Ecology, 82(10), 2001, pp.

2882–2896
q 2001 by the Ecological Society of America

HABITAT USE BY BLACK RAT SNAKES (ELAPHE OBSOLETA OBSOLETA)

IN FRAGMENTED FORESTS
GABRIEL BLOUIN-DEMERS1 AND PATRICK J. WEATHERHEAD2
Department of Biology, Carleton University, 1125 Colonel By Drive, Ottawa, Ontario K1S 5B6, Canada

Abstract. Declining nest success of forest birds in fragmented habitat has been attri-
buted to increased nest predation. Better understanding of this problem and potential so-
lutions to it require information on why nest predators are attracted to habitat edges. Toward
this end we investigated habitat use by black rat snakes (Elaphe obsoleta obsoleta), an
important avian-nest predator in eastern deciduous forests. We radio-tracked 52 black rat
snakes for periods of 3–41 mo from 1996 to 1999. All black rat snakes exhibited a strong
preference for edge habitats. Consistent with edges being used because they facilitate
thermoregulation, gravid females associated more strongly with edges than did males and
nongravid females, and sites used by snakes when shedding were significantly associated
with habitat edges. Gravid females lost an average of .20% of their body mass, while
nongravid females and males did not lose mass, suggesting that edges were not used because
they offered high success in foraging. Similarly, an increase in use of edge habitat through
the season by all rat snakes was inconsistent with the snakes being attracted principally to
hunt: avian prey would have been more abundant in spring when birds were breeding, and
the density of small mammals in edges did not vary seasonally. Also, snakes moved longer
distances and were found traveling more often when located in forests. Because our results
collectively are most consistent with the hypothesis that rat snakes use edges for ther-
moregulatory reasons, the negative impact of the snakes on nesting birds may be coinci-
dental; the snakes primarily use edges for reasons other than foraging but opportunistically
exploit prey they encounter there. Rat snakes appeared to respond to the edge structure
rather than to how the edge was created (natural vs. artificial). Thus, fragmentation of
forests by humans has created habitat structurally similar to that preferred by rat snakes
in their natural habitat, thereby inadvertently increasing contact between the snakes and
nesting birds.
Key words: black rat snake; breeding birds; ecdysis; edge effects; Elaphe obsoleta obsoleta;
fragmentation; habitat use; Ontario Canada; nest predation.

INTRODUCTION 1995). Here we investigate the extent of, and reasons

underlying, use of habitat edges by black rat snakes
Habitat loss and fragmentation are the two main
(Elaphe o. obsoleta), and consider the implications of
threats to biodiversity (Wilson 1992). Habitat loss has
these results for the role of black rat snakes as impor-
direct consequences for species abundance and diver-
tant nest predators of forest birds.
sity because it reduces the available living space. The
Most research on forest fragmentation has focused on
consequences of fragmentation are more complex and
species that are adversely affected, often using birds as
depend upon the degree of fragmentation, the shape
the model system (Murcia 1995, McCollin 1998, Tewks-
and spatial arrangement of the fragments, and the com-
bury et al. 1998). Habitat fragmentation is thought to be
position of the separating habitat (Andreassen et al.
one of the key factors responsible for the decline of many
1998). One consequence of habitat fragmentation is an
forest-bird species (Bolger et al. 1991, Donovan et al.
increase in the proportion of edge habitat in the land-
1995). These declines may result from forest fragments
scape, which in turn can affect ecosystems by modi-
becoming too small to be suitable for forest-interior spe-
fying ecological relationships such as predator–prey
cies or because nesting success is lower in fragmented
interactions (Donovan et al. 1995, Robinson et al.
forest. The decline in nest success may be a function of
fragmentation enhancing the density or success of gen-
Manuscript received 10 February 2000; revised 21 September
2000; accepted 6 October 2000; final version received 8 Novem- eralist and edge-associated predators (Yahner and Scott
ber 2000. 1988, Andrén 1992, Harrison and Bruna 1999) or of in-
1 Present address: Department of Evolution, Ecology, and creasing brood parasitism (Brittingham and Temple 1983,
Organismal Biology, Botany and Zoology Building, The Ohio Robinson et al. 1995). Nest predation is the most signif-
State University, 1735 Neil Avenue, Columbus, Ohio 43210- icant cause of nest failure in birds (Ricklefs 1969, Martin
1293 USA.
2 Present address: Department of Natural Resources and 1988) and has been demonstrated to increase dramatically
Environmental Sciences, University of Illinois, 1102 South with proximity to some edges (Andrén and Angelstam
Goodwin Avenue, Urbana, Illinois 61801 USA. 1988, Burkey 1993, King et al. 1998). However, the rea-
2882
October 2001 HABITAT USE BY RAT SNAKES 2883

son that predation is higher in some habitat edges remains in edges than in other habitats, but their movement
poorly understood. patterns should be short and localized.
More detailed information about the use of habitat Weatherhead and Charland (1985) proposed that
edges by predators is a prerequisite for understanding black rat snakes prefer habitat edges in Ontario (the
the interrelation between landscape configuration and northern extreme of the snakes’ range) because greater
the importance of predation in local animal commu- exposure to sun along edges facilitates thermoregula-
nities (Angelstam 1986, Paton 1994, Tewksbury et al. tion. Ectothermic animals obtain heat from their en-
1998, Heske et al., in press). To meet the goal of in- vironment and therefore regulate their body tempera-
creasing our understanding of the relation between hab- ture by adjusting microhabitat use and timing of activ-
itat edges and predation, we chose black rat snakes as ity (Huey et al. 1989, Krohmer 1989, Grant 1990).
the subject of our research. Snakes have been identified Thermoregulation is probably the single most impor-
as important nest predators (Loiselle and Hoppes 1983, tant proximate factor in habitat use and timing of ac-
Robinson et al. 1995, Fritts and Rodda 1998), and Ela- tivity of terrestrial squamates (Grant 1990, Reinert
phe obsoleta is particularly well known in this regard 1993), with the probable exception of tropical species
(Fitch 1963, Jackson 1970, 1978, Stickel et al. 1980, (Shine and Madsen 1996). Several predictions follow
Weatherhead and Robertson 1990). Second, previous from this thermoregulation hypothesis. First, in addi-
work has shown that black rat snakes preferentially use tion to a general preference for edges by all individuals,
edges (Weatherhead and Charland 1985, Durner and those snakes with a greater need to maintain optimal
Gates 1993). Third, as we outline below, there are sev- body temperatures should use edges more. Gravid fe-
eral clearly testable alternative hypotheses for why males of many viviparous snake species thermoregulate
black rat snakes might prefer habitat edges. Thus, black more precisely than nongravid females and males
rat snakes provide an opportunity to identify the eco- (Charland and Gregory 1990, Cobb and Peterson 1991,
logical factors that might explain observed patterns of Schwarzkopf and Shine 1991, Brown and Weatherhead
habitat use. 2000). Although the same type of behavior has been
suggested to occur in oviparous species (Shine and
One reason that black rat snakes might prefer habitat
Madsen 1996), it has not been formally documented.
edges is that more avian prey is available in those hab-
We predicted that, prior to egg-laying, gravid female
itats. If so, then higher predation on birds’ nests in
rat snakes should use habitat edges more extensively
edge habitat would be a direct consequence of more
than do nongravid females and males. Similarly, when
birds nesting in that habitat, and/or their nests being
snakes shed their skins (ecdysis), they maintain higher
more accessible to snakes in edge habitat. Increased
body temperatures (Gibson et al. 1989, Peterson et al.
density of birds near forest edges has been documented
1993). Therefore, we also predicted that sites used for
at our study site (J. Jones, personal communication),
ecdysis should occur disproportionately along habitat
consistent with what has been documented elsewhere
edges, and in this case, we expected males, nongravid
(Gates and Gysel 1978, Paton 1994). We tested this
females, and gravid females all to use these sites.
hypothesis indirectly by examining seasonal patterns Another reason that rat snakes might be attracted to
of habitat use by black rat snakes. If rat snakes are edges is that these habitats offer greater protection from
attracted to edges to hunt for birds’ eggs and nestlings, predation. This hypothesis also predicts that gravid fe-
then use of edges should be most pronounced in May males should use edge habitat more than other snakes
and June when the vast majority of birds in our study (Cooper et al. 1990, Charland and Gregory 1995).
area are nesting (Weir 1989). However, in this case their preference for edges is pre-
An alternative version of the foraging hypothesis is dicted to be a consequence of their impaired mobility
that black rat snakes are attracted to edges to hunt for (Seigel et al. 1987), which makes them more vulnerable
prey other than birds. Of the 97 food items Fitch (1963) to predators. If gravid females use edges to avoid pred-
identified from stomachs of black rat snakes, 23 (24%) ators, then we predicted that they should be concealed
were birds and 65 (67%) were small mammals. Thus, more than other snakes when in edges.
rat snakes might be attracted to edges because small An additional goal of our study was to quantify the
mammals are more abundant. If black rat snakes use snakes’ use of natural edges (e.g., boundaries between
edges for foraging on small mammals, we predicted marsh and forest) and artificial edges (e.g., boundaries
that small mammal densities should be higher in edges between forest and cleared field). The two previous
than in other habitats, and any seasonal variation in studies of habitat use by black rat snakes (Weatherhead
mammal densities should be mirrored by seasonal var- and Charland 1985, Durner and Gates 1993) were con-
iation in use of edges by rat snakes. A corollary of ducted in landscapes where the total availability of
both foraging hypotheses is that snakes should gain edge habitat had been increased substantially by human
more mass while in edges compared to other habitats, activity (e.g., forest clearing for agriculture). Although
consistent with foraging success in edges being higher. both studies showed that black rat snakes preferred
Also, because rat snakes are active hunters, if they are edge habitats, that preference could have been an ar-
in edges primarily to forage, they should be more active tifact of the artificially fragmented study areas. Deter-
2884 G. BLOUIN-DEMERS AND P. J. WEATHERHEAD Ecology, Vol. 82, No. 10

