Revisiting old data to answer modern conservation questions: Population modeling of two species in Kingsnakes, Lampropeltis sp. in Kansas

Page created by Charlie Vargas
 
CONTINUE READING
Revisiting old data to answer modern conservation questions: Population modeling of two species in Kingsnakes, Lampropeltis sp. in Kansas
Transactions of the Kansas                                                         Vol. 123, no. 1-2
Academy of Science                                                                 p. 225-233 (2020)

Revisiting old data to answer modern conservation questions:
Population modeling of two species in Kingsnakes, Lampropeltis sp.
in Kansas
J. Daren Riedle1 and George R. Pisani2
    1. Kansas Department of Wildlife, Parks, and Tourism, Pratt, Kansas
       daren.riedle@ks.gov
    2. Kansas Biological Survey, Lawrence, Kansas gpisani@ku.edu

     There is increasing interest on the effects of take for the pet trade on reptiles in Kansas,
     although little data is available to elucidate possible impacts. The Kansas Department of
     Wildlife, Parks, and Tourism and the Kansas Biological Survey recently partnered on a
     project to digitize Henry Fitch’s nearly 60 years of data collection on Northeast Kansas
     snake populations. We utilized the mark-recapture data for two species of snakes, the
     Prairie Kingsnake Lampropeltis calligaster and the Milk Snake L. triangulum to build
     population models and identify life history stages most vulnerable to take. Population
     growth rates for L. calligaster were most sensitive to changes in both juvenile and
     adult female survivorship. Lampropeltis triangulum was most sensitive to take of adult
     females. Sustained take over 20 years exceeding 30% in L. triangulum resulted in
     population extirpation based on population viability models developed here.

     Keywords: Collections, Milk Snake, pet trade, Prairie Kingsnake, reptiles, survivorship,
     viability

Introduction                                         data on snakes collected by Henry Fitch at
                                                     his study site in northeastern Kansas. The
The commercial removal of reptiles from              digitization of nearly 60 years of field data by
wild populations for use as pets is thought          Henry Fitch (as partially summarized in Fitch
to affect reptile populations globally. A            1999) has provided the opportunity to examine
study of exports of reptiles from the United         the demographic vital rates of two species of
States showed that exports for all reptiles          kingsnakes, a group that has some commercial
increased from 1996-2012 (Robinson et al.            relevance within the pet trade. While there
2015). Unfortunately, evidence linking take          is little tracking of native snakes in the pet
of reptiles to population declines is lacking as     trade, snakes within the genus Lampropeltis,
long-term studies on many species have not           particularly the Common Kingsnake, L. getula
been conducted (Todd, Willson and Gibbons            and Milk Snake L. triangulum make up a
2010). The lack of studies for some groups           portion of the trade in reptiles in the United
such as snakes are exacerbated by the difficult      States (Fitzgerald et al. 2004; Enge 2005;
nature of studying them in the field. Snakes         Prestridge, Fitzgerald and Hibbitts 2011).
are typically thought of as poor subjects for
study due to low recapture rates and high            Although not overly abundant on his research
variability in density (Parker and Plummer           sites, Fitch did capture multiple individuals
1987; Vitt 1987). With an increasing interest        of L. triangulum and the Prairie Kingsnake
in the effects of take on native amphibians and      L. calligaster. Both species are fairly cryptic
reptiles in Kansas by resource management            and not always easily observable. Richardson,
agencies, the authors revisited the long-term        Weatherhead, and Brawn (2006) noted that
Revisiting old data to answer modern conservation questions: Population modeling of two species in Kingsnakes, Lampropeltis sp. in Kansas
226                                                                                   Riedle and Pisani

