Southeastern Naturalist
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
334
2019 SOUTHEASTERN NATURALIST 18(2):334–345
Neonate Cottonmouth Spatial Ecology and Habitat
Selection: From Parturition to Hibernation
Zackary J. Delisle1,2,*, Dean Ransom1, and Johanna Delgado-Acevedo1
Abstract - Understanding neonate ecology is imperative for effective knowledge of lifehistory
stages, which have historically been based upon adult ecology. We telemetered 12
neonate Agkistrodon piscivorus (Cottonmouth) and monitored their spatial use, activity, and
habitat selection. Neonates did not have large spatial requirements, nor did they disperse far
before hibernation. Neonates selected habitat edges, and within edges they used areas with
thick vegetative cover. We suggest that the habitat selection and limited use of space by
neonate Cottonmouths are products of the migration their mothers make before parturition.
Habitat edges appear to be important for both parturition and hibernation. Our study offers
valuable insight into the initial life-history stages of Cottonmouths and presents a good
baseline for future research on their ontogenetic ecological development.
Introduction
Numerous studies of the spatial ecology and habitat selection of pit vipers have
been conducted over the last 40 years since the advent of radio telemetry (e.g.,
Reinert 1984, Roth 2005, Sutton et al. 2017, Weatherhead and Prior 1992). In
earlier studies, even the smallest available radio transmitters were too large to be
implanted in small-bodied snakes. Because of the inability of researchers to track
smaller snakes, most of the information known on the life history of snakes is based
on adults (Jellen and Kowalski 2007). This historic bias is concerning because of
the ontogenetic ecological differences that are exhibited by most species (Jellen and
Kowalski 2007).
Technological advances have led to smaller radio transmitters, and therefore
an increase in telemetric studies involving larger-bodied neonates (see Cobb et al.
2005, Figueroa et al. 2008, Gross 2017, Hileman et al. 2015, Howze et al. 2012,
Jellen and Kowalski 2007, Muellman et al. 2018, Pizzatto et al. 2009, Rivas et
al. 2016, Smith 2014). Although small radio transmitters are now available, most
telemetric neonate studies used external radio attachment via an adhesive, but have
suffered high detachment rates (Cobb et al. 2005, Howze et al. 2012, Jellen and
Kowalski 2007, Muellman et al. 2018). Even so, these studies revealed interesting
neonatal behavioral traits (e.g., conspecific scent trailing and arboreal behavior)
and intraspecific ontogenetic ecological differences (Cobb et al. 2005, Figueroa et
al. 2008, Gross 2017, Hileman et al. 2015, Howze et al. 2012, Jellen and Kowalski
2007, Muellman et al. 2018, Pizzatto et al. 2009, Rivas et al. 2016, Smith 2014).
1Department of Biological Sciences, Texas A&M-Commerce, Commerce, TX 75428. 2Current
address - Department of Forestry and Natural Resources, Purdue University, West
Lafayette, IN 47907. *Corresponding author - zdelisle@purdue.edu.
Manuscript Editor: Brad Glorioso
Southeastern Naturalist
335
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
Further telemetric studies involving the neonates of other snake species could yield
additional life-history knowledge that is imperative for understanding all of the
ontogenetic stages of a focal species (Jellen and Kowalski 2007).
Agkistrodon piscivorus (Lacépède) (Cottonmouth) is a semiaquatic pit viper
species native to southeastern United States (Gloyd and Conant 1990). Few Cottonmouth
field studies have incorporated neonates, and none used a telemetric
approach. For these reasons, we intraperitoneally implanted radio transmitters into
neonate Cottonmouths. Our objectives were to (1) discover if intraperitoneal implantation
was more efficacious than external attachment methods, and (2) monitor
neonate Cottonmouth activity, spatial use, and habitat selection. We predicted that
(1) intraperitoneal implantation would result in a smaller percentage of neonates
experiencing radio loss when compared to previous studies using external attachment,
and (2) our telemetered neonates would select areas with more vegetative
ground cover, unlike past non-telemetric neonate Cottonmouth habitat selection
studies that used opportunistic sighting methods (Eskew et al. 2009). We additionally
recorded the habitat characteristics of parturition sites, and monitored the
amount of time that gravid female Cottonmouths spent at parturition sites.
