2006 SOUTHEASTERN NATURALIST 5(1):175–183 Monthly and Daily Activity of a Fossorial Lizard, Neoseps reynoldsi KYLE G. ASHTON1,2,*, AND SAM R. TELFORD, JR.3 Abstract - Here we evaluate monthly and daily activity patterns of Neoseps Reynoldsi (Sand Skink), through examination of museum specimens and population studies at Archbold Biological Station (ABS), Highlands County, FL, and in Ocala National Forest (ONF), Marion County, FL. Sand Skinks are active throughout the year, with highest captures in February through May, and August through October. Museum specimen data suggest a peak in December, but this reflects collector bias. Number of captures is significantly negatively related to precipitation and positively related to temperature for Sand Skinks at ABS, but this relationship is due to capture of hatchlings from July through October. At ONF, no relationship between temperature and precipitation and number of captures exists. The number of Sand Skink tracks per day at ABS is not related to temperature, but is significantly negatively related to precipitation. Introduction Temperature and precipitation influence activity of lizards on a daily, monthly, and seasonal basis (e.g., Grant and Dunham 1990, Huey and Pianka 1977, Pianka 1986). In tropical areas, which have relatively aseasonal temperatures but highly seasonal patterns of precipitation (wet and dry seasons), activity of lizards tends to be greatest during the wet season (e.g., James 1994, Lister and Aguayo 1992). In contrast, in temperate regions, which have relatively aseasonal patterns of precipitation but highly seasonal temperatures, activity of lizards tends to be limited by very warm or very cold temperatures (e.g., Fitch 1958, Grant and Dunham 1990, Pianka 1970). In addition to temperature and precipitation, the timing of reproductive events also influences activity of lizards. In particular, lizards tend to be more active during the breeding season (Fitch 1958, Lister and Aguayo 1992, Pianka 1970), but it is difficult to separate the influence of temperature and precipitation from reproductive phenology because the timing of reproduction is itself cued by temperature and precipitation (Duvall et al. 1982). Regardless, it is well established that temperature, precipitation, and timing of reproductive events (particularly breeding) influence activity of lizards. 1Archbold Biological Station, PO Box 2057, Lake Placid, FL 33862. 2Current address - Division of Biological Sciences, University of California at San Diego, 9500 Gilman Drive, MC 0116, La Jolla, CA 92093. 3The Florida Museum of Natural History, University of Florida, Gainesville, FL 32611. *Corresponding author - neoseps@hotmail.com. 176 Southeastern Naturalist Vol. 5, No. 1 Here we evaluate monthly and daily activity patterns of Neoseps reynoldsi Stejneger (Sand Skink) relative to temperature, precipitation, and timing of reproductive events. Activity patterns of Sand Skinks are of interest for at least two reasons. First, most research on the effects of temperature and precipitation on activity patterns of lizards has been conducted in temperate or tropical regions. Sand Skinks occur in a subtropical environment. Subtropical environments differ from temperate and tropical environments because they exhibit seasonality in both precipitation and temperature, providing an opportunity to evaluate lizard activity patterns in response to both environmental cues. Second, most studies of activity have focused on terrestrial or arboreal lizards. Sand Skinks are fossorial and, despite the presence of fossorial lizards on at least three different continents (Pianka 1986), activity patterns of fossorial lizards are poorly known. Methods Patterns of monthly activity To evaluate monthly patterns of activity of Sand Skinks, we used three datasets. The first dataset consists of museum specimens. We examined all available museum specimens (list available from K.G. Ashton) and compiled month of collection for specimens with known collection dates (n = 225). For this dataset, individuals collected throughout the range of Sand Skinks are lumped together. The second dataset is from a population study conducted from February 2002 through January 2003 near the southern extent of the range of Sand Skinks at Archbold Biological Station (ABS), Highlands County, FL (464758, 3008716 UTM). Fifty-six bucket traps