ARTICLE IN PRESS Journal of Thermal Biology 28 (2003) 595–600 Effect of temperature on metabolic rate of the mudturtle (Kinosternon subrubrum) Jacqueline D. Litzgusa,b, William A. Hopkinsa,* a Savannah River Ecology Laboratory, University of Georgia, Drawer E, Aiken, SC 29802, USA b Department of Biological Sciences, University of South Carolina, Columbia, SC 29208, USA Received15 June 2003; accepted5 August 2003 Abstract Temperature plays an important role in various aspects of the life history andphysiology of ectotherms. We examinedthe effect of temperature on standardmetabolicrate in the mudturtle, Kinosternon subrubrum. We measured O2 consumption andCO 2 production at 20 C and30 C using a flow through respirometery system. Standard metabolic rate was significantly higher at 30 C (9.25 ml O2/h, 6.35 ml CO2/h) comparedto 20 C (2.10 ml O2/h, 1.96 ml CO2/h). The Q10 value for O2 was 5.10, andfor CO 2 was 3.40. Our findings generally agree with those of other studies of metabolism in vertebrate ectotherms. Publishedby Elsevier Ltd. Keywords: Carbon dioxide production; Kinosternon subrubrum; Oxygen consumption; Q10; RQ; Respiratory quotient; Standard metabolic rate; Temperature; Turtle 1. Introduction development of models that describe the energy budgets of organisms (Lillywhite, 1987; Secor andNagy, 1994 ; Temperature plays an important role in various Beaupre, 1995, 1996), which in turn strongly influence aspects of the life history, ecology, andphysiology of life histories (Congdon et al., 1982; Dunham et al., reptiles andother ectotherms ( Angilletta et al., 2002). 1989). Growth rates (ArnoldandPeterson, 1989 ; Avery, 1994; An examination of variation in individual metabolic Litzgus andBrooks, 1998a ), reproduction (Schwarzkopf response to environmental variation (i.e., temperature andShine, 1991 ; Litzgus andBrooks, 1998b ; Rock and changes) can elucidate mechanistic explanations for Cree, 2003), seasonal activity patterns andhabitat use broadecological andevolutionary patterns ( Congdon (Webb andShine, 1998 ; Whitaker andShine, 2002 ), and et al., 1982). In addition, information on physiological geographic distribution (Castonguay et al., 1999) are all adaptations can be gleaned from intraspecific and influencedby environmental temperatures. Physiological interspecific comparative analyses (Garlandand processes such as metabolic rate generally increase with Adolph, 1994; Zaidan, 2003). Only a handful of studies temperature (e.g., Gatten, 1974; Bennett andDawson, on turtles have directly tested the effect of temperature 1976; Beaupre et al., 1993; Karasov andAnderson,1998 ; on metabolic rate. Oxygen consumption rates increased McNab, 2002); however, a few reptile species are known with temperature in the slider turtle (Trachemys scripta), to have plateaus of temperature-independent metabo- box turtle (Terrapene ornata)(Gatten, 1974), African lism (Waldschmidt et al., 1987). Information on how tortoise (Kinixys spekii)(Hailey andLoveridge,1997 ), temperature affects metabolic rates is useful for the andin hatchling snapping turtles ( Chelydra serpentina) (Steyermark andSpotila, 2000 ). Only one of these *Corresponding author. Tel.: +1-803-952-7427; fax: +1- studies examined both O2 consumption andCO 2 803-725-3309. production (Steyermark andSpotila, 2000 ). The purpose E-mail address: [email protected] (W.A. Hopkins). of this paper was to examine the effect of temperature on 0306-4565/$ - see front matter Publishedby Elsevier Ltd. doi:10.1016/j.jtherbio.2003.08.005 ARTICLE IN PRESS 596 J.D. Litzgus, W.A. Hopkins / Journal of Thermal Biology 28 (2003) 595–600 O2 consumption andCO 2 production rates of mud measurements were correctedfor standardtemperature turtles (Kinosternon subrubrum), andto compare our andpressure by the Oxymax ER-10 software. Turtles calculated Q10 andrespiratory quotient (RQ, the ratio of were weighed( 71 g) at the endof the experiment, and CO2 produced to O2 consumed) values to those reported returnedto their tote bins until release at their sites of for other reptile species. capture. From the VO2 andVCO 2 data collected, we estimated standard metabolic rate (SMR), the metabolic rate of a 2. Materials and methods post-absorptive animal at rest at a given temperature during the inactive part if its diel cycle (Bennett and Mudturtles are relatively small freshwater turtles, Dawson, 1976; Waldschmidt et al., 1987). Mudturtles, typically less than 12 cm in shell length (Ernst et al., like most vertebrate ectotherms, exhibitedspontaneous 1994). The species ranges from Long Island, New York, activity during respirometery trials (Fig. 1). In an effort southwardthrough Floridato the Gulf Coast, to exclude periods of activity from our estimates, westwardto central Texas, andnorthwardinto the SMR