Hibernation and overwinter body temperatures in free-ranging thirteen-lined ground squirrels, Ictidomys tridecemlineatus.
Abstract: Free-ranging, juvenile thirteen-lined ground squirrels (Ictidomys tridecemlineatus) in southwestern Michigan were fitted in late summer or fall with external skin-temperature loggers. Data were obtained the following spring for five males and three females. During the heterothermal period, all squirrels exhibited 11-29 prolonged ([bar.x] = 9.4 d) torpor bouts punctuated by typically brief ([bar.x] = 14.3 h) arousal bouts, with mean monthly torpor bouts becoming longer and deeper until Feb. and reversing thereafter. Torpor-bout duration increased as minimum skin and soil temperatures decreased. On average, males initiated the first torpor bout later in fall, terminated the last torpor bout significantly earlier in spring and thus spent less time in the heterothermal period than did females. Three males displayed relatively short torpor bouts and long arousal bouts as they approached the end of hibernation. Squirrels gained weight variably in fall and spring, and one female lost 39% of body mass during hibernation.
Article Type: Report
Subject: Hibernation (Research)
Body temperature (Research)
Squirrels (Physiological aspects)
Authors: Kisser, Brooke
Goodwin, H. Thomas
Pub Date: 04/01/2012
Publication: Name: The American Midland Naturalist Publisher: University of Notre Dame, Department of Biological Sciences Audience: Academic Format: Magazine/Journal Subject: Biological sciences; Earth sciences Copyright: COPYRIGHT 2012 University of Notre Dame, Department of Biological Sciences ISSN: 0003-0031
Issue: Date: April, 2012 Source Volume: 167 Source Issue: 2
Topic: Event Code: 310 Science & research
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 287956860
Full Text: INTRODUCTION

Hibernation has been intensively studied in ground-dwelling squirrels (family Sciuridae) within the tribe Marmotini. In captivity, marmotine and other mammalian hibernators studied at controlled ambient temperatures in cold chambers universally display a characteristic body-temperature profile during hibernation: multiple, prolonged bouts of torpor punctuated by much shorter bouts of arousal when the animal returns to euthermy. Above ~0 C ambient, minimum body temperature during torpor scales directly with ambient temperature; and duration of torpor scales inversely with both ambient and body temperature, as well as with metabolic rate (Twente and Twente, 1965; Pivorun, 1976; Geiser and Kenagy, 1988; Geiser et al., 1990; Buck and Barnes, 2000). Below ~0 C ambient, animals typically defend body temperature, decrease duration of torpor and increase metabolic rate (Pivorun, 1976; Geiser and Kenagy, 1988; Buck and Barnes, 2000).

Under field conditions with seasonally variable soil temperatures, these body-temperature patterns likewise hold, at least qualitatively, for free-ranging hibernators within Tamias (chipmunks-Humphries et al., 2003), Marmota (marmots and woodchucks-Florant et al., 2000; Zervanos and Salsbury, 2003; Lee et al., 2009; Zervanos et al., 2010), Cynomys (prairie dogs-Lehmer and Biggins, 2005), Spermophilus [Old World ground squirrels-Hut et al., 2002 (studied under semi-natural conditions)] and Urocitellus (Holarctic ground squirrels-Wang, 1973, 1979; Young, 1990; Michener, 1992; Buck et al., 2008; Frank et al., 2008). However, comparisons of free-ranging and captive hibernators sometimes reveal quantitative differences in body-temperature patterns (Geiser et al., 2000). For example, free-ranging arctic ground squirrels (Urocitellus parryii) display average torpor bouts in mid-winter nearly 10 d longer than predicted based on studies of captive animals of the same species (Buck et al., 2008).

Furthermore, field studies document ecologically-relevant variation in the phenology of and body temperatures within hibernation across species, along altitudinal or latitudinal gradients within a species (Lehmer and Biggins, 2005; Zervanos et al., 2010) and across age and sex classes within a population (Young, 1990; Michener, 1992; Zervanos and Salsbury, 2003; Buck et al., 2008). They also demonstrate that under field conditions, hibernation (the period of sequestration below ground) and the heterothermal period (the period between onset of first torpor and arousal from final torpor) are not necessarily equivalent; squirrels may spend several days euthermic but sequestered below ground before and especially after the heterothermal period (Michener, 1992; Barnes and Ritter, 1993). To avoid confusion, we follow the distinction between hibernation and the heterothermal period in this paper.

Here, we report overwinter skin temperatures (as a proxy for body temperatures) for free-ranging thirteen-lined ground squirrels (Ictidomys tridecemlineatus), the first such study for the genus Ictidomys. For each individual, we document attributes of the heterothermal period shown to vary seasonally within hibernation, or between age cohorts (juveniles, adults) or sexes, in other free-ranging marmotines (Young, 1990; Michener, 1992; Zervanos and Salsbury, 2003; Buck et al., 2008). These include timing and duration of the heterothermal period, duration and depth of torpor bouts, lowest skin temperature during hibernation, duration of arousal bouts, and the proportion of the heterothermal period in torpor. We also examined correlations among torpor-bout duration, minimum skin temperature during torpor, and soil temperatures (measured regionally); explored differences between sexes and across months during the heterothermal period of hibernation; and obtained body masses before and after hibernation to estimate growth patterns, and weight loss during hibernation.

