Jump to ContentJump to Main Navigation
Ecological and Environmental Physiology of Amphibians$

Stanley S. Hillman, Philip C. Withers, Robert C. Drewes, and Stanley D. Hillyard

Print publication date: 2008

Print ISBN-13: 9780198570325

Published to Oxford Scholarship Online: April 2010

DOI: 10.1093/acprof:oso/9780198570325.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy).date: 22 March 2018

Physiological Adaptations to Extreme Environments

Physiological Adaptations to Extreme Environments

(p.281) 4 Physiological Adaptations to Extreme Environments
Ecological and Environmental Physiology of Amphibians

Stanley S. Hillman

Philip C. Withers

Robert C. Drewes

Stanley D. Hillyard

Oxford University Press

Abstract and Keywords

This chapter explores what is known about how amphibians deal with extreme environmental situations, beginning with aquatic environments, both fresh water and marine. It then describes life in extreme terrestrial and underground environments. The next section deals with overwintering in both aquatic and terrestrial environments, and discusses the particular stresses of hypoxia and low temperature. The chapter concludes with a discussion of metabolic depression as an intrinsic metabolic mechanism that extends energy reserves when environmental conditions preclude activity.

Keywords:   extreme environmental situations, aquatic environments, fresh water, sea water, terrestrial, soil, arid, hibernation, hypoxia tolerance, freeze tolerance

4.1 Salt and Osmotic Sensitivity

4.1.1 Low Salt Environments

Many aquatic species occur in streams or ponds with extremely low Na+ and Cl concentrations. Krogh (1937) showed that Rana esculenta are able to independently absorb Na+ and Cl from solutions as dilute as 10−5 M. This study is often cited to support the assumption that cutaneous salt absorption is essential for maintaining the plasma Na+ and Cl concentrations. However, in this same study Krogh observed that frogs sprayed with distilled water survived for up to 12 weeks. More recently, McAfee (1972) kept Rana pipiens without food in flowing deionized water for 30–45 days without appreciable loss of Na+. Like Krogh (1937), he suggested the role for cutaneous absorption is minimal for aquatic species with sufficient food available. In a similar study, Bentley (1973) maintained three anuran species for 15 days in dilute tap water (0.25 mmol/l NaCl) with 10 µM amiloride, an inhibitor of epithelial Na+ transport. Plasma Na+ concentration of the aquatic and semiaquatic species (Xenopus laevis and Rana pipiens) was not significantly reduced but declined by approximately 10% in the more terrestrial toad, Bufo marinus. As noted in Chapter 3, NaCl absorption from very dilute media requires the energetic contribution of both the Na+–K+ pump in the basolateral membrane of principal and MR cells and active H+ transport by the V-ATPase in the apical membrane of MR cells. The density of MR cells increases when frogs (R. temporaria) or toads (Bufo bufo and B. marinus) are exposed to dilute NaCl solutions (Voute and Meier 1978; Budtz et al. 1995; Devuyst et al. 1991). This is may explain the capacity of R. temporaria to maintain plasma Na+ and Cl concentrations but appears to be insufficient for bufonids species that appear to lose these ions in dilute media. Isotopic measurements with intact animals and isolated skin have shown terrestrial species to have a lower affinity for NaCl absorption than aquatic species in conjunction with a greater level of water permeability (Greenwald 1972; Mullen and Alvarado 1976). Alternatively, the reduction of plasma Na+ concentration in terrestrial species could result from excessive water gain since they display greater levels of cutaneous water permeability under control conditions (Table 3.3).

(p.282) 4.1.2 Hypersaline Environments

Amphibians are usually found in freshwater habitats where salt retention is of primary concern. However, many species have been reported in brackish water that for a few may approach the salinity of seawater (Neil 1958). Field observations and laboratory. experiments that determine salinity tolerance use equivalent NaCl concentrations, often in parts per thousand (g/L = ppt). Table 4.1 gives the millimolar concentrations and equivalent osmotic concentrations over the range of salinities where amphibians are known to occur, using an osmotic coefficient of 0.93 for NaCl. The term brackish water covers a range of NaCl concentrations defined by the European Environment Agency as between 5 and 18 ppt.

Balinsky (1981) identified 61 species in 13 families of anurans and 12 species in four families of urodeles that inhabit or tolerate brackish water. Three anuran species, Rana cancrivora, Bufo viridis and Xenopus laevis, have been shown to tolerate salinities in excess of what is commonly considered brackish. The most salinetolerant frog, R. cancrivora, inhabits mangrove swamps in southeast Asia and can be acclimated to salinities of 29 ppt NaCl (Gordon et al. 1961). In the laboratory, frogs showed a progressive rise in plasma Na+, Cl and urea as the external salinity was increased from freshwater to 250, 500 and 800 mOsm/l (Fig. 4.1). The plasma osmolality was regulated to be greater than the external solution while plasma K+ concentration remained relatively stable.

A similar pattern of elevated plasma Na+, Cl and urea concentrations was observed with Bufo viridis and Xenopus laevis. B. viridis occur in wide range of habitats from the middle eastern deserts to Scandinavia were it has been found in (p.283) salinities of 20 ppt (Gislen and Khoury 1959). Katz (1973) was able to acclimate B. viridis to NaCl concentrations as high as 800 mOsm/l (c. 25 ppt) by gradual exposure to increasing salinity. In arid regions, B. viridis also accumulates urea to offset water loss while burrowed in dry soil (see below). X. laevis is an entirely aquatic frog that may encounter drying ponds and, like B. viridis, is able to accumulate urea and endure elevated Na+ and Cl concentrations to offset water loss to substrates with low water potential. X. laevis is also saline tolerant and can be acclimated to NaCl concentrations as high as 600 mOsm/l.

Table 4.1 Summary of NaCl concentration values in units of salinity, molarity and osmotic concentration

Salinity (ppt NaCl)

NaCl concentration (mmol/l)

Osmotic concentration mOsm/l





Sea water



















| Brackish water















Fresh water

The capacity of R. cancrivora, X. laevis and B. viridis to tolerate elevated salinity is mediated by a greater tolerance of elevated Na+ and Cl concentrations in the extracellular fluid and the ability to accumulate urea. Katz (1973) observed that the ability of B. viridis to accumulate urea was highest during the summer that is consistent with observations, noted in Chapter 3, that corticosteroid levels are greatest during this time (Jolivet-Jaudet et al. 1983). The accumulation of urea is a function of synthesis and retention. The accumulation of urea, and also free amino acids, during exposure to elevated salinity has been linked to upregulation of ornithine urea cycle (OUC) enzymes as an osmoregulatory response (reviewed by Balinsky 1981). In contrast, Jorgensen (1997) suggests the upregulation of OUC enzymes is a response to stress and that non-stimulated rates of urea synthesis are adequate to explain the accumulation seen in saline-acclimated animals. Rather, it is the balance between urea synthesis and excretion that allows accumulation. Wright et al. (2004) acclimated juvenile R. cancrivora from fresh water to 20 and 25 ppt NaCl and found the OUC enzyme carbamoyl phosphate synthase to be upregulated and also the excretion of urea to be unchanged despite the greater (p.284) gradient for loss to the outside, suggesting an increase in cutaneous absorption and hepatic synthesis of urea. Studies with isolated skin have demonstrated active urea influx in the aquatic frog Rana esculenta and the toads B. bufo and B. viridis (Garcia Romeu et al. 1981; Katz et al. 1981). Of interest, active urea absorption across the skin was only observed in saline-acclimated R. esculenta and increased as the NaCl concentration was increased from 240 to 400 mOsm/l. Active urea uptake by toad skin was observed in freshwater and saline-adapted animals.