individuals of L. calligaster spent as much as         population minus the influence of fecundity and
73% of their time underground. Lampropeltis            is here on referred to as Pradel’s λ. Pradel’s λ
triangulum also spends considerable time               was calculated in Program MARK.
underground, often showing high site fidelity
to exposed rock outcrops that are used for             Using apparent survivorship values calculated
thermoregulation (Dyrkacz 1977; Row and                in Program MARK, and reproductive data
Blouin-Demers 2006a,b). Despite their cryptic          summarized within Fitch (1999), we developed
nature, Fitch made 331 captures of L. calligaster      population viability models in Program Vortex
and 279 captures of L. triangulum between              (Lacy and Pollack 2014). Age at first reproduction
1949-1997, primarily using funnel traps and            for L. calligaster was 3 yrs and mean clutch size
drift fence arrays (Fitch 1999). An additional 71      was 9 (range 6-13). Age at first reproduction for
captures of L. calligaster and 17 captures of L.       L. triangulum was 4 yrs and mean clutch size
triangulum were made between 1997-2003, after          was 7 (range 5-9). Age and clutch size values
the publication of Fitch (1999). While the overall     were used to calculate baseline deterministic rates
captures for both species appear low for a 54 yr       for each species within Vortex. Deterministic
time span, they represent the most robust data         rates used here include population growth rate
set for either species in Kansas. Our objective        (λ), net reproductive rate (average number of
was to subject each updated data set to a more         age class zero offspring produced by an average
robust analysis of demographic vital rates and         newborn during its life time; R0) and mean
identify those life history traits most sensitive to   generation time (Τ). We then hand calculated the
perturbation, such as take. Using these models,        reproductive value, or to what extent do members
natural resource managers can best extrapolate         of a given age group contribute to the next
how take might impact populations.                     generation between birth and death (Fisher 1930).
                                                       Reproductive value is defined as:
Materials and Methods

Basic population models focus on the interaction
between birth rates and mortality, so we focused
on female survivorship and fecundity when
constructing viability models for the two species
of Lampropeltis (Pianka 1994). Demographic             The term lt/lx represents the probability of
vital rates were constructed from 402 captures         living from age x to age t, and mt is the average
of 286 individuals of L. calligaster and 296           reproductive success of an individual at age t
captures of 198 individuals of L. triangulum.          (Pianka 1994).
Apparent survivorship (Φ) and recapture
probability (p) were calculated using Cormack-         We then ran a series of sensitivity tests for 3
Jolly-Seber models in Program MARK (White              parameters: percent of females breeding/year,
and Burnham 1999). We generated models to              juvenile mortality, and adult female mortality.
test whether Φ and p were best explained by            Sensitivity tests within Vortex work by creating
time, group (male, female), or were independent        a series of scenarios in which each parameter
of group level influences. Model selection was         can be varied. Scenarios were set so that values
based on corrected Akaike Information Criterion        for each parameter varied from 0-100%. For
(AICc) values, with lower values denoting              each scenario a value is randomly chosen for
greater parsimony (Burnham and Anderson                each of the three parameters from across the
2002). We then reversed the encounter rates            range specified. Then the mean value of the
to determine the probability of an individual          output variable was graphed in what is referred
entering the population (Pradel 1996). Pradel          to as spider plots by Vortex. All models were
models estimate the realized growth rate (λ) of a      based on a simulated timespan of twenty years.
Revisiting old data to answer modern conservation questions: Population modeling of two species in Kingsnakes, Lampropeltis sp. in Kansas
Transactions of the Kansas Academy of Science 123(1-2), 2020                                     227

Table 1. Demographic parameters ± standard error for Lampropeltis calligaster and L. triangulum
in northeast Kansas. Parameters include weighted mean apparent survivorship (Φ) and recapture
probability (p) based on AICc weights, Pradel’s λ and actual λ, reproductive replacement rate (R0)
and mean generation time (Τ).

Results                                             slight when compared to models with group
                                                    interactions, but weighted averages and standard
Apparent survivorship for L. calligaster were       errors of the most parsimonious model overlap
similar between sexes (Table 1). The most           (Table 1), suggesting the top model truly
parsimonious model explaining apparent              reflects apparent survivorship and recapture
survivorship in L. calligaster was constant         probabilities. Fitch did have enough recaptures
survivorship and recapture probability              of pre-reproductive aged L. calligaster to
independent of group or time interactions           calculate apparent survival, which was calculated
(Table 2). Variation between AICc scores were       independent of adult models. (Table 1).

Table 2. Comparison of Cormack-Jolly-Seber models for apparent annual survival (Φ) and recapture
rates (p) for Lampropeltis calligaster in northeastern Kansas. Models differ in whether Φ and p are
assumed to be constant (.), fully time dependent (t), or differ between sexes (g), and whether there
are interactions (*) among these factors. The model with the lowest AICc score is considered the
most parsimonious.
228                                                                               Riedle and Pisani

Table 3. Comparison of Cormack-Jolly-Seber models for apparent annual survival (Φ) and
recapture rates (p) for Lampropeltis triangulum in northeastern Kansas. Models differ in whether
Φ and p are assumed to be constant (.), fully time dependent (t), or differ between sexes (g), and
whether there are interactions (*) among these factors. The model with the lowest AICc score is
considered the most parsimonious.