Field-Site Description
Our study site was a constructed wetland (TAMUC Wetlands) owned by Texas
A&M University-Commerce located in Commerce, TX. This site was a hay field
before being constructed into a wetland in the spring of 2007. Construction consisted
of digging several levees and basins of differing size and shape that naturally
filled with water. The eastern side of the wetland was immediately adjacent to Texas
State Highway 24 and transitioned into a mesic bottomland forest with sporadic
patches of grazed cattle pasture to the west. The northeastern section of the property
was comprised of native prairie, and the southern border of the property was
immediately adjacent to grazed cattle pasture (Fig. 1).
Methods
Surgical implantation of transmitters
We obtained neonate Cottonmouths through a separate telemetric project on
adult Cottonmouths within the same study site (Delisle et al. 2019). When we
found a postpartum female aggregated with her neonates at a parturition site,
we captured a select few of the neonates and brought them back to the lab for surgical
implantation of radio transmitters. We anesthetized neonates using isoflurane
(Clipper Distributing Company, LLC, St. Joseph, MO) via an anesthetic vaporizer
(ASA Isoflurane, Harvard Apparatus Limited, Kent, UK), and surgery commenced
immediately upon the loss of the righting reflex. We intraperitoneally implanted
radio transmitters (modified model LPI-2012-COVERT, ~1g, Wildlife Materials
Inc. Murphysboro, IL) into neonates, similar to Smith (2014), following the procedures
described in Reinert and Cundall (1982). Radio transmitters were always
less than 5% of the subject’s body weight. We sanitized surgical instruments using
Southeastern Naturalist
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
336
Figure 1. The Texas A&M-Commerce constructed wetlands shown with all 6 cover types
and every neonate Cottonmouth parturition site (n = 4) and hibernacula site (n = 7).
Southeastern Naturalist
337
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
an autoclave prior to surgery and used isopropanol to maintain sterile conditions
during surgery. Upon completion of the surgery, we used a 4-0 synthetic absorbable
suture (CP Medical Inc. Norcross, GA) to seal the incision, and subsequently
administered 3M vetbond tissue adhesive (3M Animal Care Products, St. Paul, MN)
over the suturing. Snakes were given bedding, water ad libitum, and monitored for
24 h before being released at their point of capture, which was always with their
mother at the parturition site. We refrained from removing radio transmitters before
hibernation because of the correlation between mortality and radio implantation
surgery immediately prior to hibernation, and instead we marked hibernacula locations
and attempted to find them the following spring (Rudolph e t al. 1998).
Home range, dispersal, and activity
We tracked neonates an average of 3 times per week from 0800 to 1800 using
an R1000 radio receiver (Communications Specialist Inc. Orange, CA) and a
3-pronged yagi antenna (Wildlife Materials Inc. Murphysboro, IL). Once a telemetered
snake was located, we used a GPSMAP 64 global positioning system (Garmin
Ltd. Olathe, KS, error = 3.6 m) to record the snake’s location. Upon locating each
snake, we recoreded their activity as either coiled, active (extended or moving), or
fossorial (defined as underground in small-mammal burrow).
We quantified neonate home ranges using the 95% kernel density estimate
(KDE; 95% isopleths) with the least-squares cross-validation bandwidth estimation
in Geospatial Modelling Environment (Beyer 2012, Seaman and Powell 1996).
We also generated 100% minimum convex polygon (MCP) home ranges in order
to compare our results to those of previous telemetric studies on neonate pit vipers
(Howze et al. 2012), and because KDEs can overestimate the home ranges of
herpetofauna (Row and Blouin-Demers 2006). Home ranges were only calculated
for snakes with more than 20 unique locations. We used Arc-Map version 10.2 to
measure the maximum distance dispersed from parturition sites and the distance
between a neonate’s parturition site and hibernacula (defined as the small-mammal
hole that the respective neonate overwintered in). We only measured dispersal for
snakes that were telemetered for a minimum of 30 days, and only measured distances
between parturition sites and hibernacula for snakes that were telemetered
until mid-November (when all neonates remained in the direct proximity of the
small-mammal burrow that they overwintered in).
