TY - JOUR
AU - Jusup,, Marko
AB - Abstract Why is postnatal parental provisioning so rare in ectothermic vertebrates while prolonged parental care is almost ubiquitous in endotherms? We argue that the scarcity of postnatal parental care is a result of ectothermy itself. While almost all endothermic young require prolonged postnatal care due to thermal constraints, ectothermic physiology does not pose the same constraint. Most ectothermic young are thus independent from birth. Ectothermic mothers are better off investing in future reproductive events than to continue investing into independent young, because the cost of feeding young does not outweigh the benefits. Ectothermy further releases the constraint on offspring size resulting in offspring of ectothermic vertebrates often being much smaller than their parents. When parents and offspring differ greatly in size, both tend to specialize on different diets, making the feeding of young by much larger individuals not impossible, but less likely. Additionally, when the size difference between parents and offspring is significant, both are likely to live in different habitats. Such spatial segregation is also less conducive to the evolution of parental care. In those species where parents and offspring are not spatially separated and parental care does occur, it is mainly restricted to the guarding of eggs or juveniles; parents of very few species provide their offspring with food. We conclude that parental care beyond the guarding of eggs or young is much less likely to evolve in ectothermic vertebrates compared with endothermic vertebrates, unless there are exceptional circumstances that strongly select for parents to provide for their offspring. INTRODUCTION Ectothermic vertebrates, the reptiles, amphibians, and fishes, hardly ever provide care to their young beyond the guarding of eggs or juveniles (Shine 1988; Clutton-Brock 1991; Crump 1995; Greene et al. 2002; Webb and Manolis 2002; O’Connor and Shine 2004; Balshine 2012; Ferrara et al. 2014). Such lack of parental provisioning in ectothermic vertebrates is in sharp contrast to the almost ubiquitous presence of parental provisioning in birds and mammals (Farmer 2003; Balshine 2012). The stark difference in parental care between ectothermic and endothermic vertebrates requires an explanation (Shine 1988). It could be argued that the ectothermic vertebrates are too diverse a group to consider as one, especially given their long separate evolutionary histories. The living “reptiles” are a polyphyletic group; the 4 main lineages are usually described as reptiles–crocodilians, turtles, squamates (snakes and lizards), and tuatara. They are very different animals that have evolved independently since the Triassic (Hedges and Poling 1999). Modern amphibians are monophyletic and comprise 3 lineages while fishes are classified into 5 monophyletic groups containing as many as 11 lineages (Meyer and Zardoya 2003). However, despite their independent evolutionary histories, all reptiles, amphibians, and fishes share one important characteristic: they mostly rely on environmental sources for heat gain (Pough 1980). As we will argue, ectothermy allows for greater flexibility in life-history traits compared with endothermy, mainly because ectotherms have low energy demands (Pough 1980). This greater flexibility, in turn, does not easily select for postnatal care of offspring in the form of the provisioning of young. To illustrate how energy demands affect behavioral and physiological flexibility of ectotherms and endotherms, we place species along a supply-demand spectrum (Kooijman 2014; Lika et al. 2014). A supply-demand spectrum is based on the rules of physiological energetics; a formal, physics-like study of how environmental conditions (chiefly food availability and temperature (Nisbet et al. 2000; Sousa et al. 2008; Sousa et al. 2010; Jusup et al. 2017a, 2017b)) affect physiological performance in terms of growth, maturation, and reproduction (Jusup et al. 2014; Marn et al. 2017a, 2017b). The position of a given species on the supply-demand spectrum is determined by the balance between energy assimilation and maintenance needs. Assimilated energy is the result of ingestion after accounting for digestion inefficiencies and conversion losses (Nisbet et al. 2012). Maintenance needs is an umbrella term for energy expenditures that keep organisms alive and allow them to perform daily functions. A systematic analysis of the balance between energy assimilation and maintenance needs reveals that species are distributed along the supply-demand spectrum, with typical supply-side species possessing evidently different characteristics from demand-side ones (Table 1) (Kooijman 2014; Lika et al. 2014). The defining property of supply-side species is that energy assimilation and maintenance needs exhibit periods of severe imbalances. In favorable conditions, assimilation may exceed maintenance several orders of magnitude, whereas in unfavorable conditions, assimilation may, in fact, cease completely, leaving maintenance to be satisfied from metabolic reserves. Supply-side species, accordingly, build large reserves and lead a more sessile lifestyle to conserve energy and survive periods of starvation. The defining property of demand-side species is that energy assimilation and maintenance are closely balanced. Instead of waiting for food, demand-side species search or pursue it actively, thus leading a much more motile and energetically intensive lifestyle. An example of a demand-side organism is tuna fish, famously dubbed “energy speculators” due to their high energy turnover despite inhabiting poor pelagic environments (Korsmeyer et al. 1996). Table 1 Comparison between supply and demand systems (modified from [Kooijman 2014] and [Lika et al. 2014]) Supply Demand Eat what is available Eat what is needed Can handle large range of intake Can handle small range of intake Reserve density varies wildly Reserve density varies little Low peak metabolic rate High peak metabolic rate Large intraspecies body size range Small intraspecies body size range Energetic birth control Behavioral birth control No acceleration of ageing Acceleration of ageing Ancestral Derived Has demand components (maintenance) Has supply components (some food must be available) Supply Demand Eat what is available Eat what is needed Can handle large range of intake Can handle small range of intake Reserve density varies wildly Reserve density varies little Low peak metabolic rate High peak metabolic rate Large intraspecies body size range Small intraspecies body size range Energetic birth control Behavioral birth control No acceleration of ageing Acceleration of ageing Ancestral Derived Has demand components (maintenance) Has supply components (some food must be available) Ectotherms are generally found on the supply-side of the spectrum, while the physiology of endotherms is more demand-driven (Sousa et al. 2010). See text for further details. View Large Table 1 Comparison between supply and demand systems (modified from [Kooijman 2014] and [Lika et al. 2014]) Supply Demand Eat what is available Eat what is needed Can handle large range of intake Can handle small range of intake Reserve density varies wildly Reserve density varies little Low peak metabolic rate High peak metabolic rate Large intraspecies body size range Small intraspecies body size range Energetic birth control Behavioral birth control No acceleration of ageing Acceleration of ageing Ancestral Derived Has demand components (maintenance) Has supply components (some food must be available) Supply Demand Eat what is available Eat what is needed Can handle large range of intake Can handle small range of intake Reserve density varies wildly Reserve density varies little Low peak metabolic rate High peak metabolic rate Large intraspecies body size range Small intraspecies body size range Energetic birth control Behavioral birth control No acceleration of ageing