morphology, reproduction, habitat use, and hibernation of

150

Morphology reproduction habitat use and hibernation of Red-bellied Snake (Storeria occipitomaculata) near its northernrange limitNicholaS a caiRNS1 2 PaMela l RutheRfoRd1 and dRew J hoySak1

1Department of Biology Brandon University 270 18th Street Brandon Manitoba R7A 6A9 Canada2Department of Biology Queenrsquos University 116 Barrie Street Kingston Ontario K7L 3N6 CanadaCorresponding author rutherfordpbrandonuca

Cairns NA PL Rutherford and DJ Hoysak 2018 Morphology reproduction habitat use and hibernation of Red-belliedSnake (Storeria occipitomaculata) near its northern range limit Canadian Field-Naturalist 132(2) 150ndash162 httpsdoiorg1022621cfnv132i22054

Abstract Northern regions limit ectotherms to relatively short periods of feeding and breeding interrupted by long periods of inactivityThis may force cool-climate ectotherms into different ecological or demographic trade-offs than their southern conspecificsOur aim is to examine demography morphology reproduction habitat use and hibernation by populations of Red-belliedSnake (Storeria occipitomaculata) near their northern range limit This research was conducted in southwestern Manitobaand data on summer activity were collected from April to September 2007ndash2009 using coverboard and pedestrian surveysHibernation sites were monitored over three winters (2007ndash2008 2008ndash2009 and 2009ndash2010) and thermal profiles of Formicaant mounds were collected in 2008ndash2009 and 2009ndash2010 Females reached sexual maturity at a smaller size than most otherpopulations that have been reported but appear to have similar clutch sizes to the rest of the range The majority of adult femalescaptured at our summer sites were gravid (96) suggesting annual reproduction and activity patterns suggest fall breedingNear its northern range maxima this species appears to use relatively warm habitat have rapid reproduction and co-opt antmounds to survive in a difficult climatic environment Much remains unknown and future studies should further examine thevariation in size at maturity and the relationship between body size and clutch size In addition little is known about diet benefitsof fall mating use of open prairie habitats and late-season migration by S occipitomaculata Key words Red-bellied Snake Storeria occipitomaculata cool-climate temperate Manitoba thermoregulation brumation

copyThe Ottawa Field-Naturalistsrsquo Club (2018)

IntroductionRegions with extended cold seasons present chal-

lenges for many ectothermic species which is reflect-ed by the limited species diversity of snakes and othersquamates at higher latitudes (Kiester 1971) Theseregions have short summers and long cold winters thatlimit individuals to relatively short periods of feedingand breeding interrupted by long periods of forced in -activity (Gregory 2009) Because short active seasonscan limit the time available to allocate resources cold-climate individuals may have different ecological trade-offs than their lower latitude conspecifics (Gregory2009) As such aspects of growth (Blouin-Demers etal 2002) reproduction (Gregory 2009) and habitat use(Weatherhead et al 2012) can differ within speciesacross a range reflecting local selective pressures (Fitch1981)

Sexual size dimorphism (SSD) has been examinedin numerous snake species (reviewed in Shine 1994)Sexual size dimorphism is common but can vary geo-graphically with northern populations They are oftenless dimorphic than southern populations which is oftenattributed to phylogenetic conservatism (Shine 1994Cox et al 2007) Size at maturity can be conservedacross the range of most species but the age at whichan individual reaches mature size may be more plastic(Blouin-Demers et al 2002) Cold-climate reptiles are

expected to grow more slowly and have delayed matu-rity (Atkinson 1994 Berrigan and Charnov 1994) Thisis not the case for all species some populations are ableto compensate for shorter growing seasons and havesimilar or higher reproductive rates compared to south-ern conspecifics (Tuttle and Gregory 2012 2014)

The ability of a female to successfully reproduce inany given year is typically dependent on body condi-tion Most northern reptiles are viviparous with leci -thotrophic embryos requiring provisioning predevel-opment The time required to replenish energy storesmay delay subsequent reproduction and lead to less thanannual reproduction In some species yolk is supple-mented by limited placental nourishment allowinggravid snakes some control of provisioning during dev -elopment (King 1993) In these cases prolonged ges-tation would allow for larger or more neonates This hasa cost as prolonged gestation limits the ability of themother to feed that in an already short season may leadto tradeoffs in energy expenditure not required in lessstressful environments

In high-latitude regions many snakes are not alwaysable to maintain ideal body temperatures even duringthe active season (Blouin-Demers and Weatherhead2001) Reproductive requirements of females also con-tribute to differences in habitat use and temperatureselection between sexes Thermal requirements to grow

2018 CAIRNS et al RED-BELLIED SNAKE ECOLOGY IN MANITOBA 151

developing embryos may cause females to select dif-ferent habitats than males and alter their thermoregu-latory behaviour (Shine et al 2006) The cold-climatehypothesis predicts that warmer temperatures will beselected by gravid female squamates in cold climates toensure rapid embryo development (Tinkle and Gibbons1977) This may lead to increased use of more thermal-ly attractive open or edge habitat than in other parts of aspeciesrsquo range

Cold-climate reptiles can avoid inclement conditionsand limit energy expenditures through hibernation (Gre-gory 2009) Failure to select appropriate hibernaculaleads to death but selecting purely for overwinter survi -val also has associated costs (Gienger and Beck 2011)Time spent in hibernation subtracts from time spent for-aging therefore hibernacula are often chosen to extendthe length of the active season (Gienger and Beck 2011)Features often associated with successful hibernationsites allow for access to soil below the frost line stabil-ity in humidity and temperature and access to the watertable (Harvey and Weatherhead 2006a) The availabil -ity of suitable hibernacula is likely the most importantform of habitat selection for snakes at high latitudesand may limit the distribution of some species (Harveyand Weatherhead 2006b)

Red-bellied Snake (Storeria occipitomaculata) occu-pies most of eastern North America and is widely distri -buted in Canada reaching the northwestern edge of itsrange in eastern Saskatchewan (Ernst and Ernst 2003)Across its range there have been a number of studies ofthis species (eg Blanchard 1937 Lang 1969 Seml-itsch and Moran 1984 Willson and Dorcas 2004 Brownand Phillips 2012) but there has been little research atits northern range limit with the exception of two nat-ural history observations (Criddle 1937 Gregory 1977)Storeria occipitomaculata is a small-bodied crypticsnake that comes in several ground-shaded dorsal col -our morphs and has a bright reddish-orange ventralsurface They are viviparous likely with some limitedplacental provisioning (Blackburn et al 2009) Thisspecies like other members of the genus Storeria feedsprimarily on molluscs (Trapido 1944 Brown 1979Semlitsch and Moran 1984 Pisani and Busby 2011) al -though not exclusively (Ernst and Ernst 2003) Withfew species of slugs and snails native to the northwest-ern edge of its range local diet remains unknown Thisspecies is considered uncommon to rare throughoutmuch of its range (Ernst and Ernst 2003) but it can beabundant in suitable habitat (Lang 1971) Little is knownabout population densities for this species althoughmost studies have recorded low recapture rates (Blan-chard 1937 Gregory 1977 Semlitsch and Moran 1984)Typically S occipitomaculata inhabit thick grassyherbaceous and shrubby vegetation and also can befound in wetlands riparian areas and along forest edgesand in old field and prairie habitats (Wright and Wright1957 Cook 1984 Pisani and Busby 2011) as well asdrier grasslands habitats (Lang 1969 Brown and Phil -

lips 2012) Criddle (1937) and Lang (1969) both foundthis species to use abandoned (or nearly so) ant moundsof species of the genus Formica as overwintering siteswith the snakes often found hibernating at high den-sities

