ArticlePDF Available

Snakes on the plains: The impacts of habitat structure on snake communities in Illinois grasslands

Wiley
Wildlife Society Bulletin
Authors:

Abstract

Snakes are an integral component of grassland ecosystems, though their occurrence in high densities is often discouraged because of their role as nest predators of declining grassland birds. The effects of habitat structure on snake communities in grasslands remains poorly understood, hindering management efforts. We used 3 years of data from a series of restored grasslands in southern Illinois to examine how habitat structure across 3 spatial scales (local, patch, landscape) affected snake relative abundance, diversity, and species‐specific occupancy in grasslands. We found that snake community metrics were strongly and positively related to an increase in woody plant cover at the local (within‐patch) scale. Snake relative abundance was also positively related to an increase in grass cover and a decrease in forb cover, though our occupancy results suggest that this was primarily driven by an increase in black kingsnakes ( Lampropeltis nigra ). At the patch scale, relative abundance and diversity of snakes were both positively related to the proportion of patch edge composed of roads. Habitat structure at the landscape scale had the smallest impact on snakes in this study, though the proportion of trees in the landscape was positively related to snake diversity. We suggest that managers and conservationists interested in manipulating snake abundance in grasslands focus on within‐patch vegetation structure and composition. Decreasing woody cover in grasslands, or increasing the ratio of forbs to grasses, may reduce the presence of snakes, while maintaining a woody component could encourage both snake abundance and diversity.
Received: 26 January 2022
|
Revised: 17 May 2022
|
Accepted: 17 May 2022
DOI: 10.1002/wsb.1366
RESEARCH ARTICLE
Snakes on the plains: The impacts of habitat
structure on snake communities in Illinois
grasslands
Alex Glass
1
|Michael W. Eichholz
2
1
Cooperative Wildlife Research Laboratory,
Southern Illinois University Carbondale, 1125
Lincoln Drive, Carbondale, IL 62901, USA
2
Cooperative Wildlife Research Laboratory
and Center for Ecology, Southern Illinois
University Carbondale, 1125 Lincoln Drive,
Carbondale, IL 62901, USA
Correspondence
Alex Glass, Cooperative Wildlife Research
Laboratory, Southern Illinois University
Carbondale, 1125 Lincoln Drive, Carbondale,
IL 62901, USA.
Email: alexander.glass@siu.edu
Funding information
Illinois Department of Natural Resources,
Grant/Award Number: Federal Wildlife Aid
Grant W106R
Abstract
Snakes are an integral component of grassland ecosystems,
though their occurrence in high densities is often discouraged
because of their role as nest predators of declining grassland
birds. The effects of habitat structure on snake communities in
grasslands remains poorly understood, hindering management
efforts. We used 3 years of data from a series of restored
grasslands in southern Illinois to examine how habitat structure
across 3 spatial scales (local, patch, landscape) affected snake
relative abundance, diversity, and speciesspecific occupancy
in grasslands. We found that snake community metrics were
strongly and positively related to an increase in woody plant
cover at the local (withinpatch) scale. Snake relative abun-
dance was also positively related to an increase in grass cover
and a decrease in forb cover, though our occupancy results
suggest that this was primarily driven by an increase in black
kingsnakes (Lampropeltis nigra). At the patch scale, relative
abundance and diversity of snakes were both positively related
to the proportion of patch edge composed of roads. Habitat
structure at the landscape scale had the smallest impact on
snakes in this study, though the proportion of trees in the
landscape was positively related to snake diversity. We suggest
that managers and conservationists interested in manipulating
snake abundance in grasslands focus on withinpatch vegeta-
tion structure and composition. Decreasing woody cover in
Wildlife Society Bulletin 2022;e1366. wileyonlinelibrary.com/journal/wsb
|
1of20
https://doi.org/10.1002/wsb.1366
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and
reproduction in any medium, provided the original work is properly cited.
© 2022 The Authors. Wildlife Society Bulletin published by Wiley Periodicals LLC on behalf of The Wildlife Society.
grasslands, or increasing the ratio of forbs to grasses, may
reduce the presence of snakes, while maintaining a woody
component could encourage both snake abundance and
diversity.
KEYWORDS
Coluber constrictor, community occupancy, habitat management,
Lampropeltis calligaster, North American racer, prairie kingsnake, snake
abundance, snake diversity, tallgrass prairie
Extensive habitat loss and fragmentation have transformed the grasslands of North America into one of the continent's
most endangered resources (Rickletts et al. 1999,Samsonetal.2004). The transformation has largely been driven by the
conversion of grasslands to humancentric land uses such as agriculture or urban development (Warner 1994,Whiteetal.
2000, Rosenberg et al. 2016). Habitat loss and fragmentation contributed to the steady decline of many wildlife taxa
historically dependent on grasslands, including birds (Herkert et al. 2003), mammals (Miller and Cully 2001, Sackett et al.
2012),andreptilessuchassnakes(Cagle2008)andturtles(Dodd2001). Displacement of grassland environments by
agriculture and urbanization is especially pronounced in Illinois, USA (Warner 1994), leading the state to make the
protection and restoration of the state's grasslands a priority (State of Illinois 2015).
From a wildlife conservation perspective, the role of snakes in grasslands is complicated, as they perform
important ecological functions but can cause negative impacts at high population densities. On one hand, snakes are
an integral part of grassland communities, and as predators of small mammals (Lidicker 1989) and invertebrates
(Klimstra 1959), they can indirectly affect seed dispersal and vegetation structure (McCauley et al. 2006, Reiserer
et al. 2018). Anecdotal evidence and Illinois state survey records, summarized by Cagle (2008), suggest that
grassland snake abundance has generally declined in the region since the early twentieth century. Continued
declines of snakes may lead to a sustained increase in smallmammal populations, which can negatively affect plant
diversity and structure in grasslands (Geier and Best 1980; Howe et al. 2002,2006), and cause an increase in
transmission of zoonotic diseases such as hantaviruses and Lyme disease (Ostfeld et al. 2006, Hofmeester et al.
2017). On the other hand, snakes can be problematic for land managers and biologists when they occur in
grasslands at high densities. Snakes are common avian nest predators in many grassland systems (Renfrew and Ribic
2003, Klug et al. 2010, Lyons et al. 2015). Grassland birds are the most rapidly declining bird guild in North America
(Rosenberg et al. 2019), and nest predation by snakes may exacerbate the population declines of threatened or
sensitive bird species (Chalfoun et al. 2002, Conkling et al. 2012, DeGregorio et al. 2014).
Because of these dynamics, maintaining snake populations at a sustainable level is an important aspect of
grassland management. However, management for snakes remains hindered by a lack of information on the basic
natural history of many species (Dodd 1993), including how they interact with their environment (Gibbons et al.
2000). The cryptic nature of most snake species has made it difficult to gain insight into their habitat selection
processes (Seigel and Collins 1993). A greater understanding of how the habitat structure of grasslands may
influence snake communities is necessary for both snake conservation and effective management of grasslands.
The goal of our study was to estimate habitat association patterns of snakes in midwestern grasslands to guide
management practices aiming to increase or decrease local snake populations. Our specific objectives were to
identify structural habitat features of grasslands and their surrounding landscape that influence snake relative
abundance (RA) and diversity during the bird breeding season (early Maylate Jul). Because habitat selection by
snakes likely occurs in a hierarchical manner (Johnson 1980, Robson and BlouinDemers 2021), snake habitat use
may be affected by variables across multiple spatial scales (Harvey and Weatherhead 2006, Moore and Gillingham
2006). Therefore, we considered habitat features at 3 different scales: local (within an individual grassland), patch
(grassland size, shape, edge composition), and landscape (400m buffer around grassland patch).
2of20
|
GLASS AND EICHHOLZ
At the local scale, we predicted that snake RA and diversity would be higher in grassland patches with features
indicative of later successional stages, such as a deeper litter layer, greater vegetation height and density, less bare
ground, and more woody vegetation. Most snakes are cryptic species, and rely on dense vegetation and litter to
remain hidden from their prey and avoid predation (Fitch 1963, Martino et al. 2012), whereas woody vegetation can
serve as important refugia for snakes and provide a beneficial thermoregulatory environment (Harvey and
Weatherhead 2006, Klug et al. 2010, Martino et al. 2012). We also predicted that snake activity would be positively
associated with grass cover and negatively associated with forb cover because voles (Microtus spp.), an important
prey item for grassland snakes (Fitch 1978, Rosen 1991), may primarily consume grasses (Zimmerman 1965,
M'Closkey 1975, Howe et al. 2006).
At the patch scale, we expected that the ratio of edge to patch size (hereafter edgeinterior ratio) and edge
composition would influence snake RA and diversity. Snakes may be attracted to edges for feeding or
thermoregulating (BlouinDemers and Weatherhead 2001, Richardson et al. 2006, Row and BlouinDemers 2006).
We therefore predicted that snake RA and diversity would be positively related to edgeinterior ratio. Attraction or
aversion to edge habitat can depend on edge composition (Patrick and Gibbs 2009), so we also considered the
proportion of different edge types surrounding a patch. We predicted that agriculture edges would negatively
affect snake RA, whereas other edge types (roads, forest, water) would have a positive effect. Snakes are
understood to avoid agriculture because of the homogeneity of the vegetation structure and lack of plant litter,
resulting in a poor thermoregulatory environment, and frequent disturbances in the form of farming practices
(Richardson et al. 2006, Cagle 2008, Martino et al. 2012, King and Vanek 2020). Road, forest, or water edges,
meanwhile, may offer advantages such as beneficial thermoregulatory microclimates (Rosen and Lowe 1994,
BlouinDemers and Weatherhead 2001) and increased prey density (Weatherhead and BlouinDemers 2004). We
predicted that patch size would be negatively related to snake RA and diversity because an increase in patch size
often relates to a decrease in edgeinterior ratio. Though species such as prairie kingsnakes (Lampropeltis calligaster)
may prefer grassland interiors (Richardson et al. 2006), we suspected that most snake species would decrease in
density because of the smaller proportion of edge habitat.
At the landscape scale, we predicted that snake RA would be negatively associated with agriculture, and
positively associated with both grassland and water. Previous studies have shown that prairie kingsnakes and North
American racers (Coluber constrictor), 2 common grassland snakes in our study area, associate strongly with
grassland (Fitch 1978, Richardson et al. 2006) and riparian zones (Rosen 1991, Martino et al. 2012), respectively.
We also predicted that the amount of forest and shrubland in the surrounding landscape would positively affect
snake diversity by reducing the dominance of grassland snakes like garter snakes (Thamnophis spp.) and prairie
kingsnakes (Richardson et al. 2006, Patrick and Gibbs 2009), and encouraging the presence of more generalist or
forestadapted species such as rough greensnakes (Opheodrys aestivus) and black ratsnakes (Pantherophis obsoletus).