Home-range habitat preferences
We delineated and mapped emergent herbaceous wetlands (19.05% of study
site), pasture/hay (7.77% of study site), prairie (15.84% of study site), deciduous
forest (21.42% of study site), open water (3.36% of study site), and edge habitat
(32.55% of study site; delineated by a 15-m buffer on the edge of each cover
type [Blouin-Demers and Weatherhead 2001, Homer et al. 2004]). We evaluated
home-range habitat selection within the entire study site, similar to Johnson (1980)
second-order selection, through a use versus availability approach. We generated
Manly selection ratios (MSR; Manly et al. 2002) to compare the proportion of
each available cover type within the study site to the proportion of each cover type
Southeastern Naturalist
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
338
within each snake’s 95% KDE. MSRs were generated using the “adehabitatHS”
package in Program R (R Core Team 2017). Preferred cover types generate an MSR
greater than one, and avoided cover types generate an MSR less than 1 (Manly et
al. 2002). We considered the selection of a certain cover type to be significant if the
95% confidence intervals of that MSR did not overlap one.
Microhabitat selection
We assessed microhabitat selection, similar to Johnson (1980) fourth-order
selection, through a use–availability approach. Upon locating each snake, we
measured ground temperature, ambient temperature, and ground cover at the
snake location and subsequently again at a corresponding random location. We
identified random locations using a random-number generator to compute a bearing
(0°–360°) and distance (1–50 m) from the snake’s location (Table 1). In order
to not further disturb neonates, we flagged both locations and returned later to
record other vegetative microhabitat variables inside a 1-m2 quadrat placed over
the locations (Table 1). Coinciding snake and random microhabitat variables were
always measured on the same day within 5 minutes of each other. Thus, we measured
ground temperature, ambient temperature, and ground cover on the same
day at both locations, and the rest of the microhabitat variables within a week
later on the same day at both locations. We only measured microhabitat variables
if the individual was visually observed above ground and had moved more than
1 m from the previous location.
To evaluate microhabitat preferences, we generated a priori binary generalized
linear models in Program R (Table 2; R Core Team 2017) where “1” represented
Table 1. Microhabitat variables recorded at each unique neonate Cottonmouth location and corresponding
random location. The variable “ground cover” was objectively quantified at random
locations by putting a flexible rubber hose in the exact same position that the corresponding neonate
was found in. All variables were measured at neonate and random locations. *variables that were
measured at parturition sites.
Variable Abbreviation Description
Distance from water* dfw Distance (m) to the nearest water > 5 cm deep
Ground temperature gt Temperature (C) of ground within 2 cm of snake
Ambient temperature at Temperature (C) taken 1.3 m above the location
Ground cover gcv % of snake's body concealed by vegetation < 1.3 m
Max. veg. height mvh Tallest vegetation (mm) within the m2 quadrat
Canopy cover* cc % of canopy cover > 1.3 m
Litter* ltr % of litter cover within the m2 quadrat
Coarse woody debris* cwd % of coarse woody debris within the m2 quadrat
Bare ground* bg % of bare ground within the m2 quadrat
Water* wtr % of water cover within the m2 quadrat
Forb* frb % of forb cover within the m2 quadrat
Grass* grs % of grass cover within the m2 quadrat
Sedge and rush* sr % of sedge and rush cover within the m2 quadrat
Woody* wdy % of woody cover within the m2 quadrat
Southeastern Naturalist
339
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
snake locations and “0” represented random locations. We developed a priori models
through our ecological knowledge of Cottonmouths and the study site, which
allowed us to further test biological hypotheses (Burnham and Anderson 2002). For
instance, each of our models were based on a cover type (e.g., edge, forest, wetland,
etc.) or specific habitat-selection reasoning (e.g., thermal, cover, cryptic, etc.;
Table 2). We assessed models using an information theoretic approach. Best-supported
models were determined using Akaike’s information criterion adjusted for
small sample sizes (AICc), the difference between the most supported AICc model
and each following model (Δi), Akaike weights (ωi), and evidence ratios (ωi/ωj)
(Burnham and Anderson 2002). We calculated the standardized β-coefficients of
each parameter within the most supported model using the ‘reghelper’ package in
Program R (R Core Team 2017) and considered a parameter significant if the probability
of receiving a z-statistic greater than the absolute value of the parameter’s
z-statistic was less than 0.05.