The objective of this study is to examine demogra-phy morphology reproduction habitat use and hiber-nation in populations of S occipitomaculata near theirnorthern range limit The stressful environment at highlatitudes may force these populations to make trade-offsin their growth reproduction and habitat use We askedthe following questions 1) What are the demographicsmorphology and reproductive traits of this populationof S occipitomaculata and 2) Under what conditionsdo they hibernate

MethodsStudy sites

Field work in summer habitat use took place in south -western Manitoba Canada at the following locations(Figure 1a) Spruce Woods Provincial Park (SWPP497108degN 992528degW) Assiniboine Corridor WildlifeManagementArea (ACWMA 496675degN 995614degW)Oak Lake (496644degN 1007133degW) and CanadianForces Base Shilo (CFB Shilo 497381degN 995183degW)A wide variety of habitats were investigated includingagricultural areas beaver ponds mixed forest andmixed-grass prairie Most sites included water andormixed grass prairie Investigation into hibernation wasconducted at the Souris River Bend Wildlife Manage-ment Area (SRBWMA 494883degN 998775degW Fig-ure 1b) Habitat at this area includes mixed forest andgrassland bordered by marsh agricultural fields andgravel roads Details on the habitat types are providedbelowDemography and morphometrics

We recorded the following measurements for allsnakes date time snout-vent length (SVL in mm) taillength (mm) head width (mm) head length (mm) mass(g) sex age and reproductive condition We classifiedall individuals into three age classes young-of-year(YOY) juvenile and adult Storeria occipitomaculatawere classified into age classes based on their SVLnatal scars and the minimum breeding SVL of fe -males YOY (lt110mm) juvenile (100ndash154 mm) andadult (gt154 mm) For the age-class analysis we onlyincluded the first capture for individuals that were re -captured multiple times within the same season

To determine reproductive condition two researchersindependently counted enlarged ovarian follicles bypalpation (Fitch 1987) and the average of their countswas recorded All captured individuals were scale-clipped for identification and released at their capturesite within 15 minutes of capture The activity period(27 April to 8 September) was classified by two-weekperiods numbered 1 through 10 No snakes in this studyreceived more than three clipped ventral scales

152 THE CANADIAN FIELD-NATURALIST Vol 132

fɪɢᴜʀᴇ 1 a Map of summer sites (grey circle) and hibernation sites (black triangle) surveyed for Red-bellied Snake (Storeriaoccipitomaculata) 2007ndash2010 in southwestern Manitoba b Map of hibernation study site at Souris River Bend Wildlife Manage-ment Area surveyed 2007ndash2010 showing den sites (black triangle) where traps were set and locations of individual captures(grey circle) outside of the traps Base map from World Imagery Accessed 7 October 2018 httpssupportesricomentechnical-article000012040


Summer activityField sites were surveyed every 2ndash3 weeks in 2007ndash

2009 from April to September of each year Surveyswere done at various times of day to compare dailyhabitat usage as the season progressed We used thesame search corridor each time a site was visited tostandardize search effort These routes travelled aroundor through potentially suitable habitat often followingbarriers such as creeks park paths or tree lines Ani-mals were captured by hand as they moved in the openor when located under cover objects both natural andartificial (eg plywood sheet metal)

We recorded the following environmental and loca-tion data at each capture site universal transverse mer-cator coordinates capture location (in the open or undercover cover type and cover thickness) and aspect(0ndash360o) Collection localities were recorded using aGarminreg GPS60 handheld global positioning system(GPS) receiver (Olathe Kansas USA WAAS enabledaccuracy lt3 m 95 typical) GPS data were used toclassify the capture locations according to land coverclasses defined by the Manitoba Land Initiative (2012)The following seven classes were used agriculturebogs coniferous forest deciduous forest mixedwoodforest grassland and roads (including trails) Assign-ments were made using ArcGIS version 10 (ESRI2011)Hibernation field data collection

This study took place from September 2007 to June2010 targetting spring and fall (from 12 April to 5 Junein 2008 from 6 May to 2 June in 2009 from 4 April to20 May in 2010 from 15 September to 4 October in2007 from 6 September to 25 October in 2008 and 20September to 4 October in 2009) No traps were used in2009 (see below for a further description of trapping)

Five abandoned ant mounds were monitored through-out the study (Figure 1b) We replaced two of themounds that were monitored in 2007 and 2008 withtwo new mounds in 2008 2009 and 2010 because themounds monitored in 2007 and 2008 were destroyedAll ant mounds were ringed with 40 cm tall aluminumflashing Three openings were cut in the aluminumflashing and a mesh funnel trap made of hardware clothwas placed in each In fall two traps were placed onthe inside of the flashing at every mound to captureindividuals entering the mounds and one trap wasplaced on the outside to capture individuals exiting themounds the reverse occurred in the spring trappingsessions Traps were checked every 24ndash48 h and allindividuals were removed from traps Visual searcheswere also conducted around the trap sites when trapswere checked All captured individuals were measured(as described above) given a unique scale-clip and re -leased at the capture site Individuals caught in trapswere released in their direction of travel

Temperature profiles of the ant mounds were createdusing iButtonreg thermal probes (San Jose CaliforniaUSA) attached to plastic landscaping rods iButtonsreg

were placed in Ziplocreg bags and attached to the polesusing duct tape They were deployed over two wintersfrom 27 September 2007 to 29 May 2008 and 20 Sep-tember 2008 to 3 June 2009 The iButtonsreg were set atfour distances from the surface 40 80 120 160 cm inthe 2007ndash2008 hibernation period and at 0 40 80120 cm in the 2008ndash2009 hibernation period Tempera-tures were recorded every 3 h In 2007ndash2008 all iBut-tonsreg deployed at 160 cm below the surface failed be -cause they were below the water table We did notde ploy any iButtonsreg at this depth in 2008ndash2009 In2008ndash2009 air temperature (15 m above the surface)was also recorded Statistical analysis