Thirteen-lined ground squirrels appear to be obligate hibernators across their range, with hibernation demonstrated as far north as Manitoba (Criddle, 1939) and southward to Texas (McCarley, 1966). A previous study documented the typical pattern of torpor and arousal during hibernation under semi-natural conditions, but it did not report skin or body temperatures (Scheck and Fleharty, 1980). Vernal emergence from hibernation typically takes place in Mar. and Apr. (on average, males usually before females) and autumnal immergence into hibernation extends from Jul, to Oct. or even early Nov. (on average, adult males typically before adult females, which on average precede juveniles-Wade, 1927; Criddle, 1939; Beer, 1962; Rongstad, 1965; McCarley, 1966; Clark, 1971; Scheck and Fleharty, 1980).

MATERIALS AND METHODS

SQUIRREL CAPTURE AND OBSERVATION

Fifteen thirteen-lined ground squirrels were initially captured between 22 Aug. and 2 Sept. 2007 at Rosehill Cemetery in Berrien Springs, Michigan (41[degrees]55.97N, 86[degrees]20.47W) using Sherman live traps (9.0 x 7.5 x 23.5 cm) baited with oats and peanut butter. A resident cat subsequently killed two squirrels, and their collars (see below) were retrieved and placed on two additional animals captured 7 Sep. and 12 Oct. 2007. Traps were set opportunistically adjacent to burrows that we observed squirrels to enter, were checked every 0.5-1.0 h once they were set and were promptly moved to shade (if necessary) when a squirrel was captured. Captured animals were handled and released within 1 h after we noted their capture.

Each ground squirrel was sexed, inspected for sores or lesions, and weighed to the nearest gram with hand-held hanging scales, and its capture-location coordinates were recorded. Each animal was fitted with a temperature-sensitive iBcollar data logger (2.8-3.1 g for logger and collar; data loggers obtained from Alpha Mach Inc., Mont St-Hilaire, Quebec) set to record skin temperature every 2 h. Data loggers were initially attached around the neck with thin wire; but after we observed one animal with irritated skin under the wire upon recapture, we modified all collars upon first subsequent recapture to cover the wire with plastic insulation stripped from standard electrical wire. A unique insulation color was used for each day that we modified collars on recaptured squirrels, and this facilitated subsequent field identification. Collars were uniquely numbered to allow firm identification at capture, and collar numbers always matched field identifications based on insulation color.

After ground squirrels were fitted with collars, we visited the colony 1-3 d/wk (except 914 Sept.) for, on average, 2.8 h/d (ranged from 0.5 to 7 h/d) until 1 wk after aboveground activity ceased (last visit 18 Nov.). Daily time at the colony (h/d) is a minimum estimate; it was not recorded directly but was estimated from times of first and last sightings or captures of collared animals. We recorded the presence and location of all collared animals that were observed and opportunistically recaptured collared animals to reweigh them.

We assigned ground squirrels to age cohorts (juvenile and adult) based on body mass at first capture in late Aug. or early Sept., using for comparison data presented by Hohn and Marshall (1966) for a colony in Minnesota. They recorded minimum adult body masses in Aug. of ~160-170 g and maximum juvenile body masses of ~170 g. Thus, we conservatively identified squirrels that weighed less than 150 g in late Aug. or early Sept. as juveniles.

Beginning 24 Mar. 2008 when the first squirrel of the spring (not collared) was observed aboveground, the colony was checked for activity 3-5 d/wk until 1 d after the last of the collared squirrels active during spring was recaptured (last visit 6 May). Because we estimated daily time at the colony from times of first and last sightings or captures, we do not have estimates before 4 Apr. on account of inadequate sightings (0-1 sightings/d). Beginning 4 Apr., we usually made >1 sightings/d, and spent, on average, 4.6 h/d (ranged from 0.5 to 9 h/d) at the colony.

We initially attempted to recapture collared ground squirrels using the same procedure as in fall, but this was often ineffective. Instead, we built a portable fence that was set up around a burrow after a collared squirrel entered and placed a heavy board with attached string within the fence and adjacent to the burrow. When the squirrel emerged and moved away from the burrow, the string was manually pulled to shift the board over the burrow opening, and the squirrel was captured with gloved hands. Within 1 h of spring capture, squirrels were transported 4.6 km to the laboratory where collars were removed and squirrels were weighed. Because study animals were used in a broader study to examine how dental tissues reflect body temperature variation over winter, they were euthanized by rapid application of compressed C[O.sub.2] into a closed chamber. All procedures were reviewed and approved by the Andrews University Animal Use Committee (approved 21 Jun. 2007) and followed published guidelines of the American Society of Mammalogists' Animal Care and Use Committee (Gannon et al., 2007).