mining whether the snakes differentiate between nat- The hibernacula had been located by radio-tracking
ural and artificial edges will provide additional insight snakes to their hibernation sites, both as part of pre-
into why the snakes use edges and may suggest ways vious studies of black rat snakes (Weatherhead and
in which that use could be modified. Both previous Charland 1985, Weatherhead and Hoysak 1989) and as
studies (Weatherhead and Charland 1985, Durner and part of the present study. To capture emerging rat
Gates 1993) also involved relatively small numbers (7 snakes, we constructed 1.5 m high perimeter fences
and 31 snakes, respectively) followed for relatively around hibernacula openings. We buried the bottom
short periods of time (3 and 5 mo, respectively). In our edge of the plastic fences and installed a funnel trap
study we collected data from 52 individuals tracked for in one corner of the fence. Traps were installed in early
periods of up to 41 mo. April and we visited sites beginning prior to the start
Our final goal was to relate our results to the man- of emergence each spring (Blouin-Demers et al. 2000b)
agement of black rat snakes. Our study population of and continuing until the last snake emerged (late May).
black rat snakes occurs as a separate population at the Upon capture, we determined the sex of snakes by gent-
northern extreme of the species’ range in eastern On- ly probing for the presence of hemipenes, measured
tario. Recently this population has been classified as their snout–vent length (SVL) to the nearest 1 mm with
threatened in Canada (Prior and Weatherhead 1998). a metric tape, weighed them to the nearest 1 g with a
Conservation efforts aimed at improving habitat for calibrated spring scale, and marked them by subcuta-
this species that rely on the two existing studies of neous injection of a passive integrated transponder (PIT
habitat use (Weatherhead and Charland 1985, Durner tag). The experimental animals we selected for this
and Gates 1993) would presumably involve preserving study were captured centrally in the study area. This
or even creating edge habitats. Such an approach would increased the likelihood that the snakes would remain
conflict directly with conservation efforts for many in the study area during the active season, although we
threatened or endangered forest species that are hurt modified the exact boundaries of the study area to ac-
by forest fragmentation. The first step toward resolving commodate the snakes as necessary.
this potential conflict is to be certain that a conflict Our choice of study animals from among the animals
exists. By determining habitat preferences of black rat we captured was based on sex and size. Snakes had to
snakes under natural conditions and the factors that be large enough to carry a radio transmitter (maximum
influence habitat use, this study will provide an em- ratio of transmitter mass:body mass 5 0.025:1) and,
pirical basis necessary for sound management. because females reproduce every second to third year
on average (G. Blouin-Demers, unpublished data), we
METHODS had to track more females than males to obtain an ad-
Study area equate sample of gravid females. The transmitters we
used weighed 8.6 g and had a battery life of 20 mo at
We conducted this research from 1996 to 1999 in the 208C (Model SI-2T, Holohil Systems Incorporated,
immediate vicinity of the Queen’s University Biolog- Carp, Ontario). From May 1996 to October 1999, we
ical Station near Chaffey’s Locks in eastern Ontario followed 23 males and 41 females for periods ranging
(448349 N, 768199 W). The study area was ;10 by 3 from 16 d to 41 mo. We followed 52 of those snakes
km and encompassed the Biological Station. The dom- (18 males and 34 females) for .3 mo and used only
inant geological feature of the area is the exposed these 52 individuals in our analysis of habitat use.
southeastern extension of the Canadian Shield referred
to as the Frontenac Axis. This area is characterized by Radiotelemetry
strongly rolling terrain with ridges of granite outcrops To implant transmitters we modified the surgical
alternating with valleys approximately every 500 m and technique described by Reinert and Cundall (1982).
numerous small lakes and wetlands. The study area is This involved using sterile techniques to implant sur-
mostly second growth forest dominated by deciduous gically a radio transmitter in the body cavity of the
trees (Acer, Quercus, Ostrya, Populus, Fraxinus, Car- snake, suturing the transmitter to a rib to prevent its
ya, Ulmus, Fagus, Tilia, Betula), although some co- migration inside the coelom, and leaving the antennae
niferous trees were present in the drier, sandier soils between the epidermis and the outer body wall. For the
(Pinus, Thuja, Juniperus, Tsuga, Abies). In addition, surgery we anesthetized snakes using isoflurane gas
the ongoing abandonment of marginal farmland has delivered via a precision vaporizer (Blouin-Demers et
resulted in old fields and scrub habitats (Zanthoxylum, al. 2000c). While the snakes were waking up, we in-
Juniperus) being common. Some of the more produc- jected them with sterile fluids (0.9% lactated Ringer’s
tive fields are still hayed. solution at a dosage of 50 mL/kg) into the coelom to
help healing and to avoid damage caused by the anti-
Study animals
biotic if the snake was dehydrated. Then we gave the
To obtain experimental animals, we captured snakes snakes a subcutaneous injection of gentamicin sulfate
at 13 communal hibernacula in the spring and oppor- (reptile dose 5 2.5 mg/kg). A single subcutaneous in-
tunistically throughout the rest of the active season. jection maintains therapeutic blood levels of antibiotic
October 2001 HABITAT USE BY RAT SNAKES 2885

TABLE 1. Structural variables used in the analysis of habitat use by Ontario black rat snakes
in 1996 only and from 1996 to 1999 with associated abbreviations and sampling radii.

Radius
Variable (m) Variable
Variables sampled in 1996–99
DROCK 30 Distance (m) to nearest rock ($20 cm length)
DLOG 30 Distance (m) to nearest log ($7.5 cm diameter)
DOVER 30 Distance (m) to nearest overstory tree ($7.5 cm dbh)
DUNDER 30 Distance (m) to nearest understory tree (,7.5 cm dbh, .2 m height)
DEDGE 100 Distance (m) to nearest edge
7.5–15 10 Number of trees $7.5 and ,5 cm dbh in plot
15–30 10 Number of trees $15 and ,30 cm dbh in plot
30–45 10 Number of trees $30 and ,45 cm dbh in plot
.45 10 Number of trees $45 cm dbh in plot
%ROCK 2 Coverage (%) of rocks within plot
%LEAF 2 Coverage (%) of leaf litter within plot
%LOG 2 Coverage (%) of logs within plot
%GRASS 2 Coverage (%) of grass within plot
%SHRUB 2 Coverage (%) of shrubs within plot
CANCLO 45† Canopy closure (%) within cone
Variables sampled only in 1996
DSNAG 30 Distance (m) to nearest snag
DBHSNAG 30 Dbh (cm) of nearest snag ($30 cm dbh)
DECSNAG 30 Decay state (scale from 1 to 7) of nearest snag
DBHOVER 30 Dbh (cm) of nearest overstory tree
LROCK 30 Length (cm) of nearest rock ($20 cm)
LLOG 30 Length (m) of nearest log ($7.5 cm diameter)
DLOG 30 Mean diameter (cm) of nearest log ($7.5 cm)
NUNDER 5 Number of understory trees (,7.5 cm dbh, .2 m height)
%SOIL 2 Coverage (%) of bare soil within plot
%HERBS 2 Coverage (%) of herbs (non-woody) within plot
HGRDVEG 2 Height (m) of ground vegetation (shrubs and herbs)
NWOODY 2 Number of woody stems
HCAN 2 Height (m) of canopy
Note: Dbh 5 diameter at breast hieght.
† Measured in degrees.