All data analyses were done in R 321 (R Develop-ment Core Team 2017) with α = 005 We report meansSE and ranges Data were log-transformed or nonpara-metric tests were used when assumptions of paramet-ric tests were not satisfied Unless otherwise indicatedadult females include both gravid and non-gravid indi-viduals Unpaired t-tests (package stats function ttest)were used to compare SVLs within and between sexesand those captured in the summer versus at the hiberna-tion site (spring and fall) Chi-square analyses (packagestats function chisqtest) were used to compare amongcategorical variables (see Table 1 for all comparisons)Linear models (package stats function lm) were used totest for morphological and reproductive relationships(see Table 2 for detailed models) For P values between005 and 01 power analyses were conducted (packagepwr) with reference power = 080

We calculated the mean temperature at each depthfor all five dens for each 3 h period within each year Insome cases we did not have temperatures for all depthsat all den sites due to failure of some of the iButtonsregIn 2007ndash2008 one iButtonreg failed at a den site at boththe 80 cm and 120 cm depths In 2008ndash2009 one iBut-tonreg failed at the 80 cm depth and three iButtonsregfailed at the 120 cm depth Air temperature and surfacetemperature (0 cm) were only recorded in 2008ndash2009

ResultsSummer demography morphometrics and habitat use

Over the three years we captured 88 individuals(Table S1) Most individuals were adults (n = 81 92)with only five juveniles and two YOY captured Therewas no significant difference in the numbers of cap-tures in each age class (YOY juvenile adult) amongthe three years (Table 1) and no significant difference inthe numbers of adult males and females captured amongyears (Table 1) Adult sex ratios varied during the ac -tive season (Table 1 Figure 2) with more males cap-tured in late summer (after 3 August)

Adult female S occipitomaculata tended to be longer(female 1842 plusmn 22 mm male 1753 plusmn 40 malefemale body size = 095 t313 = 194 P = 006) Sampleswere unequal (female n = 53 male n = 21) and powerwas low (050 package pwr function pwrt2ntest)


tᴀʙʟᴇ 1 Chi-square test statistics for comparisons among categories for summer sites and hibernation sites for Red-belliedSnake (Storeria occipitomaculata)

Category Comparison χ2 df PSUMMERage class (YOY J AD) year 404 4 0400adult sex year 402 2 0130adult sex activity (10 periods) 1844 9 0030activity (10 periods) land-cover use 6235 27 lt 0001adult sex land-cover use 134 3 0720HIBERNATIONage class (YOY J AD) spring versus fall 8628 2 lt 0001adult sex spring versus fall 017 1 0680species capture location (trap mound road and grassland) 2833 9 lt 0001direction of movement (entering

versus exiting) spring versus fall 040 1 0530

YOY = young-of-year J = juvenile AD = adult

tᴀʙʟᴇ 2 Linear models to test for the effects of snout-vent length (SVL) and sex of Red-bellied Snake (Storeria occipitomaculata)on tail length head width head length natural log-transformed mass and number of follicles Models were done separatelyfor summer sites (S) and the hibernation site (H) The final linear model tests for the effect of year and SVL on the number offollicles in adult females captured at summer sites

Model Site type F df P Adjusted r2

tail length = sex + SVL + sex times SVL S 395 369 0010 011 H 5134 339 lt 0001 078 sex S 683 169 0010 H 3854 139 lt 0001 SVL S 376 169 0060 H 11528 139 lt 0001 sex times SVL S 126 169 0260 H 020 139 0660head width = sex + SVL + sex times SVL S 425 369 0008 012 H 1618 339 lt 0001 052 sex S 319 169 0080 H 116 139 0290 SVL S 829 169 0005 H 4735 139 lt 0001 sex times SVL S 128 169 0260 H 001 139 0910head length = sex + SVL + sex times SVL S 918 362 lt 0001 027 H 1643 326 lt0001 061 sex S 302 162 0090 H 115 126 0290 SVL S 2330 162 lt 0001 H 4395 126 lt 0001 sex times SVL S 121 162 0270 H 419 126 0050log(mass) = sex + SVL + sex times SVL S 6012 369 lt 0001 071 H 11920 339 lt0001 089 sex S 3976 169 lt 0001 H 225 139 0140 SVL S 14016 169 lt 0001 H 35032 139 lt 0001 sex times SVL S 045 169 0500 H 515 139 0030no follicles = year + SVL + year times SVL S 953 568 lt 0001 038 year S 1180 233 lt 0001 SVL S 2261 133 0040 year times SVL S 012 233 0890


Sample sizes of 60 in each group would have detecteda significant difference (ɑ lt 005) with power = 080Males had significantly longer tails than females (Table2 Figure 3a) Males tended to have wider and longerheads (Table 2) Females were significantly heavierthan males (Table 2 Figure 3b)

The majority of adult females captured were gravid(n = 46 96) Enlarged follicles were detected from14 May to 11 August Gravid females had 4ndash12 follicles(modes = 5 6 7 and 10 median = 7) The number offollicles increased with SVL (Table 2 Figure 4) andwas significantly different among the three years (2007modes = 0 4 and 7 median = 6 2008 mode = 6 medi-an = 6 2009 mode = 10 median = 9 Table 2 Figure 4)

Individuals were most commonly found in grasslandhabitat (n = 51 59) followed by deciduous forest (n= 22 25) roads (n = 9 10) and mixedwood forest(n = 5 6) Land-cover use differed across the 10 twoweek periods (Table 1 Figure 5) Mixedwood forest

fɪɢᴜʀᴇ 2 Numbers of adult male and female Red-belliedSnake (Storeria occipitomaculata) captured during summersurveys 2007ndash2009 at Spruce Woods Provincial Park (SWPP)Assiniboine Corridor Wildlife Management Area (ACWMA)Oak Lake and Canadian Forces Base Shilo (CFB Shilo) Theseason was divided into ten two-week periods and the cap-ture week indicates the start of the two-week period

fɪɢᴜʀᴇ 3 a Snout-vent length (SVL) versus tail length and b natural log-transformed SVL versus natural log-transformedmass for adult male (open circle dashed line) and female (filled circle solid line) Red-bellied Snake (Storeria occipitomac-ulata) captured during summer surveys 2007ndash2009 at Spruce Woods Provincial Park (SWPP) Assiniboine CorridorWildlife Management Area (ACWMA) Oak Lake and Canadian Forces Base Shilo (CFB Shilo) c SVL versus tail length andd natural log-transformed SVL versus natural log-transformed mass for adult male (open circle dashed line) and female (filledcircle solid line) S occipitomaculata captured during spring and fall surveys 2007ndash2010 at the hibernation site (Souris RiverBend Wildlife Management Area)


was used in spring (11 May to 15 June) while roadsand grassland were most commonly used in late sum-mer (3ndash10 August) There was no difference in land-cover use of males and females (Table 1)

Approximately a third of captures were on flat ground(n = 23 30) Individuals that were captured on slopeswere captured on slopes with a mean aspect of 182 plusmn104o (south facing) Most captures were under cover(n = 73 87) Preferred cover types were plywood(n = 41 56) natural log (n = 9 12) railway tie (n =8 11) plank (n = 6 8) and other (n = 9 13)Mean cover thickness was 365 plusmn 52 mm (range = 1ndash220 mm median = 19 mm) Sampling effort for slopeand cover types was not quantified Therefore the abovevalues do not necessarily reflect habitat selection andmay simply reflect a bias in availability Hibernation demography morphometrics and habitatuse

