Positive or negative? The shell alters the relationship among behavioral defense strategy, energy metabolic levels and antioxidant capacity in freshwater turtles
The relationships among energy metabolic levels, behavioral and other physiological traits help to determine the trade-off of energy allocation between different traits and the evolution of life-history driven by natural selection. However, these relationships may be distinctive in selected animal taxa because of their unique traits. In the present study, the relationships among energy metabolic levels, behavioral defense strategies, and antioxidant capacity were explored in three freshwater turtle species with different shell morphologies, by assessing responses to attack, righting time, shell morphology, whole-organism metabolic rates, tissue metabolic enzyme activities and antioxidant levels.
The Chinese three-keeled pond turtles, Chinemys reevesii, showed a passive defense strategy, relatively larger shells, a higher resting metabolic rate (RMR) and higher antioxidant levels compared to the snapping turtle, Chelydra serpentina, or the Chinese soft-shelled turtle, Pelodiscus sinensis. These latter two species both showed an active defense strategy, a higher factorial aerobic scope and better muscle anaerobic metabolic capacity but relatively smaller shells, lower RMR and antioxidant capacity.
Our results indicate a negative relationship between RMR and activity levels in behavioral defense strategies along small-big shell continuum among the three turtle species. We also found a positive relationship between antioxidant capacity and energy metabolism but a negative one between antioxidant capacity and activity levels in defense strategies. The present study indicated a role of turtle shell in forming unique relationship between energy metabolic levels and behaviors in freshwater turtle taxa and a possible trade-off between the maintenance of physiological homeostasis and activity levels in energy allocation.
KeywordsAntioxidant capacity Behavioral defense strategy Freshwater turtles Resting metabolic rate Shell
Basal metabolic rate
Respiratory complex III
Cytochrome c oxidase
Factorial aerobic scope
- H/W ratio
Carapace height/width ratio
High performance liquid chromatography
Maximum metabolic rate
Net aerobic scope
- P/C ratio
Plastron/carapace flat area ratio
Resting metabolic rate
Response to attack
Total antioxidant capacity
Energetics and animal behaviors (e.g. predator avoidance strategies, boldness or exploration) are related to each other because both are associated with the slow-fast life-history continuum [1, 2, 3]. Thus, the relationship between energy metabolism and specific behavioral traits can show the consequences of natural selection as well as suggest the evolution of life histories and can also be expanded to link with other physiological or morphological traits [2, 4, 5]. For example, basal metabolic rate (BMR) was shown to correlate positively with animal activity levels at an interspecific level among most bird species [6, 7, 8] and also correlate positively with antioxidant capacity, an important redox balance parameter ; both indicated evolution along the slow-fast continuum. However, the relationship between energy metabolic levels and animal behaviors is not consistent across all animal taxa. In addition to behavioral traits, the links between energy metabolism and physiological or other traits also matter because they reflect the principle of energy allocation to other different functions based on the activity levels . Natural selection generally acts most directly on behaviors and/or energy metabolism, but much less directly on lower-level physiological or other traits [10, 11]. Thus, exploring the relationship among energy metabolic level, behavioral and other traits in specific animal taxa can help our understanding of how interspecific variations are maintained among related species by revealing how selection affects behaviors, energy metabolism and other traits and also reveal which traits are dominant therein.
The unique traits of freshwater turtles raise questions about their evolution, behaviors and physiology. They have peculiar characteristics as their unique shell  and their long lifespan . The shell offers extra protection when turtles are attacked by predators or encounter environmental stressors but shell formation would incur extra energy costs during development and growth. Thus, the presence of the shell may result in special relationship between energy metabolic levels and turtle behaviors under the selection of predation or environmental stress, which may further affect other physiological traits. Among freshwater turtle species, interspecific variations in shell morphology, behavioral traits (e.g. righting response) and physiological traits (e.g. antioxidant capacity or immunity) have been widely observed [14, 15, 16, 17, 18, 19]. How these variations were shaped, maintained and linked may be explained by the relationship between energy metabolism and these traits although this relationship is not yet well understood.
