Seasonal changes in the diel surfacing behaviour of the bimodally respiring turtle Rheodytes leukops

1614 Seasonal changes in the diel surfacing behaviour of the bimodally respiring turtle Rheodytes leukops Matthew A. Gordos, Craig E. Franklin, and Colin J. Limpus Abstract: The purpose of this study was to determine whether a relationship existed between the diel surfacing trends of the bimodally respiring freshwater turtle Rheodytes leukops and daily fluctuations in specific biotic and abiotic factors. The diel surfacing behaviour of adult R. leukops was recorded over four consecutive seasons (Austral autumn 2000 summer 2001) within Marlborough Creek, central Queensland, Australia, using pressure-sensitive time depth recorders. Additionally, diurnal variations in water temperature and aquatic PO 2 level, as well as the turtle s behavioural state (i.e., active versus resting), were monitored. In autumn and summer, surfacing frequency increased significantly during the daylight hours, with peak levels normally occurring around dawn (0500 0700) and dusk (1700 1900). However, no consistent diel surfacing trend was recorded for the turtles in winter or spring, owing to considerable variation among individual R. leukops. Diurnal surfacing trends recorded for R. leukops in autumn and summer are attributed to periods of increased activity (possibly associated with foraging) during the daylight hours and not to daily variations in water temperature or aquatic PO 2 level. Turtles generally remained at a depth greater than 1 m throughout the day, where the effect of diel fluctuations in water temperature (<0.5 C) and aquatic PO 2 level (<15 mm Hg (1 mm Hg = 133.322 Pa)) was considered to be negligible. Résumé : Notre étude vise à déterminer s il existe une relation entre les patterns journaliers de remontée en surface de la tortue d eau douce à respiration mixte Rheodytes leukops et les fluctuations journalières de certains facteurs biotiques et abiotiques. Des chronobathymètres enregistreurs sensibles à la pression nous ont permis d enregistrer les comportements de remontée en surface de tortues adultes au cours de quatre saisons consécutives (automne austral 2000 été 2001) à Malborough Creek, Queensland central, Australie. Nous avons aussi suivi les variations journalières de la température et du niveau de PO 2 de l eau et noté le comportement des tortues (i.e. activité ou repos). En automne et en été, la fréquence des remontées augmente significativement durant les heures d éclairement, avec habituellement des maximums aux environs de l aube (0500 0700) et du crépuscule (1700 1900). En hiver et au printemps, cependant, il n y a pas de pattern journalier uniforme de remontée en surface à cause de la variation individuelle importante des déplacements des tortues. Les patterns journaliers observés en automne et en été s expliquent par des périodes d activité accrue (peut-être associée à la recherche de nourriture) durant les heures d éclairement, mais non par des variations de température ou de niveau de PO 2 de l eau au cours de la journée. Les tortues passent généralement la journée à des profondeurs supérieures à1moùleseffetsdesfluctuations journalières de température (<0,5 C) et de PO 2 (<15 mm Hg (1 mm Hg = 133.322 Pa)) de l eau semblent négligeables. [Traduit par la Rédaction] Gordos et al. 1622 Introduction Rheodytes leukops is a bimodally respiring freshwater turtle that utilizes highly modified cloacal bursae to extract a significant portion of its total oxygen requirements from the water (Priest 1997). The ability of R. leukops to supplement its pulmonary oxygen reserves via aquatic respiration results in a reduced surfacing frequency compared with primarily Received 26 February 2003. Accepted 13 August 2003. Published on the NRC Research Press Web site at http://cjz.nrc.ca on 14 October 2003. M.A. Gordos 1 and C.E. Franklin. Department of Zoology and Entomology, University of Queensland, Brisbane, QLD 4072, Australia. C.J. Limpus. Queensland Parks and Wildlife Service, P.O. Box 155, Brisbane, QLD 4002, Australia. 1 Corresponding author (email: mgordos@zen.uq.edu.au). air-breathing turtles (Gordos and Franklin 2002; Priest and Franklin 2002). However, the surfacing behaviour of R. leukops is dependent not only upon the turtle s physiological capabilities but also upon seasonally fluctuating abiotic and biotic factors (Priest and Franklin 2002; Gordos et al. 2003). Increases in water temperature from winter to summer levels result in elevated metabolic rates (as measured by oxygen consumption) for aquatic turtles, which in turn leads to diminished dive times (Gatten 1978, 1980; Ultsch and Jackson 1982; Herbert and Jackson 1985; Santos et al. 1990; Priest and Franklin 2002). Additionally, increasing temperatures reduce the oxygen capacitance of water, which decreases the effectiveness of aquatic respiration and increases the surfacing frequency of bimodally respiring turtles (Dejours 1994; Ultsch 1989; Stone et al. 1992; Prassack et al. 2001; Priest and Franklin 2002). The behavioural state of turtles (i.e., resting versus active) also varies seasonally, with periods of activity generally increasing with tempera- Can. J. Zool. 81: 1614 1622 (2003) doi: 10.1139/Z03-153