Acceleration of ageing Ancestral Derived Has demand components (maintenance) Has supply components (some food must be available) Ectotherms are generally found on the supply-side of the spectrum, while the physiology of endotherms is more demand-driven (Sousa et al. 2010). See text for further details. View Large Our interest in the supply-demand spectrum is that species found on the “demand” side have, to a large extent, homeostatic control to buffer their metabolism from changes in the environment. This control is multifaceted and manifests itself, for example, as a diminished variability in energy intake and expenditure, as well as in body temperature over time. Endothermic vertebrates are typically demand-side organisms. In contrast, ectothermic vertebrates are generally situated more on the supply side of the spectrum and therefore relatively unconstrained by their energy budgets (Pough 1980; Sousa et al. 2010). We argue that it is this freedom from energy constraints that does not easily select for the evolution of parental provisioning in ectothermic vertebrates. As we will show, 3 aspects are particularly important in this respect: small offspring size, the production of offspring that are independent from birth or hatching, and relatively low energy requirements. All three appear to limit the ability of ectothermic parents to increase their reproductive output by feeding or securing food for their offspring and hence hinder the evolution of parental provisioning. When size matters—effect of juvenile size on the evolution of parental care The evolution of parental care should benefit offspring (and hence increase the fitness of the parents), but studies that have theoretically explored the ways by which offspring benefit from care are rare. One such study showed that the ability of parents to control the rate of development of their offspring alone can favor the evolution of parental care and/or alter the life-history conditions under which care will be selected for (Klug and Bonsall 2014). The rate of development, or more precisely the rate of growth, seems particularly pertinent in endotherms, chiefly because heat loss places a limit on how small offspring can be (Figure 1) (Ricklefs 1979). In birds, survival is directly related to growth rate because faster growing chicks pass through the vulnerable developmental stage (where they could succumb to heat loss) more quickly (Ricklefs 1979). Given the thermal constraints on offspring size, in theory parents could follow 2 strategies to alleviate such constraints: 1) migrate seasonally to a much warmer habitat for reproduction to reduce the heat loss experienced by their offspring, or 2) produce larger offspring. Both strategies imply considerable energetic investments. Perhaps the best example of strategy 1) is found in tuna from the subgenus Thunnus in which the evolution of regional heterothermy (as opposed to homeothermy) allowed an expansion from subtropical or tropical to highly productive temperate ocean waters (Graham and Dickson 2004). Here, regional heterothermy refers to an ability to maintain different temperatures in different regions of the body. Despite their successful range expansion, tunas are forced to annually migrate over huge distances to return to warmer spawning grounds simply because the juveniles would not be able to survive in colder water because of their relatively small size (eggs are about 1 mm in diameter and larvae hatch about 3 mm) (Jusup et al. 2011). When migrating to warmer environments is unfeasible, strategy 2) calls for an increase in allocation of energy per offspring to improve their chances of survival. Such an increased allocation results in larger size at birth or hatching, but necessarily decreases the number of offspring unless energy available is unlimited (Ricklefs 1979). Postnatal parental care in the form of parental provisioning becomes obligatory to improve the offspring’s chances of survival, particularly when larger size at birth or hatching is still insufficient for offspring to immediately attain full control of their body temperature. Hence, feeding of young by parents is obligatory in all mammals and altricial birds (altricial development is probably ancestral in birds (see Ligon and Burt 2004)). Figure 1 View largeDownload slide What does it take to be an endotherm? Heat production is spatially distributed and scales with the cube of some characteristic length, L. By contrast, heat is lost through the surface separating the organism from the environment, which scales with the square of L. This difference in scaling suggests that heat loss dominates when body size is small because of an unfavorable surface-area-to-volume ratio. However, the details are more subtle. Using the “ellipsoid” model of Porter and Kearney (2009), we distinguish a heat-producing metabolic core and, if present, insulation (e.g., fur) or other heat-conserving structures (e.g., rete mirabile). Both core and insulation transport heat via conduction. A convective boundary layer forms at the surface between insulation and the surrounding fluid. If the surrounding fluid is air, radiation may act as a considerable heat loss mechanism in parallel with convection. A temperature gradient, Tc-Tf, between the core and the fluid may form due to heat production, Q, and conductive, convective, and radiative thermal resistances, Rc, Ri, and Rconv+rad: Tc-Tf = Q × (Rc + Ri + Rconv+rad). Heat production is an extensive process and scales with L3, while the thermal resistances scale with L-1 or L-2. Consequently, the temperature gradient between core and fluid tends to zero in the limit of small organismal size. In other words, when body size becomes too small, it is impossible to maintain a body temperature above ambient. Figure 1 View largeDownload slide What does it take to be an endotherm? Heat production is spatially distributed and scales with the cube of some characteristic length, L. By contrast, heat is lost through the surface separating the organism from the environment, which scales with the square of L. This difference in scaling suggests that heat loss dominates when body size is small because of an unfavorable surface-area-to-volume ratio. However, the details are more subtle. Using the “ellipsoid” model of Porter and Kearney (2009), we distinguish a heat-producing metabolic core and, if present, insulation (e.g., fur) or other heat-conserving structures (e.g., rete mirabile). Both core and insulation transport heat via conduction. A convective boundary layer forms at the surface between insulation and the surrounding fluid. If the surrounding fluid is air, radiation may act as a considerable heat loss mechanism in parallel with convection. A temperature gradient, Tc-Tf, between the core and the fluid may form due to heat production, Q, and conductive, convective, and radiative thermal resistances, Rc, Ri, and Rconv+rad: Tc-Tf = Q × (Rc + Ri + Rconv+rad). Heat production is an extensive process and scales with L3, while the thermal resistances scale with L-1 or L-2. Consequently, the temperature gradient between core and fluid tends to zero in the limit of small organismal size. In other words, when body size becomes too small, it is impossible to maintain a body temperature above ambient. Ectothermy removes the described energetic constraints on offspring size, and as a result, offspring of ectothermic vertebrates (and even heterotherms such as temperate tuna) are often much smaller than offspring of endothermic vertebrates relative to the size of the adults (Case 1978). Compared with mammals and birds, ectothermic vertebrates thus often have broods consisting of numerous small offspring that are independent from birth or hatching (Shine 2005). Because offspring are independent, meaning they do not die without further parental investment, further investment by parents seems unnecessary. Producing offspring that are much smaller than parents potentially poses another constraint in