Animals holding and experimental process
Chinese soft-shelled turtles, P. sinenesis (n = 15, mass = 112.6 ± 8.3 g), were obtained from a turtle hatchery facility (Yutian County, Hebei province, PR China). C. reevesii (n = 15, mass = 108.2 ± 5.9 g) and C. serpentina (n = 15, mass = 148.5 ± 5.2 g) were obtained from an aquarium market (Beijing Guanyuan market, PR China). All turtles were raised in tanks (70 × 50 × 40 cm with water depth of about 6 cm) at 26 ± 1 °C (a suitable temperature for the feeding and growth of all three species) for at least 6 weeks. All turtles were given commercial feed daily. Photoperiod was kept at 12 L/12 D.
After acclimation, all turtles were assessed for (a) behavioral traits: righting time (RT) and response to attack (RTA), (b) morphological traits: carapace height/width ratio (H/W ratio), plastron/carapace flat area ratio (P/C ratio), and (c) metabolic traits: RMR, maximum metabolic rate (MMR), net aerobic scope (NAS) and factorial aerobic scope (FAS). Then turtles were transferred back to their original tanks for another two weeks. Afterwards, all individuals of each species were divided into two groups. One group was defined as the control group without treatment (n = 7). The other group, which was defined as the electrical stimulated group (ESG, n = 8), experienced forced movement by electrical stimulation (30-50 V, 20 hz, 10 msec, 5 min or to exhaustion) [27, 28, 29]. After treatment, all turtles were sacrificed to excise liver, heart and skeletal muscle of the left forelimb. Liver and heart were weighed to provide liver index and heart index. All tissues were frozen in liquid nitrogen immediately and stored at − 80 °C. Activities of selected enzymes were measured. These included (a) energy metabolism related mitochondrial respiratory enzymes: succinate dehydrogenase (SDH), respiratory complex III (C3), cytochrome c oxidase (CO) and lactate dehydrogenase (LDH), and (b) antioxidant enzymes: TAC, superoxide dismutase (SOD) and catalase (CAT). Levels of lactic acid (LA) and ascorbic acid (AA) in tissues were also measured.
Righting time reflects the response to being overturned and the hiding time until the predator leaves [14, 30]. The response to attack reflects the choice between different behavioral defense strategies of freshwater turtles when disturbed. The water temperature for all behavioral assays was maintained at 26 °C and a 30 min acclimation period to testing area was applied for each turtle before treatments. For the righting time survey, each turtle was placed in one thermostatic chamber (diameter = 20 cm, water depth = 5 cm). After acclimation, turtles were turned upside down in the chamber to start the test. The righting process was recorded with a camera to the nearest second starting from the moment that the turtle was placed upside down to the moment that the animal righted itself completely . For response to attack, a similar method was used as described by Chang et al. . Each turtle was placed in one tank (140 × 60 × 70 cm, water depth = 5 cm) and was tapped on the carapace with a tweezer from a hidden position. The immediate response was recorded and graded with the following criteria: (1) run away, (2) stay still, (3) turn to attack, or (4) retraction into the shell. Each individual was tested twice, as in previous related studies [14, 31].
Shell morphology, liver and heart index
Maximum carapace height and carapace straight width were measured with a digital caliper (Pro skit, PR China) to the nearest millimeter. According to the model of Domokos and Várkonyi , H/W ratio of turtles was used to judge if the turtle species belonged to “flat turtles” (H/W under approx. 0.6), “medium turtles” (H/W between 0.6 and 0.8) or “tall turtles” (H/W above approx. 0.8) groups. Flat areas of carapace and plastron were measured with a camera and assessed using image analysis software (Digimizer, Belgium). The P/C ratio was used to compare the degree of shell coverage of the three turtle species. Liver index and heart index were calculated as the ratio of tissue wet weight to fresh animal body weight.