for 72 h. Because we gave a second injection before disturbed by our approach, and to make our study com-
release (72 h postsurgery), gentamicin sulfate gave a parable to other studies of snake habitat use (Reinert
total of 144 h of antibiotic coverage. The snakes were 1984, 1992). This accounted for only 285 of 3847
kept in captivity, provided warmth, and monitored for (;7%) locations. We also excluded instances where
3 d following surgery. Water was not provided because snakes were found in buildings because these sites
if the snakes soaked themselves, premature melting of could not be characterized adequately with our habitat
the stitches could occur. sampling scheme (see below). Snakes were in buildings
After releasing each snake at its capture location, we in 270 of 3847 (;7%) relocations. When snakes were
located the snakes on average every 48 h from their perched $3 m in trees (1126 of 3847 or 29% of re-
emergence in late April until they re-entered their hi- locations), we did not measure the habitat variables at
bernacula in early October. Upon locating a snake, we ground level (Table 1) because ground characteristics
recorded its location, position, and behavior (con- are unlikely to affect an individual in this situation, but
cealed, resting/basking, or traveling). All locations of instead substituted individual means for those variables
snakes were flagged and later mapped using a global (Stevens 1996) to be able to use these locations in our
positioning system (GPS) unit with submeter accuracy analyses (see below). For each snake, we only quan-
in the field (MC-V Asset Surveyor Version 3.16 in a tified the habitat at every second relocation to keep the
GPS Pathfinder with a Probeacon, Trimble Navigation habitat sampling manageable. We quantified the habitat
Limited, Sunnyvale, California). The Universal Trans- at a given position on average 7 d after the individual
verse Mercator coordinates (North American Datum of had moved to another location to prevent disturbance
1983) of each location were used to calculate the dis- and at the same time minimize phenological changes
tance moved between relocations. between occupancy and sampling. Locations at which
a snake was observed more than once were only in-
Habitat characterization cluded once in the analysis of habitat use. To quantify
We did not conduct habitat characterization at lo- the available habitat, which is necessary to determine
cations where snakes were found actively traveling to whether the snakes were using habitat nonrandomly,
avoid including instances where snakes may have been we repeated the same habitat analysis and character-
2886 G. BLOUIN-DEMERS AND P. J. WEATHERHEAD Ecology, Vol. 82, No. 10

ization at sites selected by walking a randomly deter- ther sampling. From 1997 to 1999, we only measured
mined distance (10 to 200 paces, determined by a die the 15 structural variables retained after the preliminary
with 20 faces and multiplying by 10) in a randomly analysis (Table 1).
selected direction (18 to 3608, determined by spinning At each snake or random location we also recorded
the bearing dial disc on a compass) from snake loca- the general habitat type as forest, field, marsh, natural
tions. While we concede that our random points are edge, or artificial edge. We defined an artificial edge
not a true random sample of the available habitat be- as having been created by humans (e.g., edge between
cause they are in the vicinity of snake locations, our field and forest), whereas natural edges were not the
random points were representative of the habitats result of human activity (e.g., edge between marsh and
among which the snakes could choose (Keller and Hes- forest). We considered a snake to be in an edge when
ke 2000) because snakes regularly moved .200 paces it was within 15 m of the boundary between any open
between relocations. To keep the sample size manage- habitat (marsh, open water, rock outcrop, field, road)
able and approximately the same for random locations and forest. We identified shedding sites of our study
and each snake reproductive group, we only determined animals by monitoring their skin condition and, in
random habitat points for every fourth snake location. many instances, by finding their shed skin. The skin of
To characterize the habitat at snake and random lo- black rat snakes becomes dull and bluish (this is most
cations, we measured 28 structural variables within cir- noticeable when the eyes become clouded) in the days
cular plots of different radii, depending on the variable before shedding, whereas the skin appears dark and
of interest, all centered at the snake or random location very shiny immediately following shedding. Identify-
(Table 1). We measured distance variables to the nearest ing shedding sites was made easier by the fact that
centimeter with a 50-m measuring tape. To evaluate black rat snakes spend several days inactive at their
percent ground cover and canopy closure, we used a shedding site prior to shedding.
sighting tube (a 50 3 2.5 cm piece of piping) with a
cross wire at one end. This was a modified version of Small mammal trapping
Winkworth and Goodall’s (1962) apparatus. We aimed To quantify mammalian prey densities in different
the tube randomly 50 times in a 2 m radius plot and habitats, we captured small mammals live from June
recorded the type of ground cover ‘‘hit’’ in the cross to August of 1997 to 1999 using 45 mouse live traps
wire. We then multiplied the number of ‘‘hits’’ for each (model 101, Tomahawk Live Trap Company, Toma-
cover type by two and recorded this as the percent hawk, Wisconsin, USA) baited with peanut butter and
ground cover. The same procedure was used 20 times provided with cotton for use as bedding. We trapped
at an angle .458 from horizontal to determine the per- small mammals in edges between forests and fields
cent canopy closure. We defined an edge as the bound- (artificial edges) and in edges between forests and rock
ary between an ‘‘open’’ or two-dimensional habitat outcrops (natural edges) because these are the two most
(e.g., hay fields, rock outcrops, marshes) and a common types of edges in the study area. We placed
‘‘closed’’ or three-dimensional habitat (e.g., deciduous the traps 10 m apart in lines of 15 traps (thus each trap
forest, coniferous forest). These are the types of edges line covered 150 m). We placed one trap line in the
that are associated with a higher density of breeding forest 30 m from the edge, one trap line at the edge,
birds (Gates and Gysel 1978), and that also provide a and one trap line in the open habitat 30 m from the
forest snake with thermoregulatory opportunities be- edge. We left the traps in a given location for a period
cause they offer simultaneous access to shady and sun- of 2 wk and then moved them to another location. The
ny microhabitats. An additional benefit of this defini- traps were checked every morning and because we
tion was that all edges (or habitat boundaries) were needed food items for another experiment, we used
obvious features of the landscape and it made mea- removal trapping to derive an index of prey density.
suring distances from edges an objective process. Species that are prey of rat snakes (e.g., Peromyscus,
After the 1996 field season, we ran a preliminary Blarina, Tamias, Microtus) were euthanized in the field
multivariate analysis of variance (MANOVA) and as- using cervical dislocation following capture. Live-trap-
sociated discriminant function analysis (DFA) to iden- ping followed by euthanasia (as opposed to lethal cap-
tify which variables were contributing to the multi- ture using snap traps) allowed us to release nonprey
variate group differences between random sites and species (e.g., Rana, Bufo) inadvertently captured. To
sites used by snakes (see below). From 1997 to 1999, index the abundance of small mammals, we calculated
we discontinued sampling the 13 discriminant variables the average number of captures per number of trap-
that had pooled within-group correlations with the dis- days in each habitat for each trapping day (where one
criminant functions #0.10 (Table 1). This was a con- trap-day is equal to one working trap set for one day).
servative approach because variables with loadings
#0.40 are usually deemed unimportant to discrimina- Analyses
tion in DFA (Stevens 1996, Clark and Shutler 1999). We divided the snakes into three groups based on
We used a conservative approach to ensure that all their reproductive status: males, nongravid females,
potentially meaningful variables were retained for fur- and gravid females. The reproductive state of females
October 2001 HABITAT USE BY RAT SNAKES 2887