Male L. triangulum had higher apparent              deterministic values also suggest stable
survivorship than females (Table 1), and the        populations for both species. Both true λ and
most parsimonious model states survivorship         the population replacement rate (R0) are near a
varied by group, but recapture probability          value of 1. Due to an earlier age of maturity in
did not vary by group or time (Table 3).            L. calligaster mean generation time was 1.5 yrs
Again, the differences between models was           shorter than for L. triangulum (Table 1).
slight, but weighted averages plus standard
errors of apparent survivorship between sexes       Based on the reproductive values provided
show little overlap, suggesting differences in      by Fitch (1999) and the deterministic rates
survivorship between sexes may be real. Since       provided here, populations for both species of
recaptures of pre-reproductive L. triangulum        Lampropeltis within Fitch’s study site appear
were lacking, apparent survival values for          to be stable. When we ran simulations varying
L. calligaster were substituted for use in          deterministic rates we observed differences in
population viability models.                        how variation influenced population growth
                                                    rates. There were strong interactions between
When considering population growth rates,           female and juvenile mortality and the percent
a value of λ=1 denotes a stable population,         of females breeding in the L. calligaster models
while higher or lower values signify increasing     (Fig. 1). Fitch (1999) noted that dependent on
or decreasing populations. Pradel’s λ, or           available resources, roughly 1/3 of the female
the number of new animals entering the              population may not reproduce in a given
population, suggest stable populations              year. This temporal reduction in reproductive
for both species in this study (Table 1).           females increases the importance of juveniles
When survivorship values are coupled with           transitioning into reproductive age classes.
reproductive rates within Program Vortex,
Transactions of the Kansas Academy of Science 123(1-2), 2020                                       229

Figure 1. Sensitivity plots showing the influence of percent change in measured parameters on
population size for Lampropeltis calligaster in northeast Kansas.

The L. triangulum models placed more                Discussion
importance on female mortality followed by
juvenile mortality (Fig. 2). Considering that       While the populations of the two species of
L. triangulum reach sexual maturity a year          Lampropeltis on Fitch’s study site in northeast
later and have slightly smaller clutch sizes        Kansas appear to be stable, adult survivorship
as compared to L. calligaster, maintaining          for both species was low based on mark-
reproductive age females is important for           recapture data. Low recapture rates may have
population maintenance. Sustained mortality         been influenced by both species cryptic nature
or removal of adult females exceeding 30%           and the fact that Fitch primarily used funnel
would result in population extinction.              traps as his sampling method. Even with
                                                    those considerations, the duration of Fitch’s
In all population viability models, particularly    study provided a fairly robust data set for both
those for L. triangulum, female mortality           species, including data on known age cohorts,
played an important role in population              which allowed for more accurate estimation of
persistence. The reproductive value of females      timing of important life history stages.
(the probability of future offspring offset with
age-specific mortality rates) increased with age    For both species, population persistence was
peaking at 6 yrs for L. calligaster and 7 yrs for   dependent upon the presence of breeding age
L. triangulum (Fig. 3).                             females within the population. The peak in
                                                    reproductive value calculated here coincides
230                                                                                 Riedle and Pisani

Figure 2. Sensitivity plots showing the influence of percent change in measured parameters on
population size for Lampropeltis triangulum in northeast Kansas.

with the age at which females reach mean adult        variation between species. Timber Rattlesnakes
size and begin producing larger clutches of eggs      Crotalus horridus, is considered a typical
before annual mortality rates begin to offset         k-adapted species with late age to maturity,
reproductive advantages. It appears that high         and long reproductive life cycles coupled with
annual mortality coupled with not breeding            a low number of life-time reproductive efforts
every year (L. calligaster) or delaying sexual        (Brown 2016). On the other end of the spectrum
maturity (L. triangulum) have the greatest            the Northern Water Snake Nerodia sipedon
influence on population growth rates. Increasing      grows fast and reach sexual maturity by 3y of
the number of female snakes’ breeding in a            age (Brown and Weatherhead 1999). Mortality
given year may offset low adult survivorship          greatly increases for N. sipedon after reaching
in L. calligaster but is not as likely to influence   maturity. The much smaller bodied Ring-necked
population growth rates in L. triangulum.             Snake also exhibits a young age to maturity,
Ensuring juvenile snakes transition into              with high mortality rates for pre-reproductive
reproductive age classes for both species is as       individuals and higher survivorship for adults
important as maintaining female survivorship.         (Riedle 2014). The models presented within this
                                                      manuscript demonstrated variation between two
While Fitch’s unprecedented work on snake             species of sympatric, congeneric kingsnakes.
communities took place within the confines            Both species of kingsnakes studied here are
of a protected natural area, the demographic          wide ranging, so it is expected that vital rates
information it provides is an important baseline      will vary greatly across their distributions. For
for consideration of the possible impacts of take.    the transcontinental rat snake Pantherophis
A general rule is variation within reproductive       obsoletus, latitude appears to play a significant
constraints influences population growth rates,       role in growth rates, age at maturity, and life
and it is important to elucidate the sources of       spans (Blouin-Demers et al. 2002).
Transactions of the Kansas Academy of Science 123(1-2), 2020                                   231