STUDY AREA
Our study took place at Burning Star State Fish and Wildlife Area (37°52N, 89°12W, hereafter Burning Star), a
former surface coal mine composed of 1,824 ha of both reclaimed and undisturbed land in northeast Jackson
County, Illinois (Figure 1). Landcover types present at Burning Star include forest (616 ha), shrubland (320 ha),
agriculture (456 ha), wetland (66 ha), restored tallgrass prairie (110 ha), and several freshwater lakes (223 ha; Illinois
Department of Natural Resources 2018). Fieldwork occurred on 10 restored prairie patches (hereafter sites) within
Burning Star, which ranged in size from 1.8 ha to 35.9 ha (
x
¯
± SD = 10.99 ± 10.57 ha). Common grasses on the
restored prairie sites included warmseason natives such as big bluestem (Andropogon gerardii), Indian grass
(Sorghastrum nutans), and switchgrass (Panicum virgatum), and nonnative grasses such as smooth brome (Bromus
inermis), Kentucky bluegrass (Poa pratensis), and foxtail (Setaria spp.). Common forbs included Canada goldenrod
(Solidago canadensis), annual ragweed (Ambrosia artemisiifolia), and nonnatives such as sericea lespedeza
GRASSLAND SNAKES
|
3of20
(Lespedeza cuneata) and sweet clover (Melilotus spp.). Management actions such as prescribed fire and ungulate
grazing have mostly been absent at these sites since restoration was initially conducted in the mid1990s, though
2 sites were burned in recent years (site 10 in Feb 2020 and site 8 in Mar 2021). The general lack of management has
resulted in the establishment and encroachment of woody shrubs such as eastern red cedar (Juniperus virginiana),
blackberry (Rubus spp.), honey locust (Gleditsia triacanthos), and nonnative autumn olive (Elaeagnus umbellata).
METHODS
Snake and vegetation surveys
We estimated snake RA using grids of coverboards at each grassland site that we checked once a week for the
duration of our field seasons (early May to late Jul 20192021). We used 0.6m
2
plywood sheets for our
coverboards, arranged in a grid of 4 parallel rows of 5 boards each, for a total of 20 boards per site. Rows were
spaced 50 m apart, and boards within rows were placed at 15m intervals. We identified snakes under coverboards
to species, or listed them as unknown if we could not identify the species. We calculated snake RA for each site by
dividing the number of snakes encountered (all species combined) by the number of boards checked (Klug et al.
FIGURE 1 Map of Burning Star State Wildlife Management Area in De Soto, IL, USA, where snake communities
were surveyed via coverboards across 10 field sites from MayJul 20192021.
4of20
|
GLASS AND EICHHOLZ
2009). We also calculated the Shannon diversity index of snake species at each site using the vegan package
(Oksanen et al. 2019) in program R (R Core Team 2021). The Shannon index was based on the number of
encounters of each species recorded at a site. We did not include snake encounters of unknown species in diversity
calculations.
We performed 2 rounds of vegetation structure surveys each field season from 2019 to 2021. The first round
of surveys took place in midMay, and the second in midJuly. We sampled vegetation characteristics following
methods described in Glass and Eichholz (2021). Briefly, we conducted vegetation surveys along a series of
transects composed of 10 sampling points in each grassland site. At each sampling point, we measured vegetation
density and height using a Robel pole (Robel et al. 1970). We took Robel pole readings at every sampling point from
each cardinal direction, at a viewing distance of 4 m and a height of 1 m (Fisher and Davis 2010). We averaged the
4 readings into 1 value representing the sampling point. We considered vegetation height the point of highest
contact of vegetation on the Robel pole, and estimated vegetation density by recording the lowest point on the
Robel pole that could be seen through the vegetation. We also estimated the percent cover of grasses, forbs, bare
ground, and woody vegetation within a 20 × 50cm quadrat, and measured litter depth (cm) with a standard ruler.
We estimated vegetation composition in late August of each year by revisiting each sampling point, identifying
all plant species present, and estimating their percent cover within a 20 × 50cm quadrat. We conducted vegetation
composition surveys separately from the vegetation structure surveys in midMay and midJuly because the warm
season grasses and prairie asters (Solidago canadensis,Eupatorium spp.) in our study area were more easily
distinguished during the late summer when they were in bloom. We used these data to estimate plant diversity at
each site.
Data analysis
We quantified habitat variables representing 3 spatial scales: local (withinpatch), patch, and landscape (Table 1).
Local variables included mean litter depth, vegetation height and density, bare ground, woody cover, grass cover,
forb cover, and plant diversity. Plant diversity was represented by the Shannon diversity index of plant species at
each site, whereas we derived the other variables from average values for vegetation structure measurements
within a patch.
Patch variables included patch size (ha), edgeinterior ratio, and the proportion of different edge types (tree,
agriculture, water, or roads) for each site. We estimated patch size, edgeinterior ratio (perimeter [m]/area [ha]), and
proportion of edge types in ArcMap 10.6 (ESRI 2018). We calculated the proportion of each edge type by
measuring the length of each edge type present in a site and dividing that value by the site's total perimeter.
We estimated landscape variables using the proportion of 5 different landcover types within a 400m buffer around
each site: trees (forest and shrubland), agriculture, water, grassland (including buffers around agricultural fields and roads),
anddevelopment(roads,buildings,parkinglots,andmowed lawns). We limited buffers around patches to 400 m to
minimize redundancy in landscape variable values because patches are located in close proximity to each other (Figure 1).
We calculated percent cover of different landcover types by digitizing the study site and surrounding area in ArcMap 10.6
at a 1:1500 scale, using Maxar basemap imagery from 2017 provided by ArcGIS.
We analyzed data in RStudio 4.0.4 (RStudio Team 2020). We used generalized linear mixed models created in
the glmmTMB package (Brooks et al. 2017) to investigate how snake RA and diversity relate to habitat variables
across several spatial scales. All models included year and site as random effects to compensate for unmodeled
temporal and spatial variability. We evaluated models using Akaike's Information Criterion corrected for small
sample size (AIC
c
; Hurvich and Tsai 1989) in the R package MuMIn (Bartoń2019). We employed a stepwise model
selection process based on AIC
c
, beginning with the null model (Yamashita et al. 2007). We performed this process
separately for each of the 3 sets of variables representing the 3 spatial scales. At each step, we added each
predictor variable sequentially to the model. The model with the lowest AIC
c
value would pass on to the next step
GRASSLAND SNAKES
|
5of20
and the process would repeat, until arriving at the most parsimonious model (in other words, when adding new
variables no longer lowered AIC
c
). We did not include variables that were highly correlated (|r| > 0.6) together in any
model.
Though stepwise model selection is a useful process for exploratory analyses where many different variables
are being considered (Keinath et al. 2017), one of the principal advantages of an informationtheoretic AIC
approach over traditional hypothesis testing is the ability to draw inference from multiple models receiving support
from a data set (Anderson 2008:105). Therefore, to avoid excluding models beyond the most parsimonious that
were supported by our data, at each step we retained models within 2 ΔAIC
c
of the top model for the following
TABLE 1 Habitat variables representing 3 spatial scales used as predictors of snake relative abundance and
diversity in Burning Star State Fish and Wildlife Area in De Soto, Illinois, USA, from 2019 to 2021. Local scale is
withinpatch habitat structure. Patch scale includes the physical structure of the patch and the proportion of
different edge types. Landscape scale refers to the proportion of different landcover types within 400 m of a patch.
The range of values column lists the original range of values for each variable before being standardized to have a
mean of 0 and standard deviation of 1.
Variable name Definition Range of values
Local scale
Grass cover % grass cover within a site 7.1566.03
Forb cover % forb cover within a site 4.0864.08
Bare ground % bare ground within a site 0.1312.3
Woody cover % woody vegetation within a site 014.75
Litter depth Average litter depth (cm) of a site 0.626.28
Vegetation height Average height of vegetation (cm) within a site 51.22103.49
Vegetation density Average density of vegetation within a site 22.7162.01
Plant diversity Shannon (H) index of plant diversity for a site 1.632.82
Patch scale
Patch size Size of site (ha) 1.835.9
Edgeinterior ratio Perimeter (m)/area (ha) of a site 129.8593.72
Tree edge % of a site's perimeter that is composed of trees (forests or tree rows) 3.2491.78
Road edge % of a site's perimeter composed of roads 034.1
Agriculture edge % of a site's perimeter composed of agriculture 089.15
Water edge % of a site's perimeter composed of open bodies of water 051.53
Landscape scale
% trees % of landscape composed of forests, tree rows, or shrubland within a
400m buffer of a site
8.5975.15
% agriculture % of landscape composed of agriculture within a 400m buffer of a site 8.6254.3
% water % of landscape composed of rivers, streams, or open bodies of water
within a 400m buffer of a site
5.5231.85
% grassland % of landscape composed of grassland, including grass buffers around
roads and agricultural fields
3.122.53
% development % of landscape composed of buildings, roads, parking lots, and
residential lawns
0.099.82
6of20
|
GLASS AND EICHHOLZ
step, though at the following step we only added new variables to the top model. In cases where the top model was
nested within the model being considered, the cutoff for retainment was 1 ΔAIC
c
to avoid the presence of
uninformative variables in the candidate model set (Anderson 2008, Arnold 2010). When the stepwise model
selection process was completed for a set of variables, we used the most parsimonious model along with any
competitive models (meaning models within 2 ΔAIC
c
,or1ΔAIC
c
for nested models) to form a candidate model set.
We examined the potential for crossscale models by repeating the stepwise selection process while
considering together all variables that were present in competitive models at each spatial scale. Our candidate
model set ultimately consisted of competitive models for each of the 3 spatial scales, plus crossscale models. If
stepwise selection for a set of variables determined that the null model was the most parsimonious, with no other
model being competitive, then no models from that set of variables appeared in the candidate model set. We used
this process to create 2 candidate model sets: one with snake RA as the response variable, and one with snake
diversity as the response variable.
We tested the adequacy of our models by examining residual plots, and by calculating R
2
values. For mixed
models, R
2
consists of 2 different values: marginal R
2
calculates the amount of variation in the data explained by the
fixed variables in the model, whereas conditional R
2
calculates the amount of variation in the data explained by both
the fixed and random variables. We examined modelaveraged regression coefficients (β) and 95% confidence
intervals to evaluate the effect size and direction of variables that appeared in the candidate model set. We
standardized predictor variables and their corresponding regression coefficients to have a mean of 0 and a standard
deviation of 1 to improve model fit, and to better compare effect sizes of variables that were originally measured in
different units.
To gain insight into speciesspecific responses to habitat structure, we performed a post hoc analysis using a
hierarchical Bayesian community occupancy model (Dorazio and Royle 2005, Homyack et al. 2016, Guzy et al.
2019), which estimates speciesspecific occupancy and detection probabilities, and responses to habitat
parameters, informed by occurrence data from all snake species combined. Our hierarchical modeling approach
generates more precise parameter estimates than frequentist approaches for individual species because it considers
speciesspecific parameter responses in the context of communitylevel (all snake species combined) responses,
while also accounting for imperfect detection (Dorazio and Royle 2005, Zipkin et al. 2009).
We first created observation matrices for each speciesacross12samplingoccasions(foreachofthe
12 weeks in the field season) and 30 siteyear combinations, where detection = 1 and nondetection = 0. We
excluded snakes of unknown species from this analysis. We considered each siteyear combination as an
individual site for this analysis because the sitespecific variables that we used to estimate occupancy often
changed within a site from year toyear.Wedefinedthetrueoccupancystatusofspeciesiat site jas z
i,j
.If
species iwas present at site j,thenz
i,j
=1, otherwise z
i,j
= 0. We considered the occupancy status for each
species as a Bernoulli random variable, z
i,j
~Bern(Ψ
i,j
), where Ψ
i,j
gives the probability that species iis present at
site j. We denoted the detection of species iat site jduring survey kas y
i,j,k
,wherey
i,j,k
= 1 if species iwas
detected at site jduring survey k,otherwisey
i,j,k
= 0. We also considered detection for each species as a
Bernoulli random variable, y
i,j,k
~Bern(p
i,j,k
×z
i,j
), where p
i,j,k
is the probability that species iwas detected at site
jduring survey k. We estimated occupancy probabilities for each species as a linearlogit function of the habitat
predictor variables that were present in the final candidate model set for either snake RA or diversity, and
estimated detection for each species as a linearlogit function of the following surveyspecific sampling
covariates: day of survey, temperature, and humidity. We retrieved average temperature (°C) and humidity (%)
for each day a coverboard survey occurred from archived records of the Southern Illinois Airport weather
station (37.78°N, 89.25°W), located approximately 14 km from our study area.