Metabolic rates and aerobic scope
RMR of each turtle was measured in air at 26 °C using an airtight chamber (diameter = 20 cm, height = 6.5 cm) and oxygen sensor (FireStingO2, Pyro science, Germany). Turtles were fasted for 24 h and kept still in the chamber. Each individual was acclimated to the experimental environment for 30 min with the cover off. Then the cover was closed tightly, the survey was started and the change in oxygen percentage (%) over 30 min was recorded. To measure the MMR, turtles were forced to run for 5 min or until exhaustion in response to electrical stimulation (30-50 V, 20 hz, 10 msec). Then the turtle was quickly placed in the chamber to measure the change in oxygen percentage (%) over the following 30 min. The change in oxygen percentage inside the chamber during the initial 100 s was chosen to calculate the MMR. The air volume was measured by filling the chamber with water in the presence of the animal and then measuring the volume of water to the nearest milliliter. FAS (defined as the MMR/RMR ratio) and NAS (defined as difference between MMR and RMR) were calculated .
After sampling, tissues were homogenized in phosphate buffered solution (9 g/L NaCl, 726 mg/L Na2HPO4-7H2O, and 210 mg/L KH2PO4, pH 7.2) for biochemical assays. Reagent kits (Nanjing Jiancheng, PR China) were used to measure the activities of SDH, C3, CO, SOD, CAT, LDH, TAC and LA level according to the instruction manual for each kit.
Gradient centrifugation was employed to isolate mitochondria from tissue homogenates for the SDH, C3 and CO activity assays. To do this, homogenates were centrifuged at 1000 g for 10 min, then supernatants were removed and recentrifuged at 12000 g for 15 min, followed by retrieving pellets, that were then washed and resuspended (all steps were done at 4 °C). SDH activity was measured using a 2,6-dichlorophenolindophenol (DCPIP) reduction reaction. The DCPIP reduction speed was determined by the change in absorption value at 600 nm . C3 activity was measured using a reaction in which the C3 complex reduced oxidized cytochrome c to the reduced form in the presence of a hydrogen donor and the increase in absorbance at 550 nm was measured for 2 min . CO activity was measured using a reaction that converted cytochrome c to its oxidized form and the decrease in absorbance value at 550 nm was measured for 1 min .
Portions of each tissue homogenate were centrifuged at 4500 g and 4 °C for 10 min to provide supernatants for antioxidant enzyme activity assays, LDH activity and LA content. TAC was measured by determining the absorption value at 520 nm when Fe3+ was reduced to Fe2+ in the presence of antioxidants . SOD activity was measured using a cytochrome c reduction inhibition reaction in a xanthine-xanthine oxidase system and the degree of inhibition was measured by the absorption value at 550 nm . CAT activity was measured by determining the rate of H2O2 decomposing in 1 g protein per second with the change of absorption measured at 405 nm. LDH activity was measured using the reverse reaction in which LDH catalyzes the conversion of lactate to pyruvate using NAD+ as the hydrogen donor. Then 2,4-dinitrophenylhydrazine was used to determine pyruvate product levels, monitored by absorption change at 440 nm. LA content was measured following the method of Buttery et al. . LA was transformed to pyruvate with the generation of NADH, which reacted with phenazine methosulfate; then the products reduced p-iodonitrotetrazolium to its reduced form which was assessed by the absorption value at 530 nm. Total protein content was measured according to the Coomassie blue dye-binding method with bovine serum albumin as the standard .
We employed the method described in Chen et al.  to measure tissue AA content, which is an essential antioxidant in turtles [41, 42]. Tissues were homogenized with 15% cold-phosphoric acid at a ratio of 1:19 (w/v) and then homogenates were centrifuged at 20,000 g for 20 min. The supernatants were used for high performance liquid chromatography (HPLC) analysis with electrochemical detection (Waters 2695/2487, USA) and a 5 μm analytical column (4.6 × 250 mm, Aglient, USA). Injection volume was 20 μL, flow rate was 1 mL per min and ultraviolet detection was set as 243 nm.