the evolution of parental provisioning. Being much smaller than the parents may result in spatial segregation between juveniles and adults if the difference in body size results in different needs. For example, juveniles of most terrestrial amphibians are aquatic. Such spatial segregation can reduce the possibility of parents caring for their young. Not only does a significant size difference between juveniles and adults potentially affect where both live, and hence how much contact there is between adults and juveniles, it also limits the type of care parents can provide, even when parents and juveniles are not spatially segregated. Because of their much smaller size, juveniles may need to specialize on different types of food to adults. In crocodiles and snakes, adults and juveniles often specialize on different prey simply because juveniles are too small to swallow the larger prey eaten by adults (Webb et al. 1982; Greene 1997). Many fish fry feed on microcrustaceans or tiny benthic invertebrates that are too small for the adults to feed on efficiently (Perrone and Zaret 1979). Such difference in food-size makes it difficult for adults to collect and provide food in a mass compact enough to allow efficient feeding of fry. Having young that are much smaller than adults both reduces the ability of parents to provide any care and, when care does occur, often restricts parental care to other forms of care such as guarding offspring. The evolutionary benefits of parental provisioning Despite their independence from hatching or birth, the feeding of ectothermic young can, in theory, provide fitness benefits to ectothermic offspring. To explore how the availability of food affects growth and reproduction in ectotherms, we apply the framework of Dynamic Energy Budget (DEB) theory (Nisbet et al. 2000; Kooijman 2010) to loggerhead turtles (Caretta caretta), a species we have a particularly detailed model for (Marn et al. 2017a, 2017b). DEB models describe how individuals acquire and divide energy available to them between maintenance, growth, and reproduction. By coupling DEB models with population dynamic models (Nisbet et al. 2000), we can relate population growth rate to growth, maturation, and reproduction of individuals (Jusup et al. 2017b), which in turn are affected by environmental conditions, primarily food availability and temperature (Jusup et al. 2017a). An increased food intake yields fitness benefits in loggerhead turtles, both in terms of survival and fecundity (Figure 2). The same framework also predicts faster growth rates, younger age at sexual maturity, and higher fecundity when food is plentiful. As a result, provided that survival remains unaffected, population growth will increase as illustrated in our loggerhead turtle example (Figure 2). Figure 2 View largeDownload slide Being well-fed increases fitness. Using a full life-cycle model for loggerhead turtles (Caretta caretta), built on the principles of physiological energetics and calibrated with data from all life stages (Marn et al. 2017a, 2017b), we show the extent to which environment–feeding and temperature–affects growth and reproduction of individuals, and by extension population growth rate. (A) Growth of loggerhead turtles at 3 constant functional responses f (i.e., the percentages of the maximum ingestion rate for a given turtle size). When f = 0.81, the model nicely fits the available data on loggerhead turtles (Marn et al. 2017a, 2017b). Final turtle size, measured as straight carapace (SC) length, is directly proportional to the realized functional response (i.e., food intake). Increased feeding does not cause large differences in size during the early developmental stages, but does significantly affect size at sexual maturity (around 10 years of age). Survival indicated by stars. (B) Biannual fecundity of loggerhead turtles plotted in relation to functional response. Less fed individuals not only reach sexual maturity at a much later age (almost 20 years vs. 10 years of age at f = 0.7 vs. f = 0.9, respectively), but also produce considerably fewer eggs. The reason for using biannual fecundity is that the remigration period of loggerhead turtles often takes 2 years. (C) The population growth rate, λ, of loggerhead turtles as a function of environmental conditions. More food and higher temperatures both lead to higher energy acquisition and consumption rates, thus increasing the population growth rate provided survival (the green curve in panel A (Monk et al. 2011)) remains relatively constant. If λ < 1 (λ > 1) population size decreases (increases). The black line indicates the average conditions (f = 0.81, T = 21.8 °C) experienced by wild loggerhead turtles. Interestingly, under these conditions, the population growth rate is comfortably above 1. (D) Population growth rate plotted against functional response at 3 different temperatures. We include T = 21.8 °C, the temperature at which the model nicely fits the available data. When food is sufficient, the population growth rate is a slightly concave function of the functional response. However, as food gets scarcer, the curvature of this concave function increases because there is less and less energy to mature and reproduce, which exerts a disproportionate negative impact on age at sexual maturity and fecundity. Modeling details can be found in the Modeling methods section. Figure 2 View largeDownload slide Being well-fed increases fitness. Using a full life-cycle model for loggerhead turtles (Caretta caretta), built on the principles of physiological energetics and calibrated with data from all life stages (Marn et al. 2017a, 2017b), we show the extent to which environment–feeding and temperature–affects growth and reproduction of individuals, and by extension population growth rate. (A) Growth of loggerhead turtles at 3 constant functional responses f (i.e., the percentages of the maximum ingestion rate for a given turtle size). When f = 0.81, the model nicely fits the available data on loggerhead turtles (Marn et al. 2017a, 2017b). Final turtle size, measured as straight carapace (SC) length, is directly proportional to the realized functional response (i.e., food intake). Increased feeding does not cause large differences in size during the early developmental stages, but does significantly affect size at sexual maturity (around 10 years of age). Survival indicated by stars. (B) Biannual fecundity of loggerhead turtles plotted in relation to functional response. Less fed individuals not only reach sexual maturity at a much later age (almost 20 years vs. 10 years of age at f = 0.7 vs. f = 0.9, respectively), but also produce considerably fewer eggs. The reason for using biannual fecundity is that the remigration period of loggerhead turtles often takes 2 years. (C) The population growth rate, λ, of loggerhead turtles as a function of environmental conditions. More food and higher temperatures both lead to higher energy acquisition and consumption rates, thus increasing the population growth rate provided survival (the green curve in panel A (Monk et al. 2011)) remains relatively constant. If λ < 1 (λ > 1) population size decreases (increases). The black line indicates the average conditions (f = 0.81, T = 21.8 °C) experienced by wild loggerhead turtles. Interestingly, under these conditions, the population growth rate is comfortably above 1. (D) Population growth rate plotted against functional response at 3 different temperatures. We include T = 21.8 °C, the temperature at which the model nicely fits the available data. When food is sufficient, the population growth rate is a slightly concave function of the functional response. However, as food gets scarcer, the curvature of this concave function increases because there is less and less energy to mature and reproduce, which exerts a disproportionate negative impact on age at sexual maturity and fecundity. Modeling details can be found in the Modeling methods section. The implications