At the hibernation site (SRBWMA) we found 57individuals representing all three age classes for Soccipitomaculata Adults were the most common (n= 35 61) followed by similar percentages of juve-niles (n = 12 21) and YOY (n = 10 18) Therewas a significant difference in the numbers caught byage class (YOY juvenile adult) between spring and fall(Table 1) In the fall there were more adults (n = 2571 of all adults) and YOY (n = 10 100 of all YOY)captured and more juveniles captured in the spring (n= 11 89 of all juveniles) There was no significantdifference in the numbers of males and females cap-tured in the spring versus the fall (Table 1) Adult fe -male S occipitomaculata captured at the hibernationsite were significantly larger than males (1939 plusmn 44mm and 1789 plusmn 54 mm respectively t261 = 217 P =004) Males had significantly longer tails than females

(Table 2 Figure 3c) There was no difference betweenhead widths (Table 2) or head lengths of males andfemales (Table 2) There was no difference in mass be -tween males and females (Table 2 Figure 3d)

Storeria occipitomaculata shared the hibernationsite with three other species of snakes Plains GarterSnake (thamnophis radix) Red-sided Garter Snake(t sirtalis parietalis) and Smooth Greensnake (Opheo -drys vernalis) that used both the surrounding area andden sites Over three years we captured 166 individuals(Table 3) Most were found during fall surveys (n =114 69) Trapping percentages were highest in Overnalis and t radix (Table 4) Individuals were firstcaptured in traps on 16 September 2007 13 September2008 and 23 September 2009 Timing of the first springcaptures in traps was more variable 7 May 2009 and9 April 2010 More individuals were captured duringour visual searches either in the grassland (n = 5030) on the road (n = 28 17) or inside aluminumflashing on ant mounds (n = 13 8) The remainingindividuals (n = 75 45) were captured inside trapsThere were significant differences among the four spe -cies in capture sites (Table 1) Storeria occipitomaculataand O vernalis were captured significantly more oftenon roads than thamnophis spp Most snakes found inthe grassland were t sirtalis parietalis and there wereno O vernalis found in this habitat Most snakes foundon ant mounds were thamnophis spp

There was no significant difference in direction ofmovement (entering or exiting the ant mound) in fall ascompared to spring (Table 1) Few S occipitomaculatat radix and t sirtalis parietalis were recaptured with-in the same season (11 in total Table S2) There wereonly three individuals (2 all thamnophis spp) recap-tured between seasons one of these individuals had

fɪɢᴜʀᴇ 4 Snout-vent length (SVL) versus number of folliclesfor adult gravid female Red-bellied Snake (Storeria occipit-omaculata) captured during summer surveys 2007ndash2009 atSpruce Woods Provincial Park (SWPP) Assiniboine Corri-dor Wildlife Management Area (ACWMA) Oak Lake andCanadian Forces Base Shilo (CFB Shilo) Linear regressionsare shown separately for each year 2007 (open circle dashedline) 2008 (filled circle solid line) and 2009 (cross dottedline)

fɪɢᴜʀᴇ 5 Habitat use of Red-bellied Snake (Storeria occip-itomaculata) captured during summer surveys 2007ndash2009 atSpruce Woods Provincial Park (SWPP) Assiniboine CorridorWildlife Management Area (ACWMA) Oak Lake and Cana-dian Forces Base Shilo (CFB Shilo) The season was dividedinto ten two-week periods and the capture week indicates thestart of the two-week period


also been recaptured within a season No O vernaliswere recaptured Few individuals were found dead onthe road (3) or dead in a trap or within the aluminumflashing (3) When individuals were recaptured with-in the season they were most frequently first capturedin a trap entering the ant mound and secondly capturedin a trap exiting the ant mound (n = 11 61) All otherrecaptures (first capturesecond capture) were grasslandroad (t sirtalis parietalis n = 2) grasslandgrassland(t sirtalis parietalis n = 2) roadroad (S occipitomac-ulata n = 1) moundgrassland (t sirtalis parietalisn = 1) and moundmound (t sirtalis parietalis n = 1)A single t sirtalis parietalis was recaptured both timesin a trap that exited the moundThermal profiles of den sites

Air temperature and surface temperature (0 cm) werehighly variable and both stayed below freezing from1 November to 1 March (Figure 6) Temperatures at adepth of 40 cm were more stable but were below freez-

ing from 1 December to 1 April Temperatures at a depthof 80 cm and 120 cm below the surface were also sta-ble but largely stayed above freezing In 2007ndash2008the temperature at the 80 cm depth dipped to minus01degC in February but stayed above freezing in 2008ndash2009Summer versus hibernation morphometrics

Adult females tended to be larger at the hibernationsite than at summer sites (hibernation 1939 plusmn 44 mmsummer 1842 plusmn 22 mm t233 = 199 P = 006) Sam-ple sizes were unequal (summer n = 53 hibernationn = 17) and power was low (054 package pwr func-tion pwrt2ntest) Sample sizes of 47 in each groupwould have detected a significant difference (ɑ lt 005)with power = 080 There was no significant differencein SVL of adult males captured at the hibernation sitecompared to the summer sites (hibernation 1789 plusmn54 mm summer 1753 plusmn 40 mm respectively t262 =055 P = 059) For adult females there was no sig-

tᴀʙʟᴇ 3 Numbers of each species captured in different capture locations (trap ant mound road or grassland) at the SourisRiver Bend Wildlife Management Area study site 2007ndash2010 Percentages within each category are indicated in bracketsSpecies are Smooth Greensnake (Opheodrys vernalis) Northern Red-bellied Snake (Storeria occipitomaculata) Plains GarterSnake (thamnophis radix) and Red-sided Garter Snake (thamnophis sirtalis parietalis) Only visual searches were conductedin fall 2009

Year Capture Smooth Northern Plains Red-sided season location Greensnake Red-bellied Snake Garter Snake Garter Snake Total2007 fall Trap 0 16 4 2 22 (71)

Mound 0 0 1 4 5 (16)Road 1 1 0 0 2 (6)Grassland 0 1 1 0 2 (6)Total 1 18 6 6 31

2008 spring Trap 1 0 1 0 2 (50)Mound 0 0 0 0 0Road 0 0 0 0 0Grassland 0 0 1 1 2 (50)Total 1 0 2 1 4

2008 fall Trap 4 4 5 24 37 (46)Mound 1 1 0 5 7 (9)Road 1 16 1 3 21 (27)Grassland 0 1 1 12 14 (18)Total 6 22 7 44 79

2009 spring Trap 0 3 0 0 3 (13)Mound 0 0 0 0 0Road 0 2 1 0 3 (13)Grassland 0 4 1 13 18 (75)Total 0 9 2 13 24