                   Physiological Adaptations to Extreme Environments

Fig. 4.1 Increasing plasma Na+, Cl and urea concentrations of R. cancrivora acclimated to NaCl solutions with salinities of 0, 250, 500 and 800 mOsm/l. A similar pattern is seen with B. viridis and X. laevis (plotted from tabular data in Balinsky 1981).

In addition to cutaneous uptake, urea retention is mediated by a reduction in urea excretion by the kidneys and absorption across the urinary bladder. Urea excretion can be evaluated as the ratio of urea concentration in the urine relative to the plasma (U/P ratio) where a higher ratio represents greater excretion (Schmid 1968) while a value at or below unity indicates retention of urea in the body if water is absorbed across the bladder (Jorgensen 1997). U/P ratios of 0.5, 0.3–1.0 and 0.94 have been reported, respectively, for R. cancrivora, B. viridis and X laevis (Schmidt-Nielsen and Lee 1962; Katz et al. 1984; McBean and Goldstein 1970). The equilibration of urine with plasma appears to be mediated by a facultative urea transport protein (UT) that has been localized in the apical membranes of cells in the early distal tubule and urinary bladder of B. marinus (Konno et al. 2006). These results are consistent with earlier observations that cystectomized R. cancrivora are unable to accumulate urea (Chew et al. 1972).

Although considered less tolerant, a number of anuran species have been anecdotally observed to occur in habitats where brackish water might be encountered (Niel 1958). Bufo marinus, as its name implies, is found in estuarine areas (Zug and Zug 1979) and can tolerate salinities of 15 ppt (Liggins and Grigg 1985). Bufo calamita inhabit estuarine habitats in costal regions of the Baltic sea and Great Britain and are able to tolerate salinities of 16–17 ppt for up to 4 days (Mathias 1971). Sinsch et al. (1989) found that inland populations of B. calamita were less tolerant of elevated salinity than coastal populations. However, Sinsch et al. (1992) attributed this to seasonal variation and noted a maximal level of NaCl tolerance to be 454 mOsm/l (c. 14 ppt) in active, feeding toads. Bufo bufo is considered more of a mesic species but is able to tolerate salinities of 150 mmol/l NaCl (Ferriera and Jesus 1973).

The capacity of urodeles to tolerate elevated salinity by accumulating Na+, Cl and urea has been investigated in Ambystoma tigrinum and several species of Batrachoseps (Plethodontidae). A. tigrinum is found in arid regions and is able to endure drying of ponds by burrowing. As with other burrowing species, urea accumulates and contributes to the lowering of water potential so the animals retain water in dry soil (Delson and Whitford 1973). Romspert and McClanahan (1981) were able to acclimate A tigrinum to NaCl solutions with osmolality has high as 450 mOsm/l during which the plasma urea concentration increased from 15 to 182 mmol/l along with increases in Na+ and Cl to maintain the osmolality above that of the external solution. Licht et al. (1975) collected Batrachoseps (p.285) relictus from a coastal area of California where exposure to seawater was a likely event. Its capacity to acclimate to salinity was compared with that of B. attenuatus from inland areas. Tolerance was judged by survival of animals placed in shallow solutions of FW or diluted SW. Both species tolerated direct exposure 30% SW (c. 300 mOsm/l); however, B. relictus survived longer when placed directly into 40 or 50% SW (c. 400 and 500 mOsm/l) and recovered better from brief exposure to full SW as might occur during a tidal event. After 24 hour’s exposure to 50% SW the plasma urea concentration of B. relictus rose modestly from 38 to 57 mmol/l while that of B. attenuatus rose from 19 to 35 mmol/l. Under the experimental conditions the animals were able to drink, which appeared to be a cause of excessive salt accumulation and death This was addressed by Jones and Hillman (1978) who gradually acclimated B. attenuatus and B. major to elevated salinity. The animals were placed on paper towels saturated with NaCl solutions that included F W and 250 mmol/l NaCl (c. 465 mOsm/l). After 25 days all animals had survived and showed an elevation of plasma urea from 48 mmol/l in FW compared with 218 mmol/l in NaCl. Clearly, acclimation to elevated salinity requires gradual exposure. Similar examples can be found in the anuran literature. Bufo viridis that were acutely transferred to 500 mOsm/l NaCl solutions survived for only 2–4 days (Gordon 1962; Katz 1973).

From the above examples, it is obvious that behavioral avoidance will benefit species that are found in habitats where available hydration sources include those with elevated salinity. As shown in Chapter 3, dehydrated toads will exhibit water absorption response behavior (WR) in which the seat patch is pressed to a wet surface. If the surface is moistened with NaCl solutions the frequency of initiating WR behavior is sustained until the osmotic concentration for water absorption decreases to near zero (Fig. 4.2). Brekke (1991) showed a similar avoidance of hyperosmotic urea solutions and to lightly moistened filter paper with a low water potential. Nagai et al. (1999) tested the hypothesis that the skin serves a chemosensory function that allows toads to detect and avoid unfavorable hydration sources in a manner analogous to taste cells in the tongue of vertebrates that drink orally. It was found that spinal nerves make connections with epithelial cells in the skin and that exposure of the skin to NaCl concentrations at or above 200 mmol/l produced a burst of activity in afferent nerve fibers that corresponds with the behavioral avoidance. Toads consistently avoided WR behavior on 200–250 mmol/l NaCl and the neural response increased dramatically between 200 and 300 mmol/l NaCl. The sensory ability of bufonids that have been studied (Brekke et al. 1991; Nagai et al. 1999; Maleek et al. 1999) correlates well with the ranges for acute physiological tolerance of salinity that are summarized by Balinsky (1981). The neural response was minimal at lower NaCl concentrations (less than 200 mmol/l) that most species can tolerate.

There are examples of amphibians capable of detecting dilute salt solutions. Rana temporaria embryos and larvae are more sensitive than adults to elevated (p.286) salinity. Survival is reduced by 20% in salinities of 2.3 ppt (56 mOsm/l) and further declines to about 80% with 130 mOsm/l (4.5 ppt) NaCl solutions (Viertel 1999). Because the larvae have little or no ability to change habitat, the adults must select an appropriately dilute body of water for oviposition. The greatest salinity where oviposition was observed in natural habitats was 0.9 ppt indicating the frogs select sites well within the larval salinity tolerance. Reboreda and Segura (1984) were able to record neural responses from isolated legs of Bufo arenarum that were transferred between deionized water and hyperosmotic NaCl solutions. A response was recorded during transfer from either solution to the other, indicating the existence of water and salt-sensitive fibers. Recent experiments with intact toads have obtained a neural response to deionized water that is much slower in developing than the response to salt, which might explain why it was not observed in earlier studies (Hillyard et al. 2007). Salt-specific and water-specificfibers have been characterized in the glossopharyngeal nerve from frog tongue and are connected to salt and water-sensitive taste cells in the tongue (Sato and Beidler 1975; Okada et al. 1993). The skin of adult amphibians doesn’t have identifiable taste cells (Fox 1984) so characterizing the sensory pathway remains an area for future study.