was assessed in June by externally palpating the ovi- was consistent with the hypothesis that rat snakes pre-
ducts for the presence of eggs and confirmed in July fer edges for foraging, we examined if the preference
by nesting activity. Because female black rat snakes for edges was constant over the duration of the active
do not reproduce every year, some individuals changed season. We divided the snake locations into edge and
groups from one year to the next. Because female black non-edge habitat and determined the proportion of
rat snakes very seldom moved while gravid, fewer lo- snake locations that were in edges for each month of
cations were obtained for this group. We used MAN- the main active season (May to August). Observations
OVA to determine if there was a significant difference from earlier and later in the season were not included
in habitat centroids of each group and DFA to examine in this analysis because there were too few to analyze.
along which axes the groups differed and which vari- To contrast the distances moved between relocations
ables contributed most to separation among groups. To for the three snake groups during the period prior to
facilitate the interpretation of discriminant functions, egg laying (1 June to 15 July), we divided the distances
we limited our interpretation to the five variables with moved between relocations into four classes (0 m, 1 to
the highest correlations (in absolute values) with each 10 m, 11 to 100 m, and 101 to 1000 m). We also divided
function. the behavior of all snakes located prior to the time
An assumption common to all analysis of variance females started egg laying as either traveling, resting/
(ANOVA) models is that observations are sampled at basking, or concealed. When comparing the distances
random. While truly random samples of organisms are moved and behavior of all snakes in edges and forests,
extremely hard to obtain in nature, this assumption is we used the same categories but included relocations
particularly dubious when radiotelemetry is used be- from the whole active season.
cause many observations are derived from relatively The analyses were conducted on JMP Version 3.2
few individuals. In such a case, an aberrant individual (SAS Institute 1997) and SPSS Version 6.1 (SPSS
sampled repeatedly can severely bias the conclusions 1995) on a Macintosh desktop computer. We inspected
one would reach when treating each observation as an box plots to determine if the assumptions of normality
independent sample. In our study, the individual snake and homogeneity of variance were upheld. The con-
sampled the most only accounted for 7.2% of the total tinuous habitat structure variables were log-trans-
number of snake locations, so no individual had the formed to improve their adherence to the assumption
opportunity to unduly bias the group means. Another of normality. Significance of statistical tests was ac-
potential solution to this problem would have been to cepted at a 5 0.05, but marginally nonsignificant re-
use mean habitat vectors for each individual as the basis sults are reported when deemed important. All means
for analyses. However, using this approach usually does are reported 6 one standard error.
not change the conclusions of the analyses (Reinert
1984) while not making use of all the information avail- RESULTS
able, such as the variation found within individuals. From 1996 to 1999, we sampled habitat character-
One assumption specific to MANOVA is the ho- istics at 165 random locations, 195 locations of males,
mogeneity of covariance matrices, usually tested using 190 locations of nongravid females, and 81 locations
Box’s test. This assumption is rarely met with habitat of gravid females. Snakes often remained in the same
use data because it would require that each segment of location for many days, and often returned to previ-
the population responds similarly to the different hab- ously used locations. In addition, we only sampled ev-
itat variables (Reinert 1984). If this assumption is vi- ery second relocation for each individual. Thus, the
olated and the sample sizes for each group differ sub- number of locations we sampled (466) was much less
stantially, biased tests of significance can result (Ste- than the total number of times we located snakes
vens 1996). Because female black rat snakes are not (3847). Snakes clearly used their habitat nonrandomly
gravid each year and because they are very stationary (x2 5 61.2, df 5 15, P , 0.001), preferring edge hab-
while gravid, the number of locations for this group itats and avoiding water bodies (Table 2).
was approximately half that of the other groups. To Before we used MANOVA, we tested the assumption
ensure that this difference in sample sizes did not un- of homogeneity of covariance matrices. As expected
duly bias our tests of significance, we also conducted with biological data, the Box’s test indicated that the
our analyses using a computer-generated, randomly se- covariance matrices were heterogeneous (Box’s M 5
lected subset of our data designed to achieve equal 798.36, F 5 2.11, df 5 360, 351883.2, P , 0.001).
sample sizes in each group. Because the results for all However, many authors have defended the heuristic
analyses remained qualitatively unchanged (all signif- value of multivariate methods despite the common vi-
icant relationships remained significant, no new sig- olation of this assumption with ecological data (Pi-
nificant relationship appeared, and the variables con- mentel 1979, Stevens 1996). The mean scores on each
tributing the most to group differences were the same), variable for the random locations and the locations of
we only present the results of the analyses using the the three snake groups are presented in Table 3. We
complete data set. included all shedding sites (except sites located in
To determine if the use of edges by black rat snakes buildings) in the calculations because these sites also
2888 G. BLOUIN-DEMERS AND P. J. WEATHERHEAD Ecology, Vol. 82, No. 10

TABLE 2. Habitat types where the random points were located and where male, nongravid
female, and gravid female black rat snakes followed by radiotelemetry in Ontario from 1996
to 1999 were located.

Nongravid
Habitat type Random Males females Gravid females
Artificial edge 19 (11.5) 31 (15.9) 30 (15.8) 22 (27.2)
Natural edge 51 (30.9) 86 (44.1) 83 (43.7) 32 (39.5)
Field 5 (3.0) 2 (1.0) 2 (1.0) 2 (2.4)
Forest 69 (41.8) 74 (37.9) 69 (36.3) 24 (29.6)
Wetland 7 (4.2) 2 (1.0) 6 (3.2) 1 (1.2)
Water body 14 (8.5) 0 (0.0) 0 (0.0) 0 (0.0)
Total 165 195 190 81
Note: The percentages of the total for each column are given in parentheses.

were used by snakes when they were not shedding. The in-group correlations of habitat variables indicated that
overall MANOVA indicated that the habitat character- DOVER, DUNDER, DLOG, and DEDGE contributed
istics of the three groups of snakes and the randomly strongly and positively, whereas %LOG contributed
sampled points were significantly different (Wilk’s L strongly but negatively, to the first function (Table 5).
5 0.630, F 5 6.80, df 5 45, 1821.9, P , 0.001). This function can be interpreted as a gradient from sites
Distances between group centroids in the discriminant far from trees, logs, and edges with low ground cover
space showed that males, nongravid females, and grav- of logs to sites close to trees, edges, and logs with high
id females all used habitat nonrandomly (Table 4). ground cover of logs (Fig. 1). For the second function,
Among the three classes of snakes, males and non- .45 and DROCK contributed strongly and positively,
gravid females did not differ significantly from one whereas 15–30, %ROCK, and 7.5–15 contributed
another, whereas gravid females were significantly dif- strongly but negatively (Table 5). The second function
ferent from both other groups (Table 4). In addition, indicated a gradient from sites close to rocks with many
gravid females used habitat that was the least available small and medium trees, few large trees, and high
in the study area (largest distance in the discriminant ground cover of rocks toward sites far from rocks with
space from the random group), followed by males, and few small and medium trees, many large trees, and low
then by nongravid females. ground cover of rocks (Fig. 1).
The DFA derived three discriminant functions that Separation of the random group from the three snake
summarized multivariate differences among our four groups along the first discriminant function (Fig. 1)
types of locations. Only the first discriminant function reflects that the available habitat was mostly far from
(x2 5 283.3, df 5 45, P , 0.001) and the second trees and edges with low ground cover of logs, whereas
discriminant function (x2 5 59.0, df 5 28, P , 0.001) snakes tended to be found close to trees and edges with
accounted for a significant amount of the total varia- extensive ground cover of logs. Gravid females had the
tion, with the first discriminant function explaining strongest preference for sites close to edges and trees
81.7% of the total variation (Table 5). The pooled with- with high ground cover of logs as indicated by their

TABLE 3. Means of variables used in the analysis of habitat use for the random locations and
locations of male, nongravid female, and gravid female black rat snakes followed by radio-
telemetry in Ontario from 1996 to 1999.