ANALYSIS OF BODY TEMPERATURE DATA

Temperature data were downloaded from each data logger and transferred to a Microsoft Excel spreadsheet. Data loggers were individually pre-calibrated by the supplier at three known temperatures (0.1, 15.0, 30.0 C), and calibrated temperature readings were all within [+ or -]0.2 C of known temperatures. The supplier provided a calibration routine specific for each data logger to translate raw values into temperature (C). Skin temperatures were plotted against time and the resulting profile inspected for the timing of onset and termination of the heterothermal period and patterns of skin-temperature variation during the heterothermal period. We marked the commencement and termination of the heterothermal period by onset of the first deep torpor bout and arousal from the last deep torpor bout, respectively. Total time spent in the heterothermal period (d) was calculated from these endpoints.

The number of torpor bouts during the heterothermal period was determined for each ground squirrel and the duration of each torpor and arousal bout recorded. The onset and termination of bouts of torpor and arousal typically are defined by when body temperature crosses an arbitrary threshold, usually 30 C (Young, 1990; Michener, 1992; Buck and Barnes, 2000; Zervanos and Salsbury, 2003; Karpovich et al., 2009; Zervanos et al., 2010) but sometimes a higher temperature (Lehmer and Biggins, 2005). We could not use this method to compare all squirrels because three study animals (2 males and 1 female) never achieved skin temperatures [greater than or equal to]30 C during 1-3 well-defined arousal bouts. Thus, we followed an alternative published method (Hut et al., 2002): torpor-bout duration was calculated from the recording just before the first drop in a sustained skin-temperature decline going into torpor until the recording just before arousal at the end of that torpor bout. Arousal bout duration was calculated from the recording just before the first increase in body temperature during arousal to the recording just before body temperature dropped with the onset of torpor. In addition, to facilitate comparison with previous studies, we compared torpor-bout durations obtained by both methods (paired t-tests) for five squirrels with all arousal bouts reaching skin temperatures [greater than or equal to]30 C.

We graphically explored seasonal trends in torpor-bout duration, minimum skin temperature during torpor (torpor depth), and arousal-bout duration by first binning, for each individual, each torpor and arousal bout to the month that includes its midpoint. Then, we calculated and plotted the mean monthly values per individual through the heterothermal period.

We compared male and female ground squirrels for duration and date of entry into and exit from the heterothermal period; number of torpor bouts during the heterothermal period; lowest skin temperature during hibernation; overall and monthly average torpor- and arousal-bout durations; monthly average minimum skin temperatures during torpor; and body mass at first capture (for squirrels first captured 22-27 Aug.). Body masses measured within 5 d of first torpor or final arousal were used as estimates of body mass at entry into and termination of hibernation, respectively. If both pre- and posthibernation body masses were available for a single individual, we estimated mass loss during hibernation as a percentage of prehibernation mass.

We examined associations among torpor-bout duration, minimum skin temperature per torpor bout, and soil temperature measured at ~10-cm depth beneath sod. Soil temperatures were obtained from the closest weather station to record these data (Wanatah, IN, 74 km SW of the study site). Unfortunately, this station did not record soil temperatures at depths >10 cm, which would be more relevant to the soil conditions experienced by hibernating squirrels. We also examined associations between prehibernation body mass and Julian day, and between posthibernation body mass and number of days after final arousal.

Graphical and statistical analyses were done with SPSS 19.0 (IBM, Armonk, New York) using non-parametric tests for sample comparisons due to small sample sizes. Male and female values were compared with a Mann-Whitney U-test, and monthly values were compared with the Kruskal-Wallis test. Linear regression was used to estimate relationships between correlated variables that co-varied linearly. In two cases (prehibernation body mass predicted by Julian d, and minimum skin temperature during torpor predicted by soil temperature), these relationships were descriptive but were not evaluated statistically because cases were not independent (all squirrels experienced multiple torpor bouts and some were weighed multiple times). All differences were considered significant at P < 0.05.

[FIGURE 1 OMITTED]

RESULTS

All thirteen-lined ground squirrels collared in this study were juveniles in 2007 with animals weighing 91-137 g at first capture (22 Aug.-12 Oct. 2007). Males and females first captured 22-27 Aug. did not differ in body mass at initial capture (males: [bar.x] = 108 g, SD = 13.2, n = 9; females: [bar.x] = 101 g, SD = 12.4, n = 4; P > 0.4). Animals that survived to hibernation gained an average of 0.53 g/d during late summer and early fall (Fig. 1A; not evaluated for significance because cases were not independent). Animals with relatively low body mass at first capture tended to gain mass progressively (e.g., collared males 1 and 14, female 15); animals with average or greater-than-average body mass at first capture initially gained mass but subsequently maintained, or even lost, body mass (e.g., collared male 6, females 3 and 4b; Fig. 1A). Three animals last captured within 5 d of entering the heterothermal period weighed 132-144 g at last capture (Fig. 1A).

Eight of nine ground squirrels (5 males and 3 females) recaptured in spring had functional data loggers with overwinter skin-temperature records. Six of these eight squirrels were last sighted aboveground 2-5 d before onset of torpor, three on the last day the colony was observed before they became torpid. Two squirrels (male with collar 12, female with collar 15) were last sighted 9 and 15 d, respectively, before onset of torpor, and they were not seen during 2 and 5 observation days, respectively, between last sighting and onset of heterothermy.