                   Physiological Adaptations to Extreme Environments

Fig. 4.2 Dehydrated B. punctatus will consistently initiate the water absorption response (WR) behavior on surfaces moistened with deionized water and with NaCl solutions up to approximately 125 mmol/l, which is approximately iso-osmotic with the plasma. WR behavior declines rapidly at higher concentrations.

The chemosensory function of the skin has also been shown to integrate physiological responses to salinity. Petriella et al. (1989) showed that B. arenarum immersed in 115 or 400 mmol/l NaCl solutions for 5 min reduced urine production before any measurable change in plasma concentration was observable. (p.287) This response was inhibited by the guanethidine, indicating the response is mediated as a sympathetic reflex. Viborg and Rosenkilde (2004) found that the increase in cutaneous blood flow to the seat patch observed in dehydrated toads (Chapter 3) required skin contact with a dilute solution; dehydration alone was insufficient. The increase in blood flow was mimicked by the β-adrenergic agonist isoproterenol and inhibited by the antagonist propranolol, indicating a sympathetic reflex in response to water potential receptors in the skin. Thus, behavioral and physiological mechanisms rely on the ability of the skin to detect dilute as well as concentrated solutions.

4.2 Arid Environments

Most amphibians lose water freely from their skin by evaporation (Section 3.2), they are unable to produce urine that is hyperosmotic to their blood (Section 3.7), and they have limited tolerance to high ambient temperatures (Section 3.4). Consequently, we might not expect amphibians to cope well with the characteristically low water availability and high temperatures of arid environments. Nevertheless, many deserts of the world have an abundant, if not diverse, amphibian fauna, consisting predominantly of anurans.

Arid and semiarid environments are characterized by low water availability, often in conjunction with high ambient temperatures (“hot” deserts) but sometimes low ambient temperatures (“cold” deserts). For example, under Köppen’s climate classification scheme, an arid climate (category B) occurs where Pann <10 Pth (Pann is annual rainfall, mm, and Pth is a function of the mean annual temperature, Tann, °C; Kottek et al. 2006). Steppe climate (BS) is 5 Pth < Pann <10 Pth and desert climate (BW) is 5 Pth < Pann <10 Pth. Arid climates are also categorized as hot (h, Tann ≥18) and cold (k, Tann <18). Desert and steppe climates generally occur about 30° above and 30° below the equator, and include the major deserts of the world (Fig. 4.3).

The deserts of Australia have species of burrowing Neobatrachus, Cyclorana, Notaden, Uperoleia, Pseudophryne, Limnodynastes and arboreal frogs Litoria (Cogger 2000; Seymour and Lee 1974; Tyler 1999). North American deserts have burrowing anurans (Scaphiopus, Spea, Bufo) and treefrogs (Hyla, Pternohyla); the salamanderAmbystoma mexicanum is also found throughout the southern deserts, at least where there is permanent water. South American arid and semiarid regions, particularly the Cerrado-Caahinga-Chaco have a diverse amphibian fauna, including 98 anurans and a single aquatic caecilian (Chthonerpeton perisodus); many anurans burrow and aestivate in termitaria (Dermatonotus, Elachistocleis) or rodent burrows (Bufo, Leptodactylus), or form a cocoon (Lepidobatrachus, Ceratophrys; Duellman 1999). The cold Patagonian deserts have Bufo, Pleurodema and Zachaenus, and the very arid Atacama has just Bufo and a poison-arrow frog (p.288) Colostethus (Duellman 1999). Deserts of southern Africa have various burrowing species (Pyxicephalus, Tomopterna) and scansorial species (Leptopelis and Chiromantis), and sub-Saharan Africa has the aquatic Xenopus as well as Bufo, Hoplobatrachus and Ptychadena (Poynton 1999). North African deserts similarly have Xenopus, Bufo, Hoplobatrachus and Ptychadena as well as Tomopterna and Rana (Borkin 1999). The Sinai and Arabian deserts have a few anuran species, including Bufo, Ptychadena, Rana and Hyla. The “desert” salamanders Salamandra salamandra and Triturus vittatus are found in very xeric environments (Warburg 1988). The Eurasian and central Asian deserts similarly have a diverse anuran fauna, of Bufo, Batrachuperus, Euphlyctis, Tomopterna, Hoplobatrachus, Paa, Rana, Hyla and Pelobates (Borkin 1999).

                   Physiological Adaptations to Extreme Environments

Fig. 4.3 World map of the Köppen-Geiger climate classification, showing cold semiarid (BSk), hot semiarid (BSh), cold arid (BWk) and hot arid (BWh) zones; major deserts of the world are indicated. Climate zones modified from Kottek et al. 2006.

The adaptations of these various amphibians from the semiarid and arid zones include behavioral avoidance of the harsh environmental conditions as well as specific physiological adaptations that increase desiccation tolerance, increase thermal tolerance, and decrease cutaneous EWL (e.g. Warburg 1988; Loveridge 1977; Seymour and Lee 1974; Shoemaker et al. 1992; Rome et al. 1992; Hutchison and Dupre 1992; Pinder et al. 1992; McClanahan et al. 1994). Bladder urine is an important store of water for amphibians to buffer changes in body fluid solute and osmotic concentrations during dehydration (see Section 3.1). The low concentration of urine in the bladder (very dilute compared to body fluids) and the large volume of urine stored in the bladder (up to 50% of body mass) provides a substantial water store to buffer changes in body fluid solute and osmotic (p.289) concentrations during aestivation. Over time, Na+, Cl, urea and osmotic concentrations increase in both plasma and urine, but the urine urea and osmotic concentrations increase much more than Na+ and especially Cl in absolute mM difference. The urine-to-plasma (U/P) ratio is 0.5–0.9 for total osmotic concentration and 0.6–1.9 for urea. The amount of urine remaining in the bladder is low (e.g. McClanahan 1972), indicating that the bladder stores are becoming depleted as well as isosmotic. Most amphibians have a high body water content, about 77–83% of total body mass (bladder empty), and have a high tolerance to dehydration, particularly the more terrestrial species. Desert amphibians store a considerable amount of relatively dilute urine in their bladder (see Section 3.1); up to 50% or even more of their body mass can be bladder urine (Bentley 1966; Mayhew 1968; Main and Bentley 1964).