2009 fall Road 1 1 0 0 2 (50)Grassland 0 2 0 0 2 (50)Total 1 3 0 0 4

2010 spring Trap 2 4 2 3 11 (46)Mound 0 0 1 0 1 (4)Road 0 0 0 0 0Grassland 0 8 1 3 12 (50)Total 2 12 4 6 24

Total Trap 7 (64) 27 (42) 12 (57) 29 (41) 75 (45)Mound 1 (9) 1 (2) 2 (10) 9 (13) 13 (8)Road 3 (27) 20 (31) 2 (10) 3 (4) 28 (17)Grassland 0 16 (25) 5 (24) 29 (41) 50 (30)Total 11 (7) 64 (39) 21 (13) 70 (42) 166


nificant difference in tail or head length (Table 4) forindividuals captured at the hibernation site comparedto those captured at summer sites Adult females at thehibernation site had wider heads than those from sum-mer sites (Table 4) For adult males there was no sig-nificant difference in tail length head width or headlength (Table 4) for individuals captured at the hiber-nation site compared to those captured at summer sitesBoth adult females (Table 4 Figure 3b) and adult males(Table 4 Figure 3d) were significantly heavier for theirbody size at summer sites compared to the hibernationsite

DiscussionOur study examined demography morphology re -

pro duction habitat use and hibernation in populationsof S occipitomaculata in southwestern Manitoba Thisarea is relatively arid and has an average daily tem-perature across the year of only 22degC (EnvironmentCanada 2016) This is a stressful environment but theexplosive productivity of summer may be enough tocompensate (Tuttle and Gregory 2014) Variation be -tween populations we observed and other parts of thisspeciesrsquo range may indicate how they make up for thisstress

The demographics of S occipitomaculata were sim-ilar to the findings of Blanchard (1937) with adultsaccounting for the vast majority of observations Wesuspect this may simply reflect the difficulty of findingjuveniles using walking surveys When using funneltraps at hibernation sites we collected more YOY andjuveniles However Semlitsch and Moran (1984) alsoobserved adult biased demographics using passive trap-ping

Although we observed female biased SSD in thispopulation differences were not large Male to femalebody size ratios were similar to populations from Mich -igan Minnesota and interestingly South Carolina butless dimorphic than populations in Virginia Indiana orPennsylvania (reviewed in Meshaka and Klippel 2011)Body sizes at our study sites were smaller than mostother locales including other northern populations(Criddle 1937 Meshaka and Klippel 2011) The mini-mum size of gravid snakes in our study was 154 mm

fɪɢᴜʀᴇ 6 Mean temperatures from iButtonsreg placed at fiveden sites at Souris River Bend Wildlife Management Areaover winter (20 Septemberndash22 May) in 2007ndash2008 and 2008ndash2009 Separate lines are shown for each winter air tempera-ture and surface temperature (0 cm) were only recorded in2008ndash2009

tᴀʙʟᴇ 4 Linear models comparing tail length head width head length and natural log-transformed mass of Red-bellied Snake(Storeria occipitomaculata) between summer and hibernation sites (site variable) Models were done separately for each sexSVL = snout-vent length

Model Sex F df P Adjusted r2

tail length = site + SVL F 024 264 0790 002 site 015 164 0700 SVL 058 164 0570 site + SVL M 213 231 0140 006 site 019 131 0670 SVL 408 131 0050head width = site + SVL F 249 264 0090 004 site 461 164 0040 SVL 037 164 0550 site + SVL M 712 231 0003 027 site 051 131 0480 SVL 1372 131 lt 0001head length = site + SVL F 580 254 0005 015 site 008 154 0780 SVL 1152 154 0001 site + SVL M 378 225 0040 017 site 004 125 0840 SVL 75 125 0010log(mass) = site + SVL F 3661 264 lt 0001 052 site 3677 164 lt 0001 SVL 3646 164 lt 0001 site + SVL M 2158 231 lt 0001 056 site 450 131 0040 SVL 3866 131 lt 0001

larger than those in South Carolina (Semlitsch andMoran 1984) and similar to Virginia (Mitchell 1994)but smaller than most of the rest of the range (Meshakaand Klippel 2011) The literature suggests while bodysize in S occipitomaculata varies average clutch sizesare similar across its range In our study gravid femaleshad similar clutch sizes to other populations (7ndash9 youngMeshaka and Klippel 2011) Northern reptiles are oftenthought to be constrained having slower growth and de -layed maturity relative to southern populations (Atkin-son 1994) but this is not always the case (Angilletta etal 2004 Tuttle and Gregory 2012 2014) Early matu-rity is usually accomplished by increased growth to sim-ilar minimum size (Tuttle and Gregory 2012 2014)Storeria occipitomaculata appears to mature at a small-er size in the southeast and northwest portions of itsrange while maintaining similar clutch sizes to physi-cally larger populations elsewhere

Although spring breeding has been widely reportedin this species (Semlitsch and Moran 1984 Ernst andErnst 2003 Meshaka and Klippel 2011) the abundanceof males in late summer suggests August breedingactivity (Blanchard 1937 Trapido 1940 Willson andDorcas 2004) Although we did not directly observe anyfall mating it has been reported previously in Manitoba(Gregory 1977) Also we found evidence of primaryvitellogenesis pre-hibernation in two road-killed spec-imens collected in October 2007 and captured gravidfemales as early as 14 May In our study the majority ofadult females captured during the summer were gravid(96) suggesting annual reproduction similar to pop-ulations in the south (Semlitsch and Moran 1984) Re -production should be limited by available resources(Aldridge 1979) and biennial reproduction appears tobe common in northern snakes (Larsen et al 1993 Gre-gory 2009) Mating in late summerfall would providefemales with a longer foraging and gestation period thatwould allow for greater provisioning of offspring inutero and may explain this populationrsquos ability to repro-duce annually despite the short season This is likelyan important life history trait for a short lived specieslike S occipitomaculata (Snider and Bowler 1992)

The habitat use by S occipitomaculata we observedin Manitoba was similar to that found in Minnesota(Lang 1971) and Illinois (Brown and Phillips 2012)but differed from findings in Kansas (Pisani and Bus-by 2011) We found disproportionate use of grasslandhabitat on flat land or with south facing aspects Pisaniand Busby (2011) found the majority of their sampleassociated with moister habitats avoiding open habitatson north west and east slopes and usually associatedwith water These differences may reflect local avail-ability local adaptation or perhaps a trade-off made bynorthern S occipitomaculata sacrificing osmotic pref-erences for thermal ones This may be compensated forby microhabitat selection In our study most captureswere under cover (87) and more than half were foundunder plywood This pattern may reflect our sampling

methods rather than true preferences although a pref-erence for cover objects is common in small snakes(Halliday and Blouin-Demers 2015 Gregory and Tut-tle 2016)