of the above example for parental provisioning are 2-fold. First, offspring fed by parents grow and mature faster to the point at which independent feeding is possible. This means that, at a minimum, the period during which vulnerability to predation, diseases, and other causes of natural mortality is highest is shortened by parental feeding. An additional benefit, however, is that parent-fed offspring are in a considerably better condition (in the sense of, e.g., higher Fulton’s condition factor (Nash et al. 2006)) at the moment of gaining independence, which is likely to translate into prolonged feeding success afterwards, and ultimately a much higher population growth as illustrated in our example. In reptiles, feeding of juveniles potentially increases the juveniles’ growth and survival because an increase in food intake at an early stage in life significantly increases growth rate (Madsen and Shine 2000), and feeding history directly affects fitness because clutch size is often related to a female’s body size (Doughty and Shine 1998; Shine 2005). Hence, as in the case of loggerhead turtles, we expect a female reptile that has been well fed when young to grow to be bigger and have larger clutches. Yet feeding of juveniles does not occur (Somma 2003), even though such feeding would benefit the offspring. The consequences of having young that are independent with low energy requirements Energy requirements of mammals and birds, both altricial and precocial, are orders of magnitude higher than the energy requirements of similar-sized ectotherms (Pough 1980; Shine 2005). To maintain such high rates of metabolism, mammals and birds need high and relatively constant rates of food intake, pointing to a lack of metabolic flexibility in endotherms (Lika et al. 2014). To illustrate how the different energy requirements of ectotherms and endotherms affect growth and reproduction, we again apply the DEB framework (Nisbet et al. 2000) to loggerhead turtles (C. caretta). Specifically, we make an endothermic version of our loggerhead turtle model by setting the body temperature to 37 °C thus upregulating all energy flows to values suitable for an endotherm. Our model shows that the lack of metabolic flexibility of endotherms is caused by 2 complementary effects. First, faster metabolism leads to considerably shorter time to reserve depletion during starvation (≈3.2× shorter in the example in Figure 3; see modeling methods). Second, if the endotherm is forced to spend prolonged time outside of its thermo-neutral zone, the supply stress (i.e., the ratio of maintenance costs to energy assimilation) increases, thus negatively affecting ultimate size, age at sexual maturation and fecundity, and consequently population growth rate (assuming no effect on survival) (Figure 3). This lack of metabolic flexibility is in sharp contrast to a similar-sized ectotherm, which can tolerate more uncertainty in food intake. Here it is essential to compare similar sized ectotherms and endotherms because since Kleiber’s work (Kleiber 1932), it has been widely accepted that body size is a key determinant of metabolic rates. Figure 3 View largeDownload slide Benefits, and perils, of being an endotherm. We use the full life-cycle model for loggerhead turtles in a hypothetical situation wherein all metabolic rates are upregulated as if loggerheads were endotherms (i.e., as if their average temperature were 37 °C instead of 21.8 °C). (A) Ectotherm (blue line; diamonds) versus endotherm (red lines; circles): comparison of growth curves. The upregulation increases all metabolic rates equally, resulting in much faster growth for the hypothetical endotherm. Provided this endotherm stays mostly within the bounds of its thermo-neutral zone (TNZ; the temperature range within which the rate of heat production is in equilibrium with the rate of heat loss to the external environment), its ultimate size is unaffected. However, moving outside of the TNZ results in a smaller ultimate size because of the extra energy required to maintain a stable body temperature (represented here by the loss of body size called the heating length, LT). (B) Ectotherm (blue line; diamonds) versus endotherm (red line; circles): comparison of fecundities. Under identical conditions, endotherms not only grow faster, but also mature earlier and produce either more offspring of the same size (as shown here when LT equals 0) or an equal number of larger offspring (not shown). The cost of staying outside of the TNZ may negatively affect both maturation age and fecundity of the endotherm, as reflected in the reduction in size (measured as heating length LT). (C) Ectotherm (blue line; diamonds) versus endotherm (red lines; circles): comparison of energetic restrictions. The upregulation of metabolic rates itself leaves the supply stress (i.e., the ratio of maintenance costs to energy assimilation) unaffected, but the endotherm is still much more vulnerable to starvation as indicated by a considerably shorter time to reserve depletion (see main text). The situation worsens if the endotherm spends much time out of the TNZ. (D) Ectotherm versus endotherm: comparison of the population growth rates. Endotherms benefit greatly from faster growth (which also improves survival), earlier maturation, and higher fecundity assuming the same offspring size. Even spending a lot of time outside of the TNZ does not erase all the benefits of endothermy. Consequently, if having larger offspring is advantageous due to, for example, a better chance of survival, an evolutionary tradeoff between the number and size of offspring is likely. See Modeling methods for details. Figure 3 View largeDownload slide Benefits, and perils, of being an endotherm. We use the full life-cycle model for loggerhead turtles in a hypothetical situation wherein all metabolic rates are upregulated as if loggerheads were endotherms (i.e., as if their average temperature were 37 °C instead of 21.8 °C). (A) Ectotherm (blue line; diamonds) versus endotherm (red lines; circles): comparison of growth curves. The upregulation increases all metabolic rates equally, resulting in much faster growth for the hypothetical endotherm. Provided this endotherm stays mostly within the bounds of its thermo-neutral zone (TNZ; the temperature range within which the rate of heat production is in equilibrium with the rate of heat loss to the external environment), its ultimate size is unaffected. However, moving outside of the TNZ results in a smaller ultimate size because of the extra energy required to maintain a stable body temperature (represented here by the loss of body size called the heating length, LT). (B) Ectotherm (blue line; diamonds) versus endotherm (red line; circles): comparison of fecundities. Under identical conditions, endotherms not only grow faster, but also mature earlier and produce either more offspring of the same size (as shown here when LT equals 0) or an equal number of larger offspring (not shown). The cost of staying outside of the TNZ may negatively affect both maturation age and fecundity of the endotherm, as reflected in the reduction in size (measured as heating length LT). (C) Ectotherm (blue line; diamonds) versus endotherm (red lines; circles): comparison of energetic restrictions. The upregulation of metabolic rates itself leaves the supply stress (i.e., the ratio of maintenance costs to energy assimilation) unaffected, but the endotherm is still much more vulnerable to starvation as indicated by a considerably shorter time to reserve depletion (see main text). The situation worsens if the endotherm spends much time out of the TNZ. (D) Ectotherm versus endotherm: comparison of the population growth rates. Endotherms benefit greatly from faster growth (which also improves survival), earlier maturation, and higher fecundity assuming the same offspring size. Even spending a lot of time outside of the TNZ does not erase all the benefits of endothermy. Consequently, if having