The hierarchical clustering algorithm was generated with all data according to a Euclidean method. The clustering analysis and the heatmap were conducted with R v. 3.5.0 (Development Core Team 2018). Then, all data were checked for normality by the Kolmogorov-Smirnov test and homogeneity of variance. For metabolic rates, a general linear model was employed to conduct interspecies comparation with mass as covariate. For enzyme activities and LA level, intraspecific differences between the control group and the treatment group (ESG) were assessed via a t-test. If no significant difference between the two groups was found, data of the two groups were combined for interspecific comparison. Interspecific variations in all parameters, except metabolic rates, were detected using one-way ANOVA followed by Tukey HSD post hoc test, or Kruskal-Wallis test followed by Mann-Whitney U post hoc test for data that did not fit the check for normality or variance homogeneity. P < 0.05 was set as the significance level. The correlations among different morphological and physiological parameters were analyzed using Pearson’s correlation analysis, while the correlations between behavioral parameters, which are discrete variables, and others were analyzed using Spearman’s correlation analysis.
In the righting test, C. reevesii spent the longest time whereas P. sinensis spent the shortest time in righting themselves (RTC.reevesii = 66.6 ± 16.0 s, RTC. serpentina = 63.3 ± 33.0 s and RTP. sinensis = 1.3 ± 0.2 s, χ2 = 22.746, df = 2, P < 0.001, Fig. 3a). The three turtle species also showed different responses to attack (χ2 = 16.941, df = 2, P < 0.001, Fig. 3b). More than half of C. reevesii retracted into shell whereas the others ran away. All P. sinensis except one ran away when faced with attack and most C. serpentina stayed still (Fig. 3b).
Metabolic rates, aerobic scopes and metabolic enzyme activities
C. reevesii showed the highest RMR compared with the other two species (F2,41 = 10.438, Pspecies < 0.001, Pmass = 0.259 and Pinteraction = 0.551, Fig. 4a). Maximum metabolic rate (MMR) of these three turtle species showed no difference (F2,41 = 0.169, Pspecies = 0.845, Pmass = 0.020 and Pinteraction = 0.886, Fig. 4a). Factorial aerobic scope (FAS) of both C. serpentina and P. sinensis were higher than that of C. reevesii (F2,41 = 4.592, Pspecies = 0.016, Pmass = 0.427 and Pinteraction = 0.978, Fig. 4a).
In the liver, SDH activity of C. reevesii was higher than the other two species (F2,42 = 70.002, P < 0.001, Fig. 4b). Hepatic C3 activities of C. reevesii and P. sinensis were higher than that of C. serpentina (χ2 = 10.758, df = 2, P = 0.005, Fig. 4c). P. sinensis had the highest muscular LDH activity, while that of C. reevesii was the lowest (F2,42 = 25.267, P < 0.001, Fig. 4d). Forced movement only induced muscle LA accumulation in C. serpentina and P. sinensis but not in C. reevesii (Pc. serpentina < 0.001; Pp. sinensis = 0.045 and Pc. reevesii = 0.722, Fig. 4e).
Morphological traits, liver and heart index
C. reevesii showed the highest carapace H/W ratio (F2,42 = 611.948, P < 0.001), P/C ratio (F2,42 = 382.29, P < 0.001), liver index (χ2 = 32.949, df = 2, P < 0.001) and heart index (F2,40 = 14.24, P < 0.001) among the three turtle species (Fig. 5). P. sinensis showed a higher P/C ratio but lower values for other morphological parameters than those of C. serpentina (Fig. 5). Both H/W ratio and P/C ratio positively correlated with RMR among the three species (RMR vs. H/W ratio: r = 0.322, P = 0.031; RMR vs. P/C ratio: r = 0.510, P < 0.001).
In hepatic antioxidant defense, C. reevesii showed the highest TAC among the three species (F2,42 = 20.515, P < 0.001, Fig. 6a). The SOD activity in the liver of C. serpentina was the lowest among the three turtles (χ2 = 29.4123, df = 2, P < 0.001, Fig. 6b). In the heart, C. reevesii showed the highest TAC as well as CAT activity (TAC: χ2 = 23.279, df = 2, P < 0.001, Fig. 6c; CAT: χ2 = 31.529, P < 0.001, Fig. 6d).