Maternal activity
If we found a postpartum female (females from Delisle et al. 2019) aggregated
with her neonates, we immediately recorded the location and measured several
different vegetative variables at the parturition site (Table 1). Following this, we
recorded how many days the mother remained at the natal area with her neonates.
All results are presented ± 95% CI.
Results
We telemetered 12 neonates from 4 different females (female 1 = 4 neonates,
female 2 = 4 neonates, female 3 = 3 neonates, and female 4 = 1 neonate) from 17
August 2017 to 8 December 2017 for a total of 256 unique locations. The mean
number of unique locations per individual was 21. Unknown predators depredated
2 neonates, and we lost the signal of 2 radio transmitters (due to unknown causes).
One of our telemetered neonates excreted their radio (known because this individual
was observed after we found the radio). Seven of our telemetered specimens
were telemetered until the end of the study. At least 1 neonate from every mother
survived until the end of the study.
Table 2. A priori binary generalized linear models for neonate Cottonmouth microhabitat selection.
Refer to Table 1 for information on microhabitat variable abbreviations.
Model name Microhabitat variables included
Thermal gt + at
Cover ltr + bg + wtr + frb + grs + sr + wdy + cwd
Cryptic gcv + cc + mvh
Wetland sr + wtr + dfw + gt
Forest cc + cwd + grs + wdy + at + bg + dfw
Edge cc + grs + wdy + gcv + cwd + sr + at + dfw
Prairie ltr + frb + grs + wdy + gt
Global dfw + gt + at + gcv + cc + ltr + bg + wtr + frb + grs + sr + wdy + cwd + mvh
Southeastern Naturalist
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
340
Home range, dispersal, and activity
We computed fall home-range sizes for 7 individuals. The mean (± 95% CI)
100% MCP was 0.93 ha (± 0.33–1.53) and the mean (± 95% CI) 95% KDE was
1.15 ha (± 0.60–1.70). The mean (± 95% CI) maximum distance dispersed from
a parturition site was 159.71 m (± 106.27–213.15) and the mean (± 95% CI)
distance from a neonate’s hibernacula to their parturition site was 104.47 m (±
15.16–193.78). The amount of times we telemetered neonates to locations above
ground, either active or coiled, decreased from September to November. We never
observed neonates above ground during the month of December.
Habitat selection and maternal activity
Neonates 95% KDE contained significantly more edge cover (MSR = 1.717
± 1.2556–2.1792) and significantly less prairie cover (MSR = 0.053 ± -0.0876–
0.1946) than what was proportionally available to them within the study site.
No other cover types showed a significant selection. We collected microhabitat
data from 155 neonate locations and 155 corresponding random locations. The
most supported model was the edge model (AICc = 187.28, Δi = 0 , ω i = 0.70,
ωi/ωj = 2.34). The only other model with any support was the global model
(AICc = 188.98, Δi = 1.70, ωi = 0.3). Within the edge model, neonates selected areas
with significantly more canopy cover (β = 3.120, P(>|t|) = 0.004) and ground
cover (β = 4.691, P(>|t|) = < 0.001; Table 3). No microhabitat variables were significantly
selected against.
Gravid females (n = 4) selected parturition sites that averaged 29.6 m from water
and had high amounts of canopy cover, sedge and rush cover, and coarse woody debris
cover. Postpartum females stayed at the natal area for an average of 18.3 days.
Discussion
Intraperitoneal implantation methods resulted in only 8.3% of our radios being
separated from the subject (1 individual excreted radio). Previous telemetric
Table 3. Standardized β-coefficients of the model parameters within the most supported model—the
edge model—and their standard errors, and z-statistics. We considered parameters significant if the
probability of receiving a z-statistic greater than the absolute value of the parameter’s z-statistic was
less than 0.05. * = statistically significant parameters, β = standardized regression coefficient, SE =
standard error, and P(>|z|) = probability of receiving a value > |z|.
