and thermal ecology of diamondback terrapins (Ma/ac/emys
terrapin) in a South Carolina alt marsh. J. North CaroUna Acad.
ci. 123:154-162.
HEPPELL,
.
. 1998. Application of life hi tory theory and population
model analysis to turtle conservation. Copeia J998:367-375.
HOYLE, M. E., AND j. W. GIBBO S. 2000. U e of a marked population
of diamondback terrapins (Malaclemys terrapin) to determine
impact of recreational crab pots. Chelon. Conserv. BiDI. 3:735-
737.
MCAULEY, R. H., JR. 1945. The Reptiles ofMaryland and the District of
Columbia. Privately printed, Hager town, Maryland. 194 pp.
MITRO, M. G. 2003. Demography and viability analy es ofa diamond-
back terrapin population. Can. J. Zool. 81:716-726.
ROOSENBURG, W. M. 1991. The diamondback terrapin: habitat require-
ments, population dynamic, and opportunitie for conservation.
[n A. Cbaney, and j. A. Mihursky (ed .),
ew Perspectives in the
Che apeake Bay System: A Re earch and Management Partner-
hip, pp. 227-234. Proceedings of a sympo ium. Chesapeake Bay
Con ortium Publication No. 137, Solomons, Maryland.
--.2004. The impact ofcrab pot fisheries on terrapin (MaLac/emys
terrapin) population : where are we and where do we need to go?
[n C. warth, W. M. Roo enburg, and E. Kiviat (ed .J, Conservation
and Ecology ofTurtle of the Mid-Atlantic Region: A Symposium,
pp. 23-30. Bibliomania, Salt Lake City, Utah.
--,W. CRESKO. M. MODESITIE, A'D M. B. RonBI
. 1997. Diamondback
terrapin (Malaclemys terrapin) mortality in crab pots. Conserv.
BioI. 11:1166-1172.
--, K. L. HALEY, AND . M GUIRE. 1999. Habitat selection and move-
ments of diamondback terrapins, Malaclemys terrapin, in a Mary-
land e tuary. Chelon. Con erv. BioI. 3:425-429.
SEBER, G. A. F. 1973. The E timalion of Animal Abundance. Hafner
Pres,
ewYork,
ewYork. 506 pp.
E1GEL, R. A. 1993. Apparent long-term decline in diamondback ter-
rapin populations at the Kennedy pace Center, Florida. Herpetol.
Rev. 24:102-103.
--, AD]. W. GIBBONS. 1995. Workshop on the ecology, statu, and
management of tile diamondback terrapin (Malac/emys terrapin),
avannah River Ecology Laboratory, 2 August 1994: final results
and recommendations. Chelon. Conserv. Biol.l:240-243.
OKAL, R. R., AD F. J. ROHLF. 1995. Biometry: The Principle and Prac-
tice of tati tic in Biological Re earch. 3nl ed. W. H. Freeman,
ew
York,
ewYork. 887 pp.
H~rp(tological &uiroJ 42(2 ).201J
TUCKER, A. D., ). W. GIBBONS,
D). L. GREE E. 2001. Estimate of adult
survival and migration for diamondback terrapin: con ervation
insight from local extirpation within a metapopulation. Can. ).
Zool. 79:2199-2209.
WARNER, W. W. 1977. Beautiful wimmers. Penguin Book,
ewYork,
ewYork. 304 pp.
WOlAK, M. E., G. W. GILCHRIST, V. A. RUZICKA, D. M.
AllY, A 0 R. M.
CHAMBER. 2010. A contemporary, ex-limited change in body size
of an e tuarine turtle in response to commercial fi hing. Con erv.
BioI. 24:1268-1277.
HerpetologiCflllleuiew. 2011. 42(2). 177-1BO.
(;) 2011 by ociety forthe Study ofAmphibians and Reptiles
ARTICLES 177
WOOD, R. C. 1997. The impact of commercial crab trap on northern
diamondback terrapin , Malaclemys terrapin terrapin. 1/1 J.
an
Abbema (ed.), Proceedings of a yrnposium: Con ervation, Re -
toration, and Management of Tortoises and Turtle -An Interna-
tional Conference, pp. 46-53.
ewYorkTurtle and Tortoise ociety,
ewYork,
ewYork.