Gordos et al. 1615 ture from winter to summer (Georges 1982; Chessman 1988; Dall Antonia 2001; Prassack et al. 2001). Although aquatic respiration may be sufficient to sustain resting metabolic rates in certain species, elevated rates of oxygen consumption associated with periods of activity tend to require increased bouts of air-breathing in bimodally respiring turtles (Girgis 1961; Gatten 1974, 1978; Butler et al. 1984; Bagatto and Henry 1999; Hays et al. 2000). Aquatic temperature and PO 2 level, as well as the behavioural state of turtles, also vary on a daily scale (Hynes 1970; Georges 1982; Chessman 1988; Dall Antonia et al. 2001; Wetzel 2001). In shallow lotic systems, water temperature typically decreases to a daily minimum just before sunrise before warming to a maximum value in the late afternoon (Hynes 1970; Wetzel 2001). Diel aquatic PO 2 fluctuations are chiefly influenced by plant photosynthesis and respiration, with daily values normally at their lowest in the early morning and peaking in the evening hours (Hynes 1970; Wetzel 2001). However, the effect of daily fluctuations in biotic and abiotic factors upon the diel surfacing trends of freshwater turtles has yet to be investigated. Therefore, the aim of this study was to record the diel surfacing behaviour of R. leukops over a seasonal scale to investigate the relationship between surfacing frequency and daily variations in water temperature, aquatic PO 2 level, and the turtle s level of activity. Additionally, insights into the general ecology of R. leukops inferred from seasonal changes in diel trends will be discussed. Materials and methods Study-site description Four month long seasonal investigations (Austral autumn 2000 summer 2001) were conducted within two pools (>1 km in length, 30 60 m wide, and 1 3 m deep) of Marlborough Creek (22 57 34 S, 149 52 08 E), a tributary of the Fitzroy River in central Queensland, Australia. Water height (Queensland Department of Natural Resources water station No. 1300009a), water velocity (<2 cm s 1 ), and water clarity (<1.5 m) generally remained constant within the pools throughout the seasons except for a week during a summer flood event. Diel water temperature and PO 2 profiles (YSI model 55 dissolved oxygen temperature system) were recorded seasonally within a pool located upstream of the study sites, so as not to disturb the diving behaviour of tagged turtles. Three times per season, six repeat measurements of water temperature and aquatic PO 2 level were recorded at each of three depths (surface and 1 and 2 m) every sixth hour of the day (starting at 0600) within the middle of the sampling pool. Diel water temperature and PO 2 profiles recorded within the study sites after the completion of time depth recorder (TDR) logging followed similar trends as observed in the sampling pool, suggesting that results presented herein are representative of the study sites. Animal capture and handling All turtles were captured within the study-site pools and were considered mature as determined by mass and straightline carapace length (Legler and Cann 1980; see Gordos et al. 2003). Given that R. leukops nest between September and November (Legler and Cann 1980), female turtles were palped during the spring season to avoid using gravid individuals. A TDR (55 mm 16 mm, 1ginwater, model LTD-10; Lotek Marine Technologies Inc., St. John s, Nfld.) and two single-stage radiotransmitters (25 mm 15 mm 12 mm, 11.2 ± 1.4 g, 40 mm 30 mm 15 mm, 27.6 ± 1.4 g; Titley Electronics, Ballina, Australia) were attached to the postmarginal scutes of each turtle as described in Gordos and Franklin (2002). TDRs logged water pressure (± 4.0 cm) every 4 s and temperature (±0.1 C) every 5 min for up to 21 days. However, TDRs did not start logging until 10 days after the release of a turtle at its initial capture site to minimize the effects of handling stress (Gordos and Franklin 2002). After the completion of TDR logging, turtles were recaptured, equipment was removed, and TDRs were calibrated. All turtles were notched to avoid double sampling during subsequent seasons (Limpus et al. 2002). Collection and field experimentation of R. leukops were conducted in accordance with the principles and guidelines of the Canadian Council on Animal Care. Analysis of results Downloaded TDR dive data (TagTalk, version 0.991.93; Lotek Marine Technologies Inc.) were converted from pressure (pounds per square inch) to depth (centimetres) values prior to analysis. Turtles were considered to be diving if TDR depth exceeded 40 cm (threshold = distance from TDR to turtle s nares (30 cm) + drift in the zero line of a TDR during deployment (10 cm)). However, because of an inherent fluctuation in all TDRs of ±4.0 cm, dives with a maximum depth <45 cm were not scored, with the length of these dives being added to the previous surfacing event. Custom-written programs (M.A. Gordos) initially analysed TDR data for the proportion of surfacing events per hour per