Parameter β SE z value P(>|z|)
Canopy cover* 1.562 0.273 5.713 less than 0.001
Grass cover* 0.720 0.254 2.831 0.005
Woody cover 0.091 0.233 0.389 0.698
Ground cover* 2.349 0.316 7.443 less than 0.001
Coarse woody debris cover 0.548 0.295 1.859 0.063
Sedge and rush cover* -0.809 0.264 -3.061 0.002
Ambient temperature* 0.635 0.209 3.040 0.002
Distance from water -0.275 0.201 -1.365 0.172
Southeastern Naturalist
341
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
research on neonate snakes that used external attachment methodology reported
higher detachment rates of 57.1% (8/14; Howze et al. 2012), 12.5% (2/16; Jellen
and Kowalski 2007), and 73.7% (16/22; Muellman et al. 2018); supporting our
first prediction. However, 2 of our neonates were depredated and 2 radio signals
were lost. Even though neonates are more vulnerable to predation than adult snakes
(Bonnet et al. 1999), it is possible that our methods subjected neonates to higher
risks of predation. One of our individuals excreted their radio, which has been
shown to occur in other species via the alimentary tract encapsulating the radio and
then ridding it through defecation (Pearson and Shine 2002).
Our study suggests that the spatial hierarchy of neonate Cottonmouth site selection
at the TAMUC wetlands was primarily based on parturition site selection by
the mother. Three of 4 gravid females made distinct migrations out of the wetland to
the edge of the neighboring mesic forest where parturition occurred in mid-August
to mid-September (Fig. 1). The parturition migrations of gravid females seemed to
benefit neonates in several ways. Forest edges have been found to support several
species of burrowing small mammals (Sekgororoane and Dilworth 1995), and many
pit viper species use these burrows as hibernacula (Brown 1982, Jellen and Kowalski
2007). Such use proved especially true for neonate Cottonmouths in our
study, as we frequently observed them underground in small-mammal burrows
during November and December. The parturition migration incurred by gravid
females relieves their neonates from having to make this migration themselves in
order to find suitable hibernacula. The likely result is increased neonate survivorship,
as neonates required to migrate from their natal grounds to a wintering area
could be more susceptible to predation (Bonnet et al. 1999). Furthermore, neonates
did not travel far from their parturition sites before entering hibernacula (mean =
104.47 m), with 2 individuals selecting hibernacula within 15 m of their mother’s
parturition site (Fig. 1). Although our sample size is limited, these short distances
could be evidence that gravid females are selecting parturition sites partially based
upon the hibernacula needs of their young.
At the micro spatial scale, gravid females selected parturition sites that had thick
canopy and coarse woody debris cover, and their neonates also selected areas that
had significantly more ground cover (Table 3). Forest edges readily provide these
features, all of which likely help keep neonates hidden from terrestrial and aerial
predators. The protective features found in forest edges could be another reason
gravid females migrated there for parturition. Other viviparous snake species also
select parturition sites that have dense vegetative cover because of how susceptible
neonates and postpartum females are to predation (Bonnet et al. 1999, Greene 1997,
Shine and Bonnet 2009). Neonate preference for thick vegetation coincides with
findings from the few existing telemetric studies on other neonate pit vipers (Howze
et al. 2012, Jellen and Kowalski 2007). However, our findings differ from those
of previous studies of habitat selection that found neonate Cottonmouths to select
areas with little vegetative cover (Eskew et al. 2009) , supporting our second prediction.
Past studies could have found differing results because their observations
were restricted to summer nights and their analyses were based on visual encounter
Southeastern Naturalist
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
342
surveys (Eskew et al. 2009), which can lead to visual biases (i.e., a surveyor may
be more likely to see an unconcealed snake then a snake concealed by vegetation).
Furthermore, Cottonmouths are often nocturnal hunters in the summer (Lillywhite
and Brischoux 2012), and neonates could be selecting less-concealed microsites
when foraging at night. However, even with minimal disturbance from data collection,
it is possible that repeated telemetry could have made neonates feel more
vulnerable than non-telemetered individuals, leading to telemetered individuals
selecting areas with greater vegetative ground cover.
