Hydric Conditions During Incubation Influence Locomotor Performance of Hatchling Snapping Turtles

Flexible-shelled eggs of turtles exchange water with the environment of the nest cavity (Cagle, 1950; Legler, 1954; Hendrickson, 1958; Montevecchi & Burger, 1975; M. J. Packard, Packard & Boardman, 1982; G. C. Packard, Paukstis, Boardman & Gutzke, 1985b). These water exchanges dramatically affect changes in mass of the eggs over the course of incubation. Eggs incubated on relatively wet substrates gain water and mass during incubation, while those incubated on relatively dry substrates lose water and mass over the course of development (Tracy, Packard & Packard, 1978; G. C. Packard, Taigen, Packard & Boardman, 1980; G. C. Packard, Packard & Boardman, 1981a, 1982; G. C. Packard, Packard, Boardman & Ashen, 1981b; Morris et al. 1983; G. C. Packard et al. 1983; G. C. Packard, Packard & Gutzke, 1985a).

Water exchanges between eggs and their environment affect the incubation and development of embryonic turtles. Embryos developing in eggs on relatively wet substrates incubate longer, catabolize their yolk more quickly, have a higher metabolic rate, and grow faster than embryos in eggs on relatively dry substrates (G. C. Packard et al. 1981b, 1985a; Morris et al. 1983; Gettinger, Paukstis & Gutzke, 1984). As a result, turtles hatching from eggs incubated on relatively wet substrates are larger, but contain a smaller mass of residual yolk, than those hatching from eggs incubated on drier substrates (G. C. Packard et al. 1980, 1981b, 1982, 1985a; Morris et al. 1983).

Larger hatchlings have been suggested to have a selective advantage, relative to smaller conspecifics, due to increased competitive ability or improved predator evasion, simply as the result of being larger (G. C. Packard et al. 1981b; Morris et al. 1983). We report here that snapping turtles hatching from eggs incubated on wet substrates exhibit improved locomotor performance well beyond that which might be expected on the basis of their larger size.

Four newly constructed snapping turtle nests were excavated on 12–14 June, 1984 at the Valentine National Wildlife Refuge, Cherry County, Nebraska. As eggs were removed from the nests they were gently brushed free of adhering sand, numbered to identify their clutch, and packed in moist peat moss in styrofoam containers. The containers were stored in a cool pump house until 15 June, when all eggs were transported by car to Colorado State University. On 16 June eggs were removed from the containers, recleaned, weighed, numbered, and assigned to treatments. This last date was regarded as day 0 of the experiment.

The four clutches of eggs were incubated half-buried in vermiculite substrates in covered containers at 29°C. Twenty-four eggs from each clutch were incubated on vermiculite with a water potential of –150kPa (relatively wet), and 24 eggs from each clutch were incubated on a substrate with a water potential of –850 kPa (relatively dry). Substrates with a water potential of –150kPa were produced by adding 333 g of water to 300g dry vermiculite, and substrates with a water potential of –850 kPa were produced by adding 54 g of water to 300 g dry vermiculite (Morris et al. 1983). Incubation under these conditions influences growth of embryos and the size of hatchlings (Morris et al. 1983).

Each egg was weighed using a Sartorius electronic balance on day 2 of incubation and at weekly intervals thereafter until the first egg was pipped on day 52 of incubation. Water was added to substrates twice each week to replace that which had been absorbed by eggs or lost through evaporation (G. C. Packard et al. 1981a, 1985a). Thus, relatively constant water content was maintained in incubation substrates for the duration of incubation.

Immediately after the first egg in a container was pipped, the eggs within the container were separated from one another with hardware cloth partitions so that hatchlings could be individually identified. On the day of emergence, hatchlings were cleaned of adhering vermiculite, weighed, and their carapace length was measured with a dial caliper. During the first week after hatching turtles remained on, or buried in, the vermiculite substrate of their artificial nests.

One week after hatching, locomotor performance of turtles from each clutch and water potential was determined while they were running on land or swimming in water. Turtles were stimulated to move as fast as possible by pinching their tails with blunt forceps. Lap times were taken at intermediate points during the exercise, and plots of distance moved as a function of time spent moving were constructed for each individual. Plots of carapace lengths (as an estimate of body length) moved as a function of time were also constructed for each individual to help adjust for differences in size among individuals. In both of these cases, distance was plotted as the dependent variable so that the slope of the plotted points would reflect turtle speed.