[FIGURE 2 OMITTED]

In spring, five of eight squirrels with functional data loggers were first sighted aboveground 1-4 d after terminal arousal. The remaining three squirrels, all males (collars 1, 6, and 14), were not seen aboveground for 8-16 d (3-6 observation days) after terminal arousal. Two of these males (collars 1 and 6) were seen sporadically over the first 1-2 wk after first sighting (on 1 of 4 observation days for collar 1; on 2 of 10 observation days for collar 6) before they were observed consistently aboveground.

All squirrels with functional data loggers in spring displayed a characteristic skin-temperature profile during the heterothermal period: a series of 11-22 torpor bouts punctuated by brief arousal bouts (Fig. 2). Male squirrels spent significantly less time in the heterothermal period than did females: they entered the first torpor bout later (not statistically significant), initiated final arousal significantly earlier and exhibited fewer torpor bouts (approached significance: P < 0.10) than did females (Table 1). Males and females did not differ in lowest skin temperature during hibernation, durations of torpor and arousal bouts or proportion of the heterothermal period spent in torpor (Table 1).

Torpor bouts obtained in this study averaged 0-1.5 h shorter than bouts delimited with the 30 C threshold (available for 5 squirrels). The difference was significant for collared animal 15 (paired t = -3.5, DF = 16, P < 0.01) but not for other squirrels (collar 4: t = 0.0, DF = 21, P = 1.0; collar 6: t = -1.1, DF = 10, P = 0.3; collar 13: t = -1.4, DF = 12, P = 0.2; collar 14: t = -0.2, DF = 16, P = 0.8).

Averaged by month, torpor-bout depth (Fig. 3A) and duration (Fig. 3B) varied systematically and significantly through the heterothermal period (Kruskal-Wallis test statistic = 44.4 and 41.2, respectively; P < 0.001, PF = 6 for both variables): from shallow, short bouts in Oct. to deep, long bouts in Feb. Average arousal-bout duration (Fig. 3C) showed no trend across the heterothermal period (Kruskall-Wallis test statistic = 6.0, P > 0.40, DF = 6). In Mar., average torpor bouts were always deep (Fig. 3A) ; but both average torpor and arousal bouts varied markedly in duration (Figs. 3B-C). In Apr., torpor bouts were shallower and shorter (Figs. 3A-B). Males and females did not differ in torpor-bout depth or duration, or in arousal-bout duration, during any month [standardized Mann-Whitney test statistic by month and variable (torpor-bout depth, torpor-bout duration, arousal-bout duration): Oct. (-0.58, -1.16, -0.58), Nov. (0.45, -1.34, 1.04), Dec. (-0.15, -1.64, 0.30),Jan. (-0.15, -0.45, 1.51), Feb. (-0.74, 0.60, 0.77), Mar. (1.34, -1.04, -1.64), Apr. (-1.34, 0.45, --); P > 0.10 in all comparisons], even during Mar. when three males showed markedly shorter average torpor bouts and longer average arousal bouts (Figs. 3B-C).

Across individual torpor bouts, regional minimum soil temperatures measured at 10 cm below ground surface accounted for 96% of variation in minimum skin temperatures per torpor bout ([R.sup.2] = 0.96, y = 1.2 x + 2.9; not evaluated statistically because cases are not independent), a pattern also evident across monthly average values (Fig. 3A). An episode of abrupt soil warming from 8-11 Jan. 2008 corresponded to an increase in torpid skin temperature, or to arousal, in all individuals (Fig. 2). Duration of torpor scaled inversely with both soil (not shown) and skin temperature, although the relationship was not linear (Fig. 4).

[FIGURE 3 OMITTED]

Three males (collars 1, 6, and 13) differed from other squirrels in multiple attributes of the heterothermal period. They initiated the first torpor bout late, terminated the heterothermal period early, exhibited 1-3 short torpor bouts (relative to skin temperature) in Nov. and experienced short torpor bouts (both absolute and relative to skin temperature) and long arousal bouts in Mar. (Figs. 2-4).

Two females captured within 5 d after final arousal weighed 80-81 g whereas one male captured 6 d after final arousal weighed 95 g (Fig. 1B). One female was weighed within 5 d of entering and terminating the heterothermal period (collar 4b), and she lost 39% of body mass during hibernation. On average, ground squirrels gained ~1.5 g/d after final arousal although one male (collar 13b) probably gained weight more rapidly (Fig. 1B).

[FIGURE 4 OMITTED]

Immediately following termination of the heterothermal period in spring, four males displayed relatively low skin temperatures for 2-13 d, followed by return to normal active skin temperatures (Fig. 2). Two of these individuals were seen aboveground during this interval.