The spatial use of neonates before their first winter was small, with 100%
MCPs and maximum dispersal distances averaging just 0.93 ha and 159.71 m,
respectively. Spatial use theory suggests that smaller home-range sizes are indicative
of essential resources being in close proximity to one another because
movement, most especially for neonates, is associated with increased predation
and energy expenditure (Bonnet et al. 1999, Greene 1997, MacArthur and Pianka
1966). Therefore, the small home-range sizes and short dispersal distances of
neonates could be further evidence that the parturition migrations made by their
mothers are beneficial to them.
The maternal behavior of Cottonmouths has been well documented (Greene et
al. 2002, Hoss and Clark 2014, Hoss et al. 2014). Postpartum adult females usually
depart from their natal areas 10–14 days after parturition when neonates complete
their first shed (Hoss et al. 2014). However, we found postpartum females to stay
at the parturition site an average of 18.3 days, with 1 female staying in the natal
area for 28 days. We also radio tracked 1 postpartum female back to her parturition
site after she had departed the area. This female used her parturition site as a hibernaculum,
and at least 1 of her neonates (snake 47) also hibernated there with her. In
the following spring, on 21 March 2018, we observed this same mother aggregated
with snake 47 at the same site. Jellen and Kowalski (2007) also found postpartum
adult female Sistrurus catenatus (Rafinesque) (Eastern Massasauga) and their neonates
to hibernate at parturition sites.
Although most of our gravid females migrated out of the wetland for parturition,
1 did not and instead this individual birthed on the edge of a pond within the wetland
(Fig. 1). Her only telemetered neonate (snake 56) remained in her natal area
for about 1 week before making a near linear eastward movement from the pond
to the forest edge. Snake 56 hibernated in the forest edge near many of the other
neonates, and we believe she perhaps scent-trailed conspecifics to find the area. The
use of conspecific scent-trailing to locate hibernacula has also been documented
in neonate Crotalus horridus L. (Timber Rattlesnake; Brown and MacLean 1983,
Cobb et al. 2005).
Our initial sample size (n = 12) suffered from predation and lost signals, resulting
in 7 individuals from 4 different litters being telemetered until the end of our
study. Small sample sizes comprised from only a few litters have been prevalent in
the telemetric literature regarding neonate snakes, and are likely due to high predation
and detachment rates, and the difficulty of finding multiple litters in a given
year (Cobb et al. 2005, Howze et al. 2012, Jellen and Kowalski 2007, Muellman et
Southeastern Naturalist
343
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
al. 2018, Pizzatto et al. 2009). Although it would have been ideal to have all of our
neonates born from different litters, finding females with their litters out in the open
is somewhat fortuitous (usually females and their litters resided under thick coarse
woody debris). Therefore, we did not want to sacrifice sample size on the uncertainty
of finding more litters. Acknowledging our low sample size from few litters,
we believe our study offers valuable insight into the initial life-history stages of
Cottonmouths and presents a good baseline for future research on the ontogenetic
ecological development of Cottonmouths.
Acknowledgments
We would like to thank P. Arron and M. Crowell for allowing us to track Cottonmouths
on their properties. We also thank interns E. Engelhardt and A. Savage for their help with
data collection, and W.I. Lutterschmidt for providing training on radio-implantation surgery,
discussions regarding random site sampling, and comments upon review of this manuscript.
Our research was funded by the TAMUC 2017 FY Unit Strategic Initiative Funding
and the TAMUC Ronald E. McNair Post-Baccalaureate Achievement Program. Research
was conducted under the Texas A&M-Commerce Institute’s Animal Care and Use committee
permit P17-036.
Literature Cited
Beyer, H.L. 2012. Geospatial Modelling Environment, Version 0.7.2.1. Available online at
http://www.spatialecology.com/gme. Accessed 15 February 2018.
Blouin-Demers, G., and P.J. Weatherhead. 2001. Habitat use by Black Rat Snakes (Elaphe
obsolete obsolete) in fragmented forests. Ecology 82:2882–2896.
Bonnet, X., G. Naulleau, and R. Shine. 1999. The dangers of leaving home: Dispersal and
mortality in snakes. Biological Conservation 89:39–50.