larger offspring is advantageous due to, for example, a better chance of survival, an evolutionary tradeoff between the number and size of offspring is likely. See Modeling methods for details. A critical link between the lack of metabolic flexibility in endotherms and parental provisioning is that parent-fed endothermic offspring mature to feed independently at a higher condition than nonfed offspring would. This higher condition gives more protection against unsuccessful feeding attempts that potentially plague inexperienced younglings that also have to satisfy strict energetic requirements. A likely additional benefit, as already mentioned, is that offspring with a better condition should enjoy prolonged feeding success after gaining independence, thus further increasing the chances of satisfying strict energetic requirements. Ultimately, parental provisioning helps secure that endothermic offspring enjoy benefits rather than face perils of being an endotherm as illustrated in our modeling example. All this is in sharp contrast to ectothermic young, whose lower energy requirements result in slow growth rates and the ability to forgo feeding altogether if circumstances become unfavorable. However, when feeding circumstances are favorable, an ectotherm will allocate less resources to reproduction than a similar sized endotherm (Figure 3). Ectothermy has thus led to an array of specific traits: young that are often numerous and small, independent from birth, and have low energy requirements. Thus, for the parents, young are cheap to produce (Figure 4). How then can an ectothermic parent best increase its reproductive output? When young are cheap, and resources for reproduction relatively limited, it seems unlikely that parents will benefit from expending extra resources on offspring after birth or hatching (Smith and Fretwell 1974) by either feeding them or by defending food. Even though providing food to ectothermic young may increase their growth rate, the costs to parents may not offset the benefits. Most ectotherms are capital breeders and thus store resources needed for reproduction prior to ovulation (Stearns 1992). While capital breeding is not unique to ectotherms, a key distinction between ectotherms and endotherms in this context is that due to the large energy requirements of offspring, increasing reproductive reserve by the same relative amount (say 20%), will result in dramatically different outputs. For a tuna fish for example, an additional allocation to reproduction of 20% would literally mean millions of eggs. In contrast, for a large bird that typically lays, say, a clutch of 4 eggs, a 20% increase in energy allocation will not yield an additional egg. As a result, a female ectotherm may be better off accumulating resources herself instead of investing in existing offspring, thereby increasing her reproductive output in the next reproductive season. The more resources a female has stored prior to ovulation, the greater the number of offspring she can produce (Bonnet et al. 1998; Doughty and Shine 1998). Hence, the only investment in the current young that ectothermic parents are selected for is to protect them from predators when predation pressure is high. Figure 4 View largeDownload slide Offspring are cheap for ectotherms. We hypothesized that energy investment into eggs differs between egg-laying vertebrates, with birds (endotherms) investing significantly more than reptiles, amphibians, and fish (ectotherms). We furthermore hypothesized that such investment would depend on species body size. (A) To test these hypotheses while controlling for the effects of body size, we performed an ANCOVA test on log-transformed data from Add-my-Pet database (Marques et al. 2018). The data indicate that egg energy content is indeed the highest in birds and depends on species body size in all groups except fish (e.g., anchovy are an order of magnitude smaller than tuna, but a typical egg diameter for both species is roughly 1 mm). Note that due to the log-transformation, units do not pertain to the displayed numerical values, but rather to their exponentials. (B) Population mean egg energy content differs significantly between the 4 examined groups of vertebrates. Birds on average invest exp(12.7)/exp(10.8)≈6.7 times more energy into their eggs than reptiles, exp(12.7)/exp(5.96)≈846 times more than amphibians, and exp(12.7)/exp(1.62)≈64,900 times more than fish (exponentials are due to log-transformed data). (C) The increase of egg energy content with body size is significantly more rapid in birds than amphibians and fish, but not reptiles. In fish, in fact, this increase is statistically indistinguishable from zero. Shaded zones represent overlap of the confidence intervals. (D) Additionally, comparing egg energy contents at body size of exp(0) = 1 cm, birds still invest significantly more energy than reptiles, amphibians, and fish. Figure 4 View largeDownload slide Offspring are cheap for ectotherms. We hypothesized that energy investment into eggs differs between egg-laying vertebrates, with birds (endotherms) investing significantly more than reptiles, amphibians, and fish (ectotherms). We furthermore hypothesized that such investment would depend on species body size. (A) To test these hypotheses while controlling for the effects of body size, we performed an ANCOVA test on log-transformed data from Add-my-Pet database (Marques et al. 2018). The data indicate that egg energy content is indeed the highest in birds and depends on species body size in all groups except fish (e.g., anchovy are an order of magnitude smaller than tuna, but a typical egg diameter for both species is roughly 1 mm). Note that due to the log-transformation, units do not pertain to the displayed numerical values, but rather to their exponentials. (B) Population mean egg energy content differs significantly between the 4 examined groups of vertebrates. Birds on average invest exp(12.7)/exp(10.8)≈6.7 times more energy into their eggs than reptiles, exp(12.7)/exp(5.96)≈846 times more than amphibians, and exp(12.7)/exp(1.62)≈64,900 times more than fish (exponentials are due to log-transformed data). (C) The increase of egg energy content with body size is significantly more rapid in birds than amphibians and fish, but not reptiles. In fish, in fact, this increase is statistically indistinguishable from zero. Shaded zones represent overlap of the confidence intervals. (D) Additionally, comparing egg energy contents at body size of exp(0) = 1 cm, birds still invest significantly more energy than reptiles, amphibians, and fish. Conclusion: ectothermy makes vertebrates independent Herein, we asked why ectothermic parents rarely provide food for their young. We have argued that ecothermic physiology allows offspring to be small, to be independent from birth or hatching, and to have low energy requirements. The size difference between adults and juveniles in particular, made possible because of a lack of thermal constraints on ectothermic young, affects the options available to parents to provide for their offspring due to their different diets. Yet young of marsupials are also orders of magnitude smaller than parents and are fed by the mother with milk, a different diet to the diet of adult marsupials. The fundamental difference between marsupials and ectotherms is that the young of the former are completely dependent on the mother for thermoregulation, food and even their immunological development (Deane and Cooper 1988). Without postnatal parental provisioning the young would simply die, whereas ectothermic young, while growing slower without parental provisioning, can look after themselves. Similarly, in birds, parents too can feed young even when the young require a different diet. Many seabird species eat prey, digest it, and then regurgitate for their offspring (Anderson and Ricklefs 1992) (incidentally, many seabirds continue to feed their young many months after fledging because young