4.2.1 Diurnal Retreat

The vast majority of these desert amphibian species are not able to withstand the high temperature and desiccating environment of a desert during the day, or during the dry season. Behavioral selection of an appropriate sheltering site to avoid dehydration and hyperthermia is consequently important; most frogs would be unable to survive even 12 h exposure to typical desert daytime conditions. The typical response of a number of amphibians to dehydrating conditions is increased activity, to seek out a more favorable location, although dehydration eventually compromises locomotor capacity and the amphibian becomes physiologically incapacitated at its debilitating critical activity point (Heatwole et al. 1971; Heatwole and Newby 1972; Putnam and Hillman 1977; Moore and Gatten 1989).

Frogs seek refuge on a daily basis either underground by burrowing into the soil, by moving deep into cracks in rocks or by sheltering in trees under bark or in hollows (e.g. Loveridge 1977; McClanahan et al. 1994). Diurnal basking behavior is relatively uncommon in amphibians (Hutchison and Dupré 1992; see Chapter 3) due to their permeable skin and hence high rate of evaporative water loss, but some anurans do bask during the day for thermoregulation (Bradford 1984; Carey 1978; Freed 1980; Lillywhite et al. 1973, 1998; Muths and Corn 1997; Pearson and Bradford 1976; Sinsch 1989; Vences et al. 2002; Tattersall et al. 2006). Only a few amphibians, notably waterproof scansorial frogs, are able to bask during the day in exposed sites (Loveridge 1970; Shoemaker et al. 1972; see Section 3.4). Most desert frogs are nocturnal; they emerge from their refuges on suitable nights to forage, etc.

4.2.2 Seasonal Aestivation

Many amphibians in arid environments escape the hot dry season by remaining underground or in tree hollows or under bark for extended periods of time (p.290) (aestivation; see Section 4.4). Many fossorial species burrow underground, where they retain water balance with the soil; other fossorial or arboreal species form a protective cocoon that reduces cutaneous evaporative water loss.

An extended period of summer dormancy is called aestivation. For example, the water-holding frog Cyclorana platycephala can remain buried underground for over 5 years if favorable surface conditions do not occur (van Beurden 1980), even though it is more typical for these frogs to be active each year after heavy cyclonic rains. The period of aestivation is a well-defined physiological state, especially for cocoon-forming species (see below). Aestivating frogs have considerably reduced activity. Non-cocooning species, such as North American spadefoot toads (Scaphiopus, Spea) and the Australian spadefoot (Notaden spp.), will remain sufficiently active to consume their skin whenever it is shed (see Section 3.2) and presumably are sufficiently mobile underground that they can burrow deeper as the soil dries, to follow the moisture and remain in water balance (McMaster 2006). Cocoon-forming species become immobile and do not consume their shed epidermis, which accumulates to form a water-proofing cocoon (see below). The metabolic rate of aestivating frogs is typically depressed to about 20–25% of their normal standard metabolic rate at that temperature (Guppy and Withers 1999; see Section 4.5). This allows endogenous energy stores to sustain metabolism for about five times longer, so that the amphibian can survive the lengthy aestivation period.

The African clawed toad (Xenopus laevis) is normally aquatic, like other pipids, but it has been reported to burrow in drying mud (Alexander and Bellerby 1938; Tasman 1956; Loveridge 1977). During this time, Xenopus accumulate urea but do not become dehydrated (Balinsky et al. 1967). However, in the laboratory Xenopus kept in mud failed to aestivate as the mud dried and hardened (Jokumsen and Weber 1980). This suggests that Xenopus “aestivating” in the field remain in sufficiently moist mud to maintain water balance, but as they do not form a cocoon they are unable to survive in dried mud. Interestingly, Xenopus are remarkably tolerant of exposure to hyperosmotic solutions, and accumulate urea as a balancing osmolyte (Section 4.1).

4.2.3 Fossoriality

Most “desert” amphibians escape from the hostile environment by burrowing underground. During favorable conditions, they burrow during the day and are active at night. During extended periods of unfavorable conditions, they remain underground and aestivate (Section 4.5). Many of these fossorial species continue to ingest their shed skin and hence do not form a cocoon. Some of these fossorial “desert” amphibians form a cocoon (see Section 3.2) during aestivation; a few salamanders (Siren spp.) also form a cocoon. Over time, the cocoon becomes a thick paper-like covering (Fig. 4.4). Cocoon-forming frogs tend to shed their skin (p.291) at a higher frequency than non-cocooning species, about every 1 to 3 days, both normally and during aestivation.

                   Physiological Adaptations to Extreme Environments

Fig. 4.4 B. Aestivating Australian trilling frog (Neobatrachus aquilonius) with a thick, parchment-like cocoon covering the entire outer body surface except for the external nares. (Photo: G. Thomson, P. Withers, V. Cartledge, K. McMaster).

The underground environment is typically more hospitable to an amphibian than the surface conditions. The thermal environment is relatively stable and benign for fossorial frogs. Temperature under the ground is lower than the daytime surface air temperatures (e.g. Fig. 4.5A), and the daily temperature cycle becomes more attenuated with greater depth underground. Annual soil temperature changes are also reduced with increasing depth (Fig. 4.5B). Soil temperature generally remains within tolerable limits for amphibians except near the surface. Even in extreme deserts, soil temperature at depth is 25–30°C (e.g. van Beurden 1980; Thompson et al. 2005) which is well within the thermal limits of desert amphibians (see Section 3.4). The temperature in Scaphiopus burrows varies from 17–22°C in autumn, 5–15°C in winter and 24–25°C prior to emergence in summer (Ruibal 1962; Shoemaker et al. 1969). The average temperature in aestivating Cyclorana platycephala burrows was 15°C with daily variations of (p.292) (p.293) about 7°C (van Beurden 1980). Temperatures range from 21–27°C for the salamander Siren intermedia burrowed in dry mud (Gehlbach et al. 1973). Nevertheless, soil temperatures in the top 10–20 cm of exposed sites can exceed 30°C during the hottest part of the year; the highest recorded burrow temperature was 39°C for a shallow Scaphiopus burrow in summer, although lower soil temperatures are found in more sheltered locations. High soil temperatures would be disadvantageous to burrowed frogs since metabolic rate, even though depressed during aestivation (see Section 4.4) would be increased because of the Q10 effect, and this would reduce the period of time that frogs could remain aestivating underground.

                   Physiological Adaptations to Extreme Environments

Fig. 4.5 (A) Soil water content (%) and temperature (°C) profiles for sand dunes where the sandhill frog (Arenophryne rotunda) occurs. The open triangle and square indicate the average depth and moisture/ temperature location of frogs. Values are mean ± SE. From Cartledge et al. (2006). (B) Annual temperature cycles at various depths underground in a desert wash at Palm Desert, California; dotted line indicates the isopotential for -1500 kPa (about 600 mOsm/l at 25°C). Modified from Muth (1980).