Correlation among behavioral traits, RMR, metabolic and antioxidant capacity
Spearman’s correlation between behavioral and primary morphological parameters as well as physiological parameters (n = 45)
Response to attack
Relationships among energy metabolic levels, behavioral and other traits link with energy constraints and trade-offs driven by natural selection. In the present study, as we predicted, a negative relationship was observed between energy metabolic levels, antioxidant capacity and activity levels in behavioral defense strategies along small-big shell continuum, indicating two possible branches in freshwater turtle taxa driven by the predation or environmental stress (Fig. 2).
Negative relationship between metabolic level and activity levels in defense strategy linked with shell morphology
The behavioral assays showed that a passive defense strategy was used by C. reevesii whereas an active defense strategy was used by C. serpentina and P. sinensis (Fig. 3). Both the long righting time and the behavioral response, retraction into the shell, showed by more than half of the C. reevesii individuals indicated that C. reevesii tended to hide in their shell and endure the disturbance until a confirmation of safety was achieved. However, the strategy of P. sinensis was the complete opposite. P. sinensis showed rapid righting and escape behaviors when disturbed, indicating high activity levels in its defense strategy (Fig. 3). Interestingly, C. serpentina showed a short righting time and did not show behavioral response when disturbed (Fig. 3). In Dodd and Brodie’s report , C. serpentina moved to keep facing the attacker with lifting its shell and biting behavior, but animals did not retract into their shell when they encountered intense attack or were disturbed. Thus, C. serpentina tend to be an active defense strategy user with a high activity level in response to intense stress. The active defense strategy of C. serpentina and P. sinensis was also linked with the higher FAS value and better muscular anaerobic capacity in both two species (Fig. 4a, d and e). The higher FAS value reflects a higher oxygen debt bearing capacity related to RMR . High FAS has been observed in fish and lizard species that frequently experience intense combat as the consequence of physiological adaptation to high activity level [44, 45, 46]. Similarly, better muscular anaerobic capacity, indicated by higher muscle LDH activity and quicker LA accumulation in C. serpentina and P. sinensis than that in C. reevesii (Fig. 4d and e), was also shaped by high activity level in life history and has been observed in animals that are good at diving, foraging and escaping [47, 48].
The three freshwater turtles also varied in their resting metabolic rates (RMR) with a higher RMR of C. reevesii versus lower RMR in C. serpentina and P. sinensis (Fig. 4a). The higher RMR of C. reevesii indicated higher energy investment in maintenance  and links with relatively larger organs, liver and heart, related to energy conversion or supply (Fig. 5c). In turn, the bigger liver also contributes to the higher RMR because the liver usually shows high mass-specific metabolic rates and can account for a large part of whole-organism metabolic rate . In addition, a higher energy metabolic capacity in C. reevesii can also be supported by higher hepatic SDH and C3 activities (Fig. 4b and c). Mitochondria consume almost 90% of cellular respiration to provide ATP to power cell functions  and this process depends on the SDH and C3 activities [51, 52]. Interestingly, a positive correlation was observed between righting time and RMR among the three species (Table 1), indicating that the species with the higher RMR showed a more passive response when being placed upside down. Thus, as we predicted in Fig. 1, C. reevesii showed a passive behavioral defense strategy while a higher RMR than the other two turtle species, which showed an active strategy and lower RMR.
The reason for the negative relationship between RMR and activity level in the behavioral defense strategies of these turtles might be their unique morphological trait, the shell. Our results indicated clear difference among the shell morphology of the three species and relationship among shell morphologies, behavioral traits and RMR. The shell of C. reevesii was higher (greater H/W ratio, Fig. 5a) and had a relatively bigger plastron (higher P/C ratio, 5B) than the other two species, which may link with the passive defense strategy (Table 1). A higher shell can offer better protection for turtles against attack from predators and protect them during hiding in shell  while a flatter shell, such as that of C. serpentina or P. sinensis, is advantageous for swimming or digging and benefits their survival in active defense strategy [15, 54]. Similarly, the larger plastron can offer a larger radius of protection for the ventral side of turtles when they are overturned. Thus, the variation in shell morphology matches the different behavioral defense strategies of turtle species and can also be observed in other animals that similarly carry a similar shell, such as land snails . At the same time, shell morphology also linked with RMR among the three turtle species with a positive correlation (RMR vs. H/W ratio: r = 0.322, P = 0.031; RMR vs. P/C ratio: r = 0.510, P < 0.001), indicating a greater energy cost to support bigger shells. The extra energy cost may represent both the energy required to grow and maintain the shell and the energy debt involved in the daily activity carrying the bigger shell. In fact, in the righting response, turtle species with a higher shell need less energy investment for their self-righting than the species with flat shell [15, 56]. Thus, the energy investment in shell maintenance or structure may be an energy budget in several turtle species, which can be reduced to compensate for the more energy-demanding activity in other species, and promote a unique trade-off of energy allocation between activity levels for behavioral defense strategies and shell morphology.