Performance in water was determined at 29 °C in a trough 88·8 cm in circumference and 5·5 cm wide containing water 4cm deep. Performance on land was determined on a substrate of wet paper towels at 23°C. Twenty-four hours prior to testing, turtles were isolated in closed paper cups containing vermiculite from their incubation box. Turtles tested on land were allowed 30 min prior to testing for their body temperature to reach 23°C.

Immediately following the first determination of performance, turtles were toeclipped on one rear toe so they could be individually identified, and placed in water 1cm deep at 21 °C for 1 week. Performance on land or in water was then redetermined as described above. Individuals performed in the same medium and at the same temperature as they had during the first determination of performance.

Speeds of turtles were analysed by repeated measures analysis of variance (Dixon, 1981). The fixed factor in these analyses was the water potential of the substrate during incubation (–150kPa vs –850kPa). Data were blocked by clutch to account for variation among the four clutches, and different estimates of performance from the same individual were treated as repeated measures.

We also determined the rate of lactate accumulation during sprint performance by turtles on land and in water. One week after hatching, turtles from each clutch (when possible) and water potential were forced to move as fast as possible using methods described above. Each animal was forced to move one of seven preselected distances. Immediately following the completion of the exercise, animals were plunged into liquid nitrogen. Whole-body lactate was subsequently determined on the frozen carcasses (Miller, 1983).

Data for lactate accumulation during locomotion were analysed by analyses of covariance (Nie et al. 1975). Substrate water potential was the fixed factor in these analyses, and, in the case of lactate accumulation on land, data were blocked by clutch to increase the power of the statistical test. Increasing distance moved by turtles was associated with a linear increase in lactate concentration (see below), so distance moved was used as a covariate to remove the effects of distance on lactate concentration for the purpose of these analyses.

Eggs incubated on the wetter substrate (–150 kPa) lost an average of 0·5 g over the course of incubation. In contrast, eggs incubated on the drier substrate ( – 850 kPa) lost an average of 3·0 g in the same interval (Table 1). Eggs on the wetter substrate also incubated longer and produced larger hatchlings than did clutchmates from eggs on the dry medium (Table 1).

Turtles from both treatments lost mass during the first week after hatching. Animals from –850 kPa lost an average of 0·35 g (4·9 % of hatching body mass) while turtles from –150kPa lost 0·25 g (2·7% of hatching body mass) during the week spent in their artificial nests prior to the first determination of locomotor performance. Animals from –850 kPa lost significantly more mass during this week than did animals from –150 kPa, as judged by analysis of covariance with egg mass on day 0 of incubation as a potential covariate to adjust for differences in size among turtles (F_{1 147} = 7·56, P = 0·007).

Both running and swimming turtles moved steadily over the relatively short distances we forced our animals to move in the first test of performance. Thus, plots of distance moved as a function of time spent moving for individuals were sufficiently linear (see Figs 1, 2) for us to characterize the performance of each individual by its speed (the slope). Hatchlings from eggs incubated at –150kPa moved substantially faster than clutchmates from eggs incubated at –850kPa (Table 2; Fig. 1). When data were expressed in units of body lengths s⁻¹, turtles from the wet environment were still faster than clutchmates from dry settings (Table 2; Fig. 2).

Turtles from both treatments gained mass while they were in water during the second week after hatching. Animals from –850kPa gained an average of 0·89g, while turtles from –150 kPa gained 0-08 g prior to the second determination of locomotor performance. Animals from –850 kPa gained significantly more mass during this week than did animals from –150 kPa, as judged by analysis of covariance with egg mass on day 0 of incubation as a potential covariate to adjust for differences in size among animals (F1,33 = 97·30, P<0·001).

All turtles were considerably faster during the second test of performance, and turtles from eggs incubated at –150 kPa were still faster, on both absolute and relative scales, than were their counterparts from eggs incubated at –850 kPa (Table 2).

The water potential of the incubation substrate also affected the rate of increase in lactate concentration during sprint performance by turtles, both while they were swimming in water (F₁,₃₄ = 8·16; P = 0·007) and running on land (F_1,47 = 5·27; P = 0·026). Hatchlings from eggs incubated on the relatively dry substrate accumulated lactate at a greater rate than their counterparts from eggs incubated on the wetter substrate, despite moving at a considerably slower speed (Fig. 3).