day, mean dive depth per hour per day, and mean TDR temperature per hour per day for each turtle. Additionally, the proportion of activity counts per hour per day was determined for individual turtles by logging temporal variations in depth of ±6.0 cm that were recorded as one activity count. Hourly surfacing, activity, depth, and temperature values were then averaged over the entire deployment period for each turtle. To achieve normality, hourly activity and surfacing proportions were square-root and arcsine transformed (p = arcsine( p)) prior to statistical analysis. For each season, the 24-h day was divided into five periods (night 1 (P1), dawn (P2), day (P3), dusk (P4), and night 2 (P5)) (Table 1) for the purpose of statistical analysis, with mean values of surfacing frequency, activity level, and dive depth being determined for each period for individual R. leukops. Selection of periods was based upon diel-trend results from a previous investigation on R. leukops (Gordos and Franklin 2002), with dawn and dusk (P2 and P4) being identified as the first and last 2hofsunlight for each season. Within-season differences among the five periods for surfacing frequency, activity level, and depth were investigated using a MANOVA. Following a significant finding (P < 0.05), three planned comparisons were run between the five periods to elucidate specific diel trends within a season: daylight versus night (P2, P3, and P4 versus P1 and P5), dawn versus dusk (P2 versus P4), and dawn dusk versus day (P2 and P4 versus P3). Because of the small sample size within

1616 Can. J. Zool. Vol. 81, 2003 Table 1. Delineation of the 24-h day into five periods for each season. Season Night 1 (P1) Dawn (P2) Day (P3) Dusk (P4) Night 2 (P5) Autumn 0000 0500 0500 0700 0700 1700 1700 1900 1900 2400 Winter 0000 0600 0600 0800 0800 1600 1600 1800 1800 2400 Spring 0000 0500 0500 0700 0700 1700 1700 1900 1900 2400 Summer 0000 0500 0500 0700 0700 1700 1700 1900 1900 2400 Note: P2 and P4 were determined as the first and last 2 h of sunlight for each season. each season, statistical differences between sexes were not investigated and are acknowledged as a possible confounding factor. Unless specified, all values hereafter are presented as mean ± SE. Results Diel diving behaviour The diving behaviour of R. leukops within each season was analysed for diel trends in surfacing frequency, periods of activity, and dive depth. In autumn, turtles were significantly more active (n = 2 females and 5 males) (F [4,24] = 8.617, P < 0.001) and surfaced more frequently (F [4,24] = 5.314, P < 0.01) during the daytime (P2, P3, and P4) than during the night (P1 and P5) (Figs. 1a and 1b). Despite peaks in surfacing frequency and activity level at crepuscular hours, differences between the three daylight periods (P2, P3, and P4) were not significant. Daytime dive depth was significantly shallower compared with nighttime levels (F [4,24] = 13.729, P < 0.001) (Fig. 1c), with hourly mean depth gradually decreasing throughout the day from 182.5 ± 11.4 cm at 0400 to a low value of 110.8 ± 9.7 cm at 1500. In winter, daily variation in diving behaviour with respect to time of day was significant only for the turtles level of activity (n = 4 females and 2 males) (F [4,20] = 4.376, P < 0.05) (Fig. 2). Rheodytes leukops were significantly more active during the day than during the night, with the peak level of activity generally occurring from the early morning to midday (0600 1200). Despite an increased surfacing frequency at dusk (1700) (Fig. 2a), differences among the five periods were not significant (P = 0.471), owing to the highly variable surfacing behaviour among turtles during winter. Rheodytes leukops generally remained at a depth of >100 cm (range 132.3 ± 13.5 to 141.9 ± 14.1 cm) throughout the day, with no significant trend being observed for the season (P = 0.090) (Fig. 2c). Diel trends in surfacing frequency and activity level were generally observed for individual R. leukops in spring; however, patterns varied considerably among turtles, resulting in nonsignificant findings for the season overall (n = 4 females and 5 males) (surfacing: P = 0.619, activity: P = 0.237) (Figs. 3a and 3b). Because of a decrease in mean dive depth from the early morning (137.4 ± 23.9 cm at 0500) to late afternoon (111.1 ± 18.2 cm at 1400), turtles in spring were significantly shallower at dusk than at dawn (F [4,32] = 6.365, P < 0.05) (Fig. 3c). However, the depth where R. leukops resided during the day (P2, P3, and P4) versus the night (P1 and P5) and during the crepuscular periods (P2 and P4) versus daylight hours (P3) was not significantly different. Significant diel trends were observed for R. leukops in summer regarding surfacing frequency (n = 4 females and 4 males) (F [4,28] = 6.761, P < 0.001), activity level (F [4,28] = Fig. 1. Diel trends (mean ± SE) for Rheodytes leukops in autumn for the proportion of total surfacings per hour (a), proportion of total activity per hour (b), and mean depth per hour (c). For purposes of analysis, the 24-h day was divided into five periods as depicted at the top of the figure (solid bar, night 1 (P1) and night 2 (P5); cross-hatched bar, dawn (P2) and dusk (P4); open bar, day (P3)). An asterisk indicates a significant difference (P < 0.05) between the daylight (P2, P3, and P4) and nighttime hours (P1 and P5). 11.871, P < 0.001), and dive depth (F [4,28] = 8.851, P < 0.001) (Fig. 4). Turtles surfaced more often during the daytime than during the night in summer (P < 0.01), but comparison between the crepuscular and daylight hours was not