Random Males Nongravid females Gravid females

Variable (n 5 165) (n 5 195) (n 5 190) (n 5 81)
DROCK 5.43 (0.64) 2.45 (0.29) 2.85 (0.28) 2.60 (0.35)
DLOG 5.64 (0.58) 3.22 (0.31) 3.04 (0.35) 2.23 (0.29)
DOVER 4.84 (0.54) 2.26 (0.19) 2.70 (0.24) 1.56 (0.26)
DUNDER 3.93 (0.55) 1.83 (0.24) 1.84 (0.21) 1.23 (0.15)
DEDGE 25.84 (4.37) 17.42 (3.41) 17.46 (3.12) 18.68 (3.39)
7.5–15 12.47 (0.82) 10.81 (0.64) 10.24 (0.66) 7.91 (0.87)
15–30 5.61 (0.42) 4.71 (0.27) 4.62 (0.31) 3.22 (0.37)
30–45 1.35 (0.12) 1.42 (0.11) 1.36 (0.12) 1.49 (0.17)
.45 0.36 (0.07) 0.32 (0.05) 0.40 (0.05) 0.78 (0.10)
%ROCK 8.33 (1.14) 16.63 (1.51) 17.81 (1.44) 11.99 (1.62)
%LEAF 27.67 (1.85) 20.61 (1.43) 23.21 (1.57) 24.22 (2.30)
%LOG 7.02 (0.97) 11.32 (0.81) 11.94 (0.89) 14.31 (1.61)
%GRASS 22.70 (1.90) 16.64 (1.29) 16.09 (1.36) 15.01 (1.98)
%SHRUB 16.62 (1.45) 20.34 (1.46) 22.26 (1.63) 26.25 (2.26)
CANCLO 55.59 (2.96) 53.53 (2.22) 49.53 (2.40) 56.67 (3.12)
Note: The standard errors are given in parentheses. For definitions of variables, see Table 1.
October 2001 HABITAT USE BY RAT SNAKES 2889

TABLE 4. Distances between the four group centroids in the discriminant space and their
statistical significance for the analysis of habitat use in Ontario black rat snakes followed
by radiotelemetry from 1996 to 1999.

Group
Random Males Non-gravid females
Group Distance (F) P Distance (F) P Distance (F) P
Males 1.40 (11.47) ,0.001
Nongravid 1.25 (8.97) ,0.001 0.44 (1.22) 0.25
females
Gravid females 2.06 (15.08) ,0.001 1.10 (4.54) ,0.001 1.09 (4.39) ,0.001
Note: The F statistic with 15 and 613 degrees of freedom is given in parentheses.

position on the first discriminant axis. The second dis- 15, P 5 0.73; Wilk’s L 5 0.445, F 5 0.91, df 5 11,
criminant function provided further separation between 15, P 5 0.57). The single female was followed in 1997
gravid females and the other two snake groups. Gravid (nongravid) and in 1998 (gravid) and there was a sig-
females were associated more with larger trees, less nificant difference in her habitat use between years
with small and medium trees, and less with rocks than (Wilk’s L 5 0.132, F 5 5.25, df 5 12, 15, P 5 0.003),
males and nongravid females. with a stronger preference for sites close to large trees
The habitat use of males and nongravid females did and edges when gravid than when nongravid.
not differ significantly, suggesting that the differences When we examined if habitat use by black rat snakes
we observed in habitat use among the three groups are varied seasonally, we found that although snakes were
not a function of the sex of the individual per se, but found in edges more often than in all other habitat types
rather, a consequence of the reproductive state of fe-
combined throughout the active season, their use of
males. To examine more formally if the difference in
edges increased significantly over the active season ( N
habitat use we observed between gravid females and
5 447, x2 5 8.90, df 5 3, P 5 0.03, Table 6). We also
the other two snake groups was due to the reproductive
condition of the females, we contrasted the habitat use compared the mass of 11 gravid females at emergence
of snakes followed in multiple years. We had sufficient from hibernation in early May to their mass following
data ($10 characterized locations in each year) to test parturition in July. All gravid females lost mass fol-
for multivariate differences in habitat use among years lowing emergence (mean loss 5 108.8 g or 21.3% of
for three males and one female. The three males were postpartum mass, paired t 5 8.56, df 5 11, P , 0.001).
followed in 1996 and in 1997 and, in each case, there During the same period, 17 nongravid females gained
were no significant differences between their habitat mass (mean gain 5 63.8 g, paired t 5 3.23, df 5 16,
use in each year (Wilk’s L 5 0.011, F 5 11.74, df 5 P 5 0.005) and 16 males maintained the same mass
2, 15, P 5 0.08; Wilk’s L 5 0.278, F 5 0.69, df 5 4, (mean gain 5 9.4 g, paired t 5 0.75, df 5 15, P 5

TABLE 5. Summary statistics for the three discriminant functions and their pooled within-
groups correlations (r) with the discriminating variables used in the analysis of habitat use
by Ontario black rat snakes followed by radiotelemetry from 1996 to 1999.

Statistic Function 1 Function 2 Function 3

Eigenvalue 0.4423 0.0754 0.0227
x2 test x2 5 283.3 x2 5 59.0 x2 5 13.9
(df 5 45, P , 0.001) (df 5 28, P , 0.001) (df 5 13, P 5 0.38)
Percentage of 81.7 13.9 4.2
variance explained
r, DROCK 0.28 0.37 20.01
r, DLOG 0.33 0.04 0.53
r, DOVER 0.47 0.02 20.15
r, DUNDER 0.38 0.03 20.01
r DEDGE 0.30 20.26 20.08
r, 7.5–15 0.15 20.28 0.07
r, 15–30 0.13 20.35 0.06
r, 30–45 20.04 0.07 0.13
r, . 45 20.18 0.59 20.07
r, %ROCK 20.21 20.29 20.08
r, %LEAF 0.04 0.19 20.13
r, %LOG 20.34 20.15 20.14
r, %GRASS 0.10 0.05 0.32
r, %SHRUB 20.30 0.24 20.19
r, CANCLO 20.15 0.11 0.29
2890 G. BLOUIN-DEMERS AND P. J. WEATHERHEAD Ecology, Vol. 82, No. 10

FIG. 1. Positions and standard errors of the group centroids of random locations and locations of gravid female, nongravid
female, and male black rat snakes on the two significant discriminant axes with pictorial interpretation of associated habitat
gradients in the analysis of habitat use by radio-implanted black rat snakes in Ontario from 1996 to 1999.

0.46). This suggests that females reduced their feeding active season black rat snakes were likely to move
when they were gravid. shorter distances between relocations when in edges
Groups also differed significantly in the distance than when in the forest (N 5 3375, x2 5 46.54, df 5
moved between successive relocations during the time 3, P , 0.001, Fig. 3). The only movement category
that females were carrying their eggs (N 5 1271, x2 5 that was more prevalent in forests than in edges was
72.38, df 5 6, P , 0.001). A posteriori tests of in- 101–1000 m (Fig. 3).
dependence revealed that this relationship was a con- We monitored 55 shedding events from 37 of our
sequence of gravid females moving significantly less radio-tracked snakes. The 37 individuals used 36 dif-
between relocations than nongravid females or males, ferent shedding sites (10 individuals used more than
with the latter two groups not differing significantly. one site and 15 sites were used by two to four radio-
Gravid females had not moved for 64% of the relo- tracked snakes). Shedding sites were regularly shared
cations and had moved 10 m or less in 75% of the by gravid females, nongravid females, and males. In
relocations (Fig. 2). By contrast, nongravid females had addition, we found the shed skins of unmarked black
moved 10 m or less in 53% of the relocations and males rat snakes at 17 of the 36 shedding sites, so the sites
in 47% of the relocations (Fig. 2). Also, during the used by the snakes we tracked were also used by other

TABLE 6. Number of black rat snake locations followed by radiotelemetry from 1996 to 1999
in edge habitats and other habitat types for the four months of the primary active season in
Ontario.