Humidity is generally high underground, reducing the rate of evaporative water loss, although this depends on the soil moisture level and temperature. For example, coastal sand dunes in Western Australia provide a favorable environment for the sandhill frog Arenophryne rotunda. Sand has a high moisture content after rainfall, and is saturated by 5 cm depth (5%); the temperature is stable by 10 cm depth at 15°C (Fig. 4.5A). In summer, the surface sand is much drier and sandhill frogs burrow much deeper to reach similar moisture levels (Tyler et al. 1980; Roberts 1990). In more extreme arid environments such as the Gibson Desert in Western Australia, soil water content is lower (1–3%) and underground temperatures are higher (25–30°C) in sand dunes where the desert spadefoot Notaden nichollsi burrows to 1 m or more, and claypans where the desert trilling frog Neobatrachus aquilonius burrows to about 30 cm (Thompson et al. 2005; Cartledge et al. 2006). In an exposed site (desert wash), the hygric environment is favorable for burrowed amphibians (>–1500 MPa) at depths near the surface in the wet season and below about 20 cm in the dry season (Fig. 4.5B).

Frogs that burrow in sandy soils generally do not form a cocoon and dig deeper as the dry season progresses, following the moisture profile. As the frogs burrow deeper, their burrow is typically loosely filled, and the frogs remain surrounded by loosely-packed sand (rather than an air space, as in some cocoon-forming frogs; see below). For example, Scaphiopus are burrowed at about 20 cm depth at the start of the dry season, are at 50 cm or more deep in mid-winter, then return closer to the surface in early summer in anticipation of summer rains (McClanahan 1967; Ruibal et al. 1969; Seymour 1973b). The day-time burrows of Scaphiopus during the summer activity period are much shallower, about 6–8 cm (Seymour 1973b). Similarly, the sandhill frog Arenophryne burrows deeper as the dry season progresses, to remain in noticeably moist soil (Tyler et al. 1980; Roberts 1984, 1990) whereas active wet-season burrows are only about 4 cm deep in the moist surface sand (Cartledge et al. 2006). Southern African Breviceps and Pyxicephalus burrow 80–150 cm during aestivation (Mayhew 1968). Australian sand-burrowing frogs such as Heleioporus and Notaden burrow 80 to 150 cm, or even more (Mayhew 1968; Thompson et al. 2005; Fig. 4.6).


                   Physiological Adaptations to Extreme Environments

Fig. 4.6 Excavated burrow of (A) desert spadefoots (Notaden nichollsi) from sand dunes (burrow at back, hole for excavation in front; burrow was loosely-filled with sand before excavation) and (B) a trilling frog (Neobatrachus aquilonius) from a clay pan (burrow depth about 30 cm; vertical burrow). (Photos: G. Thomson, P. Withers, V. Cartledge, K. McMaster)

In contrast, cocoon-forming frogs burrow in heavier clay soils, which become rock-hard when dry. They cannot move after they have formed their cocoon, without disrupting its physical integrity and are unable to follow the water profile deeper as the soil dries out. However, the cocoon provides protection against desiccation in situ (see below). The Australian cocoon-forming Cyclorana and Neobatrachus generally burrow into harder clay soils, to only about 30 cm (Main et al. 1959; van Beurden 1980; Thompson et al. 2005), but Neobatrachus aquilonius have been excavated from depths of as much as 120 cm in claypans (Thompson et al. 2005). Interestingly, the cocoon-forming N. aquilonius has also been excavated from moist sandy sites without a cocoon (Cartledge et al. 2006). Cocoon-forming frogs therefore appear to have the behavioral flexibility to delay cocoon-formation and follow the soil moisture profile as do non-cocooning frogs, prior to their eventually forming a cocoon. In the laboratory, cocoon-forming frogs will delay the onset of cocoon formation until the soil dries out and becomes a desiccating rather than a rehydrating environment (personal observations; Cartledge 2006). Booth (2006) found that Cyclorana albiguttata did not begin to form a cocoon (p.295) until after 3 months, when the soil environment became desiccating relative to the water potential of the frog.

Cutaneous–Soil Water Exchange

Fossorial amphibians are able to readily rehydrate from soil interstitial water, providing the osmotic balance is favorable (and if they do not form a cocoon). Many laboratory studies have demonstrated the capacity to absorb water from moist soil for a wide variety of salamanders and anurans (see Chapter 3). Amphibians in the field have been shown to rehydrate while burrowed underground in moist soil during their resting period (e.g. Heleioporus eyrei, Lee 1968; Scaphiopus toads, McClanahan 1972).

The water balance between a buried frog and its surrounding soil environment is complex, depending on, firstly, the osmotic balance between soil water and the frog and, secondly, on the rate at and mechanism by which water moves through the interstitial spaces of the soil (see Section 3.2). The driving force for water movement between soil and a buried frog is the difference in water potential (ψ; units of pressure e.g. MPa) between the soil and the frog, but the relationship between soil water potential (ψsoil) and water content is a complex function of soil particle size (see Fig 3.5). Sandy soils do not hold water as “tightly” as clay or silt soils, so the water potential of sandy soil is more favorable for reabsorption by a burrowed frog compared with the same percent water content as clay or silt soils, that is frogs in sandy soil can reabsorb water from relatively dry sand (water content of about 0.05%) whereas frogs in clay soil can only reabsorb water from wetter soil (about 0.25%; Fig. 4.7).

Liquid water exchange and water vapor exchange are the potential mechanisms for water exchange between a frog and soil because water in the soil interstitium is a mixture of liquid water and water vapor, except in very dry soil (only vapor) and saturated soil (only liquid water; Section 3.2; Booth 2006). The rate of liquid water exchange is determined by the soil hydraulic conductivity; this decreases as soils dry. Hydraulic conductivity is negligible for sandy soils if ψsoil less than about −0.1 MPa, so hydraulic loss of water is unimportant for frogs in sand because ψfrogis about −0.6 MPa and water loss is determined by vapor phase exchange (Fig. 4.7). For frogs in clay, hydraulic conductivity occurs until the water potential is very negative (e.g. <–1.5 MPa) so it is theoretically important even when frogs are considerably osmoconcentrated (about 600 mOsm/l). This is probably, at least in part, why frogs that burrow and aestivate in clay generally create an air space around their body, to break any liquid water connection between their skin and soil water (Section 3.2; Booth 2006); it would also prevent them from becoming solidly encased in dried clay.

The sandhill frog, Arenophryne rotunda, is able to absorb water from sand with a moisture content of 5% but not from 1 or 2% (Cartledge et al. 2006). Sandhill frogs could burrow into sand at 5% water content in the field and are in (p.296) positive water balance (Table 4.2) and while burrowed during the day replenish any water losses experienced during foraging on the dune surface at night. For Scaphiopus, the soil water tension is equivalent to about 60–200 mOsm just after burrowing, and rises to 150–200 mOsm during the dry season; soil water can be absorbed by the frog since this is less than the osmotic potential of its body fluids (Shoemaker et al. 1969; Ruibal et al. 1969). In early summer, the soil is much drier (about 450–600 mOsm equivalent) and the spadefoots must have an increased body fluid osmotic concentration to remain in balance. Similarly, the Australian spadefoot N. nichollsi can maintain positive water balance in moist sand but not in drier sand, in the field (Table 4.2). This desert sand dune frog will burrow to over 2 m depth to remain in moist sand (Thompson et al. 2005). The trilling frog, N. aquilonius, burrows in clay pans, to about 30–120 cm depth (Thompson et al. 2005), where it can be near water balance in favorable years but not in drier years (Table 4.2). However, the trilling frog is a cocoon-former; its protective cocoon (see Chapter 3) increases cutaneous resistance to prevent desiccation under conditions of unfavorable water balance with the soil.