Gordos et al. 1617 Fig. 2. Diel trends (mean ± SE) for R. leukops in winter for the proportion of total surfacings per hour (a), proportion of total activity per hour (b), and mean depth per hour (c). For purposes of analysis, the 24-h day was divided into five periods as depicted at the top of the figure (solid bar, night 1 (P1) and night 2 (P5); cross-hatched bar, dawn (P2) and dusk (P4); open bar, day (P3)). An asterisk indicates a significant difference (P < 0.05) between the daylight (P2, P3, and P4) and nighttime hours (P1 and P5). Fig. 3. Diel trends (mean ± SE) for R. leukops in spring for the proportion of total surfacings per hour (a), proportion of total activity per hour (b), and mean depth per hour (c). For purposes of analysis, the 24-h day was divided into five periods as depicted at the top of the figure (solid bar, night 1 (P1) and night 2 (P5); cross-hatched bar, dawn (P2) and dusk (P4); open bar, day (P3)). An asterisk indicates a significant difference (P < 0.05) between dawn (P2) and dusk (P4). significant (P = 0.869). Similarly, R. leukops displayed a sharp increase in activity at sunrise (0500) that was maintained throughout the day until sunset (1900) (P < 0.01). Finally, R. leukops in summer registered the shallowest mean dive depth (102.7 ± 14.4 cm), with levels decreasing significantly during the day (87.8 ± 3.1 cm) compared with the night (123.1 ± 2.0 cm). Water temperature and PO 2 trends Aquatic PO 2 and temperature profiles were recorded seasonally within a sampling pool located upstream from the study sites containing tagged turtles. Aquatic PO 2 levels decreased with depth in all four seasons (Fig. 5), with the greatest difference occurring in summer where surface levels (97.5 ± 12.7 mm Hg (1 mm Hg = 133.322 Pa)) were 3-fold higher than levels recorded at 2 m depth (28.9 ± 2.0 mm Hg). Winter aquatic PO 2 levels varied little with depth, with values remaining near normoxic throughout the season (range 118.3 ± 1.7 to 139.0 ± 9.4 mm Hg). Autumn and spring PO 2 profiles displayed similar patterns, with hypoxic conditions persisting below the water s surface. Diel changes in aquatic PO 2 level within each season followed similar patterns as observed in summer (Fig. 6). Considerable daily variation was observed at the water s surface, with fluctuations of 72.0 and 53.7 mm Hg being recorded during autumn and summer, respectively. Generally, daily lows occurred at 0600, with PO 2 level increasing throughout the day until dusk (1800). However, aquatic PO 2 levels at 1 and 2 m depth generally remained constant throughout the

1618 Can. J. Zool. Vol. 81, 2003 Fig. 4. Diel trends (mean ± SE) for R. leukops in summer for the proportion of total surfacings per hour (a), proportion of total activity per hour (b), and mean depth per hour (c). For purposes of analysis, the 24-h day was divided into five periods as depicted at the top of the figure (solid bar, night 1 (P1) and night 2 (P5); cross-hatched bar, dawn (P2) and dusk (P4); open bar, day (P3)). An asterisk indicates a significant difference (P < 0.05) between the daylight (P2, P3, and P4) and nighttime hours (P1 and P5). Fig. 5. Aquatic PO 2 profile (mean ± SE) of the sampling pool over four successive seasons during TDR deployment. Fig. 6. Diel aquatic PO 2 profile (mean ± SE) of the sampling pool for summer during TDR deployment. day during all four seasons, varying <15.0 mm Hg over 24 h. Mean water temperatures recorded for the sampling pool during autumn (23.8 ± 0.3 C) and spring (23.8 ± 0.2 C) were similar, while winter levels (15.2 ± 0.4 C) were more than 12 C cooler than summer temperatures (27.5 ± 0.2 C). Contrary to aquatic PO 2 levels, mean water temperature within a season varied little with depth (<3.5 C) throughout the year (Table 2). Diel temperature changes at the water s surface showed the greatest fluctuation of the three depths, with temperature increasing from lows at 0600 to a high value at 1800 (Fig. 7a). At 1 and 2 m depth, water temperature varied little (<0.5 C) throughout the day during all four seasons. Diel temperature trends as measured by TDRs followed similar patterns as recorded in the sampling pool, with TDR levels increasing from a minimum value in the early morning (0600 0800) to a maximum value in the late afternoon (1500 1700) (Fig. 7b). For all four seasons, mean hourly TDR temperature trends fluctuated <2 C over a daily range, with minimum and maximum values falling between the range of temperatures logged within the sampling pool (Table 2). Discussion The surfacing frequency of aquatic turtles is dependent upon seasonally changing environmental conditions such as water temperature and aquatic PO 2 level, with bimodally respiring species switching from facultative to obligate airbreathers as abiotic factors change from winter to summer levels (Gatten 1980; Ultsch and Jackson 1982, 1995; Herbert