Figure 3. Reproductive value of female Lampropeltis calligaster and L. triangulum from ages 0-10
in northeast Kansas.

Outside of a few common species,                   the state (Collins, Collins and Taggart 2010).
demographic vital rates are lacking to assess      Anecdotal information from law enforcement
possible impacts of trade nor are accurate         at the Kansas Department of Wildlife, Parks,
data available on the trade in many species.       and Tourism though suggests that L. triangulum
Most data have focused on international            is a target species for many reptile enthusiasts
exports while data on local trade is nonexistent   visiting Kansas. Most of their habitat is
(Schlaepfer, Hoover and Dodd 2005). The            found primarily on private lands within the
near extirpation of the insular snake Boa          state, which protects many populations from
constrictor imperator due to the pet trade has     sustained impact. Based on the demographic
been documented (Wilson and Cruz-Diaz 1993;        information presented here though, increased
Reed et al. 2007). Webb et al. (2002) noted        take on the minimal public lands available
dramatic declines in the Australian broad-         within the species range could or possibly
headed snake Hoplocephalus bungaroides             has resulted in the creation of localized sink
due to collection for pets. These declines         populations by reducing breeding females
were disparate between sexes, where 85% of         and thus future offspring at specific sites.
the females disappeared from the population        While neither species may be of immediate
but male declines were much less severe.           conservation concern, long-term data sets, such
The authors state collectors were selectively      as those compiled by Fitch, provide important
harvesting female snakes.                          baseline data for demographic analyses and
                                                   informing future management decisions. We
The two focal species of Lampropeltis              suggest using the baseline demographics
discussed in this manuscript are common            described in this paper as a starting point to
throughout most of the state of Kansas and are     better understand population trends and trade of
most likely not in any danger of extinction in     Lampropeltis in the state.
232                                                                              Riedle and Pisani