Turtles from eggs incubated at –850 kPa also accumulated more lactate than those from eggs incubated at –150 kPa. Turtles from the dry substrate accumulated an average of 1·01 mg lactate while running 132cm at l·2cms⁻¹, while turtles from the wet substrate accumulated only 0·69 mg running the same distance at 2·0 cms⁻¹. Further, turtles from the dry substrate accumulated an average of l·12mg lactate while swimming 266cm at 2·0cms⁻¹, while those from the wet substrate accumulated 0·34 mg lactate swimming the same distance at 3·3 cms’.¹.

Eggs incubated on the wet substrate declined 4% in mass over the course of incubation, while those incubated on the dry substrate lost 24% of their initial mass over the same interval. This pattern of response by incubating eggs to substrate water potential has been observed previously for eggs of snapping turtles (G. C. Packard et al. 1980, 1981b; Morris et al. 1983). Changes in the mass of eggs during incubation represent net transfers of water between incubating eggs and their environ-, ment. These water exchanges affect the physiology of developing turtles such that turtles incubating on relatively wet substrates incubate longer and are larger at hatching compared with those incubating on drier substrates (G. C. Packard et al. 1980, 1981b; Morris et al. 1983; Table 1). The mechanism through which water exchanges during incubation affect the physiology of developing turtles is not known (G. C. Packard & M. J. Packard, 1984). However, we have shown here that the impact of embryonic water exchanges on the physiology of developing turtles extends beyond hatching and affects the locomotor performance of young turtles.

Turtles from eggs incubated at –150 kPa moved substantially faster than clutchmates from eggs incubated at –850kPa (Table 2; Fig. 1). We fully expected turtles from the wet substrate to move faster than their counterparts from eggs incubated on the dry substrate, because of their larger size. Larger animals typically move faster than smaller members of the same species. However, when speeds were expressed relative to body size (as body lengths s⁻¹), turtles from the wet environment were still faster than clutchmates from dry settings (Table 2; Fig. 2). This finding was not anticipated, as larger animals typically move relatively more slowly (fewer body lengths s⁻¹) than smaller ones. This finding is without parallel in the literature on locomotion (see below).

The maximum speed sustained during prolonged bouts of swimming is absolutely faster but relatively slower in larger fish. However, the relative speed of burst locomotion in fish (lasting less than 15 s) is independent of body length (and about 10 lengths s⁻¹) (Bainbridge, 1960; Brett, 1965; Webb, 1975; Wu, 1977; Beamish, 1978). As a rule, the maximum running speed of mammals is absolutely faster but relatively slower in larger animals (Garland, 1983). Among reptiles, sprint speeds of lizards have been reported to increase in proportion to their body mass raised to powers ranging from 0 to 0 · 34 (Huey & Hertz, 1982; Garland, 1984 and references cited therein, 1985). This means that, at the extremes, absolute lizard speed is independent of lizard length (all lizards run at the same speed), or that relative lizard speed is independent of lizard length (a doubling of lizard length leads to a doubling of lizard speed). It has been proposed on theoretical grounds that the maximum speed of movement is independent of animal size (Thompson, 1942; Hill, 1950), and this has been observed in certain groups of mammals (Alexander, Langman & Jayes, 1977; Garland, 1983). In this case, all animals move at the same absolute speed, so smaller animals move at much faster relative speeds. In no case are larger animals predicted or observed to move at greater relative speeds.

Turtles hatching from eggs incubated at –850 kPa were probably dehydrated compared with those incubated at –150 kPa (Morris et al. 1983). Furthermore, animals from eggs incubated at –850 kPa lost more mass during the first week after hatching, which they spent in their artificial nests, and gained more mass during the second week after hatching, which they spent in water, compared with turtles from eggs incubated at –150 kPa. These observations suggest that turtles from eggs incubated on the dry substrate were slightly dehydrated during the first test of locomotor performance, compared with those from eggs incubated on the wet substrate, and this may explain some of the difference in performance between the two groups (Claussen, 1974; Beuchat, Pough & Stewart, 1984; Feder & Londos, 1984; Crowley, 1985; Stefanski, Gatten & Pough, 1985). All of our animals improved their performance during the second test, and turtles from eggs incubated on the dry substrate improved their performance more than those from eggs incubated on the wet substrate (Table 2). This suggests that dehydration was indeed a factor affecting the performance of both groups, and the magnitude of the difference in performance between the groups. Alternatively, the improved performance of animals in both groups during the second test of their locomotor performance may reflect an ontogenetic change in their ability due to the greater age of turtles during the second test of performance (Pough, 1978; Taigen & Pough, 1981).