DISCUSSION

Free-ranging thirteen-lined ground squirrels alternated bouts of torpor with short bouts of arousal throughout the heterothermal period (Fig. 2) with average monthly duration and depth of torpor-bouts generally increasing from onset of the heterothermal period to Feb. and decreasing thereafter (Fig. 3). A similar pattern characterizes other obligate hibernators within Marmotini for which overwinter body temperatures have been obtained under field conditions (Wang, 1973, 1979; Young, 1990; Michener, 1992; Barnes and Buck, 2000; Florant et al., 2000; Hut et al., 2002; Humphries et al., 2003; Zervanos and Salsbury, 2003; Lehmer and Biggins, 2005; Buck et al., 2008; Frank et al., 2008; Lee et al., 2009; Zervanos et al., 2010). Duration of torpor varied inversely with minimum skin temperature during torpor (Fig. 4) and ambient temperature through most of hibernation. This is consistent with previous studies of hibernation in captivity (Twente and Twente, 1965; Pivorun, 1976, 1977; Twente et al., 1977; French, 1982; Geiser and Kenagy, 1988; Geiser et al., 1990; Barnes and Buck, 2000; Buck and Barnes, 2000; Strijkstra et al., 2008) and the field (Michener, 1992; Barnes and Buck, 2000) at ambient soil temperatures >0 C, below which animals typically defend body temperature and decrease duration of torpor (Pivorun, 1976; Geiser and Kenagy, 1988; Buck and Barnes, 2000). Some species of Urocitellus tolerate skin temperatures [Richardson's ground squirrels (Urocitellus richardsonii)--Michener, 1992] or even core temperatures below 0 C during torpor (arctic ground squirrels-Barnes, 1989), but we have no evidence for this in thirteen-lined ground squirrels (lowest skin temperature recorded was 1.0 C).

The inverse relationship between duration of torpor and skin temperature broke down for three of five male ground squirrels late in hibernation (Fig. 3B) when their torpor bouts were substantially shorter than expected based on skin temperatures (Fig. 4), a pattern often displayed by both male and female Columbian (Urocitellus columbianus) and Richardson's ground squirrels (Young, 1990; Michener, 1992). Also, these three males exhibited exceptionally long late-hibernation arousal bouts (Fig. 3C) as is common for free-ranging arctic ground squirrels (Buck et al., 2008) and woodchucks (Marmota monax-Zervanos and Salbury, 2003). Decrease in duration of torpor or increase in duration of arousal late in hibernation appear to be endogenously controlled and may allow squirrels to assess environmental conditions more frequently in preparation for final arousal (Michener, 1992). We do not know why only three of five males, and no females, displayed these patterns in thirteen-lined ground squirrels.

Skin temperatures closely tracked soil temperatures, measured regionally (but not locally) at 10-cm depth, throughout hibernation (Fig. 3A). A short episode of abrupt soil warming in early Jan. reached the depth of hibernacula and elicited transient increases in torpid skin temperatures, or arousal, in all individuals (Fig. 2). Thus, hibernacula remained sensitive to transient but abrupt increases in winter temperature even though thirteen-lined ground squirrels reportedly excavate their burrows to just below winter frost line and plug them when entering hibernation (Rongstad, 1965; Desha, 1966).

All thirteen-lined ground squirrels captured in our study were juveniles; apparently most or all adults had entered hibernation before commencement of our study in late Aug. Adults of this species commonly cease aboveground activity by Aug. or Sept. with juveniles persisting as late as early Nov. (McCarley, 1966; Clark, 1971), consistent with our finding that juveniles commenced the heterothermal period in Oct. and early Nov. Adults typically enter hibernation before juveniles in marmotine hibernators (Michener, 1984) although late persistence of adult males has been reported for arctic ground squirrels (McLean and Towns, 1981; Buck et al., 2008) and one population of thirteen-lined ground squirrels (Rongstad, 1965). Late entry of juvenile males into the heterothermal period [on average, 3 wk later than juvenile females in our study (Table 1)] has been observed for Richardson's ground squirrels and may allow juvenile males to gain sufficient body mass to compete with adult males for breeding rights the following spring (Michener, 1992). We have insufficient data to rigorously test this hypothesis, but the three squirrels with largest body masses at last fall capture were all males (Fig. 1A).

Despite late onset of the heterothermal period in fall, juvenile male thirteen-lined ground squirrels terminated heterothermy, on average, 3.5 wk before juvenile females in spring, and they thus spent, on average, about 6.5 wk less time in the heterothermal period (149.5 d versus 193.5 d, respectively-Table 1). Richardson's ground squirrels display a similar pattern: juvenile males terminate heterothermy, on average, several weeks before juvenile females and spend substantially less time in the heterothermal period (114 d versus 181 d, respectively, in one study season--Michener, 1992). Early spring arousal is common for young male ground squirrels that are reproductive after their first winter (reviewed in Michener, 1984), thus juvenile males in our study probably were reproductively mature in spring. In contrast, young males that remain nonreproductive after their first winter typically end hibernation later in their first spring and thus spend substantially more time in the heterothermal period [e.g., Columbian (Murie and Harris, 1982; Young, 1990) and arctic ground squirrels (Buck and Barnes, 1999; Buck et al., 2008)]. Early spring arousal allows reproductive males to mature, establish territories and monitor the area for emerging females so that mating occurs soon after female emergence (Michener, 1984).