Acknowledgments                                    Fisher, R.A. 1930. The genetical theory of
                                                       natural selection. Clarendon Press, Oxford,
We are grateful to Kansas Dept. of Wildlife,           England.
Parks, and Tourism Chickadee Checkoff              Fitch, H.S. 1999. A Kansas snake community:
Program for financial support of the Fitch             composition and changes over 50 years.
Snake Data Archive project. GRP thanks                 Krieger Publishing, Malabar, Florida. 165 pp.
Kansas Biological Survey (KBS) for providing       Fitzgerald, L.A., Painter, C.W., Reuter, A. and
work space and computational support for               Hoover, C. 2004. Collection, trade, and
an ongoing variety of research efforts. KBS            regulation of reptiles and amphibians of the
additionally houses the Fitch Data Archive             Chihuahuan Desert Ecoregion. TRAFFIC
and is custodian of the electronic databases           North America. Washington D.C.: World
constructed by GRP. We also acknowledge                Wildlife Fund.
the exceptional efforts of Henry S. Fitch          Lacy, R.C. and Pollak, J.P. 2014. Vortex: a
(deceased) in amassing a phenomenal record of          stochastic simulation of the extinction
ecological data on eastern Kansas snakes.              process. Version 10.0. Chicago Zoological
                                                       Society, Brookfield, Illinois, USA.
Literature Cited                                   Parker, W.S. and Plummer, M.V. 1987.
                                                       Population ecology. Pp. 253–301 In Snakes:
Blouin-Demers, G., K.A. Prior and P.J.                 Ecology and Evolutionary Biology. Seigel,
   Weatherhead. 2002. Comparative                      R.A., J.T. Collins, and S.S. Novak (Eds.).
   demography of Black Rat Snakes (Elaphe              Macmillan, New York, New York, USA.
   obsoleta) in Ontario and Maryland. Journal      Pianka, E.R. 1994. Evolutionary Ecology.
   of Zoology 256:1–10.                                Harper Collins, New York, New York, USA.
Brown, G.P. and Weatherhead, P.J. 1999.            Pradel, R. 1996. Utilization of capture-mark-
   Demography and sexual size dimorphism in            recapture for the study of recruitment and
   Northern Water Snakes, Nerodia sipedon.             population growth rate. Biometrics 52:703-709.
   Canadian Journal of Zoology 77:1358–1366.       Prestridge, H.L., Fitzgerald, L.E. and Hibbitts,
Brown, W.S. 2016. Lifetime reproduction                T.J. 2011. Trade in non-native amphibians
   in a northern metapopulation of timber              and reptiles in Texas: lessons for better
   rattlesnakes (Crotalus horridus).                   monitoring and implications for species
   Herpetologica 72:331-342.                           introduction. Herpetological Conservation
Burnham, K.P. and Anderson, D.R. 2002.                 and Biology 6:324-339.
   Model selection and multimodel inference.       Reed, R.N., Boback, S.M., Montgomery, C.E.
   Springer-Verlag, New York, 488 pp.                  Green, S., Stevens, Z. and Watson, D. 2007.
Collins, J.T., S.L. Collins and Taggart, T.T.          Ecology and conservation of an exploited
   2010. Amphibians, Reptiles, and Turtles in          insular population of Boa constrictor
   Kansas. Eagle Mountain Publishing, Eagle            (Squamata: Boidae) in the Cayos Cochinos,
   Mountain, UT, 318pp.                                Honduras. pp. 389-403 in Biology of the
Dyrkacz, S. 1977. The natural history of the           Boas and Pythons. Henderson, R.W. and R.
   eastern milk snake (Reptilia, Serpentes,            Powell (eds.). Eagle Mountain Publishing,
   Colubridae) in a disturbed environment.             Eagle Mountain, UT.
   Journal of Herpetology 11:155-159.              Richardson, M.L., Weatherhead, P.J. and
Enge, K.M. 2005. Commercial harvest of                 Brawn, J.D. 2006. Habitat use and activity
   amphibians and reptiles in Florida for the          of prairie kingsnakes (Lampropeltis
   pet trade. pp. 198-211 in Amphibians and            calligaster calligaster) in Illinois. Journal
   reptiles: status and conservation in Florida.       of Herpetology 40:423-428.
   Krieger Publishing, Malabar, Florida.
Transactions of the Kansas Academy of Science 123(1-2), 2020                                  233

Riedle, J.D. 2014. Demography of an urban           Todd, R.D., Wilson, J.D. and Gibbons,
   population of ring-necked snakes (Diadophis          J.W. 2010. The global status of reptiles
   punctatus) in Missouri. Herpetological               and causes of their decline. pp. 47-67
   Conservation and Biology 9:278-284.                  in Ecotoxicology of Amphibians and
Robinson, J.E., Griffiths, R.A., St. John, F.A.V.       Reptiles, Second Edition. Sparling D.W.,
   and Roberts, D.L. 2015. Dynamics of                  C.A. Bishop, S. Krest (eds.). CRC Press,
   the global trade in live reptiles: shifting          Pensacola, FL, USA.
   trends in production and consequences for        Vitt, L.J. 1987. Communities. pp. 335–365 in
   sustainability. Biological Conservation              Snakes: Ecology and Evolutionary Biology.
   184:42-50.                                           Seigel, R.A., J.T. Collins, and S.S. Novak
Row, J.R. and Blouin-Demers, G. 2006a.                  (eds.). Macmillan, New York, New York.
   Thermal quality influences habitat               Webb, J.K, Brook, B.W. and Shine, R. 2002.
   selection at multiple scales in milksnakes.          Collectors endanger Australia’s most
   Ecoscience 13:443-450.                               threatened snake, the broad-headed snake
Row, J.R. and Blouin-Demers, G. 2006b.                  Hoplocephalus bungaroides. Oryx 36:170-
   Kernels are not accurate estimators of               181.
   home-range size for herpetofauna. Copeia         White, G.C. and Burnham, K.P. 1999.
   2006:797-802.                                        Program MARK: survival estimation from
Schlaepfer, M.A., Hoover, C. and Dodd Jr.,              populations of marked individuals. Bird
   C.K. 2005. Challenges in evaluating the              Study 46(Suppl.): S120–139.
   impact of the trade in amphibians and            Wilson, L.D. and Cruz-Diaz, G. 1993. The
   reptiles on wild populations. BioScience             herpetofauna of the Cayos Cochinos,
   55:256-264.                                          Honduras. Herpetological Natural History
                                                        1:13-23.
You can also read