Gordos et al. 1619 Table 2. Minimum and maximum (mean ± SE) water temperature recorded for each season within the sampling pool and by TDRs attached to Rheodytes leukops. Pool temperature ( C) TDR temperature ( C) Season Minimum Maximum Minimum Maximum Autumn 23.1±0.9 25.2±1.2 24.5±0.4 25.1±0.5 Winter 14.4±2.6 17.8±2.8 16.1±0.4 16.7±0.4 Spring 22.9±0.2 25.9±1.5 23.9±0.1 25.1±0.2 Summer 26.6±1.3 29.6±0.4 26.8±0.1 28.3±0.3 Note: TDR temperature values are based upon mean hourly values calculated from all turtles within a given season. Fig. 7. Diel water temperature profile (mean ± SE) of the sampling pool for summer during TDR deployment (a) and hourly water temperatures (mean ± SE) recorded during summer by TDRs attached to R. leukops (b). and Jackson 1985; Ultsch 1989; King and Heatwole 1994; Crocker et al. 2000; Prassack et al. 2001; Gordos et al. 2003). However, on a daily scale, fluctuations in water temperature and aquatic PO 2 level do not appear to directly influence the diel surfacing trends of R. leukops. The absence of a relationship is attributed to negligible daily variations of the two abiotic factors at the depth (>1 m) where R. leukops primarily resided throughout the year. At 1 m depth within the sampling pool, water temperature and aquatic PO 2 levels varied <0.5 C day 1 and <15.0 mm Hg day 1, respectively. Additionally, water temperature as recorded by TDRs attached directly to R. leukops generally varied <2 C day 1 within each season. On a cautionary note, however, recordings of aquatic PO 2 levels at predetermined depths (i.e., surface and 1 and 2 m) may not be reflective of the levels where a turtle resides. Turtles situated at a particular depth can experience considerably different PO 2 levels depending upon the selected habitat (e.g., within a weed bed versus decaying organic matter) and the time of day (noon versus midnight). Diurnal surfacing trends observed for R. leukops in autumn and summer are instead attributed to periods of increased activity during the daylight hours. Increased activity results in elevated rates of oxygen consumption and hence an increased surfacing frequency in aquatic turtles (Gatten 1974, 1978; Stockard and Gatten 1983; Butler et al. 1984; Bagatto and Henry 1999; Hochscheid et al. 1999; Hays et al. 2000). Within green turtles (Chelonia mydas), swimming resulted in a 2.83-fold increase in the rate of oxygen consumption compared with resting turtles, which in turn led to a nearly 3-fold decrease in dive length (Butler et al. 1984). Using a three-dimensional compass-based TDR system, Hays et al. (2000) similarly demonstrated for free-ranging green turtles that dive length decreased with increased activity during the bottom phase of dives. A limitation of the study was the dependence of activity counts upon the turtle s surfacing frequency whereby the more the turtle surfaces, the more activity counts will be logged. Hays et al. (2000) circumvented this problem by focusing upon the activity level of C. mydas during the bottom phase of dives, thus removing periods related to travel to and from the surface. Unfortunately, such an approach is inappropriate for R. leukops because of differences in the turtle s physiology as well as its environment. Unlike C. mydas, R. leukops is negatively buoyant and thus rarely separates itself from the substrate, even when surfacing (i.e., the turtle climbs rather than floats to the surface) (Legler and Cann 1980; M.A. Gordos, personal observation). Additionally, unlike the marine environment where the water column separates the surface from the substrate, the bottom in freshwater systems can include the sides of creeks (which lead up to the surface) as well as shallow sections such as riffle zones. Differentiating between activities associated with the bottom phase of dives and movement associated with surfacing is thus problematical for R. leukops and would invariably bias the selection of specific dive types for analysis (e.g., U-shaped dives). Bimodal activity patterns observed for R. leukops in autumn and summer are similarly reported for other aquatic pleurodiran (Georges 1982; Chessman 1988) and cryptodiran species (Ernst 1986a; Rowe and Moll 1991; Dall Antonia et al. 2001), with active periods often being attributed to behaviour associated with foraging (Georges 1982; Ernst 1986a; Chessman 1988; Rowe and Moll 1991; van Dam and Diez 1996). Chessman (1986) recorded a sharp increase in the stomach content volume of Emydura macquarii directly after sunrise, with a maximum volume occurring in late afternoon. Additionally, bimodal feeding patterns have been recorded via capture rates using baited traps for Sternotherus