Land-cover use differed across the 10 two weekperiods of our study Use of mixedwood forest by Soccipitomaculata occurred in spring while they usedroads and grassland most commonly in late summerand fall (at the hibernation site) Semlitsch and Moran(1984) suggested that moisture gradients and food avail-ability drove activity pulses and habitat choice in thisspecies in South Carolina Alternatively use of grass-lands in late summer may reflect habitat preferences ofgravid females because grasslands are warmer thanforest habitats in keeping with the cold climate hypoth-esis (Tinkle and Gibbons 1977) although we did notobserve a sex bias in habitat choice

Over the course of three years the first arrival ofS occipitomaculata at hibernation sites was later thanother species and was consistently associated with thefirst frost or nighttime temperatures below freezingsimilar to Lang (1971) In cold regions such as Min-nesota or Manitoba it is difficult to understand why asmall snake would arrive at a hibernaculum so late inthe season when hard frosts are likely Although Soccipitomaculata appears to be able to cope with sev-eral days of cold weather a prolonged or particularlyhard frost may cause mortality (Lang 1971) Criticalthermal minimum (CTmin) has not been determined inS occipitomaculata but ranges from 25 to 115degC inother snake species with body size species and latitu-dinal effects (reviewed in Cox et al 2018) so it likelyabove 0degC In the smaller Ring-necked Snake (Dia do -phis punctatus) CTmin was 115degC and decreased withincreasing body size suggesting that small individualsrequire higher CTmin (Cox et al 2018)

The use of ant mounds by S occipitomaculata Overnalis and t sirtalis parietalis has been previouslyreported by Criddle (1937) Carpenter (1953) Lang(1971) and Pisani (2009) These mounds do not offera thermal advantage over the winter compared to sur-rounding soil (Scherba 1962) However due to theirinsulation and ability to collect solar radiation thesesites are warmer and more stable during the active sea-son (Scherba 1962 Duff et al 2016) thereby poten-tially increasing season length The thermal profiles wegenerated at the hibernation site indicated stability in -creases with soil depth but even at 80 cm temperaturesdipped below freezing This is similar to the tempera-ture profile generated by Lang (1971) although thefrost lines at most of the ant mounds he investigatedwere slightly deeper likely due to the shallow watertable at our hibernation site The temperature profileof the mounds suggests that these snakes would haveto hibernate near or below the water table (Costanzo1989) which was ~120 cm in our study Criddle (1937)found this to be the case when he excavated a moundnear Treesbank Manitoba and found many of the snakes



at ~144 cm depth in contact with the water table With-out these abandoned ant mounds S occipitomaculatapoorly suited to true burrowing with its kinetic skulland large eyes would be unable to access such thermal-ly stable and humid refuges (Pisani 2009) These sitesor other similar fossorial retreats are likely critical forthe persistence of this species at such northern climates

Caution must be taken when comparing snapshots ofpopulationsrsquolife histories (Seigel and Fitch 1985) Thereis a great deal of variation reported across the speciesrsquorange and among individuals and like Meshaka andKlippel (2011) we found a limited effect of latitude be -tween the population we observed and the rest of therange This observation adds to our understanding ofhow this species can survive in an area that is climati-cally unsuitable for terrestrial activity for much of theyear In this area this species appears to use relativelywarm habitats rapid reproduction and abandoned antnests to persist and thrive Productive habitats such asthe northern Great Plains may allow this species toadopt a ldquofastrdquo lifestyle that favours early maturity andhigher than expected annual fecundity with most fe -males reproducing annually (Tuttle and Gregory 2014)Much remains unknown and future studies should con-firm the genetic or environmental underpinning of bodysize at maturity and clutch size along with local diettiming of breeding and the drivers and risks associatedwith late-season migration for this species

AcknowledgementsThanks to L Cairns R Cairns B Cairns J Cairns

R Dale T Dempsey A Dransfield N Gushulak CJaeger J Larkin D Macintyre and J Phillips for theirassistance in the field and study design The Ella May-Stewart Perdue Trust Fund Brandon University Re -search Committee Brandon University Student UnionFellowship and Department of Zoology provided finan-cial support Thank you to W Halliday and two anony-mous reviewers for their helpful comments to improvethe manuscript All protocols were done with the ap -proval of the Brandon University Animal Care Com-mittee (2006R05) and all necessary permits for fieldstudy were obtained (Manitoba Conservation WB06082 WB07879 WB09616 and WB11022 ManitobaParks 21739 25035 25968 and 2011-P-HQ-022)

Literature Citedaldridge Rd 1979 Female reproductive cycles of the

snakes arizona elegans and Crotalus viridis Herpetolog-ica 35 256ndash261

angilletta MJ Jr td Steury and Mw Sears 2004Temperature growth rate and body size in ectotherms fit-ting pieces of a life-history puzzle Integrative and Com-parative Biology 44 498ndash509 httpsdoiorg101093icb446498

atkinson d 1994 Temperature and organism size a biologi-cal law for ectotherms Advances in Ecological Research25 1ndash58 httpsdoiorg101016S0065-2504(08)60212-3

Berrigan d and el charnov 1994 Reaction norms forage and size at maturity in response to temperature a puz-

zle for life historians Oikos 70 474ndash478 httpsdoiorg1023073545787

Blackburn dG ke anderson aR Johnson SRknight and GS Gavelis 2009 Histology and ultrastruc-ture of the placental membranes of the viviparous brownsnake Storeria Dekayi (Colubridae Natricinae) Journalof Morphology 270 1137ndash1154 httpsdoiorg101002jmor10650

Blanchard fN 1937 Data on the natural history of the red-bellied snake Storeria occipitomaculata (Storer) in North-ern Michigan Copeia 1937 151ndash162 httpsdoiorg1023071436135

Blouin-demers G ka Prior and PJ weatherhead2002 Comparative demography of black rat snakes (elapheobsoleta) in Ontario and Maryland Journal of Zoology256 1ndash10 httpsdoiorg101017S0952836902000018

Blouin-demers G and PJ weatherhead 2001 Thermalecology of black rat snakes (elaphe obsoleta) in a thermallychallenging environment Ecology 82 3025ndash3043 httpsdoiorg1018900012-9658(2001)082[3025TEOBRS]20CO2

Brown ee 1979 Stray food records from New York andMichigan snakes American Midland Naturalist 102 200ndash203 httpsdoiorg1023072425088