Brown, W.S. 1982. Overwintering body temperatures of Timber Rattlesnakes (Crotalus
horridus) in northeastern New York. Journal of Herpetology 16:145–150.
Brown, W.S., and F.M. MacLean. 1983. Conspecific scent-trailing by newborn Timber
Rattlesnakes, Crotalus horridus. Herpetologica 39:430–436.
Burnham, K.P., and D.R. Anderson. 2002. Model Selection and Multimodel Inference: A
Practical Information-Theoretic Approach. 2nd Edition. Springer-Verlag, Berlin, Germany.
323 pp.
Cobb, V.A., J.J. Green, T. Worrall, J. Pruett, and B. Glorioso. 2005. Initial den location
behavior in a litter of neonate Crotalus horridus (Timber Rattlesnakes). Southeastern
Naturalist 4:723–730.
Delisle, Z.J., D. Ransom Jr., W.I. Lutterschmidt, and J. Delgado-Acevedo. 2019. Site-specific
differences in the spatial ecology of Northern Cottonmouths. Ecosphere 10:e02557.
DOI:10.1002/ecs2.2557.
Eskew, E.A., J.D. Willson, and C.T. Winne. 2009. Ambush-site selection and ontogenetic
shifts in foraging strategy in a semi-aquatic pit viper, the Eastern Cottonmouth. Journal
of Zoology 277:179–186.
Figueroa, A., E.A. Dugan, and W.K. Hayes. 2008. Behavioral ecology of neonate Southern
Pacific Rattlesnakes (Crotalus oreganus helleri) tracked with externally attached
transmitters. Pp. 365–376, In W.K. Hayes, K.R. Beaman, M.D. Cardwell, and S.P. Bush
(Eds.). Biology of the Rattlesnakes. Loma Linda University Press, Loma Linda, CA.
606 pp.
Southeastern Naturalist
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
344
Gloyd, H.K., and R. Conant. 1990. Snakes of the Agkistrodon complex: A monographic
review. Contributions to Herpetology, No. 6. Society for the Study of Amphibians and
Reptiles. 614 pp.
Greene, H.W. 1997. Snakes: The Evolution of Mystery in Nature. University of California
Press, Berkeley, CA. 365 pp.
Greene, H.W., P.G. May, D.L. Hardy, Sr, J.M. Sciturro, and T.M. Farrell. 2002. Parental
behavior by vipers, Pp. 179–205, In G.W. Schuett, M. Höggren, M.E. Douglas, and H.W.
Greene (Eds.). Biology of the Vipers. Eagle Mountain Publishing, LC. Eagle Mountain,
UT. 580 pp.
Gross, I. 2017. Habitat use, dispersal, and hibernation of maternal and neonatal Copperheads
(Crotalinae; Agkistrodon) in a managed southeastern forest. Ph.D. Dissertation.
Alabama Agricultural and Mechanical University, AL. 129 pp.
Hileman, E.T., D.R. Bradke, D.M. Delaney, and R.B. King. 2015. Protection by association:
Implications of scent trailing in neonate Eastern Massasaugas (Sistrurus catenatus).
Herpetological Conservation and Biology 10:654–60.
Homer, C., C. Huang, L. Yang, B. Wylie, and M. Coan. 2004. Development of a 2001
national landcover database for the United States. Photogrammetric Engineering and
Remote Sensing 70:829–840.
Hoss, S.K., and R.W. Clark. 2014. Mother Cottonmouths (Agkistrodon piscivorus) alter
their antipredator behavior in the presence of neonates. Ethology 120:933–941.
Hoss, S.K., M.J. Garcia, R.L. Earley, and R.W. Clark. 2014. Fine-scale hormonal patterns
associated with birth and maternal care in the Cottonmouth (Agkistrodon piscivorus),
a North American pitviper snake. General and Comparative Endocrinology 208:85–93.
Howze, J.M., K.M. Stohlgren, E.M. Schlimm, and L.L. Smith. 2012. Dispersal of neonate
Timber Rattlesnakes (Crotalus horridus) in the southeastern coastal plain. Journal of
Herpetology 46:417–422.
Jellen, B.C., and M.J. Kowalski. 2007. Movement and growth of neonate Eastern Massasaugas
(Sistrurus catenatus). Copeia 2007:994–1000.