1620 Can. J. Zool. Vol. 81, 2003 odoratus (Ernst 1986a), Emydoidea blandingi (Rowe and Moll 1991), Emydura krefftii (Georges 1982), and E. macquarii, Chelodina expansa, and Chelodina longicollis (Chessman 1988). Furthermore, van Dam and Diez (1996) described a diurnal feeding pattern for juvenile Hawksbill turtles (Eretmochelys imbricata) from TDR dive profiles, with foraging dives represented by continuous depth fluctuations during relatively short submergence intervals. Diel activity trends recorded for R. leukops shifted from a bimodal pattern in autumn and summer to a unimodal trend in winter. Similar seasonal changes in activity have been recorded for aquatic turtles in the field (Georges 1982) and in the laboratory (Graham and Hutchison 1979), with the shift possibly resulting from changes in foraging intensity or food availability, altered metabolic rates associated with reduced water temperatures, or a decreased photoperiod (Graham 1979; Georges 1982; Ernst 1986b; Chessman 1988; Spencer et al. 1998). However, unlike autumn and summer, diel surfacing frequency for R. leukops was not related to the activity patterns of turtles in winter. This result may be due to the reduced number of activity counts recorded for turtles during winter (Gordos et al. 2003) where, although R. leukops was more active during the day in a statistical sense, the actual metabolic cost associated with the activity was too low to influence the turtle s surfacing frequency. Additionally, bimodally respiring turtles generally increase their reliance upon aquatic respiration as water temperature decreases because of the high oxygen capacitance of the water and the depressed metabolic rate of the turtle (Gatten 1980; Herbert and Jackson 1985; King and Heatwole 1994; Prassack et al. 2001). Fewer surfacing intervals would be recorded for R. leukops in winter if turtles were able to increasingly meet their metabolic demands via aquatic respiration. In spring, individual R. leukops displayed consistent diel activity trends, with daily surfacing frequency mirroring the observed activity patterns. However, because of considerable variation among turtles, no consistent trend was observed for activity or surfacing frequency for the season. Variation in diel behavioural patterns among turtles may be due to changes in the abundance and location of food, with fluctuations in foraging intensity influencing the turtle s activity trends and hence surfacing frequency. Additionally, nesting is known to occur during the spring season (September November) for R. leukops (Legler and Cann 1980), with the influence of physiological and behavioural changes associated with breeding and nesting upon the turtle s diel rhythms unknown. For male snapping turtles (Chelydra serpentina), Brown and Brooks (1993) observed that movement and activity levels increased significantly during the breeding season. Freshwater turtles routinely bask during the daylight hours for the purpose of thermoregulation (Moll and Legler 1971; Crawford et al. 1983; Ernst 1986b; Rowe and Moll 1991; King et al. 1998; Krawchuk and Brooks 1998; Dall Antonia et al. 2001), with advantages associated with raised body temperatures including elevated growth rates (Christy et al. 1973; King et al. 1998) and decreased gut transit time (Spencer et al. 1998). Although R. leukops does not display atmospheric basking (Legler and Cann 1980; Cann 1998), several species of aquatic turtles are known to thermoregulate using warmer surface waters (Moll and Legler 1971; Spotila et al. 1984; Chessman 1987) or thermal springs (Doody et al. 2001). Movement into shallower depths, as observed for R. leukops in autumn and summer, is suggestive of aquatic basking behaviour. However, temperature data obtained from the sampling pool and from the TDRs indicate that R. leukops rarely took advantage of warmer temperatures recorded at the surface. Instead, shallower dive depths recorded during the day for turtles in autumn and summer are attributed to the increased surfacing frequency during these time periods. Moreover, daytime decreases in dive depth may also be indicative of foraging behaviour, since R. leukops within Marlborough Creek are known to feed extensively upon a ribbonweed species (Vallisnaria sp.) that is restricted to shallow water (<1 m) (Cann 1998). In summary, the diel surfacing trends observed for R. leukops within Marlborough Creek during autumn and summer are credited to the timing of active periods rather than to diurnal fluctuations in abiotic factors. Unfortunately, the measure of activity used in this study (i.e., activity counts) fails to indicate what the turtles are doing (e.g., feeding or surfacing) as well as the associated metabolic cost of the behaviour. Baseline ecological studies are warranted for determining what activities R. leukops is undertaking during the day, the adaptive significance of the timing of these active periods, and the mechanism(s) that stimulate or terminate these periods (e.g., light intensity). Although daily changes in water temperature and aquatic PO 2 level do not appear to influence the surfacing frequency of R. leukops directly, small diel temperature fluctuations are known to trigger the onset and cessation of active periods in turtles (Graham 1972). A further consideration