Brown le and ca Phillips 2012 Distribution habitatand zoogeography of the semifossorial red-bellied snakeStoreria occipitomaculata (Storer) in Illinois Illinois Natu-ral History Survey Bulletin 39 297ndash322

carpenter cc 1953 A study of hibernacula and hibernat-ing associations of snakes and amphibians in MichiganEcology 34 74ndash80 httpsdoiorg1023071930310

cook fR 1984 Introduction to Canadian Amphibians andReptiles National Museum of Natural Sciences NationalMuseums of Canada Ottawa Ontario Canada

costanzo JP 1989 Effects of humidity temperature andsubmergence behavior on survivorship and energy use inhibernating garter snakes thamnophis sirtalis CanadianJournal of Zoology 67 2486ndash2492 httpsdoiorg101139z89-351

cox RM Ma Butler and hB John-alder 2007 Theevolution of sexual size dimorphism in reptiles Pages 38ndash49 in Sex Size and Gender Roles Evolutionary Studies ofSexual Size Dimorphism edited by DJ Fairbairn WUBlanckenhorn and T Szeacutekely Oxford University PressNew York New York USA httpsdoiorg101093acprofoso97801992087840030005

cox cl Ml logan o Bryan d kaur e leung JMccormack J McGinn l Miller c Robinson JSalem J Scheid t warzinski and ak chung 2018Do ring-necked snakes choose retreat sites based upon ther-mal preferences Journal of Thermal Biology 71 232ndash236httpsdoiorg101016jjtherbio201711020

criddle S 1937 Snakes from an ant hill Copeia 1937 142httpsdoiorg1023071436960

duff lB tM urichuk lN hodgins JR young andwa untereiner 2016 Diversity of fungi from the moundnests of Formica ulkei and adjacent non-nest soils Cana-dian Journal of Microbiology 62 562ndash571 httpsdoiorg101139cjm-2015-0628

environment canada 2016 National Climate Data and In -formation Archive Environment Canada Accessed 18 May2018 httpstinyurlcomya5pe4cl

ernst ch and eM ernst 2003 Snakes of the UnitedStates and Canada Smithsonian Books Washington DCUSA


eSRi 2011 ArcGIS Desktop Redlands California USAfitch hS 1981 Sexual size differences in reptiles Univer-

sity of Kansas Museum of Natural History MiscellaneousPublications 70 1ndash72

fitch hS 1987 Collecting and life-history techniques Pages143ndash164 in Snakes Ecology and Evolutionary Biologyedited by RA Seigel JT Collins and SS Novak Mac -millan New York New York USA

Gienger cM and dd Beck 2011 Northern Pacific Rat-tlesnakes (Crotalus oreganus) use thermal and structuralcues to choose overwintering hibernacula Canadian Jour-nal of Zoology 89 1084ndash1090 httpsdoiorg101139z11-086

Gregory Pt 1977 Life history observations of three speciesof snakes in Manitoba Canadian Field-Naturalist 91 19ndash27 Accessed 11 November 2018 httpsbiodiversitylibraryorgpage28061222

Gregory Pt 2009 Northern lights and seasonal sex thereproductive ecology of cool-climate snakes Herpetologica65 1ndash13 httpsdoiorg1016550018-0831-6511

Gregory Pt and kN tuttle 2016 Effects of body sizeand reproductive state on cover use of five species of tem-perate-zone Natricine snakes Herpetologica 72 64ndash72httpsdoiorg101655HERPETOLOGICA-D-15-00021

halliday wd and G Blouin-demers 2015 Efficacy ofcoverboards for sampling small northern snakes Herpetol-ogy Notes 8 309ndash314

harvey dS and PJ weatherhead 2006a Hibernation siteselection by Eastern Massasauga Rattlesnakes (Sistruruscatenatus catenatus) near their northern range limit Jour-nal of Herpetology 40 66ndash73 httpsdoiorg10167089-05A1

harvey d and PJ weatherhead 2006b A test of the hier-archical model of habitat selection using eastern massas-auga rattlesnakes (Sistrurus c catenatus) Biological Con-servation 130 206ndash216 httpsdoiorg101016jbiocon200512015

kiester aR 1971 Species density of North American am -phibians and reptiles Systematic Biology 20 127ndash137httpsdoiorg1023072412053

king RB 1993 Determinants of offspring number and sizein the brown snake Storeria dekayi Journal of Herpetology27 175ndash185 httpsdoiorg1023071564934

lang Jw 1969 Hibernation and movements of Storeriaoccipitomaculata in northern Minnesota Journal of Her-petology 3 196ndash197

lang Jw 1971 Overwintering of three species of snakes innorthwestern Minnesota MSc thesis University of NorthDakota Grand Forks North Dakota USA

larsen kw Pt Gregory and R antoniak 1993 Repro-ductive ecology of the common garter snake thamnophissirtalis at the northern limit of its range American Mid-land Naturalist 129 336ndash345 httpsdoiorg1023072426514

Manitoba land initiative 2012 Manitoba Conservation andWater Stewardship Accessed 5 February 2012 httpsmli2govmbca

Meshaka we and aN klippel 2011 Seasonal activityreproduction and growth of the Northern Redbelly SnakeStoreria occipitomaculata occipitomaculata (Storer 1839)from Pennsylvania Journal of Kansas Herpetology 37 16ndash20

Mitchell Jc 1994 The Reptiles of Virginia SmithsonianInstitution Press Washington DC USA

Pisani G 2009 Use of an active ant nest as a hibernaculumby small snake species Transactions of the Kansas Acade-my of Science Kansas Academy of Science 112 113ndash118httpsdoiorg1016600621120215

Pisani GR and wh Busby 2011 Ecology of the SmoothEarth Snake (Virginia valeriae) and Redbelly Snake (Store-ria occipitomaculata) in Northeastern Kansas Kansas Bio-logical Survey Lawrence Kansas USA

R development core team 2017 R A language and envi-ronment for statistical computing R Foundation for Sta-tistical Computing Vienna Austria

Scherba G 1962 Mound temperatures of the ant Formicaulkei Emery American Midland Naturalist 67 373ndash385httpsdoiorg1023072422715

Seigel Ra and hS fitch 1985 Annual variation in repro-duction in snakes in a fluctuating environment Journal ofAnimal Ecology 54 497ndash505 httpsdoiorg1023074494

Semlitsch Rd and GB Moran 1984 Ecology of the red-belly snake (Storeria occipitomaculata) using mesic habi-tats in South Carolina American Midland Naturalist 11133ndash40 httpsdoiorg1023072425539

Shine R 1994 Sexual size dimorphism in snakes revisitedCopeia 1994 326ndash346 httpsdoiorg1023071446982

Shine R J webb a lane and R Mason 2006 Flexiblemate choice a male snakersquos preference for larger femalesis modified by the sizes of females encountered AnimalBehaviour 71 203ndash209 httpsdoiorg101016janbehav200504005