                   Physiological Adaptations to Extreme Environments

Fig. 4.7 Relationship between soil water potential and soil water content for sandy and clay soils; approximate Ψ range is shown for water exchange by hydraulic conduction (solid line) and water vapor diffusion (dotted line). Ψ is about -0.6 MPa for normally-hydrated frogs and -1.6 MPa for frogs aestivating in the field (see text). Modified from Booth (2006).

Accumulation of Solutes

Amphibians that aestivate in soil with a higher ψ than their body fluids can retain a positive water balance and absorb water from the soil to replace any lost by urination or evaporation when active. Urinary excretion would prevent urea (synthesized from ammonia formed by protein catabolism) from accumulating in the body fluids, but would also result in ion depletion (especially Na+ and Cl). Dilute urine stored in the bladder is also an important water store. Consequently, urination might be avoided or at least minimized even if water balance with the soil was positive. For example, spadefoot toads buried in relatively moist soil in the laboratory slowly accumulate urea and ions in their body fluids, and their total osmotic concentration increases (McClanahan 1972; Fig. 4.8). The minor mass loss after 17 months of aestivation presumably reflects dry tissue loss (e.g. fat body) rather than water loss.


Table 4.2 Summary of balance of osmotic forces in the field for burrowed North American spadefoot toads (Scaphiopus couchii), Australian sandhill frogs (Arenophryne rotunda), Australian desert spadefoots (Notaden nichollsi) that do not form a cocoon, and Australian cocooned trilling frogs (Neobatrachus aquilonius), by comparison of body fluid osmotic concentration (Ocfrog) and equivalent osmotic pressure (Ψfrog), with the actual soil water potential (Ψsoil)


OCfrog (mOsm/l)

Ψfrog (kPa)

Ψsoil (kPa)


A. rotunda





S. couchii















N. nichollsi





    sand dune 2003





    sand dune 2004





N. aquilonius *





    clay pan 2003





    clay pan 2004




near balance


*cocooned. 1000 mOsm/l ≡ −24.4 atm ≡ −2.475 MPa at 25°C. Data from Shoemaker et al. (1969) and Cartledge et al. (2006a, b).

As soil progressively dries out during aestivation, Ψsoil decreases to less than Ψfrogand water balance becomes negative. Non-cocooned frogs could move deeper underground, to retain contact with moist sand and avoid osmoconcentration. Alternatively, the total body fluid osmotic concentration can be increased such that Ψsoil is similar to Ψfrog and at least neutral if not positive water balance is maintained. Body fluid osmotic concentration could be increased by loss of water to the soil, but a better strategy would be to increase body fluid osmotic concentration by accumulation of additional solutes. This pattern of osmotic change over time is well illustrated by Scaphiopus couchii burrowed in the laboratory (McClanahan 1972; Fig. 4.8). As soil dries, the body fluid osmotic concentration increases further to retain water balance, by increases in urea concentration (Fig. 4.8). Body (p.298) mass declines slightly over time, presumably reflecting loss of metabolized tissue mass (e.g. fat body) rather than loss of water. A similar pattern of increasing ion but especially urea and thus osmotic concentration over time has been observed in the field in Scapiopus couchii (Shoemaker et al. 1969).

                   Physiological Adaptations to Extreme Environments

Fig. 4.8 Changes in plasma Na+ (hatched), urea (shaded) and total osmotic concentrations (OC) of the spadefoot toad (Scaphiopus couchii) during aestivation in soil of varying water content; (A) 8.8, (B) 5.5 and (C) 5.1 ml/100 g. The equivalent water potential, expressed in equivalent osmotic concentration, is given by the dotted line in each panel. Data from McClanahan (1972).

Many other burrowed (non-cocooned) amphibians show similar increases in body fluid solute concentration (particularly urea), and hence Ψ, over time. The Australian spadefoot (Notaden nichollsi) has a higher osmotic concentration in drier sand in the field, due to higher Na+, Cl and urea concentrations (Cartledge et al. 2006). The toad Bufo viridis initially loses mass, then stabilizes after a few days when burrowed in the laboratory (Katz et al. 1986; Katz and Hoffman. 1990; Hoffman and Katz 1997). After 7 weeks, its osmotic concentration increases to over 1000 mOsm, reflecting changes in Na+ (160 mmol/l) and Cl (205 mmol/l) but especially urea (495 mmol/l; Katz and Hoffman 1990). Urine voidance ceases, minimizing urinary urea, ion and water loss. Burrowed tiger salamanders (Ambystoma tigrinum) exceed 500 mOsm/l osmotic concentration after 9 months burrowed in soil, reflecting a large increase in urea concentration (>200 mM; Delson and Whitford 1969). This ureo-osmoconforming strategy (see Withers 1998) is not restricted to burrowed amphibians, occurring in chondrichthyeans, the coelacanth, the crab-eating frog Rana cancrivora and other amphibians acclimated to hypersaline media (Section 4.1). Some of these burrowing amphibians acclimated to hypersaline media (400–850 mOsm/l) have an increased osmotic concentration to retain water balance, and this involves large increases in urea, for example Xenopus laevis, Scaphiopus couchii, Bufo viridis and Ambystoma tigrinum (see Katz 1992). Other non-arid amphibians also show a ureo-osmoconforming response to hypersaline media, for example the crab-eating frog Rana cancrivora, treefrog Hyla arborea, and a plethodontid salamander Batrachoseps spp. (see Katz (p.299) 1992). Many anurans show a lesser osmotic/ urea accumulation response to desiccating burrow conditions, with only moderate increases in osmotic concentration (e.g. to 250–400 mOsm/l) and small changes in urea (e.g. 10–50 mmol/l), for example Bufo bufo, B. regularis, Pelobates syriacus, Rana ridibunda, R. pipiens, R. temporaria and R. esculenta (Katz 1992).

An increase in urea concentration over time is important in many aestivating amphibians for maintenance of water balance with the soil, and it is clear that the urea concentration increases faster for frogs in drier soil. The retention of urea from background levels of protein catabolism during aestivation is apparently not sufficient to explain the magnitude of the increased urea concentration, nor is osmoconcentration by loss of body water, since Na+ and Cl do not increase so dramatically, for example Scaphiopus couchii; (McClanahan 1972; Fig. 4.8), Bufo viridis (Katz and Hoffman 1990) and Ambystoma tigrinum (Delson and Whitford 1969). Similar, rapid accumulation of urea has been reported for amphibians acclimated to hypersaline media (e.g. X. laevis). Consequently, the increase in urea is often attributed to accelerated rates of urea synthesis (Balinsky et al. 1981; McClanahan 1972; Jones 1980b; Hoffman and Katz 1999; Delson and Whitford 1973). However, Barker Jorgensen (1997) suggested that the data for accelerated urea synthesis by Scaphiopus, Bufo and other ureo-osmoconforming amphibians are equivocal, with elevated urea concentrations reflecting changes in balance between synthesis and excretion.