We fit the model using the R package R2jags (Su and Yajima 2021), which interfaces with JAGS (Plummer
2003), a program for analysis of Bayesian hierarchical models that employs Markov chain Monte Carlo (MCMC)
simulation to generate samples from posterior distributions. We used N(0, 0.5) priors for communitylevel mean
parameters, and unif(0.1, 3) priors for communitylevel standard deviation parameters (Homyack et al. 2016).
GRASSLAND SNAKES
|
7of20
We ran 3 Markov chains, each of length 200,000, with the first 100,000 removed as burnin, and the remainder
thinned by a factor of 50. This yielded 6,000 samples across the 3 chains that we used to estimate posterior mean,
standard deviation, and 95% Bayesian credible intervals (BCI) for each variable in our occupancy and detection
submodels. We assessed model convergence by visually examining trace plots to confirm that chains were
appropriately mixed, and via the GelmanRubin statistic (R; Gelman and Rubin 1992) for variables in our model. The
GelmanRubin statistic represents the ratio of betweenchain to withinchain variability, so values closer to 1
indicate greater model convergence. The Rvalues for all variables in our model were <1.01, indicating no issues
with model convergence. Parameter estimates from the occupancy model are reported as the mean and 95% BCI.
We considered parameter estimates to be highly supported by the data if BCIs did not overlap zero, and moderately
supported if BCIs overlapped zero by <10% of their value range.
RESULTS
Through 3 field seasons, we encountered 108 snakes representing 8 species (Table 2), including 39 North American
racers (hereafter racers), 23 prairie kingsnakes, 17 black kingsnakes (Lampropeltis nigra), 15 common garter snakes
(Thamnophis sirtalis), 3 western ribbon snakes (T. proximus), 3 DeKay's brown snakes (Storeria dekayi), 3 smooth
earth snakes (Virginia valeriae), 2 black ratsnakes, and 3 encounters where the species was not identified. We
encountered 62 snakes in 2019, 59 snakes in 2020, and 59 snakes in 2021.
Several pairs of predictor variables were highly correlated, both within and among spatial scales (Figure S1,
available in Supporting Information). We did not include any of these highly correlated variables together in the
same model. A correlation matrix of snake encounters across all sites and years revealed no high correlations among
species (Figure S2, available in Supporting Information), which we interpret to suggest that competition between
species did not influence our results in a meaningful way. This was not unexpected, because Durso et al. 2013
reported a similar lack of influence by interspecific competition on snake community structure.
TABLE 2 Snake encounters under coverboards at each field site from 2019 to 2021 (all years combined) across
10 grassland sites at Burning Star State Fish and Wildlife Area in De Soto, Illinois, USA.
Site
North
American
racer
Prairie
kingsnake
Black
kingsnake
Common
garter
snake
Western
ribbon
snake
Black
ratsnake
DeKay's
brown
snake
Smooth
earth
snake Unknown Total
13 4 3 1 11
23 4 3 1 11
32 2 3 3 10
49 1 4 3 2 1 20
51 1 3 2 1 1 1 10
62 3 4 9
76 4 2 1 13
88 5 13
94 3 7
10 1 3 4
Total 39 23 17 15 3 2 3 3 3 108
8of20
|
GLASS AND EICHHOLZ
Relative abundance and diversity
For snake RA, the model representing woody cover was the most parsimonious among localscale variables, and
2 other models were competitive: woody cover + grass cover, and woody cover +forb cover (Table 3). Woody
cover was positively related to snake RA (Figure 2), with a standardized regression coefficient (β) of 0.48 (95%
CI = 0.10.87; Table 4). Grass cover (β= 0.33, 0.050.71) was also positively related to snake RA, whereas forb
cover and snake RA had a negative association (β=0.26, 0.60.08). At the patch scale, the null model was the
most parsimonious, though the model representing road edge was competitive. Snake RA was positively
associated with road edge (β= 0.27, 0.10.63). No landscapescale or crossscale models were competitive
with the null model.
For snake diversity, the top model at the local scale was woody cover + vegetation height, and the model
containing only woody cover was also competitive (Table 3). Snake diversity was positively associated with
woody cover (β= 0.5, 0.170.84; Figure 3), and negatively associated with vegetation height (β=0.31,
0.640.01). For patchscale variables, road edge was the most parsimonious model, and the model representing
water edge was competitive. Snake diversity was positively associated with road edge (β= 0.36, 0.020.71), and
negatively associated with water edge (β=0.3, 0.660.06). The null model was the topranked model among
landscapescale variables, though the model representing percent trees was competitive. Snake diversity and
the percentage of trees in the landscape were positively related (β= 0.28, 0.080.64). No crossscale models
TABLE 3 Model selection results estimating the influence of habitat variables on snake relative
abundance and snake diversity in Burning Star State Fish and Wildlife Area in De Soto, Illinois, USA, from
2019 to 2021. All models included year and site as random effects. We present the number of variables
(fixed and random) in each model (K), Akaike's Information Criterion corrected for small sample sizes (AIC
c
),
the difference in AIC
c
between the given model and the topranked model in its set (ΔAIC
c
), and Akaike
weight (w
i
), a measure of the evidence that a particular model is the best model in the set. We assessed
adequacy of our models by calculating the proportion of variation in the response variable explained by fixed
effectsinthegivenmodel R
()
marg
2and the variation in the response variable explained by fixed and random
effects
R
()
cond
2
.
Model KAIC
c
ΔAIC
c
w
i
R
mar
g
R
cond
Snake relative abundance
Woody cover 5 91.265 0.000 0.323 0.17 0.18
Woody cover + grass cover 6 91.419 0.154 0.299 0.25 0.26
Woody cover + forb cover 6 92.071 0.806 0.216 0.23 0.23
Null 4 93.711 2.445 0.095 0.00 0.02
Road edge 5 94.423 3.158 0.067 0.07 0.07
Snake diversity
Woody cover + vegetation height 6 89.104 0.000 0.403 0.31 0.31
Woody cover 5 89.128 0.024 0.398 0.22 0.27
Road edge 5 92.323 3.219 0.081 0.13 0.16
Null 4 93.461 4.357 0.046 0.00 0.13
Water edge 5 93.739 4.634 0.040 0.09 0.12
% trees 5 94.129 5.025 0.033 0.08 0.12
GRASSLAND SNAKES
|
9of20
were present in the candidate set. The crossscale model of woody cover and road edge outcompeted the null
model but was 1.5 AIC
c
higher than the univariate woody cover model, which suggests that the addition of road
edge to the woody cover model did not add explanatory value (Arnold 2010). Thus, woody cover + road edge
was not included in the candidate set.
FIGURE 2 Linear plots illustrating the relationships between snake relative abundance (RA; number of snakes
encountered per coverboard checked) and relevant habitat variables at Burning Star Wildlife Management Area in De
Soto, IL, USA, according coverboard surveys conducted from MayJul 20192021. The plots are based on predictive
linear models ranked by AIC
c
. Predicted RA values on theY axis and their relationship with the habitat variable on the X
axis are calculated using the top model in which each habitat variable appears. The shaded area around the regression
line represents 95% confidence intervals of the estimate. Note differing Y axis values among plots.
TABLE 4 Standardized regression coefficient estimates, with 95% lower (LCI) and upper (UCI) confidence
intervals, for habitat variables appearing in the snake relative abundance and snake diversity candidate model sets
for snakes in Burning Star State Fish and Wildlife Area in De Soto, Illinois, USA, from 2019 to 2021. We model
averaged regression coefficient estimates among all models in the candidate set in which the variable appeared.
Variable Estimate LCI UCI
Snake relative abundance
Woody cover 0.48 0.10 0.87
Grass cover 0.33 0.05 0.71
Forb cover 0.26 0.60 0.08
Road edge 0.27 0.10 0.63
Snake diversity
Woody cover 0.50 0.17 0.84
Vegetation height 0.31 0.64 0.01
Road edge 0.36 0.02 0.71
Water edge 0.30 0.66 0.06
% trees 0.28 0.08 0.64
10 of 20
|
GLASS AND EICHHOLZ
Occupancy model results
In our hierarchical community occupancy model, occupancy was estimated as a function of the following 7 habitat
variables: woody cover, grass cover, forb cover, vegetation height, road edge, water edge, and percent trees.
Speciesspecific occupancy estimates varied widely, ranging from 0.86 for prairie kingsnakes to 0.12 for smooth
earth snakes. Detection estimates were low for all species, and ranged from 0.1 for racers to 0.01 for 4 different
species (Table 5). Communitylevel and speciesspecific responses to habitat variables generally reflected results
from the snake RA and diversity model sets, though responses did vary among species (Table 6). Communitylevel
posterior occupancy parameter estimates indicated moderate support for a positive association with woody cover
(
x
¯
= 1.09, 95% BCI = 0.22.37), and a negative association with vegetation height (0.89, 1.990.13). In other
words, an increase in 1 standard deviation (about 4%) woody cover corresponded with a 1.09% increase in
occupancy probability, and 1 standard deviation (about 12 cm) increase in vegetation height corresponded with a
0.89% decrease in occupancy probability. Posterior parameter estimates were less precise for grass cover (0.85,
0.862.55), forb cover (0.27, 1.711.16), road edge (0.44, 0.641.55), water edge (0.37, 1.710.97), and
percent trees (0.49, 0.852.15).
DISCUSSION
Our model selection results suggest that snake RA and diversity are most strongly influenced by habitat structure at
local scale, as no patchor landscapescale variables were present in the most supported models (<2 ΔAIC
c
) of our
candidate model sets. Harvey and Weatherhead (2006) detected a similar pattern while investigating habitat
selection by eastern massasauga rattlesnakes (Sistrurus catenatus), with microhabitat structure having an outsized
influence relative to landscapescale factors. The high relative importance of microscale habitat relative to that at
FIGURE 3 Linear plots illustrating the relationships between snake diversity (Shannon index) and relevant
habitat variables at Burning Star Wildlife Management Area in De Soto, IL, USA, according to coverboard surveys
conducted from MayJul 20192021. The plots are based on predictive linear models ranked by AIC
c
. Predicted
Shannon index values on the Y axis and their relationship with the habitat variable on the X axis are calculated using
the top model in which each habitat variable appears. The shaded area around the regression line represents 95%
confidence intervals of the estimate. Note differing Y axis values among plots.
GRASSLAND SNAKES
|
11 of 20
greater spatial scales may be common for many ectotherm species because their need to thermoregulate should
increase their dependence on local habitat structure (Rubio and Carrascal 1994).
Among the localscale habitat variables we examined, woody cover was arguably the most influential for both
snake RA and diversity; it was present in the most supported models of both candidate model sets and had the
strongest association with both response variables. Woody cover was positively associated with occupancy for all
snake species, though we found strongest support for a positive influence of woody cover on black kingsnakes,
common garter snakes, and western ribbon snakes. The positive effect of woody cover on snakes was consistent
with our initial predictions, and corroborates results from other studies (Harvey and Weatherhead 2006, Klug et al.