Positive relationship between antioxidant capacity and metabolic levels linked with different defense strategies
Good antioxidant defenses are important for all animals to deal with oxidative stress and various chemical and xenobiotic stresses. Among freshwater turtles, our results indicated a higher tissue antioxidant capacity along with higher energy metabolic capacity (Figs. 6 and 7). On one hand, a higher RMR means a higher energy cost of self-maintenance and, in the present study, a possible cost is the investment in antioxidant defense. On the other hand, a higher RMR also may result in more reactive oxygen species (ROS) production, thereby promoting a higher tissue antioxidant capacity . For example, hepatic SOD activity correlated with the mitochondrial energetics, probably to eliminate superoxide radicals (O2−) generated in mitochondria in association with high rates of aerobic respiration among the three species (Additional file 3: Figure S2).
The variation in antioxidant capacity also linked with different behavioral defense strategies (Table 1). The passive defense strategy of C. reevesii may result in animals taking refuge in an adverse environment, such as hiding underwater to escape predators but without access to air to breathe. There turtles can experience hypoxia/anoxia stress which is associated with excessive ROS generation in animal tissues . A high constitutive antioxidant capacity favors the maintenance of redox balance without an extra energy investment when animals face with environmental stress frequently. Thus, high antioxidant capacity could be an important physiological adaption to the passive behavioral defense strategy of freshwater turtles, such as C. reevesii.
Trade-off between behavioral activities and physiological homeostasis
The relationships among energy metabolism, behavioral traits and other traits reflect trade-offs of energy allocation, which is driven by natural selection (Engqvist et al. 2015; Houston and Mcnamara 1989; Réale et al. 2010; Reznick et al. 2000). Our results provided evidence of the special negative correlation between energy metabolism, antioxidant capacity and behavioral activity levels in different defense strategies of freshwater turtles (as shown in Fig. 1b). These relationships link with the energy allocation trade-off between maintenance of physiological homeostasis (big shell, high RMR and antioxidant defense) and activity levels (high FAS and efficient muscular anaerobic capacity) (Cohen et al. 2008; Galliard et al. 2013). Physiological homeostasis can be considered as less variation in physiology over time. Repeated regulation of homeostasis is costly due to the expense of adjusting all physiological processes and producing molecules involved in homeostatic regulation (Cohen et al. 2008). Among freshwater turtle species, the size and mineralization of the shell has been shown to be positively correlated with protection against predator attack or environmental stresses [58, 59]. At the same time, a high constitutive antioxidant capacity can maintain redox balance without extra energy cost in transcription and translation of antioxidant proteins when animals encounter predator attack or environmental stresses [60, 61]. Thus, some turtle species, such as C. reevesii, invest much energy in building their shell and in high constitutive antioxidant levels while showing a passive endurance strategy under stress or attack (Fig. 2 Branch 2). This strategy further avoids frequent fluctuations in physiological processes such as energy metabolism or redox balance. Oppositely, other turtle species, including C. serpentina and P. sinensis, show a high investment in behavioral defense strategies and are good at muscle anaerobic mechanism (Fig. 2 Branch 1). They may thus form shells with lower protection and lower constitutive tissue antioxidant capacity which both can be related to a relatively lower energy investment in maintenance. However, they may benefit from their strategy because immediate and active defense responses to extrinsic factors may cause greater adaptability to a changing environment.