Month
Habitat May June July August Total
Edge 59 (51.8) 84 (59.6) 68 (66.0) 63 (70.8) 274 (61.3)
Other 55 (48.2) 57 (40.4) 35 (34.0) 26 (29.2) 173 (38.7)
Total 114 141 103 89 447
Note: The percentages of the total for each column are given in parentheses.
October 2001 HABITAT USE BY RAT SNAKES 2891

FIG. 2. (A) Distance categories moved between successive relocations and (B) behavior exhibited at relocation for radio-
implanted gravid female, nongravid female, and male black rat snakes in Ontario from 1996 to 1999. Only relocations from
the time period when females carry their eggs were included; sample sizes are shown above each bar.

snakes in the population. We divided the shedding sites cations of gravid female were in trees (DOVER or
into seven habitat categories and determined how many DUNDER 5 0) whereas nongravid females and males
sites were situated in edges (Table 7). Based on the were in trees 31% and 33% of the time, respectively.
availability of artificial and natural edges in the study However, rocks often provide retreats for black rat
area, shedding sites were significantly more likely to snakes (snakes were under rocks in 15.7% of the re-
be in edges than in other habitat types (N 5 201, x2 5 locations) and rocks were more abundant at sites where
12.64, df 5 2, P 5 0.002). males and nongravid females were located (mean ø
Greater preference for edge by gravid females could 17% rock cover, Table 3) than at sites where gravid
also be a function of edges providing safer sites. DFA females were located (mean ø 12% rock cover, Table
indicated that gravid females preferred sites with higher 3). The three groups of snakes also differed in their
ground cover of logs and were found near trees more frequency of traveling, resting/basking, or being con-
often than the other snake groups. In fact, 59% of lo- cealed during the time that females were carrying their
2892 G. BLOUIN-DEMERS AND P. J. WEATHERHEAD Ecology, Vol. 82, No. 10

FIG. 3. (A) Distance categories moved between successive relocations and (B) behavior exhibited at relocation in edges
and forest for radio-implanted black rat snakes in Ontario from 1996 to 1999. Relocations from the whole active season were
included; sample sizes are shown above each bar.

eggs (N 5 1313, x2 5 23.46, df 5 4, P , 0.001). L 5 0.716, F 5 1.92, df 5 13, 63, P 5 0.044). DFA
However, a posteriori tests of independence revealed indicated that the only difference was higher ground
that this relationship was a result of males being con- cover of rocks in natural edges (these sites were often
cealed less than gravid females or nongravid females, at the edge of rock outcrops) and more ground cover
while the latter two groups did not differ significantly of grass in artificial edges (these sites were often at the
(Fig. 2). During the active season, black rat snakes were edge of hayfields). Although natural edges were ;2.5
concealed more often and visible (basking/resting or times more abundant than artificial edges in the study
traveling) less often when relocated in edges than in area, snakes were no more likely to be found in natural
the forest (N 5 3297, x2 5 24.26, df 5 2, P , 0.001, edges than in artificial edges, based on the availability
Fig. 3). of both types of edges (N 5 354, x2 5 0.12, df 5 1,
MANOVA of random habitat samples indicated that P 5 0.73, Table 2). This was also true when we con-
natural and artificial edges differed structurally (Wilk’s sidered gravid females alone (N 5 123, x2 5 0.67, df
October 2001 HABITAT USE BY RAT SNAKES 2893

TABLE 7. Numbers of each type of shedding site, numbers of radio-implanted Ontario black
rat snakes that used each type of site from 1996 to 1999, and numbers of shedding sites that
were in artificial and natural edge habitats.

Artificial edge Natural edge

Shedding sites Sites (N) Individuals (N) (N) (N)
Old barns 3 (8.3) 5 (9.8) 0 (0.0) 0 (0.0)
Old mining machinery 1 (2.8) 2 (3.9) 0 (0.0) 1 (5.3)
Cracks in house foundations 4 (11.1) 5 (9.8) 0 (0.0) 0 (0.0)
Old hay piles 3 (8.3) 5 (9.8) 3 (37.5) 0 (0.0)
Large hollow logs 3 (8.3) 4 (7.8) 0 (0.0) 3 (15.8)
Crevices in rock outcrops 9 (25.0) 8 (15.7) 3 (37.5) 6 (31.6)
Standing hollow snags 13 (36.1) 22 (43.1) 2 (25.0) 9 (47.4)
Total 36 51 8 19
Note: The percentages of the total for each column are given in parentheses.

5 1, P 5 0.41). Shedding sites also were not more Gates 1993). Furthermore, our results indicate that the
likely to be in artificial edges than in natural edges preference for edges is equally strong in human-dis-
given the availability of the two edge types (N 5 97, turbed landscapes and in more pristine landscapes
x2 5 0.06, df 5 1, P 5 0.81). where the edges occur naturally. Finally, as we sum-
We trapped small mammals on 215 d, for a total of marize below, of the hypotheses we tested, the best
5039 trap-days, and captured 187 individuals of species explanation for why black rat snakes prefer edges is
on which rat snakes prey (126 Peromyscus, 28 Blarina, that edges facilitate behavioral thermoregulation.
22 Tamias, and 11 Microtus). Before pooling the data Weatherhead and Charland (1985) proposed that
for natural and artificial edges, we ensured that our black rat snakes might prefer edges because breeding
index of small mammal abundance did not differ be- birds make edges more profitable habitats in which to
tween edge types (t 5 0.684, df 5 86, P 5 0.49). A forage. They based that hypothesis on their observation
two-way ANOVA on the pooled data indicated that that the preference for edges by black rat snakes in
there was a significant month by habitat interaction (F their study was most pronounced early in the season.
5 3.19, df 5 2, 206, P 5 0.03). Separate one-way Studying the same population of black rat snakes, we
ANOVAs for each month showed that capture success found exactly the opposite pattern; although edges were
varied significantly between habitats only in August ( F preferred in all months, the use of edge habitat in-
5 8.25, df 5 2, 64, P , 0.001). Tukey-Kramer hsd creased from May to August. The most obvious reason
tests on the August data showed that capture success for this discrepancy is that Weatherhead and Charland
was higher in forests and edges than in the open habitats (1985) only tracked seven snakes for one year resulting
(Fig. 4). If we consider the data for edges only, ANOVA in only 118 locations of snakes. In fact, the decline in
revealed that there were no monthly differences in the the use of edges occurred in only one of the two habitats
index of small mammal abundance (F 5 2.69, df 5 2, analyzed, and this result was based on only 13 loca-
85, P 5 0.07). tions. With much more extensive data, we found that
black rat snakes used edges more later in the season,
DISCUSSION after the period of avian reproduction is over. This is
Our study confirms that black rat snakes prefer edge not consistent with the hypothesis that snakes are at-
habitat (Weatherhead and Charland 1985, Durner and tracted to edges principally to forage for avian eggs

FIG. 4. Mean index of small mammal cap-

ture success in open habitats, edges, and forest
for June, July, and August in Ontario from 1997
to 1999. Error bars indicate one standard error.
Within a month, means with the same letters are
deemed not significantly different based on Tu-
key-Kramer hsd tests; when there are no letters,
there were no significant differences for the
month. A ‘‘trap-day’’ is defined as one working
trap set for one day.
2894 G. BLOUIN-DEMERS AND P. J. WEATHERHEAD Ecology, Vol. 82, No. 10