birds require extensive experience before they can catch fish successfully (Mendez et al. 2017)). There seems no reason why ectotherms could not have evolved similar mechanisms if selection on postnatal feeding was strong enough. Are there any examples of where ectothermic vertebrates have been selected to provide for their offspring? There are 3 known examples of parental provisioning in ectothermic vertebrates. As we will illustrate it appears that extraordinary circumstances have selected for the feeding of young in all 3 cases. Some species of frog lay eggs in phytotelmata (aquatic microhabitats in leaf axils, flowers, tree holes, bamboo stumps, and nut capsules) (Duellman and Trueb 1986; Lannoo et al. 1987; Summers 1990; Caldwell 1993; Crump 1995; Kam et al. 1996). Because of the choice of habitat, frog parents need to provide their juveniles with food as they will otherwise starve (Tumulty et al. 2014). The female therefore regularly returns to the phytotelmata, in many species encouraged by the male (Tumulty et al. 2014), and deposits trophic eggs that are consumed by her offspring (Crump 1995; Chiu and Kam 2005; Chen et al. 2013; Tumulty et al. 2014). The mother of a caecilian amphibian needs to feed her altricial juveniles as they are unable to forage for themselves (Kupfer et al. 2006). Lastly, adults of some species of cichlids ingest food from the bottom of the lake and expel it in mid-water so that their fry can feed on it, whereas others have fry that feed from mucus produced externally by their parents (Clutton-Brock 1991; Jordan et al. 2013). Cichlid juveniles suffer a high predation risk and smaller individuals have a higher chance of being eaten. Growing fast reduces the length of the vulnerable stage and it may well be that this has selected for special mechanisms to feed juveniles, despite the large difference in size between adults and juveniles. The above described exceptions illustrate that if offspring are unlikely to survive without postnatal parental provisioning, then parents will be selected to provide such care. In most instances, however, the selective forces that have shaped the physiology of ectothermy has resulted in independent offspring. As such ectothermy does not allow an increase in the number of offspring produced by increasing the amount of food available to offspring. The scope for parents to increase the growth rates of current offspring does not offset the costs in terms of a reduction in resources available for future reproductive events. Instead, ectothermic vertebrates are well-adapted to boom and bust environments, where resource availability is erratic, resources are accumulated over extended periods and only used when the time is right. Their energetic efficiency and lack of thermal constraints makes ectothermic vertebrates invest in their own growth and maintenance instead of provisioning their young. As a result, the main form of parental care found in ectothermic vertebrates is the guarding of eggs or young. Only under very special circumstances, will parents be selected to provide parental care in the form of feeding of young. Modeling methods We used the standard Dynamic Energy Budget (DEB) model and adapted it to loggerhead turtles. Because the standard DEB model has been described in detail elsewhere (Sousa et al. 2008; Sousa et al. 2010) as well as its implications for loggerhead turtles (Marn et al. 2017a, 2017b), we only provide a short, self-contained overview. Schematically, the standard DEB model falls into a class of individual-based models that can be represented as dLda=G˙(i-state, e-state), where L is an organism’s size, a is its age, and Ġ = Ġ(i-state, e-state) is a growth function dependent on the organism’s state (i-state variables) and the state of the environment (e-state variables). In the case of the standard DEB model, i-state variables–in addition to L itself–include the energy in reserve, E, which does not require maintenance and can be used to power metabolic processes, and the cumulative investment into maturation, EH, which at threshold values EHb and EHp signifies transitions from embryo to juvenile stage and from juvenile to adult stage, respectively. Each i-state variable is accompanied by its own differential equation that describes the rate of change of this variable with age. Specifically, for the standard DEB model, we have: dEda=p˙A−p˙C, dLda=p˙G3L2[EG], and dEHda=p˙R, if EH<EHp, where [EG] is a model parameter interpreted as the volume-specific cost of growth (in J cm-3) and ṗ* = ṗ*(i-state, e-state) are energy flows (in J d-1). We describe energy flows below. Reserve energy is replenished via the assimilation flow, ṗA = {ṗAm}fL2, where the value of parameter {ṗAm}, called the surface-area-specific maximum assimilation rate, reflects the performance of the organism’s digestive system (in J d-1 cm-2). The dimensionless functional response, f, can be interpreted as the percentage of the maximum ingestion rate at which the organism of a given size feeds depending on the environmental food density. Energy is mobilized from reserve to power metabolic processes at a rate determined by the utilization (or catabolic) energy flow p˙C=[E]v˙[EG]L2+[p˙M]L3+{p˙T}L2[EG]+κ[E], where [E] = E/L3 is reserve density. Parameter v˙ is called energy conductance (in cm d-1) and together with the surface-area-specific maximum assimilation rate, {ṗAm}, defines the maximum reserve density, [Em] = {ṗAm}/ v˙ ⁠. In addition, parameters [ṗM] and {ṗT} reflect volume-specific (in J d-1 cm-3) and surface-area-specific (in J d-1 cm-2) maintenance costs, respectively. Finally, dimensionless parameter κ is the fraction of the utilization energy flow invested into somatic maintenance and growth. If fraction κ of utilization flow ṗC carries enough energy to satisfy somatic maintenance needs (i.e., κṗC>[ṗM]L3+{ṗT}L2) the organism grows at rate p˙G=[EG]κv˙[E]L2−[p˙M]L3−{p˙T}L2[EG]+κ[E]. The remaining fraction of the utilization flow, (1-κ)ṗC, is in embryo and juvenile stages invested into maturation such that ṗR = (1-κ)ṗC- k˙JEH ⁠, where k˙J is the maturity maintenance rate coefficient (in d-1). Maturity stops increasing when it reaches maturity at puberty (i.e., EH = EHp) and the organism enters the adult stage. In this stage, reproduction flow ṗR is used for egg production, which leads to an expression for the fecundity rate as a function of i-state variables, Ḟ = κRṗR/E0, where κR is a dimensionless efficiency with which reserve energy is converted into eggs and E0 is the energy content of a single egg. See Jusup et al. (2017a) and Kooijman (2009) for how one calculates E0. To summarize, the standard DEB model provides access to functions L = L(a) and Ḟ = Ḟ(L) via a system of differential equations described above. When survival as a function of length, S = S(L), is also known, one can estimate population growth rate. The population growth rate, ṙ, can be obtained using the Euler-Lotka equation: 1=∫daexp(−r˙a)S[L(a)]F˙[L(a)]. It is crucial to recognize that size rather than age plays the key ecological role (De Roos and Persson 2002; Persson and De Roos 2013), and that the ability of the standard DEB model to adjust growth curve L = L(a) to the prevailing environmental conditions results in a population growth rate responsive to these conditions. Such responsiveness would be impossible if function L = L(a) were obtained in a traditional manner, i.e., by means of statistical fitting to growth data. We express the results of population growth rate calculations in terms of λ = exp(ṙ×1 year), meaning that λ>1 (λ<1) indicates a growing (declining) population. When addressing evolutionary questions, growth rate ṙ1 of the native population, N1, is compared with growth rate (ṙ2) of a mutant invader, N2, using replicator dynamics such that the rate of change in the number of natives is dN1/dt = ṙ1N1. Similarly, for the number of mutants we use dN2/dt = ṙ2N2. If we define a fraction of mutants in the whole population, N = N1+N2, by