Although all turtles were considerably faster during the second test of performance, turtles from eggs incubated at -150 kPa were still faster, on both absolute and relative scales, than their counterparts from eggs incubated at –850 kPa (Table 2). Whereas some of the difference in initial performance between turtles from eggs incubated on substrates differing in water potential may have been due to differences in hydration of animals, not all of the difference between groups can be explained on this basis.

To examine the metabolic basis for the superior locomotor performance by turtles from eggs incubated at –150 kPa, we determined the rate of lactate accumulation during sprint performance by turtles. Lactate accumulation is a measure of the extent of anaerobic metabolism during locomotion by reptiles (Bennett, 1982). The rate of lactate production during a sprint gives an estimate of the energy cost of locomotion (Miller, 1983). Differences in the rate of lactate accumulation among groups reared at different substrate water potentials presumably reflect different degrees of reliance on anaerobic metabolism in support of locomotion. Turtles from eggs incubated at –850 kPa may be more reliant on lactate production to supply energy for locomotion because of lower relative tissue hydration (see above). However, differences in tissue hydration of turtles from eggs incubated at the two water potentials were probably too small to affect cellular metabolism (Hillman, 1978; Clegg, 1981, 1984). Alternatively, lower rates of locomotion and increased reliance on anaerobic metabolism of turtles from eggs incubated on the drier substrate may be due to a greater percentage of body mass due to the mass of residual yolk in these hatchlings (Morris et al. 1983). The mass of residual yolk represents body mass that does not contribute to the generation of force during movement. The body mass of hatchling snapping turtles from eggs reared at –850 kPa includes about 18% yolk, while that of hatchlings from eggs reared at –150 kPa includes only 4 % residual yolk (Morris et al. 1983). However, most of the residual yolk is metabolized by the end of the second week after hatching in snapping turtles (Wilhoft, 1985), and so, at least during the second test of performance, most of our animals should have had little residual yolk left to carry. The improved locomotor performance of turtles from eggs experiencing favourable water exchanges may, however, reflect ontogenetic changes in endurance and metabolism (Pough, 1978; Taigen & Pough, 1981), expressed as a result of longer incubation times (Morris et al. 1983). Hatchling snapping turtles from eggs incubated on a wet substrate may be developmentally more mature compared with those from eggs incubated on a dry substrate.

Increased speed on land and reduced reliance on anaerobic metabolism to provide energy for locomotion should allow hatchling snapping turtles from nests constructed in relatively wet substrates to move the distance from nest to water more quickly than turtles from nests in drier settings. Further, once hatchlings from relatively wet nests reach water, their faster swimming speeds and lower reliance on anaerobic processes to provide energy for swimming locomotion may allow them to evade predators more successfully and to take a greater variety of food items than turtles from nests constructed in drier substrates. Thus, the adaptive value of experiencing favourable water exchange during incubation may be expressed as an improved locomotor performance resulting from physiological advantages exceeding those expected on the basis of larger size alone. This may be the basis for the improved competitive ability and increased survival of larger hatchlings of reptiles Froese & Burghardt, 1974; Ferguson & Bohlen, 1978; Fox, 1978; Swingland & Coe, 1979; Ferguson, Brown & DeMarco, 1982; Ferguson & Fox, 1984).

Turtle eggs were collected under authority of permit 84-69 issued by the Nebraska Game and Parks Commission. We thank Leonard McDaniel, Assistant Manager, for assistance on the Valentine Refuge, and Bill Gutzke for providing shade during the collection of eggs and for comments that significantly improved the quality of our manuscript. This research was supported in part by National Science Foundation Grant DCB-8308555 and by supplementary funds from the programme for Participation of Small College Faculty in NSF Supported Research at Large Institutions.