2010, Martino et al. 2012). Overhead cover provided by woody vegetation can allow snakes to better avoid
detection by predators (Charland and Gregory 1995), and aid in thermoregulation (Webb and Shine 1997).
Additionally, shrubs and stump hollows may provide secure places for oviposition for egglaying snakes (Plummer
1990, Madsen and Shine 1999) and can serve as overwinter hibernacula (Prior and Weatherhead 1996, Rudolph
et al. 2007). Although our field work occurred during the summer, summer territories for snakes may be located in
close proximity to winter hibernacula (Brown and Parker 1976, Plummer and Congdon 1994). Finally, woody cover
may improve hunting efficiency for snakes by providing opportunities to conceal themselves from prey
(Theodoratus and Chiszar 2000, Tsairi and Bouskila 2004), and by promoting the presence of smallmammal and
amphibian prey, as woody cover may increase habitat suitability for Peromyscus mice (Horncastle et al. 2005,
Matlack et al. 2008, Glass and Eichholz 2021) and amphibians (Semlitsch et al. 2009).
Although woody cover was strongly and positively associated with snake RA and diversity, this relationship is
likely unimodal rather than linear. The maximum amount of woody cover at Burning Star was 15% at site 4, so we
are unable to infer how snakes may react to higher levels of woody encroachment in this study. The encroachment
of woody plants in grasslands beyond a certain point may reduce, instead of increase, grassland snake abundance
and diversity. High proportions of woody cover can negatively affect habitat suitability for snakes by reducing the
availability of basking sites (Reading and Jofré 2009), and shading out herbaceous ground cover plants that snakes
may use to remain hidden from predators and prey (Jofré et al. 2016).
TABLE 5 Mean occupancy and detection estimates, with 95% Bayesian credible intervals (BCI), for snake
species that we encountered during coverboard surveys at Burning Star State Fish and Wildlife Area in De Soto,
Illinois, USA, from 2019 to 2021. The occupancy and detection values reported here were estimated using the full
occupancy model with environmental covariates. The model estimated the occupancy of each species separately
for every siteyear combination (n= 30). The occupancy estimates and credible intervals reported here are
averaged for each species across all 30 siteyears. The model estimated detection for each species for every site
yearsurvey combination (n= 360). As with occupancy, we report detection probabilities and credible intervals as
averaged across all 360 estimates for each species.
Occupancy Detection
Species
x
¯95% BCI
x
¯95% BCI
North American racer 0.85 0.49 >0.99 0.1 0.04 0.16
Prairie kingsnake 0.86 0.49 >0.99 0.06 0.01 0.11
Black kingsnake 0.42 0.11 0.79 0.04 0.01 0.12
Common garter snake 0.44 0.13 0.84 0.04 0.01 0.10
Western ribbon snake 0.26 0.03 0.75 0.01 <0.01 0.05
Black ratsnake 0.15 <0.01 0.63 0.01 <0.01 0.03
DeKay's brown snake 0.41 0.02 0.96 0.01 <0.01 0.05
Smooth earth snake 0.12 <0.01 0.46 0.01 <0.01 0.03
12 of 20
|
GLASS AND EICHHOLZ
TABLE 6 Posterior distribution means and 95% Bayesian credible intervals for parameters used to estimate
occupancy for snake species encountered at Burning Star State Fish and Wildlife Area in De Soto, Illinois, USA,
from 2019 to 2021. Mean = mean of the posterior distribution, LCI =lower credible interval, UCI =upper credible
interval. An asterisk (*) by a mean estimate indicates that credible intervals overlap zero by less than 10% of their
value range.
Parameter Estimate
North American
racer
Prairie
kingsnake
Black
kingsnake
Common
garter snake
Woody cover Mean 1.16 0.96 1.27* 1.27*
LCI 0.48 0.83 0.15 0.2
UCI 2.96 2.64 2.95 3.01
Grass cover Mean 0.16 1.2 2.06* 0.69
LCI 2.12 1.8 0.09 3.5
UCI 2.43 4.07 4.41 1.84
Forb cover Mean 0.19 0.34 0.42 0.13
LCI 1.78 2.14 2.3 1.7
UCI 1.43 1.43 1.2 1.53
Vegetation height Mean 0.91 0.83 1* 0.96*
LCI 2.34 2.12 2.37 2.31
UCI 0.5 0.44 0.14 0.18
Road edge Mean 0.54 0.39 0.44 0.77
LCI 0.86 1.06 0.82 0.46
UCI 2.1 1.82 1.78 2.34
Water edge Mean 0.35 0.26 0.45 0.56
LCI 1.95 1.82 1.97 2.28
UCI 1.21 1.38 1.01 0.99
% trees Mean 0.29 0.4 0.76 0.43
LCI 1.6 1.38 0.75 1.2
UCI 2.32 2.32 2.74 2.45
Parameter Estimate
Western ribbon
snake Black ratsnake
DeKay's brown
snake
Smooth earth
snake
Woody cover Mean 1.45* 0.97 0.95 0.87
LCI 0.06 0.78 0.96 0.91
UCI 3.33 2.6 2.71 2.45
Grass cover Mean 0.41 1.67 1 1.47
LCI 2.22 0.54 1.3 0.82
UCI 2.8 4.15 3.41 3.92
Forb cover Mean 0.24 0.32 0.17 0.37
LCI 1.96 2.13 1.78 2.16
UCI 1.43 1.31 1.5 1.28
(Continues)
GRASSLAND SNAKES
|
13 of 20
Grass cover and forb cover within a site affected snake RA in our study. We suspect this relationship is due at
least in part to the effect of vegetation composition on vole populations. Voles are abundant in our study area
(Glass and Eichholz 2021), and tend to associate with grassdominated patches in grasslands (Howe and Lane 2004,
Howe et al. 2006, Poe et al. 2019). Voles are a common prey item for kingsnakes (Fitch 1978, Jenkins et al. 2001),
racers (Rosen 1991), and black ratsnakes (Trauth and McAllister 1995). Additionally, vole burrows are used by a
variety of snakes for shelter and thermoregulation (Huey et al. 1989, Richardson et al. 2006, Steen et al. 2010).
Thus, sites with more grass cover may contain more refuge sites, and greater prey density for snakes that rely
heavily on small mammals for their diet. As in many grasslands, an increase in forbs corresponds with a decrease in
grasses at Burning Star (r=0.81), likely contributing to the negative relationship we documented between forb
cover and snake RA. Speciesspecific occupancy results suggest that the positive effect of grass cover on snake RA
may be driven primarily by black kingsnakes, as they were the most strongly associated with grass cover. Estimates
for other species were more variable. For example, garter snake occupancy was negatively related to grass cover,
though this estimate was less precise. Unlike black kingsnakes, small mammals comprise only a small proportion of
the diet of common garter snakes (Durso et al. 2021, and citations therein), so they may not necessarily respond to
any increases in vole abundance that might result from greater grass cover.
Our results suggest that vegetation height had a negative impact on snake diversity. Vegetation height can
have a strong effect on local habitat structure and microclimate for ectotherms (Masterson et al. 2008, Mizsei et al.
2020), so taller vegetation may create a microclimate unfavorable for some snake species. Although vegetation
height was not an influential variable for snake RA, we found support for a negative association between vegetation
height and occupancy of 3 species: black kingsnakes, common garter snakes, and DeKay's brown snake.
At the patch scale, both snake RA and snake diversity were positively associated with road edge. The positive
effect of road edge on both snake community metrics may result from the opportunity these roads provide
for thermoregulation (Gibbons and Semlitsch 1987, Rosen and Lowe 1994). Though trees, shrubs, and tall herbaceous
vegetation are important for keeping snakes from overheating when temperatures rise during the day, open areas
such as roads can be advantageous in early mornings when temperatures are cool. Road avoidance by snakes has
been documented in several studies (Shine et al. 2004, Andrews et al. 2005), though roads in these studies
experienced regular vehicle use for at least part of the year. The roads at Burning Star are gravel and lightly trafficked,
TABLE 6 (Continued)
Parameter Estimate
Western ribbon
snake Black ratsnake
DeKay's brown
snake
Smooth earth
snake
Vegetation height Mean 0.89 0.86 1.05* 0.68
LCI 2.23 2.22 2.56 2.04
UCI 0.36 0.48 0.22 0.78
Road edge Mean 0.44 0.32 0.35 0.36
LCI 0.85 1.25 1.08 1.07
UCI 1.77 1.72 1.74 1.76
Water edge Mean 0.45 0.17 0.34 0.43
LCI 2.14 1.73 1.92 2.13
UCI 1.14 1.16 1.32 1.18
% trees Mean 0.64 0.25 0.52 0.71
LCI 0.92 1.56 1.22 0.86
UCI 2.74 2.2 2.64 2.65
14 of 20
|
GLASS AND EICHHOLZ
as motorized vehicles are only allowed for researchers and staff. Thus, the level of human disturbance along these
roads, and mortality risk for snakes, is minimal. A greater proportion of lightly used open roads along grassland edges
may therefore increase individual use of these grasslands by a variety of snake species. However, this would likely not
be the case for paved roads with more consistent traffic, as the increase in road mortality can reduce snake
populations (Row et al. 2007) and lead to behavioral avoidance of roads (Shine et al. 2004, Andrews et al. 2005).
Water edge was also negatively associated with snake diversity at the patch scale. This result may simply be
due to the composition of Burning Star, as water edge was significantly and negatively correlated with road edge
(r=0.52) and percent trees at the landscape scale (r=0.7), both of which were influential variables in our study
that were positively related to snake diversity. Alternatively, grasslands surrounded by a greater proportion of
water will likely have more mesic soil. This, in turn, may increase vegetation growth, potentially discouraging certain
species in our study area that avoid tall vegetation, such as black kingsnakes, DeKay's brown snakes, and common
garter snakes. Although some of Burning Star's snakes, such as Thamnophis species, are known to use riparian areas
with mesic soil (White and Kolb 1974), no snake species in our study responded positively to water edge. The water
edges of Burning Star's grasslands are all composed of manmade deepwater lakes that abruptly transition to
grassland, so the riparian streamside ecosystems that Thamnophis occasionally inhabit are not present near our
study sites, even when they abut a waterbody.
Landscapescale variables were the least impactful set of variables examined in this study. While this may indicate
a lack of influence of landscape composition compared to smaller scale habitat variables (Harvey and Weatherhead
2006), our 400m buffer is smaller than those of other landscapescale studies, and this may have limited the
inferential abilities of our models at this scale. No landscapescale variables were present in the candidate model set
for snake RA, and only percent trees was present in the snake diversity model set, though it was outperformed by the
null model. Support for the impact of trees in the landscape on snake diversity is limited compared to that of other
variables such as woody cover, though the positive relationship was consistent with our initial predictions. Many
snakes are capable of using resources in a variety of landcover types (Carfagno and Weatherhead 2006, Martino et al.
2012, França and Braz 2013), especially if they are located in close proximity to one another. Forest or shrubland
associated snake species may occasionally use nearby grasslands to forage or thermoregulate, particularly along
forestgrassland edges (BlouinDemers and Weatherhead 2001). This could lead to a higher diversity of snakes in
grasslands with forests or shrubland in the surrounding landscape. Other landcover types examined here such as
agriculture, development, and water, contain few, if any, terrestrial snake species that could use grasslands.