xi = Ni/N, we obtain dxi/dt = (ṙi-<ṙ>)xi, where <ṙ> = ṙ1x1+ ṙ2x2 is the average population growth rate. This implies that mutants increase at the expense of natives if ṙ1 < ṙ2. Generally, growth rates ṙi are subject to density effects, but a concrete form of density dependence differs according to the ecological circumstances of each species. Without considering density effects, inequality ṙ1 < ṙ2 is the necessary (but not sufficient) condition for mutants to successfully invade. Temperature dependence To examine the effects of body temperature on growth, maturation, reproduction, and life-history traits, we assume that all energy flows depend on temperature according to the Arrhenius model (Clarke 2004; Jusup et al. 2017a). Specifically, if ṗ*(T0) is an energy flow known at temperature T0, then at temperature T, we have ṗ*(T) = C(T)ṗ*(T0), where the correction function is C(T) = exp(TA/T0-TA/T). Parameter TA (in K) is called the Arrhenius temperature. Supply stress If we denote the somatic maintenance flow as ṗS = [ṗM]L3+{ṗT}L2 and the maturity maintenance flow as ṗJ = k˙JEH ⁠, the supply stress can be defined as sS = ṗJṗS2/ṗA3. This quantity is useful for an immediate evaluation of feeding conditions. Under favorable conditions, reserve should not get depleted, that is, ṗA ⩾ ṗC. Due to the splitting of flow ṗC into fractions κṗC and (1-κ)ṗC, which respectively satisfy somatic and maturity maintenance needs before growth and maturation (or reproduction) can occur, we have ṗS ⩽ κṗA and ṗJ ⩽ (1-κ)ṗA, and finally sS ⩽ (1-κ)κ2. This expression indicates that when the ratio of sS to (1-κ)κ2 approaches unity, maintenance needs are about to exceed energy supply needed for normal metabolic functioning in the sense that the organism’s survival becomes dependent on drawing energy from reserve, which is unsustainable in the long-term. Hence the name supply stress. Time to reserve depletion When fraction κ of the utilization flow is unable to satisfy somatic maintenance needs (κṗC < ṗS), the organism is in starvation mode wherein the expression for ṗC given above breaks down. We therefore assume that from the moment when κṗC = ṗS, that is, when energy in reserve is Ec = ṗSL/(κ v˙ ⁠), reserve gets mobilized at rate ṗS, which is just enough to satisfy maintenance needs and stay alive. If in starvation mode no food is available (f = 0), energy Ec will be depleted from reserve in time td given by td = Ec/ṗS = L/(κ v˙ ⁠). Larger individuals are thus more resilient to starvation. More importantly, the denominator of this expression is a function of body temperature via the energy conductance which has the dimension of time-1. Endotherms will generally have a much higher energy conductance than ectotherms due to a higher body temperature, implying a considerably shorter time until reserve depletion for an endotherm compared with a similarly sized ectotherm. Parameter values We implemented the standard DEB model adapted to loggerhead turtles (C. caretta), which nicely fits the data from all life stages of this ectothermic animal–from an egg to an adult individual and its eggs (Marn et al. 2017a, 2017b). The parameter values are: Ehb = 3.81∙104 J, Ehp = 8.73∙107 J, [EG] = 7847 J cm-3, {ṗAm} = 906.1 J d-1 cm-2, v˙ = 0.0708 cm d-1, [ṗM] = 13.2 J d-1 cm-3, {ṗT} = 0 J d-1 cm-2, κ = 0.648, k˙J = 0.002 d-1, κR = 0.95, and TA = 7000 K. FUNDING M.B. and M.T. are supported by the Australian Research Council (ARC). MJ was supported by the Research Grant Program of Inamori Foundation. Acknowledgements Special thanks to Rick Shine for sorting out some of our woolly thinking and to Nina Marn for technical help. We also thank the anonymous reviewers for constructive comments on an earlier version of the manuscript. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Beekman et al. (2019). REFERENCES Anderson DJ , Ricklefs RE . 1992 . Brood size and food provisioning in masked and blue-footed boobies (Sula spp.) . Ecology . 73 : 1363 – 1374 . Google Scholar Crossref Search ADS Balshine S . 2012 . Patterns of parental care in vertebrates. In: Royle NJ , Smiseth PT , Kölliker M , editors. The evolution of parental care . Oxford : Oxford University Press . p. 62 – 80 . Beekman M , Thompson M , Jusup M . 2019 . Data from: thermodynamic constraints and the evolution of parental provisioning in vertebrates . Dryad Digital Repository . http://dx.doi.org/10.5061/dryad.cb690j5 Bonnet X , Bradshaw D , Shine R . 1998 . Capital versus income breeding: an ectothermic perspective . Oikos . 83 : 333 – 342 . Google Scholar Crossref Search ADS Caldwell JP . 1993 . Brazil nut fruit capsules as phytotelmata: interactions among anuran and insect larvae . Can J Zool . 71 : 1193 – 1201 . Google Scholar Crossref Search ADS Case TJ . 1978 . On the evolution and adaptive significance of postnatal growth rates in the terrestrial vertebrates . Q Rev Biol . 53 : 243 – 282 . Google Scholar Crossref Search ADS PubMed Chen W , Tang ZH , Fan XG , Wang Y , Pike DA . 2013 . Maternal investment increases with altitude in a frog on the Tibetan Plateau . J Evol Biol . 26 : 2710 – 2715 . Google Scholar Crossref Search ADS PubMed Chiu CT , Kam YC . 2005 . Growth of oophagous tadpoles (Chirixalus eiffingeri: Rhacophoridae) after nest displacement: implications for maternal care and nest homing . Behaviour . 143 : 123 – 139 . Clarke A . 2004 . Is there a universal temperature dependence of metabolism? Funct Ecol . 18 : 252 – 256 . Google Scholar Crossref Search ADS Clutton-Brock TH . 1991 . The evolution of parental care . Princeton : Princeton University Press . Crump ML . 1995 . Parental care. In: Heatwole H , Sull van BK , editors. Amphibian biology, volume 2: Social behaviour chipping Norton . Chipping Norton, Australia: Surrey Beatty & Sons . p. 518 – 567 . De Roos AM , Persson L . 2002 . Size-dependent life-history traits promote catastrophic collapses of top predators . Proc Natl Acad Sci USA . 99 : 12907 – 12912 . Google Scholar Crossref Search ADS PubMed Deane EM , Cooper DW . 1988 . Immunological development of pouch young marsupials. In: Tyndale-Biscoe CH , Janssens PA , editors. The developing marsupial: models for biomedical research Berlin . Heidelberg : Springer Berlin Heidelberg . p. 190 – 199 . Doughty P , Shine R . 1998 . Reproductive energy allocation and long-term energy stores in a viviparous lizard (Eulamprus tympanum) . Ecology . 79 : 1073 – 1083 . Google Scholar Crossref Search ADS Duellman WE , Trueb L . 1986 . Biology of amphibians . New York : McGraw-Hill . Farmer CG . 2003 . Reproduction: the adaptive significance of endothermy . Am Nat . 162 : 826 – 840 . Google Scholar Crossref Search ADS PubMed Ferrara CR , Vogt RC , Sousa-Lima RS , Tardio BMR , Bernardes VCD . 2014 . Sound communication and social behavior in an Amazonian river turtle (Podocnemis expansa) . Herpetologica . 70 : 149 – 156 . Google Scholar Crossref Search ADS Graham JB , Dickson KA . 2004 . Tuna comparative physiology . J Exp Biol . 207 : 4015 – 4024 . Google Scholar Crossref Search ADS PubMed Greene HW . 1997 . Snakes . The evolution of mystery in nature . Berkeley : University of California Press . Greene HW , May PG , Hardy DL , Sciturro JM , Farrell TM . 2002 . Parental behavior by vipers. In: Schuett GW , Höggren M , Douglas ME , Greene HW , editors. Biology of the vipers . Eagle Mountain : Eagle Mountain Publ . p. 179 – 205 . Hedges SB , Poling LL . 1999 . A molecular phylogeny of reptiles . Science . 283 : 998 – 1001 . Google Scholar Crossref Search ADS PubMed Jordan LA , Herbert-Read JE , Ward AJW . 2013 . Rising costs of care make spiny chromis discerning parents . Behav Ecol & Sociobiol . 67 : 449 – 455 . Google Scholar Crossref Search ADS Jusup M , Klanjscek T , Matsuda H , Kooijman SA . 2011 . A full lifecycle bioenergetic model for bluefin tuna . PLoS One . 6 : e21903 . Google Scholar Crossref Search ADS PubMed Jusup M , Klanjšček T , Matsuda H . 2014 . Simple measurements reveal the feeding history, the onset of reproduction, and energy conversion efficiencies in captive bluefin tuna . J Sea Res . 