Urea becomes an important balancing osmolyte for burrowed, non-cocooned amphibians, whether or not its synthesis is accelerated as a response to soil water balance. Urea levels in aestivating amphibians depend on the length of the aestivation period, and the soil conditions; concentrations from 100 to 500 mM are not unusual. These high urea concentrations approach or exceed those of marine chondrichthyean fishes, whose enzymes are adapted to resist the inimical effects of high urea on protein structure and function and beneficial effects of counteracting solutes such as betaine and sarcosine (see Hochachka and Somero 2002). In aestivating amphibians, counter-acting solutes do not appear to coaccumulate with urea (Withers and Guppy 1996) and the inimical kinetic effects of urea are relatively minor, particularly compared with the effects of high ion concentrations (Grundy and Storey 1994; Cowan and Storey 2002; see Section 4.4).

The cocoon of amphibians seals their entire body surface, including the eyes, mouth and cloaca, but not the external nares which remain open for pulmonary gas exchange. A cocooned frog is therefore isolated from the external environment (apart from gas exchange) and must survive its aestivation period on internal energy and water reserves. Energy reserves are extended considerably by metabolic depression (see Section 4.4). By the time a cocooned is formed, water balance with the soil is most likely negative and the essential role of the cocoon is to minimize cutaneous EWL. Frugal use of bladder water reserves is equally important for survival.

(p.300) For cocooned amphibians, body fluids and bladder water are the only water stores to balance loss by cutaneous and pulmonary evaporation; metabolic water production is a minor source of water for aestivating cocooned frogs. The inevitable loss of water by pulmonary and cutaneous avenues during aestivation means that the body fluids will become concentrated, particularly after the bladder reserves are depleted or become equivalently concentrated as body fluids. For example, body fluids of the aestivating, cocooned Lepidobatrachus llanensis become concentrated over time, particularly with respect to Na+ and Cl (McClanahan et al. 1972). Urine solute concentrations increase even more dramatically than plasma over the course of extended aestivation, and eventually are about the same concentration as body fluids. Urine volume also declines markedly. Similar changes in plasma and urine solutes have been shown for other aestivating frogs in the laboratory, including Cyclorana platycephala and Neobatrachus aquilonius (Cartledge 2006; Cartledge et al. 2006; Table 4.4). Field measurements of N. aquilonius show no significant increases in plasma Na+, Cl, K+ or total osmotic concentration, with aestivation in relatively moist conditions (in sand, non-cocooned), but substantial increases were measured under drier conditions for cocooned frogs in clay pans (Cartledge et al. 2006).

Cocooning amphibians do not have the adaptive need to accumulate solutes to balance soil water potential during aestivation, unlike non-cocooning frogs. The cocoon seems to not be formed until the soil water potential becomes desiccating, and then the high diffusion resistance of the cocoon retards EWL. They will accumulate urea over time to detoxify ammonia formed by protein metabolism, and progressive dehydration by evaporative water loss will increase the urea concentration, but there would seem to be no selective pressure for cocoon- forming amphibians to accelerate urea synthesis during aestivation. Consequently, they generally do not accumulate urea to the high concentrations measured for urea-tolerant, non-cocooning amphibians (Withers and Guppy 1996; Table 4.3). Balancing/non-perturbing solutes do not co-accumulate during aestivation (Withers and Guppy 1996) and high concentrations of urea have lesser adverse effects on enzyme function for aestivating frogs (Fuery et al. 1997; Cowan and Storey 2002).

4.2.4 Scansoriality

Many amphibians are found in trees, on rocks and other above-ground locations. Particularly in semiarid and arid environments, a scansorial habit potentially exposes an amphibian to lethal desiccating conditions, but it also provides the potential to exploit heliothermic behavioral thermoregulation. Many scansorial amphibians remain nocturnal and shelter during the day from the desiccating and high temperature conditions, seeking shelter in tree-hollows, under bark and inside vegetation (e.g. bromeliads, etc). Some, however, have become more diurnally active. Not surprisingly, diurnal scansorial amphibians have a suite of (p.301) behavioral and physiological adaptations that in concert allow survival in hygric and thermal conditions that would be lethal for most amphibians. These adaptations for diurnal scansoriality have evolved independently a number of times in anuran amphibians (with notable absence in salamanders and caecilians), and to differing extents between typical non-waterproof frogs to extremely waterproof xeric frogs. The suite of inter-related behavioral and physiological adaptations (p.302) include reduced cutaneous evaporative water loss, basking behavior, capacity for color change (particularly blanching to a white color), and the ability to increase EWL when heat-stressed (“sweating”). Many other physiological systems are influenced by these changes, for example cutaneous gas exchange, kidney function and color change. A few waterproof frogs have even switched their nitrogenous waste disposal strategy to uricotely. None, however, have completely freed their reproductive strategy from the requirement for free water (although some fossorial frogs have direct development).

Table 4.3 Urea levels of amphibians that do not form a cocoon (Nc) with low or high tolerance, cocooning species (C), and species acclimated to hypersaline media (Hs)



Urea (mmol/l)


Bufo bufo

NC low


Schoffeniels and Tercafs (1965)

Rana temporaria

NC low


Ackrill et al. (1969)

Thoropa miliaris



Abe and Bicudo (1991)

Rana pipiens

NC low


Jungreis (1971)

Pelobates syriacus

NC low


Katz (1992)

Rana esculenta

NC low


Garcia-Romeu et al. (1981)

Siren intermedia



Asquith and Altig (1986)

Lepidobatrachus llanensis



McClanahan et al. (1976)

Bufo regularis

NC low


Katz (1992)

Siren lacertina



Etheridge (1990)

Rana ridibunda

NC low


Katz (1992)

Notaden nichollsi

NC high


Cartledge et al. (2006a)

Neobatrachus aquilonius



Withers and Guppy (1996)

Cyclorana platycephala



Withers and Guppy (1996)

Pyxicephalus adspersus



Loveridge and Withers (1971)

Bufo woodhousei

NC high


Jones (1982)

Neobatrachus kunapalari



Withers and Guppy (1996)

Hyla arborea

NC high


Katz (1992)

Xenopus laevis



Schlisio et al. (1973) Romspert (1976)

Heleioporus albopunctatus

NC high


Withers and Guppy (1996)

Neobatrachus pelobatoides



Withers and Guppy (1996)

Ambystoma tigrinum

NC high


Delson and Whitford (1973)

Neobatrachus sutor



Withers and Guppy (1996)

Batrachoseps sp.



Jones and Hillman (1978)

Cyclorana maini



Withers and Guppy (1996)

Rana cancrivora (adult)



Gordon et al. (1961)

Bufo viridis

NC high


Hoffman et al. (1988)

Scaphiopus couchii

NC high


McClanahan (1972)

Table 4.4 Plasma and urine solute and osmotic concentrations for aestivating anurans while burrowed, with the change in urine concentration (∆ urine) and final urine-to-plasma (U/P) ratio












Bufo viridis1



































N. nichollsi2



































N. aquilonius3



































1 7 weeks, Katz 1990;

2about 8 months in 2004, Cartledge et al. 2006a;

3 cocooned in clay-pan for about 18 months in 2003, Cartledge et al. 2006a.