Though we predicted that agriculture would have a negative impact on snake communities, we did not
detect such a relationship in this study, at either the patch or the landscape scale. In addition to the limited size of
our 400m buffer, the amount of agriculture at Burning Star is smaller than in many landscapes elsewhere in the
Midwest and Great Plains; agriculture only makes up 25% of Burning Star's total area. This may limit the negative
impact of agriculture on snakes in our study sites. Additionally, high levels of human disturbance are understood to
be at least partially responsible for snakes' documented aversion to agriculture (Richardson et al. 2006, Cagle 2008,
Martino et al. 2012, King and Vanek 2020). Because our field seasons took place after spring planting and before
fall harvest, human and machinery presence in agricultural fields was minimized, which may have lessened active
avoidance of agriculture by snakes.
MANAGEMENT IMPLICATIONS
As for many species, habitat use for snakes changes throughout the year because they pivot to target areas of
overwinter refugia during the fall (Rosen 1991, Harvey and Weatherhead 2006). Therefore, our findings are only
indicative of snake habitat use during the late spring and summer. We focused our study on the bird breeding
season because snakes are often of highest concern for land managers in the context of their role as avian nest
predators. We suggest that managers interested in influencing snake abundance and diversity focus on within
GRASSLAND SNAKES
|
15 of 20
patch structure, particularly woody vegetative cover, as the influence of the surrounding landscape appears
minimal. Managers wishing to increase or maintain snake abundance and diversity should encourage a moderate
woody or shrubby component in grasslands, and provide areas of minimal vegetation, like the roads at Burning Star,
to aid thermoregulation. If reducing snake abundance is the goal, decreasing the amount of woody cover through
prescribed fire or shrub removal, and planting a high ratio of forbs to grasses, may prove effective.
ACKNOWLEDGMENTS
We would like to thank C. Crawford, B. Baum, and the many field technicians that have assisted with research at
Burning Star for the immense amount of help they provided in the field. We also thank J. Dallas, J. O'Connell, M.
Miller, and A. Minor for helpful comments and feedback on early drafts of this manuscript, and S. Ballard for his
help identifying snake species. We thank J. Refsnider (Associate Editor), A. Knipps (Editorial Assistant), A.
Tunstall (Copy Editor), and J. Levengood (Content Editor) and 2 anonymous reviewers for constructive
comments that improved our manuscript. Funding for this study was provided by the Illinois Department of
Natural Resources.
CONFLICT OF INTEREST
The authors declare that there is no conflict of interest.
ETHICS STATEMENT
All field methods used in this study were approved by the Institutional Animal Care and Use Committee (IACUC) of
Southern Illinois University Carbondale (protocol number 18020).
DATA AVAILABILITY STATEMENT
Data available on request from the authors.
ORCID
Alex Glass http://orcid.org/0000-0002-1084-8625
REFERENCES
Anderson, D. R. 2008. Model based inference in the life sciences: a primer on evidence. Springer, New York, New
York, USA.
Andrews, K. M., J. W. Gibbons, and T. W. Reeder. 2005. How do highways influence snake movement? Behavioral
responses to roads and vehicles. Copeia 2005:772782.
Arnold, T. W. 2010. Uninformative parameters and model selection using Akaike's Information Criterion. Journal of Wildlife
Management 74:11751178.
Bartoń, K. 2019. MuMIn: multimodel inference. R package version 1.43.6. https://CRAN.R-project.org/package=MuMIn
BlouinDemers, G., and P. J. Weatherhead. 2001. Habitat use by black rat snakes (Elaphe obseleta obseleta) in fragmented
forests. Ecology 82:28822896.
Brooks, M. E., K. Kristensen, K. J. van Benthem, A. Mangusson, C. W. Berg, A. Nielsen, H. J. Skaug, M. Maechler, and
B. M. Bolker. 2017. glmmTMB balances speed and flexibility among packages for zeroinflated generalized linear
mixed modeling. R Journal 9:378400.
Brown, W. S., and W. S. Parker. 1976. Movement ecology of Coluber constrictor near communal hibernacula. Copeia 1976:
225242.
Cagle, N. L. 2008. Snake species distributions and temperate grasslands: a case study from the American tallgrass prairie.
Biological Conservation 141:744755.
Carfagno, G. L. F., and P. J. Weatherhead. 2006. Intraspecific and interspecific variation in use of forestedge habitat by
snakes. Canadian Journal of Zoology 84:14401452.
Chalfoun, A. D., M. J. Ratnaswamy, and F. R. Thompson, III. 2002. Songbird nest predators in forestpasture edge and
forest interior in a fragmented landscape. Ecological Applications 12:858867.
Charland, M. B., and P. T. Gregory. 1995. Movements and habitat use in gravid and nongravid female garter snakes
(Colubridae: Thamnophis). Journal of Zoology 236:543561.
16 of 20
|
GLASS AND EICHHOLZ
Conkling, T. J., T. L. Pope, K. N. Smith, H. A. Mathewson, M. L. Morrison, R. N. Wilkins, and J. W. Cain. 2012. Blackcapped
vireo nest predator assemblage and predictors for nest predation. Journal of Wildlife Management 76:14011411.
DeGregorio, B. A., S. J. Chiavacci, P. J. Weatherhead, J. D. Willson, T. J. Benson, and J. H. Sperry. 2014. Snake predation on
North American bird nests: culprits, patterns and future directions. Journal of Avian Biology 45:325333.
Dodd, C. K., Jr. 1993. Strategies for snake conservation. Pages 363393 in R. A. Seigel and J. T. Collins, editors. Snakes:
ecology and behavior. McGrawHill, Inc., New York, New York, USA.
Dodd, C. K., Jr. 2001. North American box turtles. University of Oklahoma Press, Norman, USA.
Dorazio, R. M., and J. A. Royle. 2005. Estimating size and composition of biological communities by modeling the
occurrence of species. Journal of the American Statistical Association 100:389398.
Durso, A. M., J. D. Willson, and C. T. Winne. 2013. Habitat influences diet overlap in aquatic snake assemblages. Journal of
Zoology 291:185193.
Durso, A. M, L. A. NeumanLee, G. R. Hopkins, and E. D. Brodie, Jr. 2021. Stable isotope analysis suggests that
tetrodotoxinresistant common gartersnakes (Thamnophis sirtalis) rarely feed on newts in the wild. Canadian Journal
of Zoology 99:331338.
ESRI. 2018. ArcMap GIS. Version 10.6. Environmental System Research Institute, Redlands, California, USA.
Fisher, R. J., and S. K. Davis. 2010. From Wiens to Robel: a review of grassland bird habitat selection. Journal of Wildlife
Management 74:265273.
Fitch, H. S. 1963. Natural history of the racer, Coluber constrictor. University of Kansas publications, Museum of Natural
History 15:351468.
Fitch, H. S. 1978. A field study of the prairie kingsnake (Lampropeltis calligaster). Transactions of the Kansas Academy of
Science 81:353363.
França, F. G. R., and V. S. Braz. 2013. Diversity, activity patterns, and habitat use of the snake fauna of Chapada dos
Veadeiros National Park in central Brazil. Biota Neotropica 13:7484.
Geier, A. R., and L. B. Best. 1980. Habitat selection by small mammals of riparian communities: evaluating effects of habitat
alterations. Journal of Wildlife Management 44:1624.
Gelman, A., and D. B. Rubin. 1992. Inference from iterative simulation using multiple sequences. Statistical Science 7:
457472.
Gibbons, J. W., D. E. Scott, T. J. Ryan, K. A. Buhlmann, T. D. Tuberville, B. S. Metts, J. L. Greene, T. Mills, Y. Leiden, S. Poppy,
et al. 2000. The global decline of reptiles, deja vu amphibians. Bioscience 50:653666.
Gibbons, J. W., and R. D. Semlitsch. 1987. Activity patterns. Pages 396421 in R. A. Seigel, J. T. Collins, and S. S. Novak,
editors. Snakes: ecology and evolutionary biology. MacMillan, New York, New York, USA.
Glass, A., and M. W. Eichholz. 2021. Habitat associations of small mammal communities in a restored prairie system in
southern Illinois. Journal of Mammalogy 102:789801.
Guzy, J., K. Halloran, J. Homyack, J. E. ThorntonFrost, and J. D. Willson. 2019. Differential responses of amphibian and
reptile assemblages to size of riparian buffers within managed forests. Ecological Applications 29:e01995.
Harvey, D. S., and P. J. Weatherhead. 2006. A test of the hierarchical model of habitat selection using eastern massasauga
rattlesnakes (Sistrurus c. catenatus). Biological Conservation 130:206216.
Herkert, J. R., D. L. Reinking, D. A. Wiedenfeld, M. Winter, J. L. Zimmerman, W. E. Jensen, E. J. Finck, R. R. Koford,
D. H. Wolfe, S. K. Sherrod, et al. 2003. Effects of prairie fragmentation on the nest success of breeding birds in the
midcontinental United States. Conservation Biology 17:587594.
Hofmeester, T. R., P. A. Jansen, H. J. Wijnen, E. C. Coipan, M. Fonville, H. H. T. Prins, H. Sprong, and S. E. van Wieren. 2017.
Cascading effects of predator activity on tickborne disease risk. Proceedings of the Royal Society of London B:
Biological Sciences 284:20170453.
Homyack, J. A., C. J. O'Bryan, J. E. Thornton, and R. F. Baldwin. 2016. Community occupancy of herpetofauna in roadside
ditches in a managed pine landscape. Forest Ecology and Management 361:346357.
Horncastle, V. J., E. C. Hellgren, P. M. Mayer, A. C. Ganguli, D. M. Engle, and D. M. Leslie 2005. Implications of invasion by
Juniperus virginiana on small mammals in the southern Great Plains. Journal of Mammalogy 86:11441155.
Howe, H. F., J. S. Brown, and B. ZornArnold. 2002. A rodent plague on prairie diversity. Ecology Letters 5:3036.
Howe, H. F., and D. Lane. 2004. Voledriven succession in experimental wetprairie restorations. Ecological Applications 14:
12951305.
Howe, H. F., B. ZornArnold, A. Sullivan, and J. S. Brown. 2006. Massive and distinctive effects of meadow voles on
grassland vegetation. Ecology 87:30073013.
Huey, R. B., C. R. Peterson, S. J. Arnold, and W. P. Porter. 1989. Hot rocks and notsohot rocks: retreatsite selection by
garter snakes and its thermal consequences. Ecology 70:931944.
Hurvich, C. M., and C. Tsai. 1989. Regression and time series model selection in small samples. Biometrika 76:297307.
Illinois Department of Natural Resources. 2018. Burning Star State Fish and Wildlife Area Management Brief. Illinois
Department of Natural Resources, Springfield, USA.
GRASSLAND SNAKES
|
17 of 20
Jenkins, L. N., T. J. Thomasson IV, and J. G. Byrd. 2001. A field study of the black kingsnake, Lampropeltis getula nigra.
Herpetological Natural History 8:5767.
Jofré, G. M., M. R. Warn, and C. J. Reading. 2016. The role of managed coniferous forest in the conservation of reptiles.
Forest Ecology and Management 362:6978.
Johnson, D. H. 1980. The comparison of usage and availability measurements for evaluation of resource preference.
Ecology 61:6571.
Keinath, D. A., D. F. Doak, K. E. Hodges, L. R. Prugh, W. Fagan, C. H. Sekercioglu, S. H. M. Buchart, and M. Kauffman. 2017.