94 : 144 – 155 . Google Scholar Crossref Search ADS Jusup M , Sousa T , Domingos T , Labinac V , Marn N , Wang Z , Klanjšček T . 2017a . Physics of metabolic organization . Phys Life Rev . 20 : 1 – 39 . Google Scholar Crossref Search ADS Jusup M , Sousa T , Domingos T , Labinac V , Marn N , Wang Z , Klanjšček T . 2017b . The universality and the future prospects of physiological energetics: Reply to comments on “Physics of metabolic organization” . Phys Life Rev . 20 : 78 – 84 . Google Scholar Crossref Search ADS Kam YC , Chuang ZS , Yen CF . 1996 . Reproduction, oviposition-site selection and larval oophagy of an aboreal nester, Chirixalus eiffingeri (Rhacophoridae), from Taiwan . J Herpetol . 30 : 52 – 59 . Google Scholar Crossref Search ADS Kleiber M . 1932 . Body size and metabolism . Hilgardia . 6 : 315 – 351 . Google Scholar Crossref Search ADS Klug H , Bonsall MB . 2014 . What are the benefits of parental care? The importance of parental effects on developmental rate . Ecol Evol . 4 : 2330 – 2351 . Google Scholar Crossref Search ADS PubMed Kooijman SA . 2009 . What the egg can tell about its hen: embryonic development on the basis of dynamic energy budgets . J Math Biol . 58 : 377 – 394 . Google Scholar Crossref Search ADS PubMed Kooijman SALM . 2010 . Dynamic energy budget theory for metabolic organisation . Cambridge (UK) : Cambridge University Press . Kooijman SALM . 2014 . Energy budgets. In: Hastings A , Gross L , editors. Sourcebook in Theoretical Ecology . Oakland: University of California Press . p. 249 – 258 . Korsmeyer KE , Dewar H , Lai NC , Graham JB . 1996 . The aerobic capacity of tunas: adaptation for multiple metabolic demands . Comp Biochem Physiol . 113 : 17 – 24 . Google Scholar Crossref Search ADS Kupfer A , Müller H , Antoniazzi MM , Jared C , Greven H , Nussbaum RA , Wilkinson M . 2006 . Parental investment by skin feeding in a caecilian amphibian . Nature . 440 : 926 – 929 . Google Scholar Crossref Search ADS PubMed Lannoo JM , Townsend DS , Wassersug RJ . 1987 . Larval life in the leaves: arboreal tadpole types, with special attention to the morphology, ecology, and behavior of the oophagous Osteopilus brunneus (Hylidae) larva . Fieldiana Zool . 38 : 1 – 31 . Ligon JD , Burt B . 2004 . Evolutionary origins. In: Koenig WD , Dickinson JL , editors. Ecology and evolution of cooperative breeding in birds . Cambridge : Cambridge University Press . p. 5 – 34 . Lika K , Augustine S , Pecquerie L , Kooijman SA . 2014 . The bijection from data to parameter space with the standard DEB model quantifies the supply-demand spectrum . J Theor Biol . 354 : 35 – 47 . Google Scholar Crossref Search ADS PubMed Madsen T , Shine R . 2000 . Silver spoons and snake sizes: prey availability early in life influences long-term growth rates of free-ranging pythons . J Anim Ecol . 69 : 952 – 958 . Google Scholar Crossref Search ADS Marn N , Jusup M , Legović T , Kooijman SALM , Klanjšček T . 2017a . Environmental effects on growth, reproduction, and life-history traits of loggerhead turtles . Ecol Model . 360 : 163 – 178 . Google Scholar Crossref Search ADS Marn N , Kooijman SALM , Jusup M , Legović T , Klanjšček T . 2017b . Inferring physiological energetics of loggerhead turtle (Caretta caretta) from existing data using a general metabolic theory . Mar Environ Res . 126 : 14 – 25 . Google Scholar Crossref Search ADS Marques GM , Augustine S , Lika K , Pecquerie L , Domingos T , Kooijman SALM . 2018 . The AmP project: comparing species on the basis of dynamic energy budget parameters . PLoS Comput Biol . 14 : e1006100 . Google Scholar Crossref Search ADS PubMed Mendez L , Prudor A , Weimerskirch H . 2017 . Ontogeny of foraging behaviour in juvenile red-footed boobies (Sula sula) . Sci Rep . 7 : 13886 . Google Scholar Crossref Search ADS PubMed Meyer A , Zardoya R . 2003 . Recent advances in the (molecular) phylogeny of vertebrates . Ann Rev Ecol Syst . 34 : 311 – 338 . Google Scholar Crossref Search ADS Monk MH , Berkson J , Rivalan P . 2011 . Estimating demographic parameters for loggerhead sea turtes using mark-recapture data and a multistate model . Pop Ecol . 53 : 165 – 174 . Google Scholar Crossref Search ADS Nash RD , Valencia AH , Geffen AJ . 2006 . The origin of Fulton’s condition factor - setting the record straight . Fisheries . 31 : 236 – 238 . Nisbet RM , Jusup M , Klanjscek T , Pecquerie L . 2012 . Integrating dynamic energy budget (DEB) theory with traditional bioenergetic models . J Exp Biol . 215 : 892 – 902 . Google Scholar Crossref Search ADS PubMed Nisbet RM , Muller EB , Lika K , Kooijman SALM . 2000 . From molecules to ecosystems through dynamic energy budget models . J Anim Ecol . 69 : 913 – 926 . Google Scholar Crossref Search ADS O’Connor D , Shine R . 2004 . Parental care protects against infanticide in the lizard Egernia saxatilis (Scincidae) . Anim Behav . 68 : 1361 – 1369 . Google Scholar Crossref Search ADS Perrone M , Zaret TM . 1979 . Parental care patterns of fishes . Am Nat . 113 : 351 – 361 . Google Scholar Crossref Search ADS Persson L , de Roos AM . 2013 . Symmetry breaking in ecological systems through different energy efficiencies of juveniles and adults . Ecology . 94 : 1487 – 1498 . Google Scholar Crossref Search ADS PubMed Porter WP , Kearney M . 2009 . Size, shape, and the thermal niche of endotherms . Proc Natl Acad Sci USA . 106 : 19666 – 19672 . Google Scholar Crossref Search ADS PubMed Pough FH . 1980 . The advantages of ectothermy for tetrapods . Am Nat . 115 : 92 – 112 . Google Scholar Crossref Search ADS Ricklefs RE . 1979 . Adaptation, constraint, and compromise in avian postnatal development . Biol Rev Camb Philos Soc . 54 : 269 – 290 . Google Scholar Crossref Search ADS PubMed Shine R . 1988 . Parental care in reptiles. In: Gans C , Huey RB , editors. Biology of the Reptilia . New York : Alan R. Liss, Inc. p. 276 – 329 . Shine R . 2005 . Life-history evolution in reptiles . Ann Rev Ecol Evol Syst . 36 : 23 – 46 . Google Scholar Crossref Search ADS Smith CC , Fretwell SD . 1974 . The optimal balance between size and number of offspring . Am Nat . 108 : 499 – 506 . Google Scholar Crossref Search ADS Somma LA . 2003 . Parental behavior in lepidosaurian and testudinian reptiles . Malabar (FL) : Krieger Publishing Company . Sousa T , Domingos T , Kooijman SA . 2008 . From empirical patterns to theory: a formal metabolic theory of life . Philos Trans R Soc Lond B Biol Sci . 363 : 2453 – 2464 . Google Scholar Crossref Search ADS PubMed Sousa T , Domingos T , Poggiale JC , Kooijman SA . 2010 . Dynamic energy budget theory restores coherence in biology . Philos Trans R Soc Lond B Biol Sci . 365 : 3413 – 3428 . Google Scholar Crossref Search ADS PubMed Stearns SC . 1992 . The evolution of life histories . Oxford : Oxford University Press . Summers K . 1990 . Paternal care and the cost of polygyny in the green dart-poison frog . Behav Ecol & Sociobiol . 27 : 307 – 313 . Google Scholar Crossref Search ADS Tumulty J , Morales V , Summers K . 2014 . The biparental care hypothesis for the evolution of monogamy: experimental evidence in an amphibian . Behav Ecol . 25 : 262 – 270 . Google Scholar Crossref Search ADS Webb G , Manolis C . 2002 . Australian crocodiles: a natural history . Sydney : Reed New Holland . Webb GJW , Mandis SC , Buckworth R . 1982 . Crocodylus johnstoni in the McLinlay River area, N.T. I. Variation in the diet, and a new method of assessing the relative importance of prey . Aust J Zool . 30 : 877 – 899 . Google Scholar Crossref Search ADS © The Author(s) 2019. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Thermodynamic constraints and the evolution of parental provisioning in vertebrates
JF - Behavioral Ecology
DO - 10.1093/beheco/arz025
DA - 2019-06-13
UR - https://www.deepdyve.com/lp/oxford-university-press/thermodynamic-constraints-and-the-evolution-of-parental-provisioning-RaEYqO4sm8
SP - 583
VL - 30
IS - 3
DP - DeepDyve
ER -