The majority of ground squirrels in our study (6 of 8) showed little lag between last aboveground sighting in fall and the onset of torpor ([less than or equal to]5 d), or between terminal arousal and first spring sighting (5 of 8; [less than or equal to]4 d), but other individuals exhibited substantial lags [1415 d in fall (a male and female), 8-16 d in spring (3 males)]. We are not confident that these lags indicate extended sequestration before or after the heterothermal period because several squirrels were not observed for 2-3 consecutive observation days prehibernation, and two of three males with substantial lags between final arousal and first aboveground sighting posthibernation were observed infrequently for 1-2 wk after first observation. Early arousing males appear to have been relatively inactive before 1 Apr. although we did not quantify h/d spent on the colony early in the active season.

Perhaps consistent with low levels of aboveground activity early in spring, four of five males exhibited relatively low skin temperatures for 2-13 d after terminating the heterothermal period (Fig. 2). We initially hypothesized this to represent male sequestration underground between termination of torpor and emergence from hibernation, a behavior reported for several species of Urocitellus (Columbian ground squirrels-Young, 1990; Richardson's ground squirrels--Michener, 1992; arctic ground squirrels-Barnes and Ritter, 1993). In these species, males, but not females, eat cached food for several days before emerging (Michener, 1992). This hypothesis was contradicted by our observation that two of these individuals were aboveground during their episodes of low skin temperatures.

Alternatively, low skin temperatures after hibernation may reflect inadequate skin-collar contact due to hibernation weight loss, but it is not clear why this would differentially affect males. Variable skin-collar contact may cause skin temperatures to underestimate body temperatures by 0-8 C in active mammals, although skin temperatures more closely track body temperatures ([+ or -]2 C) in mammals curled up to conserve heat during torpor (Humphries et al., 2003). This may explain why skin temperatures sometimes did not reach 30 C during arousal bouts within hibernation (Fig. 2).

Juvenile thirteen-lined ground squirrels in our study gained body mass, on average, less rapidly late in the active season (~0.5 g/d from late Aug. to early Nov.; Fig. 1A) than previously reported early in the season from Minnesota (~1.4-1.9 g/d, depending on year, between Jun. and late Aug.; Hohn and Marshall, 1966). Growth trajectories varied among individuals: mass tended to increase throughout fall for animals with below-average body mass at first capture but tended to stabilize, or even decrease, late in the season for animals with average or above-average body mass at first capture (Fig. 1A). The latter pattern is common for ground squirrels as they approach hibernation (Davis, 1976).

Overwinter weight loss estimated for one juvenile female thirteen-lined ground squirrel (39% of prehibernation body mass) was similar to weight loss reported for juvenile females of Richardson's ground squirrels [40-48% depending on season (Michener and Locklear, 1990; Michener, 1992)]. Animals captured within 10 d after final arousal all weighed <100 g at capture, but squirrels gained weight rapidly over the first 1.5 mo of the active season (Fig. 1B). In summary, this paper incorporated an additional marmotine genus--Ictidomys--into comparative studies of overwinter body temperatures among free-ranging hibernators. Thirteen-lined ground squirrels exhibited several universal features of hibernation: alternating bouts of prolonged torpor and brief arousal, with torpor depth and duration varying systematically through the heterothermal period and inversely with ground temperature. All study animals were juveniles; most adults probably immerged into hibernation before our study commenced in late Aug. Compared to females, males entered the heterothermal period later, terminated heterothermy earlier, and spent significantly less time in the heterothermal period. Some males, but no females, exhibited exceptionally short torpor bouts and long arousal bouts late in the heterothermal period. We also presented preliminary data on body-mass changes before, during and after hibernation in thirteen-lined ground squirrels.

Acknowledgments.--Andrews University supported this research with a Faculty Research Grant (to HTG) that included a graduate research stipend (for BK). Wayne Spletzer and his crew allowed us to use the Rosehill Cemetery grounds as our study site. Shaun Kisser assisted with animal capture, collaring, and observation throughout this project. Two anonymous reviewers provided helpful, critical feedback on an earlier draft of this paper.

LITERATURE CITED

BARNES, B. M. 1989. Freeze avoidance in a mammal: body temperatures below 0 degrees C in an Arctic hibernator. Science, 244:1593-1595.

--AND C. L. BUCK. 2000. Hibernation in the extreme: burrow and body temperatures, metabolism, and limits to torpor bout length in arctic ground squirrels, p. 65-72. In: G. Heldmaier and M. Klingenspor (eds.). Life in the cold: eleventh international hibernation symposium. Springer, Berlin.

--AND D. RITTER. 1993. Patterns of body temperature change in hibernating arctic ground squirrels, p. 119-130. In: C. Carey, G. L. Florant, B. A. Wunder and B. Horwitz (eds.). Life in the cold: ecological, physiological, and molecular mechanisms. Westview Press, Boulder, Colorado.

BEER, J. R. 1962. Emergence of thirteen-lined ground squirrels from hibernation. J. Mammal., 43:109.