and nestlings. Also, mammalian prey densities did not itat by gravid females is consistent with their need to
vary seasonally and were not higher in edges than in maintain higher body temperatures. Among viviparous
the forest. Therefore, if black rat snakes were selecting snakes, effective thermoregulation is important for
habitat principally for foraging on small mammals, then gravid females because the phenotype and fitness of
the snakes should have been found in forests as often their offspring are dependent upon the body tempera-
as in edges throughout the active season, a pattern tures they experience during development (Alberts et
clearly different from what we observed. This evidence al. 1997, Shine et al. 1997, Blouin-Demers et al.
suggests that edges are not selected primarily for for- 2000a). Black rat snakes are oviparous, so the impor-
aging, although this does not mean that edges are not tance of maintaining body temperature for them must
good foraging habitats or that the snakes do not forage be associated with general metabolic function being
in edges. more efficient at optimal body temperatures. Mobiliz-
Also inconsistent with the foraging hypothesis was ing the nutrients for eggs seems likely to be facilitated
the loss of mass we documented for gravid females. by the female maintaining her body temperature. At
Gravid females used edges more extensively than males the northern extreme of the species’ distribution where
and nongravid females. If gravid females used edges we studied black rat snakes, the short active season
to forage, they should have at least maintained their must place a premium on developing and laying the
mass from emergence to postegg laying. However, the eggs quickly so they have adequate time to incubate.
mean loss of .21% of body mass suggests that females The trade-off between behavioral thermoregulation
forage very little when gravid. This result, combined and other activities may favor careful thermoregulation
with the general inactivity we documented for gravid more in gravid females, but nongravid females and
females, suggests that gravid females were not using males should also benefit from some thermoregulation.
edges to forage. It does not necessarily follow that Ecdysis is something that all snakes do, and snakes
males and nongravid females used habitat on the same increase their body temperatures during ecdysis (Gib-
basis as gravid females. In fact, Elaphe obsoleta ac- son et al. 1989, Peterson et al. 1993). Thus, the fact
tively search for prey (Jackson 1970, 1978) and, al- that shedding sites were located disproportionately in
though they were concealed more often in edges than edge habitat is also consistent with edges being used
in forest, short localized movements most likely to be for increasing body temperatures. Although our obser-
associated with foraging (11–100 m) were more fre- vations are consistent with the thermoregulation hy-
quent in edges than in the forest. However, our data pothesis, all our tests are indirect. To confirm that pat-
for gravid females, combined with the seasonal patterns terns of habitat use by black rat snakes and the diver-
of habitat use by all snakes, point to some reason other gence in habitat use between reproductive females and
than foraging explaining the general preference for the other groups is linked to thermoregulation, we will
edges. need to determine if snakes thermoregulate differently
We also rejected the hypothesis that edges were pre- in edges than in other habitat types. Our expectation
ferred by gravid females because they provided supe- is that edges should be ideal habitats to maintain op-
rior protection from predators. The behavioral analysis timal body temperatures because they allow simulta-
suggested that gravid females were not actively avoid- neous access to shaded microhabitats to decrease body
ing predators more than nongravid females and, al- temperatures and sunny microhabitats to increase body
though we did not quantify retreat sites per se, our temperatures.
impression was that features such as snags and rock Our results have several implications for conserva-
crevices were abundant throughout our study area, and tion. Weatherhead and Charland (1985) had hypothe-
did not occur primarily in edge habitat. We did find sized that the ideal habitat for black rat snakes is a
that some natural edges appeared to offer better retreat mosaic of field and forest where the scale of fragmen-
sites than artificial edges for snakes because of their tation is sufficiently small that each individual snake
rocky ground cover. Because the snakes did not dif- can include both forest and edges within its home
ferentiate between natural and artificial edges, however, range. The results we presented here support this hy-
we conclude that it was a general structural feature of pothesis and extend its implications to more pristine
edges that attracted them and that edges were not used landscapes. The ideal pristine habitat for black rat
primarily to avoid predation. snakes would also be a mosaic of forest and open hab-
Weatherhead and Charland (1985) also proposed that itats, but the open habitats would be natural (e.g.,
black rat snakes might prefer edges because increased marshes, rock outcrops) instead of man-made (e.g.,
solar radiation allows snakes to bask to increase body fields). In both agricultural and pristine habitat, the
temperatures. Presumably, edges also facilitate ther- scale of forest fragmentation is important. An area with
moregulation because they provide simultaneous ac- several large fields and several large forest patches
cess to open sunny habitats to increase body temper- would provide much less edge for the snakes than the
atures and shaded forests to decrease body tempera- same area divided into many smaller patches of field
tures. Our observations are generally consistent with and forest. As such, we would expect the former sit-
this hypothesis. First, the clear preference for edge hab- uation to support fewer black rat snakes than the latter
October 2001 HABITAT USE BY RAT SNAKES 2895

situation. Of course from the snakes’ perspective, either Andrén, H., and P. Angelstam. 1988. Elevated predation rates
situation is superior to even larger scale deforestation, as an edge effect in habitat islands: experimental evidence.
Ecology 69:544–547.
a phenomenon hypothesized to be the cause of the al- Angelstam, P. 1986. Predation on ground-nesting birds’ nests
most complete disappearance of black rat snakes from in relation to predator densities and habitat edge. Oikos 47:
southwestern Ontario (Prior and Weatherhead 1998). 365–373.
Our results indicated that black rat snakes do not Blouin-Demers, G., K. J. Kissner, and P. J. Weatherhead.
discriminate between artificial and natural edges. Thus, 2000a. Plasticity in preferred body temperature of young
snakes in response to temperature during development.
if conservation of black rat snakes is the goal, it should Copeia 2000:841–845.
be possible simply to create more edge. However, in- Blouin-Demers, G., K. A. Prior, and P. J. Weatherhead. 2000b.
creasing edge habitat through forest fragmentation is Patterns of variation in spring emergence by black rat
detrimental to many species of forest birds because snakes (Elaphe obsoleta obsoleta). Herpetologica 56:175–
rates of nest predation are higher near some edges 188.
Blouin-Demers, G., P. J. Weatherhead, C. M. Shilton, C. E.
(Gates and Gysel 1978, Andrén and Angelstam 1988). Parent, and G. P. Brown. 2000c. Use of inhalant anesthetics
Where black rat snakes occur, they may be responsible in three snake species. Contemporary Herpetology 2000:4.
for some of the increased nest predation, even though Bolger, D. T., A. C. Alberts, and M. E. Soulé. 1991. Occur-
their foraging in edges seems likely to be coincidental rence patterns of bird species in habitat fragments: sam-
to their preference for that habitat. Thus, conservation pling, extinction, and nested species subsets. American
Naturalist 137:155–166.
efforts intended to protect black rat snakes by preserv- Brittingham, M. C., and S. A. Temple. 1983. Have cowbirds
ing or creating their preferred edge habitats seem likely caused forest songbirds to decline? Bioscience 33:31–35.
to contribute to declines of forest-interior species of Brown, G. P., and P. J. Weatherhead. 2000. Thermal ecology
birds. and sexual size dimorphism in northern water snakes, Ner-
Several research directions are suggested by our re- odia sipedon. Ecological Monographs 70:311–330.
Burkey, T. V. 1993. Edge effects in seed and egg predation
sults. First, the thermoregulation hypothesis should be at two neotropical rainforest sites. Biological Conservation
tested directly by determining how strongly habitat use 66:139–143.
and thermoregulation are linked in black rat snakes. Charland, M. B., and P. T. Gregory. 1990. The influence of
Second, because snakes have been identified as im- female reproductive status on thermoregulation in a vivip-
portant nest predators in fragmented forest habitat arous snake, Crotalus viridis. Copeia 1990:1089–1098.
Charland, M. B., and P. T. Gregory. 1995. Movements and
(Robinson et al. 1995), it would be informative to know
habitat use in gravid and nongravid female garter snakes
whether all snakes respond to edges in the same way (Colubridae: Thamnophis). Journal of Zoology, London
and for the same reasons. This will improve the ability 236:543–561.
to predict the consequences for either future fragmen- Clark, R. G., and D. Shutler. 1999. Avian habitat selection:
tation or attempts to remedy current fragmentation. In- pattern from process in nest-site use by ducks? Ecology
teresting in this regard is evidence that tropical snakes 80:272–287.
Cobb, V. A., and C. R. Peterson. 1991. The effects of preg-
may benefit from forest fragmentation (Henderson and nancy on body temperature variation in free-ranging west-
Winstel 1995). It seems unlikely that edge habitat is ern rattlesnakes. American Zoologist 31:78A.
important for thermoregulation in tropical snakes. If Cooper, W. E., Jr., L. J. Vitt, R. Hedges, and R. B. Huey.
different snake species respond to habitat fragmenta- 1990. Locomotor impairment and defense in gravid lizards
tion for different reasons, the challenge that will arise (Eumeces laticeps): behavioral shift in activity may offset
costs of reproduction in an active forager. Behavioral Ecol-
will be to develop a general theory that accounts for ogy and Sociobiology 27:153–157.
that variation. Donovan, T. M., R. H. Lamberson, A. Kimber, F. R. Thomp-
ACKNOWLEDGMENTS son, III, and J. Faaborg. 1995. Modeling the effects of
habitat fragmentation on source and sink demography of
Many field assistants provided valuable help with the cap- Neotropical migrant birds. Conservation Biology 9:1396–
ture of snakes, radio-tracking, and habitat characterization: 1407.
H. McCracken, A. Moenting, E. O’Grady, J. Svec, C. Ver-
Durner, G. M., and J. E. Gates. 1993. Spatial ecology of
reault, and T. Volk. We are grateful to the Queen’s University
black rat snakes on Remington Farms, Maryland. Journal
Biological Station for the permission to use their facilities
of Wildlife Management 57:812–826.
and logistical support. Funding for this research was provided
by a NSERC research grant to P. J. Weatherhead and NSERC Fitch, H. S. 1963. Natural history of the black rat snake
postgraduate scholarships to G. Blouin-Demers. (Elaphe o. obsoleta) in Kansas. Copeia 1963:649–658.
Fritts, T. H., and G. H. Rodda. 1998. The role of introduced
LITERATURE CITED species in the degradation of island ecosystems: a case
Alberts, A. C., A. M. Perry, J. M. Lemm, and J. A. Phillips. history of Guam. Annual Review of Ecology and System-
1997. Effects of incubation temperature and water potential atics 29:113–140.
on growth and thermoregulatory behavior of hatchling Cu- Gates, J. E., and L. W. Gysel. 1978. Avian nest dispersion
ban rock iguanas (Cyclura nubila). Copeia 1997:766–776. and fledging success in field–forest ecotones. Ecology 59:
Andreassen, H. P., K. Hertzberg, and R. A. Ims. 1998. Space- 871–883.
use responses to habitat fragmentation and connectivity in Gibson, A. R., D. A. Smucny, and J. Kollar. 1989. The effects
the root vole Microtus oeconomus. Ecology 79:1223–1235. of feeding and ecdysis on temperature selection by young
Andrén, H. 1992. Corvid density and nest predation in re- garter snakes in a simple thermal mosaic. Canadian Journal
lation to forest fragmentation: a landscape perspective. of Zoology 67:19–23.
Ecology 73:794–804. Grant, B. W. 1990. Trade-offs in activity time and physio-
2896 G. BLOUIN-DEMERS AND P. J. WEATHERHEAD Ecology, Vol. 82, No. 10