The present study assessed the relationships among energy metabolism, behavioral and physiological traits in three freshwater turtle species. Our results indicated a negative relationship between energy metabolic levels and activity levels in behavioral defense strategies, which is shaped by the extra energy investment needed to support shell morphology. We also found an interspecific variation in antioxidant capacity promoted by both energy metabolic levels and behavioral defense strategies. A possible trade-off of energy allocation between the maintenance of physiological homeostasis and animal activity levels can be inferred under the selective pressures of predation and environmental stress.
This study was funded by the National Natural Science Foundation of China (Grant No.31670419). All authors declare no conflict of interest.
Availability of data and materials
Primary data in this study are available in Additional file 1.
WYZ and CJN conceived the study; WYZ and YKL collected the data; WYZ analyzed the data and wrote the initial draft of the manuscript; KBS revised the manuscript. All authors read and approved the final manuscript.
All experiments were conducted in accordance with the standards of the Ethics and Animal Welfare Committee of Beijing Normal University (Approval NO. CLS-EAW-2017-014). All applicable institutional and/or national guidelines for the care and use of animals were followed.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 1.Biro PA, Stamps JA. Are animal personality traits linked to life-history productivity? Trends Ecol Evol. 2008;23:361–8.Google Scholar
- 2.Réale D, Garant D, Humphries MM, Bergeron P, Careau V, Montiglio PO. Personality and the emergence of the pace-of-life syndrome concept at the population level. Philos T R Soc B. 2010;365:4051–63.Google Scholar
- 6.Kalcounis-Rueppell MC. Relationship of basal metabolism and life history attributes in Neotomine-Peromyscine rodents (Cricetidae: Neotominae). Ecoscience. 2007;14:347–56.Google Scholar
- 7.Wiersma P, Muñoz-Garcia A, Walker A, Williams JB. Tropical birds have slow pace of life. P Natl Acad Sci USA. 2007;104:9340–5.Google Scholar
- 8.Careau V, Orp BE, Thomas DW, Réale D, Humphries MM. Exploration strategies map along fast-slow metabolic and life-history continua in muroid rodents. Funct Ecol. 2010;23:150–6.Google Scholar
- 10.Garland T, Losos JB. Ecological morphology of locomotor performance in squamate reptiles. Chicago: University of Chicago Press; 1994.Google Scholar
- 14.Delmas V, Baudry E, Girondot M, Prevotjulliard AC. The righting response as a fitness index in freshwater turtles. Biol J Linn Soc. 2007;91:99–109.Google Scholar
- 15.Domokos G, Várkonyi PL. Geometry and self-righting of turtles. P Roy Soc B. 2008;275:11.Google Scholar
- 16.Zhang WY, Niu CJ, Chen BJ, Yuan L. Antioxidant responses in hibernating Chinese soft-shelled turtle Pelodiscus sinensis hatchlings. Comp Biochem Physiol A. 2017;204:9–16.Google Scholar
- 18.Ibáñez A, Matĺn J, Gazzola A, Pellitteri-Rosa D. Freshwater turtles reveal personality traits in their antipredatory behaviour. Behav Process. 2018;157:142–7.Google Scholar
- 20.Cheung SM, Dudgeon D. Quantifying the Asian turtle crisis: market surveys in southern China, 2000–2003. Aquat Conserv. 2006;16:751–70.Google Scholar
- 21.Dodd CK, Brodie ED. Notes on the defensive behavior of the snapping turtle, Chelydra serpentina. Herpetologica. 1975;31:286–8.Google Scholar
- 22.Niu CJ, Zhang T, Sun R. Food consumption and growth of juvenile Chinese soft-shelled turtles (Pelodiscus sinensis) in relation to body weight and water temperature. Asiatic Herpetol Res. 1999;8:81–4.Google Scholar
- 23.Polocavia N, López P, Martín J. Interspecific differences in responses to predation risk may confer competitive advantages to invasive freshwater turtle species. Ethology. 2008;114:115–23.Google Scholar