A global analysis of traits predicting species sensitivity to habitat fragmentation. Global Ecology and Biogeography 26:
115127.
King, R. B., and J. P. Vanek. 2020. Responses of grassland snakes to tallgrass prairie restoration. Restoration Ecology 28:
573582.
Klimstra, W. D. 1959. Foods of the racer, Coluber constrictor, in southern Illinois. Copeia 1959:210214.
Klug, P. E., S. L. Jackrel, and K. A. With. 2010. Linking snake habitat use to nest predation risk in grassland birds: the dangers
of shrub cover. Oecologia 162:803813.
Klug, P. E., L. L. Wolfenbarger, and J. P. McCarty. 2009. The nest predator community of grassland birds responds to
agroecosystem habitat at multiple scales. Ecography 32:973982.
Lidicker, W. Z. 1989. Impacts of nondomesticated vertebrates on California grasslands. Pages 135150 in L. F. Huenneke
and H. Mooney, editors. Grassland structure and function: California annual grasslands. Kluwer Academic Publishers,
Dordrecht, The Netherlands.
Lyons, T. P., J. R. Miller, D. M. Debinksi, and D. M. Engle. 2015. Predator identity influences the effect of habitat
management on nest predation. Ecological Applications 25:15961605.
Madsen, T., and R. Shine. 1999. Life history consequences of nestsite variation in tropical pythons (Liasis fuscus). Ecology
80:989997.
Martino,J.A.,R.G.Poulin,D.L.Parker,andC.M.Somers. 2012. Habitat selection by grassland snakes
at northern range limits: implications for conservation. Journal of Wildlife Management 76:759767.
Masterson, G. P. R., B. Maritz, and G. J. Alexander. 2008. Effects of fire history and vegetation structure on herpetofauna in
a South African grassland. Applied Herpetology 5:129143.
Matlack, R. S., D. W. Kaufman, and G. A. Kaufman. 2008. Influence of woody vegetation on small mammals in tallgrass
prairie. American Midland Naturalist 160:719.
McCauley, D. J., F. Keesing, T. P. Young, B. F. Allan, and R. M. Pringle. 2006. Indirect effects of large herbivores on snakes in
an African savanna. Ecology 87:26572663.
M'Closkey, R. T. 1975. Habitat succession and rodent distribution. Journal of Mammalogy 56:950955.
Miller, S. D., and J. F. Cully 2001. Conservation of blacktailed prairie dogs (Cynomys ludovicianus). Journal of Mammalogy
82:889893.
Mizsei, E., Z. Fejes, A. Malatinszky, S. Lengyel, and C. Vadász. 2020. Reptile responses to vegetation structure in a grassland
restored for an endangered snake. Community Ecology 21:203212.
Moore, J. A., and J. C. Gillingham. 2006. Spatial ecology and multiscale habitat selection by a threatened rattlesnake: the
eastern massasauga (Sistrurus catenatus catenatus). Copeia 2006:742751.
Oksanen, J., R. Kindt, and B. O'Hara. 2019. Vegan: community ecology package. R package version 2. 56. https://CRAN.R-
project.org/package=vegan
Ostfeld, R. S., F. Keesing, and K. LoGiudice. 2006. Community ecology meets epidemiology: the case of Lyme disease. Pages
2840 in S. Collinge, and C. Ray, editors. Disease ecology: community structure and pathogen dynamics. Oxford
University Press, Oxford, United Kingdom.
Patrick, D. A., and J. P. Gibbs. 2009. Snake occurrences in grassland associated with road versus forest edges. Journal of
Herpetology 43:716720.
Plummer, M. 2003. JAGS: a program for analysis of Bayesian graphical models using Gibbs sampling. Proceedings of the 3rd
International Workshop on Distributed Statistical Computing, 2022 March 2003, Vienna, Austria.
Plummer, M. V. 1990. Nesting movements, nesting behavior, and nest sites of green snakes (Opheodrys aestivus) revealed
by radiotelemetry. Herpetologica 46:190195.
Plummer, M. V., and J. D. Congdon. 1994. Radiotelemetric study of activity and movements of racers (Coluber constrictor)
associated with a Carolina Bay in South Carolina. Copeia 1994:2026.
Poe, N., K. L. Stuble, and L. Souza. 2019. Small mammal herbivores mediate the effects of soil nitrogen and invertebrate
herbivores on grassland diversity. Ecology and Evolution 9:35773587.
Prior, K. A., and P. J. Weatherhead. 1996. Habitat features of black rat snake hibernacula in Ontario. Journal of Herpetology
30:211218.
R Core Team. 2021. R: a language and environment for statistical computing. R Foundation for Statistical Computing,
Vienna, Austria.
18 of 20
|
GLASS AND EICHHOLZ
RStudio Team. 2020. RStudio: integrated development for R. RStudio, PBC, Boston, Massachusetts, USA.
Reading, C. J., and G. M. Jofré. 2009. Habitat selection and range size of grass snakes Natrix natrix in an agricultural
landscape in southern England. AmphibiaReptilia 30:379388.
Reiserer, R. S., G. W. Schuett, and H. W. Greene. 2018. Seed ingestion and germination in rattlesnakes: overlooked agents
of rescue and secondary dispersal. Proceedings of the Royal Society B: Biological Sciences 285:20172755
Renfrew, R. B., and C. A. Ribic. 2003. Grassland passerine nest predators near pasture edges identified on videotape. Auk
120:371383.
Richardson, M. L., P. J. Weatherhead, and J. D. Brawn. 2006. Habitat use and activity of prairie kingsnakes (Lampropeltis
calligaster calligaster) in Illinois. Journal of Herpetology 40:423428.
Rickletts, T. H., E. Dinerstein, D. M. Olsen, C. J. Loucks, W. Eichbaum, D. DellaSala, K. Kavanaugh, P. Hedao, P. T. Hurley,
K. M. Carney, et al. 1999. Terrestrial ecoregions of North America. Island Press, Washington D.C., and Covello,
California, USA.
Robel, R. J., J. N. Briggs, A. D. Dayton, and L. C. Hulbert. 1970. Relationships between visual obstruction measurement and
weight of grassland vegetation. Journal of Range Management 23:295297.
Robson, L. E., and G. BlouinDemers. 2021. Eastern hognosed snake habitat selection at multiple spatial scales in Ontario,
Canada. Journal of Wildlife Management 85:838846.
Rosen, P. C. 1991. Comparative ecology and life history of the racer (Coluber constrictor) in Michigan. Copeia 1991:
897909.
Rosen, P. C., and C. H. Lowe. 1994. Highway mortality of snakes in the Sonoran desert of southern Arizona. Biological
Conservation 68:143148.
Rosenberg, K. V., A. M. Dokter, P. J. Blancher, J. R. Sauer, A. C. Smith, P. A. Smith, J. C. Stanton, A. Panjabi, L. Helft, M. Parr,
et al. 2019. Decline of the North American avifauna. Science 366:120124.
Rosenberg, K. V., J. A. Kennedy, R. Dettmers, R. P. Ford, D. Reynolds, J. D. Alexander, C. J. Beardmore, P. J. Blancher,
R. E. Bogart, G. S. Butcher, et al. 2016. Partners in Flight Landbird Conservation Plan: 2016 Revision for Canada and
Continental United States. Partners in Flight Science Committee.
Row, J. R., and G. BlouinDemers. 2006. Thermal quality influences effectiveness of thermoregulation, habitat use and
behavior in milk snakes. Oecologia 148:111.
Row, J. R., G. BlouinDemers, and P. J. Weatherhead. 2007. Demographic effects of road mortality in black ratsnakes
(Elaphe obsoleta). Biological Conservation 137:117124
Rubio, J. L., and L. M. Carrascal. 1994. Habitat selection and conservation of an endemic Spanish lizard Algyroides marchi
(Reptilia, Lacertidae). Biological Conservation 70:245250.
Rudolph, D. C., R. R. Schaefer, S. J. Burgdorf, M. Duran, and R. N. Conner. 2007. Pine snake (Pituophis ruthveni and Pituophis
melanoleucus lodingi) hibernacula. Journal of Herpetology 41:560565.
Sackett, L. C., T. B. Cross, R. T. Jones, W. C. Johnson, K. Ballare, C. Ray, S. K. Collinge, and A. P. Martin. 2012. Connectivity
of prairie dog colonies in an altered landscape: inferences from analysis of microsatellite DNA variation. Conservation
Genetics 13:407418.
Samson, F. B, F. L Knopf, and W. R. Ostlie. 2004. Great Plains ecosystems: past, present, and future. Wildlife Society
Bulletin 32:615.
Seigel, R. A., and J. T. Collins. 1993. Snakes: ecology and behavior. McGrawHill, New York, New York, USA.
Semlitsch, R. D., B. D. Todd, S. M. Blomquist, A. J. K. Calhoun, J. W. Gibbons, J. P. Gibbs, G. J. Graeter, E. B. Harper,
D. J. Hocking, M. L. Hunter, Jr., et al. 2009. Effects of timber harvest on amphibian populations: understanding
mechanisms from forest experiments. BioScience 59:853862.
Shine, R., M. Lemaster, M. Wall, T. Langkilde, and R. Mason. 2004. Why did the snake cross the road? Effects of roads on
movement and location of mates by garter snakes (Thamnophis sirtalis parietalis). Ecology and Society 9:9.
State of Illinois. 2015. Illinois wildlife action plan 2015 implementation guide. Illinois Department of Natural Resources,
Springfield, USA. https://www.dnr.illinois.gov/conservation/iwap/pages/default.aspx. Accessed 13 Mar 2019.
Steen, D. A., J. M. Linehan, and L. L. Smith. 2010. Multiscale habitat selection and refuge use of common kingsnakes,
Lampropeltis getula, in southwestern Georgia. Copeia 2010:227231.
Su, Y.S., and M. Yajima. 2021. R2jags: using R to run JAGS'. R package version 0.71. https://CRAN.R-project.org/
package=R2jags
Theodoratus, D. H., and D. Chiszar. 2000. Habitat selection and prey odor in the foraging behavior of western rattlesnakes
(Crotalus viridis). Behaviour 137:119135.
Trauth, S. E., and C. T. McAllister. 1995. Vertebrate prey of selected Arkansas snakes. Journal of the Arkansas Academy of
Science 49:41.
Tsairi, H., and A. Bouskila. 2004. Ambush site selection of a desert snake (Echis coloratus) at an oasis. Herpetologica 60:
1323.
GRASSLAND SNAKES
|
19 of 20
Warner, R. E. 1994. Agricultural land use and grassland habitat in Illinois: future shock for midwestern birds? Conservation
Biology 8:147156.
Weatherhead, P. J., and G. BlouinDemers. 2004. Understanding avian nest predation: why ornithologists should study
snakes. Journal of Avian Biology 35:185190.
Webb, J. K., and R. Shine. 1997. Out on a limb: conservation implications of treehollow use by a threatened snake species
(Hoplocephalus bungaroides: Serpentes, Elapidae). Biological Conservation 81:2133.
White, M., and J. A. Kolb. 1974. A preliminary study of Thamnophis near Sagehen Creek, California. Copeia 1974:126136.
White, R., S. Murray, and R. Rohweder. 2000. Pilot analysis of global ecosystems: grassland ecosystems. World Resources
Institute, Washington, D.C, USA.
Yamashita, T., K. Yamashita, and R. Kamimura. 2007. A stepwise AIC method for variable selection in linear regression.
Communications in StatisticsTheory and Methods 36:23952403.