logical performance for thermoregulating desert lizards, Campbell and E. D. Brodie, editors. Biology of the pitvi-
Sceloporus merriami. Ecology 71:2323–2333. pers. Selva, Tyler, Texas, USA.
Harrison, S., and E. Bruna. 1999. Habitat fragmentation and Reinert, H. K. 1993. Habitat selection in snakes. Pages 201–
large-scale conservation: what do we know for sure? Ecog- 240 in R. A. Seigel and J. T. Collins, editors. Snakes: ecol-
raphy 22:225–232. ogy and behavior. McGraw-Hill, New York, New York,
Henderson, R. W., and R. A. Winstel. 1995. Aspects of habitat USA.
selection by an arboreal boa (Corallus enydris) in an area Reinert, H. K., and D. Cundall. 1982. An improved surgical
of mixed agriculture on Grenada. Journal of Herpetology implantation method for radio-tracking snakes. Copeia
29:272–275. 1982:702–705.
Heske, E. J., S. K. Robinson, and J. D. Brawn. In press. Nest Ricklefs, R. E. 1969. An analysis of nesting mortality in
predation and Neotropical migrant songbirds: piecing to- birds. Smithsonian Contributions in Zoology 9:1–48.
gether the fragments. Wildlife Society Bulletin. Robinson, S. K., F. R. Thompson, T. M. Donovan, D. R.
Huey, R. B., C. R. Peterson, S. J. Arnold, and W. P. Porter. Whitehead, and J. Faaborg. 1995. Regional forest frag-
1989. Hot rocks and not-so-hot rocks: retreat-site selection mentation and the nesting success of migratory birds. Sci-
by garter snakes and its thermal consequences. Ecology 70: ence 267:1987–1990.
931–944. SAS Institute. 1997. JMP statistics and graphics guide. Ver-
Jackson, J. A. 1970. Predation of a black rat snake on yellow- sion 3.2 for Macintosh. SAS Institute, Cary, North Caro-
shafted flicker nestlings. Wilson Bulletin 82:329–330. lina, USA.
Jackson, J. A. 1978. Predation by a gray rat snake on red- Schwarzkopf, L., and R. Shine. 1991. Thermal biology of
cockaded woodpeckers nestlings. Bird-Banding 49:187– reproduction in viviparous skinks, Eulamprus tympanum:
188. why do gravid females bask more? Oecologia 88:562–569.
Seigel, R. A., M. M. Huggins, and N. B. Ford. 1987. Re-
Keller, W. L., and E. J. Heske. 2000. Habitat use by three
duction in locomotor ability as a cost of reproduction in
species of snakes at the Middle Fork Fish and Wildlife
gravid snakes. Oecologia 73:481–485.
Area, Vermilion County, Illinois. Journal of Herpetology
Shine, R., and T. Madsen. 1996. Is thermoregulation unim-
34:558–564.
portant for most reptiles? an example using water pythons
King, D. I., R. M. Degraaf, and C. R. Griffin. 1998. Edge- (Liasis fuscus) in tropical Australia. Physiological Zoology
related nest predation in clearcut and groupcut stands. Con- 69:252–269.
servation Biology 12:1412–1415. Shine, R., T. R. L. Madsen, M. J. Elphick, and P. S. Harlow.
Krohmer, R. W. 1989. Body temperature relationships in the 1997. The influence of nest temperatures and maternal
lined snake, Tropidoclonion lineatum. Comparative Bio- brooding on hatchling phenotypes in water pythons. Ecol-
chemistry and Physiology 92A:541–543. ogy 78:1713–1721.
Loiselle, B. A., and W. G. Hoppes. 1983. Nest predation in SPSS. 1995. Advanced statistics. Version 6.1 for Macintosh.
insular and mainland lowland rainforest in Panama. Condor SPSS, Chicago, Illinois, USA.
85:93–95. Stevens, J. 1996. Applied multivariate statistics for the social
Martin, T. E. 1988. Habitat and area effects on forest bird sciences. Third edition. Lawrence Erlbaum Associates,
assemblages: is nest predation an influence? Ecology 69: Mahwah, New Jersey, USA.
74–84. Stickel, L. F., W. H. Stickel, and F. C. Schmid. 1980. Ecology
McCollin, D. 1998. Forest edges and habitat selection in of a Maryland population of black rat snakes (Elaphe o.
birds: a functional approach. Ecography 21:247–260. obsoleta). American Midland Naturalist 103:1–14.
Murcia, C. 1995. Edge effects in fragmented forests: impli- Tewksbury, J. J., S. J. Hejl, and T. E. Martin. 1998. Breeding
cations for conservation. Trends in Ecology and Evolution productivity does not decline with increasing fragmentation
10:58–62. in a western landscape. Ecology 79:2890–2903.
Paton, P. W. 1994. The effect of edge on avian nest success: Weatherhead, P. J., and M. B. Charland. 1985. Habitat se-
how strong is the evidence? Conservation Biology 8:17– lection in an Ontario population of the snake, Elaphe ob-
26. soleta. Journal of Herpetology 19:12–19.
Peterson, C. R., A. R. Gibson, and M. E. Dorcas. 1993. Snake Weatherhead, P. J., and D. J. Hoysak. 1989. Spatial and ac-
thermal ecology: the causes and consequences of body- tivity patterns of black rat snakes (Elaphe obsoleta) from
temperature variation. Pages 241–314 in R. A. Seigel and radiotelemetry and recapture data. Canadian Journal of Zo-
J. T. Collins, editors. Snakes: ecology and behavior. Mc- ology 67:463–468.
Weatherhead, P. J., and I. C. Robertson. 1990. Homing to
Graw-Hill, Toronto, Ontario, Canada.
food by black rat snakes (Elaphe obsoleta). Copeia 1990:
Pimentel, R. A. 1979. Morphometrics: the multivariate anal- 1164–1165.
ysis of biological data. Kendall/Hunt, Dubuque, Iowa, Weir, R. D. 1989. Birds of the Kingston region. Quarry Press,
USA. Kingston, Ontario, Canada.
Prior, K. A., and P. J. Weatherhead. 1998. Status of the black Wilson, E. O. 1992. The diversity of life. W. W. Norton, New
rat snake, Elaphe obsoleta obsoleta, in Canada. Committee York, New York, USA.
on the Status of Endangered Wildlife in Canada, Ottawa, Winkworth, R. E., and D. W. Goodall. 1962. A crosswire
Ontario, Canada. sighting tube for point quadrat analysis. Ecology 43:342–
Reinert, H. K. 1984. Habitat variation within sympatric snake 343.
populations. Ecology 65:1673–1682. Yahner, R. H., and D. P. Scott. 1988. Effects of forest frag-
Reinert, H. K. 1992. Radiotelemetric field studies of pitvi- mentation on depredation of artificial nests. Journal of
pers: data acquisition and analysis. Pages 185–197 in J. A. Wildlife Management 52:158–161.