- 24.Martín J, López P. An experimental test of the costs of antipredatory refuge use in the wall lizard, Podarcis muralis. Oikos. 1999;84:499–505.Google Scholar
- 25.Costantini D, Rowe M, Butler MW, McGraw KJ. From molecules to living systems: historical and contemporary issues in oxidative stress and antioxidant ecology. Funct Ecol. 2010;24:950–9.Google Scholar
- 26.Hermes-Lima M, Zenteno-Savĺn T. Animal response to drastic changes in oxygen availability and physiological oxidative stress. Comp Biochem Physiol C. 2002;133:537–56.Google Scholar
- 28.Gatten RE. Aerobic metabolism in snapping turtles, Chelydra serpentina, after thermal acclimation. Comp Biochem Physiol A. 1978;61:325–37.Google Scholar
- 29.Stockard ME, Gatten RE. Activity metabolism of painted turtles (Chrysemys picta). Copeia. 1983;1983:214–21.Google Scholar
- 31.Steyermark AC, Spotila JR. Body temperature and maternal identity affect snapping turtle (Chelydra serpentina) righting response. Copeia. 2001;2001:1050–7.Google Scholar
- 34.Gleason C, Huang S, Thatcher LF, Foley RC, Anderson CR, Carroll AJ, Millar AH, Singh KB. Mitochondrial complex II has a key role in mitochondrial-derived reactive oxygen species influence on plant stress gene regulation and defense. P Natl Acad Sci USA. 2011;108:10768–73.Google Scholar
- 35.Kirby DM, Thorburn DR, Turnbull DM, Taylor RW. Biochemical assays of respiratory chain complex activity. Method Cell Biol. 2007;80:93.Google Scholar
- 37.Orbea A, Ortiz-Zarragoitia M, Solé M, Porte C, Cajaraville MP. Antioxidant enzymes and peroxisome proliferation in relation to contaminant body burdens of PAHs and PCBs in bivalve molluscs, crabs and fish from the Urdaibai and Plentzia estuaries (Bay of Biscay). Aquat Toxicol. 2002;58:75–98.PubMedGoogle Scholar
- 40.Chen BJ, Niu CJ, Yuan L. Ascorbic acid regulation in stress responses during acute cold exposure and following recovery in juvenile Chinese soft-shelled turtle (Pelodiscus sinensis). Comp Biochem Physiol A. 2015;184:20–6.Google Scholar
- 41.Zhou XQ, Niu CJ, Sun RY. The effect of vitamin C on stress withstanding capability in the juvenile soft-shelled turtle (Pelodiscus sinensis). Aquac Nutr. 2015;11:169–74.Google Scholar
- 44.Alderman S. Bold fish, shy fish: it's all in the metabolism. J Exp Biol 2016, 219:1937–1937.Google Scholar
- 46.Bickler PE, Anderson RA. Ventilation, gas exchange, and aerobic scope in a small monitor lizard, Varanus gilleni. Physiol Zool. 1986;59:76–83.Google Scholar
- 47.Gleeson TT. Lactic acid production during field activity in the Galapagos marine iguana, Amblyrhynchus cristatus. Physiol Zool. 1980;53:157–62.Google Scholar
- 48.Gleeson TT, Dalessio PM. Lactate and glycogen metabolism in the lizard Dipsosaurus dorsalis following exhaustive exercise. J Exp Biol. 1989;144: 377–393.Google Scholar
- 49.Burton T, Killen SS, Armstrong JD, Metcalfe NB. What causes intraspecific variation in resting metabolic rate and what are its ecological consequences? P Roy Soc B. 2011;278:3465–73.Google Scholar
- 50.Rolfe DFS, Brand MD. Contribution of mitochondrial proton leak to skeletal muscle respiration and to standard metabolic rate. Am J Phys. 1996;271:1380–9.Google Scholar
- 53.Pritchard PCH. Encyclopedia of turtles. Copeia. 1980;3:566–7.Google Scholar
- 54.Claude J, Paradis E, Tong H, Auffray JC. A geometric morphometric assessment of the effects of environment and cladogenesis on the evolution of the turtle shell. Biol J Linn Soc. 2015;79:485–501.Google Scholar
- 58.Hu DL, Sielert K, Gordon M. Turtle shell and mammal skull resistance to fracture due to predator bites and ground impact. J Mech Mater Struct. 2012;6:1197–211.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.