Zimmerman, E. G. 1965. A comparison of habitat and food of two species of Microtus. Journal of Mammalogy 46:605612.
Zipkin, E. F., A. DeWan, and J. A. Royle. 2009. Impacts of forest fragmentation on species richness: a hierarchical approach
to community modelling. Journal of Applied Ecology 46:815822.
Associate Editor: J. Refsnider.
SUPPORTING INFORMATION
Additional supporting information may be found in the online version of the article at the publishers website.
How to cite this article: Glass, A., and M. W. Eichholz. 2022. Snakes on the plains: The impacts of habitat
structure on snake communities in Illinois grasslands. Wildlife Society Bulletin e1366.
https://doi.org/10.1002/wsb.1366
20 of 20
|
GLASS AND EICHHOLZ
... Two recent studies investigated the effect vegetation diversity had on snake abundance. Glass and Eichholz (2022) found that the relative proportion of forbs to grasses was an important predictor of snake relative abundance in Illinois, US grasslands. Mizsei et al. (2020) found that directly measured vegetation Shannon diversity had a negative relationship with abundance of meadow vipers (Vipera ursinii) and had no effect on three species of lizards in Hungary. ...
... Our covariate list ( Table 1) was composed of variables from our a priori list of hypotheses (Appendix S2) and additional expert opinion variables thought to be important to occupancy to reduce un-modeled variation and improve our ability to resolve the relationships outlined in our hypotheses. Expert opinion variables included confounding variables (quadrats_mowed_percent), those found to be important in similar studies (woody_cov-er_ppn_##_m_r) (Glass & Eichholz, 2022), or those related to potential overwintering habitat (develo-ped_##_m_r_ppn, distance_to_water_meters). Detection covariates included effort (number of board-turns), time of year, mean temperature, cloudiness, and total precipitation. Covariates on occupancy and naïve presence included vegetation richness and density, landscape configuration, and habitat area variables. ...
... Harvey and Weatherhead (2006) found that for eastern Massasaugas in Ontario, Canada, microhabitats such as refugia and hibernacula were better at describing habitat selection than landscape-scale habitat variables. Within-patch woody cover was the best predictor of snake presence in a study in southern Illinois (Glass & Eichholz, 2022). ...
Article
Full-text available
Nearly one in five reptile species is at risk of extinction. Changes in habitat area, its configuration, and vegetation diversity could affect habitat use, but their relative importance is understudied. We assessed how these factors affected reptile presence in agricultural landscapes figure in Iowa, United States, using 695 cover boards visited 16,441 times in 2015–2020. Species‐wise encounter rates ranged 0.0001–0.012. Eight of 11 species and 54.2% of individuals were species of greatest conservation need. Habitat area, configuration, and vegetation diversity influenced reptile presence similarly. Mean patch occupancy was 0.18 for common garter snake (CG, Thamnophis sirtalis ) and 0.45 for all snakes (AS). Naïve presence was explained by effort (odds ratio [OR] AS = 1.83, OR CG = 1.79), vegetation diversity (OR AS = 1.28, OR CG = 1.28), woody cover (OR AS = 1.24, OR CG = 1.41), and patch size (OR AS = 1.30). Large patch prairies were more likely to contain snakes than other conservation practices ( encounter = 0.291), and more likely to contain CG (0.098) than prairie contour strips (0.031), waterways (0.018), grass contour strips (0.016), or terraces (0.015). While we documented low overall reptile presence, their higher presence in large prairie patches underscores the importance of core nature reserves for reptile conservation.
... Additionally, previous studies suggest that nest survival for grassland-nesting birds is negatively affected by woody cover because of its association with nest predators (Bakker 2003, Graves et al. 2010, Klug et al. 2010. Previous research from our study area (Glass and Eichholz 2022) has linked snake abundance to woody cover in these grassland sites, likely due to thermoregulatory benefits and protection from overhead predators that shrubs provide to snakes. The reduction of plant litter post-fire may have also benefitted Field Sparrow nest survival, as daily survival rate for Field Sparrows was positively associated with bare ground. ...
... The edge effects on nest survival we detected at Burning Star may instead relate more to road edges than forest edges. Previous studies in Midwestern U.S. grasslands have indicated that snake and raccoon abundance may be positively associated with road edge (Barding andNelson 2008, Glass andEichholz 2022). The roads at Burning Star are lightly used by humans; thus, they may serve as travel lanes for mammalian mesopredators (Barding andNelson 2008, Wysong et al. 2020) and may attract snakes due to their beneficial thermoregulatory environment (Gibbons andSemlitsch 1987, Rosen andLowe 1994). ...
Article
Full-text available
Grassland birds are the most rapidly declining bird guild in North America, largely due to extensive habitat loss and fragmentation. Because many grassland bird species have different habitat preferences, managing grasslands to provide habitat for a range of species can be a challenge. We used four years of data from southern Illinois, USA grasslands to estimate the influence of prescribed fire and habitat structure on nest survival, nest density, and abundance of three grassland bird species with different habitat preferences: Dickcissel (Spiza americana), Field Sparrow (Spizella pusilla), and Common Yellowthroat (Geothlypis trichas). We found that Dickcissels exhibited the strongest response to prescribed fire, as nest density and nest survival both increased after previously undisturbed grasslands were burned. Fire may have also benefitted Common Yellowthroats and Field Sparrows by reducing woody cover and increasing bare ground, both of which were characteristics associated with nest survival for these birds. Dickcissel abundance was positively related to plant diversity within a grassland and agriculture in the surrounding landscape (within 400 m of a grassland patch), and negatively related to edge-interior ratio. Field Sparrows demonstrated a positive association with woody cover and proximity to forests. Common Yellowthroats were associated with tall vegetation and agriculture in the surrounding landscape. Both Field Sparrows and Common Yellowthroats associated positively with habitat characteristics that reduced nest survival, suggesting potential adaptive mismatches. Our results suggest that periodic prescribed fire, increased plant diversity, and larger patch size may simultaneously benefit a broad variety of grassland bird species with different habitat preferences.
... While it is common knowledge among herpetologists that the temporal niche varies both daily and seasonally (DeGregorio et al., 2015;Rocha et al., 2014), its daily occurrence is often not investigated. On the other hand, regarding spatial niches, recent studies have revealed that partitioning is more pronounced at the landscape level than at the microhabitat level in snake communities (Di Pietro et al., 2020;Glass & Eichholz, 2022). Regarding diet, niche partitioning by food type such as frogs, lizards, rodents rather than food species within a food type is particularly prevalent among snakes (Toft, 1985). ...
Article
Full-text available
Investigations on resource use by predators are important for understanding the mechanisms of biological coexistence. Although time, habitat, and diet are considered the three major dimensions for niche partitioning, studies that comprehensively investigate these dimensions in predator guilds are limited. Snakes are one of the predators whose populations have been reported to be declining worldwide. While diet has been considered as a fundamental variable that allows snakes to coexist, the importance of spatiotemporal resources has also been noted especially in temperate regions. To clarify the coexistence mechanisms of terrestrial snakes, we assessed the niche partitioning patterns of snakes on Sado Island, a Japanese island rich in snake species, from the perspectives of spatiotemporal and dietary resources. Specifically, we investigated the daily and seasonal occurrences as temporal niches, landscape‐level distribution as a spatial niche, and stomach content as a dietary niche. We found that niche partitioning in all three major resources occurred among snake species on the island. Daily occurrence was partitioned into three groups: completely diurnal, nocturnal, and active during both periods. Seasonal occurrence was partitioned into three groups: widely active from spring to autumn, mainly active in summer, and mainly active in autumn. Distribution was partitioned into two groups: mainly distributed in lowlands and distributed even in mountainous areas. Food habits were partitioned into three groups: rodents, frogs, and earthworms as main prey, respectively. Our results provide empirical evidence that snakes can coexist through multidimensional niche partitioning, and that spatiotemporal resources are also an important force in terrestrial snake coexistence. Furthermore, we suggest that snakes on the island coexist through subtle differences along the three major niche axes, and that conservation of a variety of niches, rather than a single niche, will increase the species diversity of local snakes.
... For example, it is estimated that for every oil well, 100 ha of road is constructed ( Riley et al. 2012 ), and some of the ecological effects of these roads may extend up to 1 km ( Benítez-López et al. 2010 ). Roads cut through otherwise intact grassland landscapes and can act as corridors for invasive vegetation ( Forman and Alexander 1998 ;Gelbard and Belnap 2003 ), alter predator communities ( Frey and Conover 2006 ;Glass and Eichholz 2022 ), and create movement barriers for grassland bird species ( Londe et al. 2022 ). Numerous studies of grassland birds have documented not only avoidance of roads ( Sutter et al. 20 0 0 ;Sliwinski and Koper 2012 ;Wellicome et al. 2014 ;Thompson et al. 2015 ) but also depressed reproductive success and increased parasitism by brown-headed cowbirds (Molothrus ater) with proximity to roads ( Miller et al. 1998 ;Bernath-Plaisted et al. 2017 ;Daniel and Koper 2019 ). ...
Article
Full-text available
North America's grassland birds remain in crisis despite decades of conservation effort s. This review provides an overview of factors contributing to these declines, as well as strategies and resources available to a diversity of stakeholders to help conserve grassland bird communities with an emphasis on the Great Plains-a grassland region of global ecological significance and a habitat stronghold for grassland birds. Grassland bird declines are driven by historical and continuing threats across the full annual cycle including grassland habitat loss, agriculture intensification, woody encroachment, and disruption of fire and grazing regimes. More recently, energy development activities, the use of neonicotinoid pesticides, and anthropogenic climate change have emerged as additional threats. While threats to grassland birds are numerous and often synergistic, possibilities for conservation are also diverse and multifaceted. Land set-aside programs, incentives and voluntary practices for producers, improved environmental management by energy and utility companies, and policy and regulation can all contribute to the conservation of these unique species. We suggest that future grassland bird research should focus on poorly studied aspects of the annual cycle, such as overwinter survival and habitat use, and the migratory period, which remains completely unexplored for many species. Filling these knowledge gaps may facilitate more sophisticated population modeling that can identify limiting factors and more effectively guide investment in conservation.
... In our study area, such species include the North American Racer (Coluber constrictor), Prairie Kingsnake (Lampropeltis calligaster), Black Kingsnake (Lampropeltis nigra), Common Garter Snake (Thamnophis sirtalis), and Black Ratsnake (Pantherophis obsoletus). We calculated a relative abundance index of snakes for each site by dividing the number of snakes encountered (all nest predator species combined) by the number of boards checked (Klug et al., 2009;Glass and Eichholz, 2022). ...
Article
Full-text available
Due to consistent population declines across the continent, grassland birds have become a guild of high conservation and management interest. Despite a large number of studies investigating grassland bird habitat associations, we know relatively little about the mechanisms through which habitat characteristics may impact grassland birds, as these mechanisms are often assumed rather than directly tested. For this study, we estimated whether the effects of habitat structure on breeding Field Sparrows are mediated through changes in predator (snake and raccoon) abundance, alternative prey availability, or arthropod biomass using structural equation models. We found no evidence of nest survival or nest density of Field Sparrows being directly influenced by nest predator abundance, alternative prey, or arthropod biomass, although habitat characteristics associated with increased nest survival were also associated with greater arthropod biomass and reduced predator abundance. We suggest that habitat structure in our study area primarily impacts breeding Field Sparrows through direct means, such as influencing nest concealment or foraging efficiency. Our results also suggest that nest success and nest density are decoupled in our study area, so Field Sparrows may be preferentially selecting nest sites with structural characteristics that do not increase nest survival. Ultimately, our findings from this study indicate that while predator avoidance and food provisioning likely play an important role in determining nest survival for grassland birds, predator abundance and arthropod biomass may not necessarily predict predation risk and foraging efficiency to the extent that is often assumed.