Snider at and Jk Bowler 1992 Longevity of reptilesand amphibians in North American collections Herpeto-logical Circulars No 21 Society for the Study of Amphib-ians and Reptiles Publications Lawrence Kansas USA

tinkle dw and Jw Gibbons 1977 The distribution andevolution of viviparity in reptiles Miscellaneous Publica-tions Museum of Zoology University of Michigan 1541ndash55

trapido h 1940 Mating time and sperm viability in Store-ria Copeia 1940 107ndash109 httpsdoiorg1023071439051

trapido h 1944 The snakes of the genus Storeria AmericanMidland Naturalist 31 1ndash84 httpsdoiorg1023072421382

tuttle kN and Pt Gregory 2012 Growth and maturityof a terrestrial ectotherm near its northern distributionallimit does latitude matter Canadian Journal of Zoology90 758ndash765 httpsdoiorg101139z2012-044

tuttle kN and Pt Gregory 2014 Reproduction of thePlains Garter Snake thamnophis radix near its northernrange limit more evidence for a ldquofastrdquo life history Copeia2014 130ndash135 httpsdoiorg101643CH-13-119

weatherhead PJ Jh Sperry Glf carfagno and GBlouin-demers 2012 Latitudinal variation in thermalecology of North American ratsnakes and its implicationsfor the effect of climate warming on snakes Journal ofThermal Biology 37 273ndash281 httpsdoiorg101016jjtherbio201103008

willson Jd and Me dorcas 2004 Aspects of the ecologyof small fossorial snakes in the western Piedmont of NorthCarolina Southeastern Naturalist 3 1ndash12 httpsdoiorg1016561528-7092(2004)003[0001AOTEOS]20CO2

wright ah and aa wright 1957 Handbook of Snakesof the United States and Canada Cornell University PressIthaca New York USA

Received 2 March 2018Accepted 25 May 2018


SuPPleMeNtaRy MateRial

taBle S1 Numbers of Red-bellied Snake (Storeria occipitomaculata) in three age classes (young-of-year [YOY] juvenileadult) for each sex that were captured during summer surveys 2007ndash2009 at Spruce Woods Provincial Park (SWPP) AssiniboineCorridor Wildlife Management Area (ACWMA) Oak Lake and Canadian Forces Base (CFB) Shilo

taBle S2 Number of captures recaptures and dead animals for individuals captured at the hibernation site (Souris River BendWildlife Management Area) during spring and fall surveys 2007ndash2010


developing embryos may cause females to select dif-ferent habitats than males and alter their thermoregu-latory behaviour (Shine et al 2006) The cold-climatehypothesis predicts that warmer temperatures will beselected by gravid female squamates in cold climates toensure rapid embryo development (Tinkle and Gibbons1977) This may lead to increased use of more thermal-ly attractive open or edge habitat than in other parts of aspeciesrsquo range

Cold-climate reptiles can avoid inclement conditionsand limit energy expenditures through hibernation (Gre-gory 2009) Failure to select appropriate hibernaculaleads to death but selecting purely for overwinter survi -val also has associated costs (Gienger and Beck 2011)Time spent in hibernation subtracts from time spent for-aging therefore hibernacula are often chosen to extendthe length of the active season (Gienger and Beck 2011)Features often associated with successful hibernationsites allow for access to soil below the frost line stabil-ity in humidity and temperature and access to the watertable (Harvey and Weatherhead 2006a) The availabil -ity of suitable hibernacula is likely the most importantform of habitat selection for snakes at high latitudesand may limit the distribution of some species (Harveyand Weatherhead 2006b)

Red-bellied Snake (Storeria occipitomaculata) occu-pies most of eastern North America and is widely distri -buted in Canada reaching the northwestern edge of itsrange in eastern Saskatchewan (Ernst and Ernst 2003)Across its range there have been a number of studies ofthis species (eg Blanchard 1937 Lang 1969 Seml-itsch and Moran 1984 Willson and Dorcas 2004 Brownand Phillips 2012) but there has been little research atits northern range limit with the exception of two nat-ural history observations (Criddle 1937 Gregory 1977)Storeria occipitomaculata is a small-bodied crypticsnake that comes in several ground-shaded dorsal col -our morphs and has a bright reddish-orange ventralsurface They are viviparous likely with some limitedplacental provisioning (Blackburn et al 2009) Thisspecies like other members of the genus Storeria feedsprimarily on molluscs (Trapido 1944 Brown 1979Semlitsch and Moran 1984 Pisani and Busby 2011) al -though not exclusively (Ernst and Ernst 2003) Withfew species of slugs and snails native to the northwest-ern edge of its range local diet remains unknown Thisspecies is considered uncommon to rare throughoutmuch of its range (Ernst and Ernst 2003) but it can beabundant in suitable habitat (Lang 1971) Little is knownabout population densities for this species althoughmost studies have recorded low recapture rates (Blan-chard 1937 Gregory 1977 Semlitsch and Moran 1984)Typically S occipitomaculata inhabit thick grassyherbaceous and shrubby vegetation and also can befound in wetlands riparian areas and along forest edgesand in old field and prairie habitats (Wright and Wright1957 Cook 1984 Pisani and Busby 2011) as well asdrier grasslands habitats (Lang 1969 Brown and Phil -

lips 2012) Criddle (1937) and Lang (1969) both foundthis species to use abandoned (or nearly so) ant moundsof species of the genus Formica as overwintering siteswith the snakes often found hibernating at high den-sities

The objective of this study is to examine demogra-phy morphology reproduction habitat use and hiber-nation in populations of S occipitomaculata near theirnorthern range limit The stressful environment at highlatitudes may force these populations to make trade-offsin their growth reproduction and habitat use We askedthe following questions 1) What are the demographicsmorphology and reproductive traits of this populationof S occipitomaculata and 2) Under what conditionsdo they hibernate

MethodsStudy sites

Field work in summer habitat use took place in south -western Manitoba Canada at the following locations(Figure 1a) Spruce Woods Provincial Park (SWPP497108degN 992528degW) Assiniboine Corridor WildlifeManagementArea (ACWMA 496675degN 995614degW)Oak Lake (496644degN 1007133degW) and CanadianForces Base Shilo (CFB Shilo 497381degN 995183degW)A wide variety of habitats were investigated includingagricultural areas beaver ponds mixed forest andmixed-grass prairie Most sites included water andormixed grass prairie Investigation into hibernation wasconducted at the Souris River Bend Wildlife Manage-ment Area (SRBWMA 494883degN 998775degW Fig-ure 1b) Habitat at this area includes mixed forest andgrassland bordered by marsh agricultural fields andgravel roads Details on the habitat types are providedbelowDemography and morphometrics

We recorded the following measurements for allsnakes date time snout-vent length (SVL in mm) taillength (mm) head width (mm) head length (mm) mass(g) sex age and reproductive condition We classifiedall individuals into three age classes young-of-year(YOY) juvenile and adult Storeria occipitomaculatawere classified into age classes based on their SVLnatal scars and the minimum breeding SVL of fe -males YOY (lt110mm) juvenile (100ndash154 mm) andadult (gt154 mm) For the age-class analysis we onlyincluded the first capture for individuals that were re -captured multiple times within the same season

To determine reproductive condition two researchersindependently counted enlarged ovarian follicles bypalpation (Fitch 1987) and the average of their countswas recorded All captured individuals were scale-clipped for identification and released at their capturesite within 15 minutes of capture The activity period(27 April to 8 September) was classified by two-weekperiods numbered 1 through 10 No snakes in this studyreceived more than three clipped ventral scales










































































































































morphology, reproduction, habitat use, and hibernation of

Documents