were placed in long-unburned rosemary scrub habitat. Traps were arranged in a 7 x 8 grid, with traps spaced 6.7 m apart (total trapping area = 40 x 47 m). Each 18.9-L bucket was sunk a few cm below the sand surface and covered with a 30- x 30-cm lid of masonite or hard plastic propped 2 cm above the bucket by four wooden pegs. Sand was placed in the bottom of each bucket. Traps were checked every two to three days, and water was added to the sand at each checking. Each Sand Skink was measured and marked (toe clipping) in the field, then released immediately after capture. The third dataset consists of individuals captured in a population study conducted from June 1997 through May 1998 near the northern extent of the range of Sand Skinks in Ocala National Forest (ONF), Marion County, FL (R25E, T12S, S4). This study took place in Pinus palustris Mill (longeaf pine)-Quercus laevis Walt (turkey oak) sandhill habitat. A 20- x 20-m enclosure was constructed of aluminum flashing buried 30 cm in the sand. Sixteen pitfall drift fences, including 2 m of aluminum flashing with two 3.8-L buckets sunk into the sand at each end, were evenly spaced within the 2006 K.G. Ashton, and S.R. Telford, Jr. 177 enclosure. An additional 3.8-L bucket was placed every 5 m along the interior and exterior of the enclosure flashing. Additional buckets were placed at 5-m intervals, extending 10 m in each direction, outside of the enclosure. The total trapping area was 40 x 40 m and included 148 buckets. Each bucket contained at least 10 cm of sand, was countersunk to a depth of 13 cm and covered by a masonite board (41 cm2) resting on the surface of the sand. Each bucket was checked once per week, and all captures were taken to a laboratory for processing. Snout-vent length was measured for each individual. All individuals were released at capture sites within three to seven days of capture. For each site, we partitioned captures into hatchlings (≤ 35 mm SVL), juveniles (> 35 mm, but < 45 mm SVL) and adults (≥ 45 mm SVL; Sutton 1996, Telford 1959). We plotted the number of individuals of each age category caught per month for each site to examine monthly variation in activity and evaluate the effect of hatchling emergence on number of captures. We used multiple regression analysis in combination with backward stepwise multiple regression to test for relationships between mean maximum temperature, total precipitation, and the total number of lizards captured per month at each site. Prior to each analysis, data that violated the assumptions of regression analysis were transformed using log (X+1) transformation, except in one case (ABS temperature data) with leftskewed data, for which we used the X2 transformation. Weather data for the ONF site (Fig. 1) were obtained from the Florida Climate Center (http://www.coaps.fsu.edu/climate_center/) by averaging data from weather stations at Ocala and Crescent City, FL, because the study site was geographically intermediate. Weather data for ABS (Fig. 1) were collected by staff at ABS. Patterns of daily activity To evaluate the effect of temperature and precipitation on daily activity, we used information from a study of Sand Skink tracks conducted from November 2001 through October 2002 at ABS (a different area and habitat than used for the population study at ABS). For this study, we used a 3- x 75-m area of scrubby flatwoods habitat that was adjacent to an 8-m wide sand firelane. Surveys of the scrubby flatwoods area were conducted for 14 days each month. Each day of monitoring consisted of one of us (KGA) walking the study plot and counting the number of Sand Skink tracks (distinct, sinusoidal tracks; Andrews 1994, Telford 1959) present. Tracks were considered distinct if discontinuous from other tracks and located > 1 m from any other track. Given that rainfall erases tracks, it is important that surveys are conducted before rains occur. On days without rain (mostly during the non-rainy season from October through May), track surveys were conducted at the end of the day, just before sunset. On days with rain (predicted from forecasts and mostly 178 Southeastern Naturalist Vol. 5, No. 1 during the rainy season from June through September), track surveys were conducted just prior to the onset of rains. In this way tracks could be Table 1. Mean number of Sand Skink tracks, mean maximum air temperature, and mean total precipitation per day for each 14-day monthly sampling period from November 2001 through October 2002 at one site at Archbold Biological Station, Highlands County, FL. For simplicity, data are presented as January through December. Tracks = mean total number of Sand Skink tracks/day; Temp. = mean daily maximum air temperature (°C); Precip. = mean total precipitation/ day (cm). Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Tracks 3.1 0.5 2.1 4.1 1.4 0 0.8 1.9 1.1 1.4 2.9 3.3 Temp. 28.0 24.6 28.2 32.0 34.2 32.5 33.7 33.4 32.6 30.7 27.4 27.3 Precip. 0.00 0.08 0.01 0.00 0.55 1.76 0.92 0.11 1.10 0.22 0.00 0.03 Figure 1. Temperature and precipitation for A) Archbold Biological Station (February 2002 through January 2003), and B) Ocala National Forest (from average of weather data from June 1997 through May 1998 for Ocala and Crescent City weather stations). For simplicity, data are presented as January through December. 2006 K.G. Ashton, and S.R. Telford, Jr. 179 checked before being erased by rainfall. If rains stopped prior to sunset, a track survey was also conducted just before sunset. In cases where more than one track survey was conducted on the same day, the number of tracks was summed to produce the total number of tracks for that day. It is assumed that Sand Skink tracks were not made during rainfall. All Sand Skink tracks were wiped away by hand as each survey was conducted. Thus, any tracks present on subsequent surveys were caused by new surface activity of Sand Skinks occurring after the last survey. Maximum air temperature and total precipitation were compiled for each day that tracks were counted. To evaluate the relationship between surface activity (as indicated by tracks), air temperature, and precipitation, we first calculated the mean number of tracks/day, mean maximum air temperature/ day, and mean total precipitation/day for each monthly monitoring period (Table 1). Prior to conducting multiple regression and backward stepwise multiple regression analyses to test the dependence of Sand Skink tracks on precipitation and air temperature across months, we log (X+1) transformed precipitation data. Results Patterns of monthly activity Museum specimens were collected during every month, indicating yearround activity (Fig. 2a). Most individuals were collected from February through May (peak in February), with relatively few individuals collected throughout the rest of the year, with the exception of December. At ABS, Sand Skinks were captured in every month except January (Fig. 2b). Most captures occurred between February and May (peak in March). Captures were few for all other months, except August. At ONF, Sand Skinks were captured from April to June (peak in May), in August and October, but not in any other month (Fig. 2c). At ABS, number of Sand Skinks captured per month was significantly related to monthly mean maximum temperature and total monthly precipitation (multiple r2 = 0.56, df = 2,9, F = 5.82, P < 0.05), with a positive relationship between monthly captures and temperature (β = +0.76) and a negative relationship between monthly captures and precipitation (β = -0.60). However, backward stepwise multiple regression analysis revealed that neither temperature nor precipitation alone were significant predictors of monthly captures at ABS. At ONF, monthly captures was not related to temperature or precipitation (multiple r2 = 0.34, df = 2,9, F = 2.27, P > 0.05). Analyses of monthly activity of Sand Skinks are confounded by captures of hatchlings in July through October (Fig. 2) because new individuals are added to the population. Thus, we also tested for relationships between monthly captures and temperature and precipitation using only captures of 180 Southeastern Naturalist Vol. 5, No. 1 Figure 2. Monthly distribution of captures based on A) museum specimens, B) a population study at Archbold Biological Station from Feb. 2002 through Jan. 2003, and C) a population study at Ocala National Forest from June 1997 through May 1998. For simplicity, data are presented as January through December. 