BUCK, C. L. AND B. M. BARNES. 1999. Annual cycle of body composition and hibernation in free-living arctic ground squirrels. J. Mammal., 80:430-442.

-- AND --. 2000. Effects of ambient temperature on metabolic rate, respiratory quotient, and torpor in an arctic hibernator. Am. J. Physiol. Regul. Integr Comp. Physiol., 279:R255-R262.

--, A. BRETON, F. KOHL, O. TOIEN AND B. M. BARNES. 2008. Overwinter body temperature patterns in free-living arctic squirrels (Spermophilus parryii), p. 317-326. In: B. G. Lovegrove and A. E.. McKechnie (eds.). Hypometabolism in animals: hibernation, torpor and cryobiology. University of KwaZulu-Natal, Pietermaritzburg, South Africa.

CLARK, T. W. 1971. Notes on the biology of the thirteen-lined ground squirrel in the Laramie Plains, Wyoming. Southwest. Nat., 15:499-502.

CRIDDLE, S. 1939. The thirteen-striped ground squirrel in Manitoba. Can. Field-Nat., 56:1-6.

DAVIS, D. E. 1976. Hibernation and circannual rhythms of food consumption in marmots and ground squirrels. Quar. Rev. Biol., 51:477-514.

DESHA, P. G. 1966. Observations on the burrow utilization of the thirteen-lined ground squirrel. Southwest. Nat., 11:408-410.

FLORANT, G. L., V. HILL AND M. D. OGILVIE. 2000. Circadian rhythms of body temperature in laboratory and field marmots (Marmota flaviventris), p. 223-231. In: G. Heldmaier and M. Klingenspor (eds.). Life in the cold: eleventh international hibernation symposium. Springer, Berlin.

FRANK, C. L., S. KARPOVICH AND B. M. BARNES. 2008. Dietary fatty acid composition and the hibernation patterns in free-ranging arctic ground squirrels. Physiol. Biochem. Zool., 81:486-495.

FRENCH, A. R. 1982. Intraspecific differences in the pattern of hibernation in the ground squirrel Spermophilus beldingi. J. Comp. Physiol. B., 148:83-91.

GANNON, W. L., R. S. SIKES AND COMMITTEE. 2007. Guidelines of the American Society of Mammalogists for the use of wild mammals in research. J. Mammal., 88:809-823.

GEISER, F., S. HIERERT AND G.J. KENAGY. 1990. Torpor bout duration during the hibernation season of two sciurid rodents: interrelations with temperature and metabolism. Physiol. Zool., 63:489-503.

--, J. C. HOLLOWAY, G. KORTNER, T. A. MADDOCKS, C. TURBILL AND R. M. BRIGHAM. 2000. Do patterns of torpor differ between free-ranging and captive mammals and birds?, p. 95-102. In: G. Heldmaier and M. Klingenspor (eds.). Life in the cold: eleventh international hibernation symposium. Springer, Berlin.

--AND G. J. KENAGY. 1988. Torpor duration in relation to temperature and metabolism in hibernating ground squirrels. Physiol. Zool., 6:442-449.

HOHN, B. AND W. MARSHALL. 1966. Annual and seasonal weight changes in a thirteen-lined ground squirrel population, Itasca State Park, Minnesota. J. Minn. Acad. Sci., 33:102-106.

HUMPHRIES, M. M., D. L. KRAMER AND D. W. THOMAS. 2003. The role of energy availability in mammalian hibernation: an experimental test in free-ranging eastern chipmunks. Physiol. Biochem. Zool., 76:180-186.

HUT, R. A., B. M. BARNES AND S. DAAN. 2002. Body temperature patterns before, during, and after semi-natural hibernation in the European ground squirrel. J. Comp. Physiol. B., 172:47-58.

KARPOVICH, S. A., O. TOIEN, C. L. BUCK AND B. M. BARNES. 2009. Energetics of arousal episodes in hibernating arctic ground squirrels, J. Comp. Physiol. B., 179:691-700.

LEE, T. N., B. M. BARNES AND C. L. BUCK. 2009. Body temperature patterns during hibernation in a free-living Alaska marmot (Marmota browen). Ethol. Ecol. Evol., 21:403-413.

LEHMER, E. M. AND D. E. BIGGINS. 2005. Variation in torpor patterns of free-ranging black-tailed and Utah prairie dogs across gradients of elevation. J. Mammal., 86:15-21.

MCCARLEY, H. 1966. Annual cycle, population dynamics and adaptive behavior of Citellus tridecemlineatus. J. Mammal., 47:294-316.

MCLEAN, I. G. AND A. J. TOWNS. 1981. Differences in weight changes and the annual cycle of male and female arctic ground squirrels. Arctic, 34:249-254.

MICHENER, G. R. 1984. Age, sex, and species differences in the annual cycles of ground-dwelling sciurids: implications for sociality, p. 81-107. In: J. O. Murie and G. R. Michener (eds.). The biology of ground-dwelling squirrels. University of Nebraska Press, Lincoln, Nebraska.