is whether the trends recorded for turtles in Marlborough Creek are reflective of the behavioural patterns of R. leukops in other habitat types, especially given that R. leukops within the main stem Fitzroy River prefer the shallow riffle zones rather than pools, as observed during this study (Legler and Cann 1980; Cann 1988; Tucker et al. 2001). Acknowledgements Collection of and field experimentation on R. leukops were approved by the Queensland Parks and Wildlife Service (permit No. C6/000094/00SAA) and by the University of Queensland s Animal Experimentation and Ethics committee (AEEC approval No. ZOO/ENT/063/00-01/URG/PHD). We thank Toni Priest, Vikki Rogers, John Cay, Cathy Pillans, Dave Putland, and Max Coyne for assistance with the capture and recapture of turtles and John Parmenter for the provision of a backup radio receiver. Additional thanks are given to the Physiological Ecology laboratory at the University of Queensland for their helpful comments on data analysis and early drafts of the manuscript. Research was supported by a Large Australian Research Council grant to C.E.F. References Bagatto, B., and Henry, R.P. 1999. Exercise and forced submergence in the pond slider (Trachemys scripta) and softshell turtle

Gordos et al. 1621 (Apalone ferox): influence on bimodal gas exchange, diving behaviour and blood acid base status. J. Exp. Biol. 202: 267 278. Brown, G.P., and Brooks, R.J. 1993. Sexual and seasonal differences in activity in a northern population of snapping turtles, Chelydra serpentina. Herpetologica, 49: 311 318. Butler, P.J., Milsom, W.K., and Woakes, A.J. 1984. Respiratory, cardiovascular and metabolic adjustments during steady state swimming in the green turtle, Chelonia mydas. J. Comp. Physiol. B Biochem. Syst. Environ. Physiol. 154: 167 174. Cann, J. 1998. Australian freshwater turtles. Beaumont Publishing, Singapore. Chessman, B.C. 1986. Diet of the Murray turtle, Emydura macquarii (Gray) (Testudines: Chelidae). Aust. Wildl. Res. 13: 65 69. Chessman, B.C. 1987. Atmospheric and aquatic basking of the Australian freshwater turtle Emydura macquarii (Gray) (Testudines: Chelidae). Herpetologica, 43: 301 306. Chessman, B.C. 1988. Habitat preferences of freshwater turtles in the Murray Valley, Victoria and New South Wales. Aust. Wildl. Res. 15: 485 491. Christy, E.J., Farlow, J.O., Bourque, J.E., and Gibbons, J.W. 1973. Enhanced growth and increased body size of turtles living in thermal and post-thermal aquatic systems. In Thermal ecology. Edited by J.W. Gibbons and R.R. Sharitz. USAEC, Springfield, Va. pp. 277 284. Crawford, K.M., Spotila, J.R., and Standora, E.A. 1983. Operative environmental temperatures and basking behavior of the turtle Pseudemys scripta. Ecology, 64: 989 999. Crocker, C.E., Graham, T.E., Ultsch, G.R., and Jackson, D.C. 2000. Physiology of common map turtles (Graptemys geographica) hibernating in the Lamoille River, Vermont. J. Exp. Biol. 286: 143 148. Dall Antonia, L., Lebboroni, M., Benvenuti, S., and Chelazzi, G. 2001. Data loggers to monitor activity in wild freshwater turtles. Ethol. Ecol. Evol. 13: 81 88. Dejours, P. 1994. Environmental factors as determinants in bimodal breathing: an introductory overview. Am. Zool. 34: 178 183. Doody, J.S., Sims, R.A., and Georges, A. 2001. Use of thermal springs for aquatic basking by the pig-nosed turtle, Carettochelys insculpta. Chelonian Conserv. Biol. 4: 81 87. Ernst, C.H. 1986a. Ecology of the turtle, Sternotherus odoratus, in southeastern Pennsylvania. J. Herpetol. 20: 341 352. Ernst, C.H. 1986b. Environmental temperatures and activities in the wood turtle, Clemmys insculpta. J. Herpetol. 20: 22 29. Gatten, R.E., Jr. 1974. Effects of temperature and activity on aerobic and anaerobic metabolism and heart rate in the turtles Pseudemys scripta and Terrapene ornata. Comp. Biochem. Physiol. A Comp. Physiol. 48: 619 648. Gatten, R.E., Jr. 1978. Aerobic metabolism in snapping turtles, Chelydra serpentina, after thermal acclimation. Comp. Biochem. Physiol. A Comp. Physiol. 61: 325 337. Gatten, R.E., Jr. 1980. Aerial and aquatic oxygen uptake by freelydiving snapping turtles (Chelydra serpentina). Oecologia, 46: 266 271. Georges, A. 1982. Ecological studies of Krefft s River Tortoise, Emydura krefftii (Gray), from Fraser Island, Queensland. Ph.D. thesis, University of Queensland, Brisbane, Queensland. Girgis, S. 1961. Aquatic respiration in the common Nile turtle, Trionyx triunguis (Forskal). Comp. Biochem. Physiol. 3: 206 217. Gordos, M.A., and Franklin, C.E. 2002. Diving behaviour of two Australian bimodally respiring turtles, Rheodytes leukops and Emydura macquarii, in a natural setting. J. Zool. (Lond.), 258: 335 342. Gordos, M.A., Franklin, C.E., and Limpus, C.J. 2003. Seasonal changes in the diving performance of the bimodally respiring freshwater turtles, Rheodytes leukops, in a natural setting. Can. J. Zool. 81: 617 625. Graham, T.E. 1972. Temperature-photoperiod effects on diel locomotor activity and thermal selection in the turtles, Chrysemys picta (Schneider), Clemmys guttata (Schneider), and Sternotherus odoratus (Latreille). Ph.D. dissertation, University of Rhode Island, Kingston. Graham, T.E. 1979. Locomotor activity in blanding s turtle, Emydoidea blandingii (Reptilia, Testudines, Emydidea): the phasing effect of temperature. J. Herpetol. 