Johnson, D.H. 1980. The comparison of usage and availability measurements for evaluating
resource preference. Ecology 61:65–71.
Lillywhite, H.B. and F. Brischoux. 2012. Is it better in the moonlight? Nocturnal activity
of insular Cottonmouth snakes increases with lunar light levels. Journal of Zoology
286:194–199.
MacArthur, R.H., and E.R. Pianka. 1966. On optimal use of a patchy environment. The
American Naturalist 100:603–609.
Manly, B.F.J, L.L. McDonald, D.L. Thomas, T.L. McDonald, and W.P. Erickson. 2002.
Resource Selection by Animals, Statistical Design and Analysis for Field Studies. 2nd
Edition. Kluwer Academic Publishers, Dordrecht, The Netherlands. 221 pp.
Muellman, P.J., O.D. Cunha, and C.E. Montgomery. 2018. Crotalus horridus (Timber
Rattlesnake) maternal scent trailing by neonates. Northeastern Naturalist 25:50–55.
Pearson, D.J., and R. Shine. 2002. Expulsion of intraperitoneally implanted radiotransmitters
by Australian pythons. Herpetological Review 33:261–263.
Pizzatto, L., T. Madsen, G.P. Brown, and R. Shine. 2009. Spatial ecology of hatchling Water
Pythons (Liasis fuscus) in tropical Australia. Journal of Tropical Ecology 25:181–191.
R Core Team. 2017. R: A Language and Environment for Statistical Computing, Version
3.4.3. Available online at http://www.R-project.org/. R Foundation for Statistical Computing,
Austria. Accessed 25 February 2018.
Reinert, H.K. 1984. Habitat separation between sympatric snake populations. Ecology
65:478–486.
Southeastern Naturalist
345
Z.J. Delisle, D. Ransom, and J. Delgado-Acevedo
2019 Vol. 18, No. 2
Reinert, H.K., and D. Cundall. 1982. An improved surgical implantation method for radiotracking
snakes. Copeia 1982:702–705.
Rivas, J.A., C.R. Molina, S.J. Corey, and G.M. Burghardt. 2016. Natural history of neonatal
Green Anacondas (Eunectes murinus): A chip off the old block. Copeia 104:402–410.
Roth, E.D. 2005. Spatial ecology of a Cottonmouth (Agkistrodon piscivorus) population in
east Texas. Journal of Herpetology 39:308–312.
Row, J.R., and G. Blouin-Demers. 2006. Kernels are not accurate estimators of home-range
size for herpetofauna. Copeia 2006:797–802.
Rudolph, D.C., S.J. Burgdorf, R.R. Schaefer, R.N. Conner, and R.T. Zappalorti. 1998.
Snake mortality associated with late season radio-transmitter implantation. Herpetological
Review 29:155–156.
Seaman, D.E., and R.A. Powell. 1996. An evaluation of the accuracy of kernel density estimators
for home range analysis. Ecology 77:2075–2085.
Sekgororoane, G.B., and T.G. Dilworth. 1995. Relative abundance, richness, and diversity
of small mammals at induced forest edges. Canadian Journal of Zoology 73:1432–1437.
Shine, R., and X. Bonnet. 2009. Reproductive biology, population viability, and options for
field management. Pp. 172–200, In S. Mullin, and R. Seigel (Eds.). Snakes: Ecology and
Conservation. Cornell University Press, Ithaca, NY. 365 pp.
Smith, K.P. 2014. The neonate ecology of the Northern Pine Snake (Pituophis melanoleucus)
in the New Jersey Pine Barrens. Ph.D. Dissertation. Drexel University, Philadelphia,
PA. 88 pp.
Sutton, W.B., Y. Wang, C.J. Schweitzer, and C.J. McClure. 2017. Spatial ecology and multiscale
habitat selection of the Copperhead (Agkistrodon contortrix) in a managed forest
landscape. Forest Ecology and Management 391:469–481.
Weatherhead, P.J., and K.A. Prior. 1992. Preliminary observations of habitat use and movements
of the Eastern Massasauga Rattlesnake (Sistrurus c. catenatus). Journal of Herpetology
26:447–452.