YEARICK, E. E, R. C. WOOD, A 0 W. . JOH 0 '. 1981. Hibernation of the
northern diamondback terrapin, Malaclemys terrapin terrapin.
E tuarie 4:78-80.
A Large-Scale Snake Mortality Event
Road mortality has been shown to constitute a consider-
able threat for a variety of herpetofaunal pecie. A number
of po
ible explanations for movement onto or across road
resulting in mortality have been proposed, including sea on-
aI movements between habitats (Bernardino and Dalrymple
1992; mith and Dodd 2003), ea onal dry-down of suitable
habitats (Are co 2005; Bernardino and Dalrymple 1992; Enge
and Wood 2002), de ire to access resources on the other ide
(Andrews and Gibbons 2005), movement to new areas for
breeding (Lebboroni and Corti 2006), and movement and for-
aging follOWing rainfall events (Ashton and Ashton 1988; Carr
1963; Cook 1983; Gibbons and Dorcas 2004; Tennant 1997).
In addition to the hazards of entering or attempting to
cross roads, it appears that aspects of snake natural hi tory
may put them at a greater ri k of road mortality than other
herpetofaunal species. Snakes are known to use warm road
surfaces for thermoregulation (Bernardino and Dalrymple
1992; Enge and Wood 2002; Ro en and Lowe 1994). Andrews
and Gibbon (2005) report that nake road kill may be mag-
nified by immobilization behavior with some nake species
stopping on the road during crossing. Some authors have
uggested that snakes are commonly intentionally targeted
by drivers (Ashley et al. 2007; Rudolph et al. 1999; Shepard et
aI.2008).
In March of2006, we ob erved a snake roadkill event along
a newly opened road leading into the Southwest Florida In-
ternational Airport located in Fort Myers, Lee Co., Florida,
USA (26.5123°
,81.7726°W). The airport developed a new
..........................................................................................................................
JENNIFER EVANS*
LAURA WEWERKA**
EDWIN M. EVERHAM III
and
A. JAMES WOHLPART
Florida GulfCoast University, Fort Myers, Florida 33965, USA
terminal, which included the opening of this new road to the
public in September 2005. The initial report of a snake road-
kiJI event came from ob ervations taken on 12 March 2006,
suggesting that the majority of dead-on-road (DOR) nake
were killed during the preceding week (5-11 March). The ob-
servations were made along 1.5 km of the new airport acces
road; no snake carcasses were found before or after this sec-
tion of the road. We documented this large-scale snake road-
kill event, which appears to have been greater in magnitude
and density than any other snake mortality event reported in
the literature for this small of an area for this short of a time
period (Beck 1938; Hellman 1956; Smith and Dodd 2003).
with the exceptions of mass snake roadkill events associated
with a hurricane (Carr 1963) and a snake migration event
(Tennant 1997).
Methods.-The airport expansion included construction
of a new four-lane roadway and creation of a canal (Figs. lA
and IB). The roadway consisted of both an east and west-
bound terminal road with a grassy median separating them.
The speed limit was 45 mph. Bike paths were located exte-
rior to the roadway and had a width of 1.6 m as compared
to a single road lane width of 3.6 m. Grassy shoulders were
located exterior to the bike paths. The southern shoulder ex-
tended 10-12 m to a low (0.5 m) berm which eparated the
houlder from a 20-25 m wide tormwater treatment swale.
A second berm (1.5 m high and 15 m wide), located south of
the swale, ran parallel to the canal which was approximately
15 m in width.
From 19-22 March, 2006, 1.5 km of the eastbound ter-
minal road (Fig. IB) heading toward the Southwe t Florida
International Airport was walked and snake species were re-
corded. The road was divided into 10 m egments. The loca-
tion of each roadkilled nake along the eastbound ide was
recorded as follows: in the left hand lane of the road near
the median, in the right-hand lane of the road near the bike
path, on the bike path, or on the grassy shoulder. The area
of grass along the road edge that could be reliably observed
H"peto1ogicol Review 42(2). 2011