Article
Full-text available
Extensive habitat loss and degradation have transformed the grasslands of North America into one of the continent’s most vulnerable ecosystems and have greatly imperiled the plants and animals that depend on them. Small mammal communities play a pivotal role in maintaining the health and function of grasslands, and managing small mammal populations is an important, though overlooked, part of grassland restoration and management. We identified habitat variables in restored tallgrass prairie patches that were most strongly associated with variation in small mammal abundance and community composition, with the goal of aiding management efforts to manipulate small mammal populations. We found that small mammal abundance at the local vegetation scale was negatively related both to litter depth and plant diversity. At the landscape scale, small mammal abundance was positively associated with the amount of water surrounding a patch, and negatively associated with the amount of grassland surrounding a patch. Variation in small mammal community composition largely was governed by differences in habitat structure at the landscape scale, rather than differences in vegetation structure at the local scale. We suggested that managers interested in influencing small mammal abundance in grasslands employ tools such as prescribed fire to decrease litter depth if increased small mammal abundance is desired, or increase plant diversity by sowing a high diversity of seeds to depress small mammal abundance.
Article
Full-text available
Toxin-resistant predators may suffer costs from eating chemically defended prey and do not feed exclusively on toxic prey. Common Gartersnakes (Thamnophis sirtalis (Linnaeus, 1758)) have been considered the drivers of an evolutionary arms race with highly toxic newts (genus Taricha Gray, 1850), which they consume with few or no deleterious effects. However, how frequently newts are consumed in nature is less clear. To address this question, we investigated the diets of Th. sirtalis at a site in central Oregon where snakes have high levels of resistance and newts have high levels of tetrodotoxin in the skin. Because snake diets are difficult to quantify using traditional means, we used stable isotopes to estimate the proportion of Th. sirtalis diets made up of newts. Our estimate for the proportion of Th. sirtalis diet made up of Rough-skinned Newts (Taricha granulosa (Skilton, 1849)) at this site is 3.2%. Mole Salamanders (genus Ambystoma Tschudi, 1838) were predicted to be the most important prey, followed by slugs, chorus frogs, and mice, with a very minor role for earthworms. Our results demonstrate that even though Th. sirtalis are physiologically capable of consuming toxic prey, they do not often do so. Generalist predators can be exposed to very strong selection from, and exert reciprocal selection on even rarely eaten, chemically defended prey.
Article
Full-text available
Grassland ecosystems are among the most threatened biomes, and their restoration has become common in nature conservation. Yet restoration is rarely applied specifically for reptiles, which are among the most threatened vertebrates. The Meadow Viper (Vipera ursinii) has become extinct in most of lowland Europe, and an endangered subspecies (Vipera ursinii rakosiensis) has been a target of habitat restoration and captive breeding in Hungary since 2004. We quantified vegetation properties and the density of reptiles that either spontaneously colonised (three species) or were reintroduced (V. ursinii) in a grassland restored specifically for this purpose. We used a fine-scale survey to estimate the cover, and compositional and vertical diversity of the vegetation. We characterised sampling units along three habitat gradients: wetness, openness and grass tussock size. Model selection based on data from replicated counts showed that Green Lizards (Lacerta viridis) responded positively to vegetation cover and negatively to tussock area and height, and number of burrows. The Sand Lizard (Lacerta agilis) responded positively to vegetation cover, vertical diversity and wetness, and negatively to openness. The Balkan Wall Lizard (Podarcis tauricus) responded positively to tussock height and negatively to vegetation cover. Finally, V. ursinii responded positively to vegetation cover and tussock height, and negatively to compositional diversity. Our results show the general importance of structural and compositional diversity of vegetation to reptiles. These results suggest that adaptive management should focus on increasing the total cover (for lizards) and the structural diversity of vegetation (for each species) to benefit reptiles in restored grasslands.
Article
Full-text available
Staggering decline of bird populations Because birds are conspicuous and easy to identify and count, reliable records of their occurrence have been gathered over many decades in many parts of the world. Drawing on such data for North America, Rosenberg et al. report wide-spread population declines of birds over the past half-century, resulting in the cumulative loss of billions of breeding individuals across a wide range of species and habitats. They show that declines are not restricted to rare and threatened species—those once considered common and wide-spread are also diminished. These results have major implications for ecosystem integrity, the conservation of wildlife more broadly, and policies associated with the protection of birds and native ecosystems on which they depend. Science , this issue p. 120
Article
Full-text available
Streamside management zones (i.e., riparian buffers; SMZs) are commonly implemented within managed forests to protect water quality but may also provide habitat for riparian‐associated wildlife. Yet, little research has rigorously addressed the value of SMZs for wildlife, particularly for cryptic species such as amphibians and reptiles. Previous studies of herpetofauna within SMZs have focused on one or a few stream‐associated species, and questions remain regarding variation among species or guilds and what role SMZs serve toward conservation of herpetofaunal diversity in managed forests. However, recent statistical advances have improved our ability to analyze large multi‐species presence–absence data sets, accounting for low detection rates typical for some herpetofaunal species. Our study represents an extensive landscape‐scale examination of herpetofaunal communities within SMZs using a multi‐species occupancy approach. We conducted four replicate surveys at 102 headwater streams, spanning a gradient of SMZ widths and adjacent forest stand ages, within the Ouachita Mountains, Arkansas, USA. We used a hierarchical Bayesian community occupancy model to estimate species richness and species‐specific occupancy responses to SMZ and overstory characteristics, accounting for variation in occupancy and detection attributable to site and sampling covariates. We documented high richness (37 species) within SMZs. Across the herpetofaunal community, occupancy and species richness were consistently positively associated with SMZ width, with maximum predicted richness of 30 species occurring at sites with buffers extending 51 m on either side of the stream. However, we documented considerable variation among groups and species within groups, underscoring the potential for different responses to forest management among taxa. Reptile predicted richness increased more rapidly up to SMZs of ~35 m, whereas maximum salamander predicted richness was not seen until an SMZ width of 55 m. Estimated salamander richness was highest within SMZs embedded in mature managed pine stands and was higher in SMZs comprised of a deciduous or mixed overstory vs. a pine overstory. Compared to salamanders, more anuran species showed high mean estimated occupancy (>75%) at narrower SMZs (<30 m). Collectively, our results indicate that SMZs surrounding small first‐order streams in intensively managed forests not only protect water quality, but also can support diverse amphibian and reptile communities.
Article
Full-text available
• Simultaneous reductions in herbivore abundance and increases in nitrogen deposition have led to radical shifts in plant communities worldwide. While the individual impacts of these human‐caused disturbances are apparent, few studies manipulate both herbivory and N, nor differentiate among herbivore guilds, to understand contingencies in the ability of these drivers to affect producer diversity and productivity. As such, understanding how the main and combined effects of increasing soil N with declining herbivores may influence plant community structure and function is critical to better understand the future of grassland ecosystems under multiple global change drivers. • In this study, we asked: (a) What are the main effects of small mammal herbivores, invertebrate herbivores, and soil N on plant community structure and function? and (b) Are the effects of invertebrate herbivores and soil N on plant community structure and function contingent on small mammal herbivory? We used a nested design, with invertebrate and soil N treatments nested within small mammal manipulations in an existing tallgrass prairie. We measured plant community structure by quantifying plant richness, evenness, diversity, and composition across two full growing seasons. We also recorded total aboveground biomass to quantify grassland productivity. • We found that small mammal herbivores strongly shaped plant diversity, species composition, and productivity. Small mammal herbivores also mediated the effects of soil N and invertebrate herbivores on grassland community structure, but not composition or productivity. Small mammal reduction lowered plant species richness while increasing aboveground biomass and altering compositional similarity. Invertebrate herbivores, in the presence of small mammals, promoted plant dominance by reducing evenness without altering compositional similarity. Additionally, soil nitrogen addition reduced plant richness, but only when small mammals were reduced, and no effects were detected on compositional similarity or productivity. • Our findings provide further evidence that temperate grasslands can be strongly influenced by consumers, and that consumers mediate the effects of resources as well as other consumer guilds on producer evenness and richness. Taken together, we provide evidence of strong contingencies in the drivers of grassland structure, with small mammals directly altering plant diversity as well as mediating the effects of soil nitrogen and invertebrate herbivory on plant richness and evenness. Therefore, we suggest it is imperative to consider how consumer guilds and resource types may interact to shape grassland plant communities.
Article
Habitat loss is the greatest contributor to the decline of species globally. To prioritize protection of imperiled species, it is important to examine habitat use at multiple spatial scales because the availability of different resources and habitat features is scale dependent. We conducted a radio‐telemetry study in the Long Point region of Ontario, Canada, in 2009 and 2010 to examine habitat selection at multiple spatial scales by eastern hog‐nosed snakes (Heterodon platirhinos), a species at risk in Canada. We documented the habitat composition of home ranges compared to the surrounding landscape, the selection of locations within home ranges based on classified satellite imagery, and the use of microhabitat features based on site characterization in the field. At the scale of the home ranges, hog‐nosed snakes avoided areas of agriculture and selected sand barrens. Within home ranges, hog‐nosed snakes selectively used areas altered by humans (e.g., residential sites, openings in tree plantations). Microhabitats used by hog‐nosed snakes had more woody debris, logs, and lower vegetative coverage than adjoining random sites. Because hog‐nosed snakes prefer open areas and require sandy soils for nesting, management efforts should focus on the conservation and maintenance of sand barrens and patches of early successional forest. © 2021 The Wildlife Society. Eastern hog‐nosed snakes select sand dunes for nesting but use early‐successional forests and anthropogenic lands for the remainder of their active season. Conservation planning for this species should focus on protecting and maintaining open land cover types.
Article
Understanding faunal responses to habitat restoration is important in assessing restoration success. We investigated occupancy and abundance of snakes at Nachusa Grasslands, a large‐scale grassland restoration in the midwestern United States. Using artificial cover objects, we sampled within a chronosequence of 12 units converted from row‐crop agriculture 2‐25 years before the start of our study. Recaptures of marked snakes revealed that movement distances differed among species in accordance with differences in body size, being least in Dekay's Brownsnakes, intermediate in the Plains Gartersnakes and Common Gartersnakes, and greatest in Eastern Foxsnakes. Consistent with this result, occupancy increased with restoration age in Dekay's Brownsnakes but was unrelated to restoration age in the three larger, more mobile species. Similarly, abundance increased with restoration age in Dekay's Brownsnake but was unrelated to restoration age in other species. The Smooth Greensnake, another small‐bodied snake with limited mobility, and an Illinois species of greatest conservation need, was not detected at Nachusa Grasslands. Given detection probabilities observed during a parallel study at a nearby large grassland‐dominated preserve, we infer that the Smooth Greensnake is truly absent from Nachusa Grasslands. Taken together, our results demonstrate that establishment of faunal components following restoration may be time‐dependent with more sedentary species colonizing only slowly (e.g., Dekay's Brownsnakes) or not at all (e.g., Smooth Greensnakes). These results emphasize the need to clearly identify faunal restoration goals and the means to achieve them. This article is protected by copyright. All rights reserved.