for VO2 andVCO 2 were determined by Mississippi Valley (Ernst et al., 1994). Mud calculating the mean of the lowest approximate 30% turtles (3 females, 4 males, mass range 97–220 g) of the data (4 of 13 data points) for each gas for each were collectedby handbetween 15 and25 February turtle. This methodof estimation eliminatedobvious 2003, shortly after emergence from hibernation, bouts of spontaneous activity (Fig. 1) andtherefore from the Francis Beidler Forest National Audubon appears to accurately represent SMR in our sample of andNature Conservancy Sanctuary near Harleyville, mudturtles. SC (33N, 80W). Beidler Forest is a 12,000-acre We calculatedthe respiratory quotient (RQ) for each nature preserve locatedin Four Holes Swamp, temperature treatment, andthe Q10s (20–30 C) for VO2 and includes cypress-tupelo blackwater swamp, hard- andVCO 2 using the SMR values. RQ (the ratio of CO2 woodbottom swamp, anduplandpine forest. Turtles produced to O2 consumed) allows inference about were brought into the lab where they were housedin aerobic catabolism (Withers, 1992). The effect of plastic tote bins (17 liters, 42 cm  30 cm  15 cm) temperature on metabolic processes can be described with shallow water, andfastedfor a minimum of 9 using the term Q10, which is defined as the change in the days to ensure post-absorptivity prior to metabolic rate of metabolism over a 10C change in temperature; determinations. Q10 normally has a value between 2 and3 ( McNab, Oxygen consumption (VO2) and carbon dioxide 2002). production (VCO2) were measuredusing a portable, To examine the effects of temperature on SMR, we flow-through respirometer (Oxymax ER-10, Columbus useda repeatedmeasures ANCOVA with bodymass as Instruments, Columbus, OH) interfacedto a laptop the covariate (proc mixedmodel,SAS Institute, Cary, computer. Each turtle was placedin a circular NC). To satisfy the assumption of linearity, both body Plexiglass chamber (volume B800 ml, diameter=15 cm, mass andSMR were log 10 transformedprior to height=5 cm) linedwith moist paper towels. Turtles statistical analyses. Statistical significance was accepted were not restrainedandcouldtherefore move freely at Po0.05. inside the chamber. Chamber volumes and leaks were measuredandcheckedby the Oxymax ER-10 software. Each chamber was attached to an independent 3. Results channel of the respirometer by tubing. The respirometer andchambers containing turtles were placedin an The effect of sex on SMR was not significant for incubator (Percival Scientific, Boone, Iowa; 24 h dark) either VO2 (F1,4=1.49, P=0.29) or VCO2 (F1,4=4.67, andmetabolic rate was measuredfor 24 h at each P=0.10); therefore, data for the sexes were pooled in all of two temperatures sequentially, 30C and20 C. further analyses. Similarly, the effect of mass on SMR Turtles were acclimatedto the treatment temperature was not significant for VO2 (F1,5=2.04, P=0.21) or while in respirometery chambers for 2 h prior to VCO2 (F1,5=4.42, P=0.09), although a positive linear commencement of metabolic rate determinations. Every trend was detected. Mean (7SE) SMR andRQs for 1.75 h, samples of air from each turtle chamber were each temperature treatment are shown in Table 1. passedover magnesium perchlorate andrates of VO 2 Temperature hada significant effect on SMR for both andVCO 2 were simultaneously determined by the VO2 (F1,6=26.94, Po0.01) andVCO 2 (F1,6=37.74, respirometer. The O2 andCO 2 sensors were purged Po0.001), such that as temperature increased, meta- with outside air passed over a column containing bolic rate increased. Spontaneous activity bouts were Drierite after each measurement. Thus, at both tem- also more frequent at the higher temperature (Fig. 1). peratures, a total of 13 measurements of each gas were Mean (7SE) Q10 (20–30 C) for VO2 was 5.1071.18, obtainedfor each turtle over the 24-h period.All andfor VCO 2 was 3.4070.52. ARTICLE IN PRESS J.D. Litzgus, W.A. Hopkins / Journal of Thermal Biology 28 (2003) 595–600 597 Fig. 1. Oxygen consumption (VO2, diamonds, A and B) and carbon dioxide production (VCO2, squares, C andD) of a mudturtle (Kinosternon subrubrum) (mass=220 g) at 20C (A andC) comparedto the same turtle at 30 C (B andD). The horizontal lines represent the standard metabolic rates (SMR) calculated as the mean of the lowest 30% of the data for each gas. The lowest 30% of the data are shown as solid symbols, and the highest 70% of the data (not used in the calculation of SMR) are represented as hollow symbols. This methodof estimation reducesthe influence of spontaneous activity on the calculation of SMR. Table 1 metabolic rates of turtles did not test the effect of Mean7SE standard metabolic rate (SMR) measured indirectly temperature, making broadcomparisons of SMR values as oxygen consumption (VO2)