--. 1992. Sexual differences in over-winter torpor patterns of Richardson's ground squirrels in natural hibernacula. Oecologia, 89:397-406.

--AND L. LOCKLEAR. 1990. Over-winter weight loss by Richardson's ground squirrels in relation to sexual differences in mating effort. J. Mammal., 71:489-499.

MURIE, J. O. AND M. A. HARRIS. 1982. Annual variation of spring emergence and breeding in Columbian ground squirrels (Spermophilus columbianus). J. Mammal., 63:431-439.

PIVORUN, E. B. 1976. A biotelemetry study of the thermoregulatory patterns of Tamias striatus and Eutamias minimus during hibernation. Comp. Biochem. Physiol. A., 53:265-271.

--. 1977. Hibernation of a southern subspecies of Tamias striatus:, thermoregulatory patterns. Am. Midl. Nat., 98:495-499.

RONGSTAD, O.J. 1965. A life history study of thirteen-lined ground squirrels in southern Wisconsin. J. Mammal., 46:76-87.

SCHECK, S. H. AND E. D. FLEHARTY. 1980. Subterranean behavior of the adult thirteen-lined ground squirrel (Spermophilns tridecemlineatus). Am. Midl. Nat., 103:191-195.

STRIJKSTRA, A. M., R. A. HUT AND S. DAAN. 2008. Does the timing mechanism of periodic euthermia in deep hibernation depend on its homeostatic need?, p. 157-162. In: B. G. Lovegrove and A. E. McKechnie (eds.). Hypometabolism in animals: hibernation, torpor and cryobiology. University of KwaZulu-Natal, Pietermaritzburg, South Africa.

TWENTE, J. W. AND J. A. TWENTE. 1965. Regulation of hibernating periods by temperature. Proc. Natl. Acad. Sci., 54:1058-1061.

--, -- AND R. M. MOY. 1977. Regulation of arousal from hibernation by temperature in three species of Citellus. J. Appl. Physiol., 42:191.

WADE, O. 1927. Breeding habits and early life of the thirteen-striped ground squirrel, Citellus tridecemlineatus (Mitchill). J. Mammal., 8:269-276.

WANG, L. C. H. 1973. Radiotelemetric study of hibernation under natural and laboratory conditions. Am. J. Physiol., 224:673-677.

--. 1979. Time patterns and metabolic rates of natural torpor in the Richardson's ground squirrel. Can. J. Zool., 57:149-155.

YOUNG, P. J. 1990. Hibernating patterns of free-ranging Columbian ground squirrels. Oecologia, 83:504-511.

ZERVANOS, S. M., C. R. MAHER, J. A. WALDVOGEL AND G. L. FLORANT. 2010. Latitudinal differences in the hibernation characteristics of woodchucks (Marmota monax). Physiol. Biochem. Zool., 83:135-141.

--AND C. M. SALISBURY. 2003. Seasonal body temperature fluctuations and energetic strategies in free-ranging eastern woodchucks (Marmota monax). J. Mammal., 84:299-310.

SUBMITTED 9 FEBRUARY 2010

ACCEPTED 4 NOVEMBER 2011

BROOKE KISSER (1) AND H. THOMAS GOODWIN (2)

Department of Biology, Andrews University, Berrien springs, Michigan 49104

(1) Present address: Biology Department, Everett Community College, Everett, Washington 98201; Telephone: (425) 388-9043; e-mail: bkisser@everettcc.edu

(2) Corresponding author: e-mail: goodwin@andrews.edu
TABLE 1.--Comparison of male (n = 5) and female (n = 3) juvenile
Ictidomys tridecemlineatus on  various attributes of the
heterothermal period during hibernation

                                    Male             Female

                              [bar.x]    SD    [bar.x]    SD

Date of first torpor          31 Oct.   12.9   10 Oct.   10.2
Date of final arousal         28 Mar.    9.2   21 Apr.    1.5
Duration of heterothermal
 period [first torpor to
 lastarousal (d)]               149.5   21.6     193.5   11.8
Number of torpor bouts           14.8    2.7      20.0    2.6
Minimum skin temperature
during hibernation (C)            1.4    0.3       1.7    0.1
Torpor-bout duration (d)          9.6    0.8       9.2    0.8
 ([diamond])
Arousal-bout duration (h)        15.8    3.5      12.9    1.1
 ([diamond])
Proportion of heterothermal
 period in torpor (%)            94.0    1.3      94.7    0.7

                               Mann-
                               Whitney
                              U[dagger]

Date of first torpor           1.64
Date of final arousal         -2.24 *
Duration of heterothermal
 period [first torpor to
 lastarousal (d)]             -2.24 *
Number of torpor bouts        -1.99
Minimum skin temperature
during hibernation (C)        -1.65
Torpor-bout duration (d)       0.45
 ([diamond])
Arousal-bout duration (h)      1.04
 ([diamond])
Proportion of heterothermal
 period in torpor (%)         -0.75

([dagger]) Standardized test statistic

* P < 0.05

([diamond]) Based on mean value across the heterothermal
period for each squirrel
Gale Copyright: Copyright 2012 Gale, Cengage Learning. All rights reserved.