13: 365 366. Graham, T.E., and Hutchison, V.H. 1979. Turtle diel activity: response to different regimes of temperature and photoperiod. Comp. Biochem. Physiol. A Comp. Physiol. 63: 299 305. Hays, G.C., Hoschscheid, S., Broderick, A.C., Godley, B.J., and Metcalfe, J.D. 2000. Diving behaviour of green turtles: dive depth, dive duration and activity levels. Mar. Ecol. Prog. Ser. 208: 297 298. Herbert, C.V., and Jackson, D.C. 1985. Temperature effects on the responses to prolonged submergence in the turtle Chrysemys picta bellii. II. Metabolic rate, blood acid base and ionic changes, and cardiovascular function in aerated and anoxic water. Physiol. Zool. 58: 670 681. Hochscheid, S., Godley, B.J., Broderick, A.C., and Wilson, R.P. 1999. Reptilian diving: highly variable dive patterns in the green turtle Chelonia mydas. Mar. Ecol. Prog. Ser. 185: 101 112. Hynes, H.B.N. 1970. The ecology of running water. University of Toronto Press, Toronto, Ont. King, P., and Heatwole, H. 1994. Partitioning of aquatic oxygen uptake among different respiratory surfaces in a freely diving pleurodiran turtle, Elseya latisternum. Copeia, 1994: 802 806. King, J.M., Kuchling, G., and Bradshaw, S.D. 1998. Thermal environment, behavior, and body condition of wild Pseudemydura umbrina (Testudines: Chelidae) during late winter and early spring. Herpetologica, 54: 103 112. Krawchuk, M.A., and Brooks, R.J. 1998. Basking behavior as a measure of reproductive cost and energy allocation in the painted turtle, Chrysemys picta. Herpetologica, 54: 112 131. Legler, J.M., and Cann, J. 1980. A new genus and species of chelid turtle from Queensland, Australia. Los Angel. Cty. Mus. Nat. Sci. Contrib. Sci. 324: 1 18. Limpus, C.J., Limpus, D.J., and Hamann, M. 2002. Freshwater turtle populations in the area to be flooded by Walla Weir, Burnett River, Queensland: baseline study. Mem. Queensl. Mus. 48: 155 168. Moll, E.O., and Legler, J.M. 1971. The life history of the slider turtle, Pseudemys scripta (Schoepff), in Panama. Bull. Los Angel. Mus. Cty. Nat. Hist. Sci. 11: 1 102. Prassack, S.L., Bagatto, B., and Henry, R.P. 2001. Effects of temperature and aquatic PO 2 on the physiology and behaviour of Apalone ferox and Chrysemys picta. J. Exp. Biol. 204: 2185 2195. Priest, T. 1997. Bimodal respiration and dive behaviour of the Fitzroy River Turtle, Rheodytes leukops. B.Sc. Honours thesis, University of Queensland, Brisbane, Queensland. Priest, T., and Franklin, C.E. 2002. Effect of water temperature and oxygen levels on the diving behaviour of two freshwater turtles: Rheodytes leukops and Emydura macquarii. J. Herpetol. 36: 555 561. Rowe, J.W., and Moll, E.O. 1991. A radiotelemetric study of activity and movements of the Blanding s Turtle (Emydoidea blandigni) in northeastern Illinois. J. Herpetol. 25: 178 185.

1622 Can. J. Zool. Vol. 81, 2003 Santos, E.A., Laitano, S.Y.T., and Genofre, G.C. 1990. Diving physiology of Chrysemys dorbignyi Dum & Bibr., 1835 (Reptilia: Chelonia). Comp. Biochem. Physiol. A Comp. Physiol. 95: 229 236. Spencer, R., Thompson, M.B., and Hume, I.D. 1998. The diet and digestive energetics of an Australian short-necked turtle, Emydura macquarii. Comp. Biochem. Physiol. A Comp. Physiol. 121: 341 349. Spotila, J.R., Foley, R.E., Schubauer, J.P., Semlitsch, R.D., Crawford, K.M., Standora, E.A., and Gibbons, J.W. 1984. Opportunistic behavioural thermoregulation of turtles, Pseudemys scripta, in response to microclimatology of a nuclear reactor cooling reservoir. Herpetologica, 40: 299 308. Stockard, M.E., and Gatten, R.E., Jr. 1983. Activity metabolism of painted turtles (Chrysemys picta). Copeia, 1983: 214 221. Stone, P.A., Dobie, J.L., and Henry, R.P. 1992. The effect of aquatic O 2 levels on diving and ventilatory behaviour in softshelled (Trionyx spiniferus), stinkpot (Sternotherus odoratus), and mud turtles (Kinosternon subrubrum). Physiol. Zool. 65: 331 345. Tucker, A.D., Limpus, C.J., Priest, T.E., Cay, J., Glen, C., and Guarino, E. 2001. Home ranges of Fitzroy River turtles (Rheodytes leukops) overlap riffle zones: potential concerns related to river regulation. Biol. Conserv. 102: 171 181. Ultsch, G.R. 1989. Ecology and physiology of hibernation and overwintering among freshwater fishes, turtles, and snakes. Biol. Rev. Camb. Philos. Soc. 64: 435 516. Ultsch, G.R., and Jackson, D.C. 1982. Long-term submergence at 3 C of the turtle Chrysemys picta bellii, in normoxic and severely hypoxic water. J. Exp. Biol. 96: 11 28. Ultsch, G.R., and Jackson, D.C. 1995. Acid base status and ionic balance during simulated hibernation in freshwater turtles from the northern portion of their ranges. J. Exp. Zool. 273: 482 493. van Dam, R.P., and Diez, C.E. 1996. Diving behaviour of immature hawksbills (Eretmochelys imbricata) in a Caribbean cliff-wall habitat. Mar. Biol. 127: 171 178. Wetzel, R.G. 2001. Limnology: lake and river ecosystems. Academic Press, San Diego, Calif.