2006 K.G. Ashton, and S.R. Telford, Jr. 181 juveniles and adults. Number of juveniles and adults collected each month at ABS (multiple r2 = 0.38, df = 2,9, F = 2.77, P > 0.05) and at ONF (multiple r2 = 0.32, df = 2,9, F = 2.08, P > 0.05) failed to show a significant relationship with temperature and precipitation. Patterns of daily activity Sand Skink tracks were most abundant in April, and November through January; tracks were least common in June and July (Table 1). Across months, the mean number of Sand Skink tracks per day was significantly related to mean maximum daily air temperature and mean total daily precipitation (multiple r2 = 0.61, df = 2,9, F = 7.06, P < 0.05). Backward stepwise multiple regression revealed precipitation as the only significant predictor of Sand Skink tracks, with a negative relationship between tracks and precipitation (β = -0.74; r2 = 0.54, df = 1,10, F = 11.83, P < 0.01). Discussion Sand Skinks can be captured every month of the year (Fig. 2). Most specimens are captured between February and May, with low numbers of captures for most other months. Many museum specimens were collected in November and December, but field study does not corroborate this result. Instead, our field studies suggest a second peak of captures in August through October. Field studies of populations in Marion County, FL (Smith 1982), and in Orange and Osceola Counties, FL (Sutton et al. 1999), show similar peaks in the number of captures (February through May, and September). At ABS (Fig. 2b), ONF (Fig. 2c), and in Orange and Osceola Counties (Sutton et al. 1999), peak captures is in March (ABS), April (Orange and Osceola Counties) or May (ONF), which may indicate geographic variation in the peak of monthly activity (ABS is the furthest south and ONF is the furthest north). The large number of museum specimens from February, November, and December reflects time of collection of mainly three researchers (N.R. Wood, R.H. Mount, and one of us, S.R. Telford) rather than indicating levels of activity. Timing of reproductive events is associated with number of captures of Sand Skinks. Sand Skinks mate between February and May (Ashton 2005, Telford 1959). The greatest numbers of Sand Skinks are captured during this time period (Fig. 2), indicating increased activity during the mating season, similar to other lizards (e.g., Lister and Aguayo 1992, Pianka 1970). Females lay eggs principally in June, and hatchlings emerge in late July (Ashton 2005, Telford 1959). The emergence of hatchlings causes an increase in number of captures of Sand Skinks from July through October (with a particularly strong effect in August at ABS). Though total monthly captures at ABS are higher in months with lower precipitation and higher temperatures, this pattern is not present when hatchlings are excluded from the analysis. Total monthly captures of Sand Skinks at ONF show no relationship with temperature and precipitation. 182 Southeastern Naturalist Vol. 5, No. 1 Given that the significant relationship at ABS can be explained by the presence of hatchlings in July through October, and data at ONF show no association with temperature or precipitation, temperature and precipitation do not have a direct effect on the number of captures of Sand Skinks each month. Daily surface activity, as indexed through tracks, depends on precipitation, with more tracks detected on days with less precipitation. Precipitation increases soil compaction, which may impede movement of Sand Skinks through sand, causing less activity on days with rain. In fact, sand with > 27% moisture impedes movement of Sand Skinks (Lee 1969). Daily activity does not depend on temperature, suggesting that environmental conditions, at least at the southern portion of its range, permit Sand Skinks to attain temperatures necessary for activity on most days throughout the year. Indeed, in our study, Sand Skinks were found active on days with maximum air temperatures as low as 20.6 °C, 8 to 12 °C lower than selected body temperatures for Sand Skinks (Andrews 1994). Activity at relatively low ambient temperatures may be facilitated by fossoriality, because temperature profiles 5 mm into sand (Sand Skink activity occurs just below the surface of sand) are higher, particularly during mid-day, than air temperatures 1 mm above sand in open areas (Andrews 1994). Even though we did not find a significant relationship between activity per day and temperature, temperature does influence daily activity cycles, with Sand Skinks showing unimodal activity (active during the day) in January, but bimodal activity (activity mainly in the evening and morning) in April and May (Andrews 1994). In sum, activity of Sand Skinks is primarily influenced by precipitation and reproductive activities. Unlike most tropical lizards (e.g., James 1994, Lister and Aguayo 1992), the highest levels of Sand Skink activity do not occur during times of greatest rainfall. In fact, activity is negatively associated with rainfall. Instead, Sand Skinks show monthly activity patterns similar to temperate zone species that breed in the spring (e.g., Fitch 1940, Tinkle 1961). But unlike temperate zone species, temperature appears to have less of an influence on Sand Skink activity than precipitation. Acknowledgments For access to museum specimens, we thank L. Ford (AMNH), J. Wiens (CM), A. Resetar (FMNH), J. Simmons (KU), K. Beaman (LACM), J. Hanken (MCZ), A. Braswell (NCSM), H. Dundee (TU), K. Krysko and M. Nickerson (UF), A. Kluge (UMMZ), and K. de Queiroz (USNM). For permitting, we thank the Florida Fish and Wildlife Conservation Commission (permits WX 01623, WX 97122). For weather data, we thank N. Deyrup and M. Griffin. For funding, we thank Archbold Biological Station, the Earthwatch Institute, USDA Forest Service, and US Fish and Wildlife Service. For field assistance at ABS, we thank A. Knipps, B. Meneken, B. Branciforte, and Earthwatch volunteers. For field assistance at ONF, we thank D. Auth, L. Lowery, R. Lowery, M. Telford, and R.S. Telford. For comments and discussion, we thank R. Andrews, M. Jennings, E. McCoy, P. Moler, H. Mushinsky, K. Penney, and H. Swain. 2006 K.G. Ashton, and S.R. Telford, Jr. 183 Literature Cited Andrews, R.M. 1994. Activity and thermal biology of the Sand-swimming Skink Neoseps reynoldsi: Diel and seasonal patterns. Copeia 1994:91–99. Ashton, K.G. 2005. Life history of a fossorial lizard, Neoseps reynoldsi. Journal of Herpetology 39:389–395. Duvall, D., L.J. Guillette, Jr., and R.E. Jones. 1982. Environmental control of reptilian reproductive cycles. Pp. 201–231, In C. Gans and F.H. Pough (Eds.). Biology of the Reptilia. Vol. 13. Academic Press, San Diego, CA. 345 pp. Fitch, H.S. 1940. A field study of the growth and behavior of the Fence Lizard. University of California Publications in Zoology 44:151–172. Fitch, H.S. 1958. Natural history of the Six-lined Racerunner (Cnemidophorus sexlineatus). University of Kansas Publications, Museum of Natural History 11:11–62. Grant, B.W., and A.E. Dunham. 1990. Elevational covariation in environmental constraints and life histories of the desert lizard Sceloporus merriami. Ecology 71:1765–1776. Huey, R.B., and E.R. Pianka. 1977. Seasonal variation in thermoregulatory behavior and body temperature of diurnal Kalahari lizards. Ecology 58:1066–1075. James, C.D. 1994. Spatial and temporal variation in structure of a diverse lizard assemblage in arid Australia. Pp. 287–317, In L.J. Vitt and E.R. Pianka (Eds.). Lizard Ecology: Historical and Experimental Perspectives. Princeton University Press, Princeton, NJ. 403 pp. Lee, D.S. 1969. Moisture toleration: A possible key to dispersal ability in three fossorial lizards. Bulletin of the Maryland Herpetological Society 5:53–56. Lister, B.C., and A.G. Aguayo. 1992. Seasonality, predation, and the behaviour of a tropical mainland anole. Journal of Animal Ecology 61:717–733. Pianka, E.R. 1970. Comparative autecology of the lizard Cnemidophorus tigris in different parts of its geographic range. Ecology 51:703–720. Pianka, E.R. 1986. Ecology and Natural History of Desert Lizards. Princeton University Press, Princeton, NJ. 208 pp. Smith, C.R. 1982. Food resource partitioning of fossorial Florida reptiles. Pp. 173– 178, In N.J. Scott (Ed.). Herpetological Communities. Wildlife Research Report 13, US Department of the Interior, Washington, DC. Sutton, P.E. 1996. A mark and recapture study of the Florida Sand Skink, Neoseps reynoldsi, and a comparison of Sand Skink sampling methods. M.Sc. Thesis. University of South Florida, Tampa, FL. 45 pp. Sutton, P.E., H.R. Mushinsky, and E.D. McCoy. 1999. Comparing the use of pitfall drift fences and cover boards for sampling the threatened Sand Skink (Neoseps reynoldsi). Herpetological Review 30:149–151. Telford, Jr., S.R. 1959. A study of the Sand Skink, Neoseps reynoldsi Stejneger. Copeia 1959:110–119. Tinkle, D.W. 1961. Population structure and reproduction in the lizard Uta stansburiana Stejnegeri. American Midland Naturalist 66:206–234.