The primary behavioral switch that would initiate the suite of physiological adaptations of waterproof frogs is basking in exposed sites on trees or rocks. Exposure, even in shade, predisposes an amphibian to desiccation and warming to air temperature, and basking directly in sunlight further increases the risk of desiccation and elevation of Tb above ambient air temperature. The two main physiological changes required to allow basking are reduced EWL and elevated thermal tolerance, and not surprisingly these are interconnected in the graded transition from a typical frog to a very waterproof frog (Table 4.5). Resistance (p.303) to evaporative water loss increases from that for a free water surface and typical amphibians (about 1 s/cm) to over 200 s/cm in very waterproof species. This is essential for a scansorial amphibian to avoid rapid dehydration to their lethal limit since ambient air in semiarid/ arid environments can have a very low humidity. There is a general correlation between cutaneous EWL and selected temperature; the very waterproof Chiromantis, Phyllomedusa and Hyperolius have a higher CTmax than less-waterproof species such as Rana and Limnodynastes (Table 4.5); within Litoria, the less-waterproof species have a slightly lower CTmax than the more waterproof species (Tracy and Christian 2005).

Table 4.5 Change in cutaneous water loss (resistance; s/ cm) and thermal tolerance in sequence from typical frogs to waterproof frogs


Resistance (s/cm)

CTmax (°C)

Limnodynastes fletcheri



Rana pipiens



Bufo marinus



Litoria meiriana



Scaphiopus spp.



Bufo cognatus



Litoria aurea



Litoria rubella



Litoria rothii



Litoria caerulea



Litoria chloris



Litoria bicolor



Litoria gracilenta



Hyperolius viridiflavus



Chiromantis spp.


>41 ??

Phyllomedusa spp.


>42 ??

Data from Lillywhite (2006), Tracy et al. (2008), Shoemaker et al. (1989) Giese and Linsenmeir (1986b) Brattstrom (1968, 1970) Buttemer (1990).

One mechanism for reduction of EWL is well understood—Phyllomedusa spp. use a stereotypic wiping behavior to cover their body surface with a waxy secretion from cutaneous glands, and Litoria spp. might do the same. Other waterproof frogs do not wipe and do not have a waxy surface covering—dried cutaneous protein/ mucous secretions might provide their waterproofing barrier. The thermal consequence of exposure to the atmosphere is warming to ambient air temperature, which in many semiarid and arid environments would easily exceed the (p.304) lethal thermal tolerance of a typical amphibian (30–35°C; see Section 3.4). Basking in direct sunlight would elevate skin and core body temperature above ambient air temperature, and further elevate the rate of cutaneous water loss.

Scansorial frogs that bask in direct sunlight often have a remarkable cap acity to change color to a grey or sometimes brilliant white. The African tree frog Chiromantis xerampelina often basks in very exposed sites in direct sunlight, and can blanch to a very bright white (Shoemaker et al. 1989; McClanahan et al. 1994; Fig. 3.10). Some Hyperolius species also blanch to a very bright white when exposed to desiccating conditions (Withers et al. 1982; Kobelt and Linsenmair 1986), as does the Australian desert tree frog, Litoria rubella (Withers 1995). This pattern of blanching to white sometimes correlates directly with a decrease in EWL, suggesting that there is a common mechanism for EWL reduction and skin blanching, for example dermal iriodiophores (Drewes et al. 1977; Bentley and Yorio 1977; Withers et al. 1982; Kobelt and Linsenmair 1986). However, it is difficult to envisage how the waterproofing barrier could involve living cells such as iridophores (or layers of waxy material in subepidermal glands), so the tight correlation of reduced EWL with color change appears to not be a mechanistic link but rather a coadaptive response to different physiological consequences of a specific environmental challenge.

Waterproof frogs will select a higher body temperature (Tb) in a thermal gradient than non-waterproof frogs. For example, Chiromantis will select Tbs about 36–37°C and sometimes > 40°C, and Phyllomedusa will select 33–40°C and sometimes > 41°C (Shoemaker et al. 1989). Interestingly, both genera will select a lower Tb when dehydrated. The adaptive increases in selected Tb and CTmax of waterproof frogs means that they must be able to thermoregulate Tb below their CTmax at high ambient air temperatures. Enhanced evaporative cooling is the major strategy for thermoregulation by terrestrial animals in the heat, and waterproof frogs use the same strategy. They control the discharge of cutaneous mucous glands, like normal frogs (cf Lillywhite 1971; Lilly white and Licht 1975) to regulate EWL and hence Tb. Both Chiromantis and Phyllomedusa will regulate Tb at about 40–42°C at higher Ta by proportional increases in EWL (Shoemaker et al. 1987). For Chiromantis xerampelina, the control of EWL is achieved by modulation of sympathetic outflow influencing β-adrenergic receptors of cutaneous mucous glands (Kaul and Shoemaker 1989). Litoria chloris and L. xanthomera have a reduced resistance to EWL at Ta >40°C (Buttemer 1990; Buttemer and Thomas 2003). Litoria spp. at high Ta (43–45°C) have a sufficiently low R that they can keep Tb below their CTmax, but the more waterproof species that increase R at lower Ta would reach or exceed their CTmax if they didn’t adaptively reduce R (Fig. 4.9; Tracy et al. 2008). However, once waterproof frogs reduce their R to evaporatively cool, they reduce the selective advantage of being waterproof, that is water conservation, and then rely on bladder water stores to maintain hydration, like non-waterproof frogs.

Only two genera of amphibians are known to excrete their nitrogenous waste as uric acid, rather than the typical form of nitrogenous waster for aquatic amphibians (p.305) (ammonia) or terrestrial amphibians (urea). Not surprisingly, these were the first two genera of frogs reported to be waterproof, Chiromantis (Loveridge 1970) and Phyllomedusa (Shoemaker et al. 1972). The water-savings accruing from uricotely are presumably only of adaptive significance for a waterproof frog; a non- waterproof frog would lose so much more water by cutaneous evaporation than urinary excretion that there would be no selective pressure to conserve urinary water loss. Nevertheless, uricotely is not necessarily adaptive for all waterproof frogs. Hyperolius nasutus is a small, very waterproof frog that is conspicuously not uricotelic (Withers et al. 1984). Despite their high cutaneous resistance, they are unable to survive desiccating conditions for more than about 14 days because of their small size and high surface-to-volume ratio. However, urea (as their main nitrogenous waste) would not accumulate to lethal levels over this time, so there is no survival advantage to being uricotelic, and incurring the additional metabolic cost of synthesizing uric acid rather than urea (Withers et al. 1984). Similarly, Giese and Linsenmair (1986) have argued that there is no selective advantage for Hyperolius viridiflavus to be uricotelic despite apparently having the biochemical machinery for uricotely—these frogs accumulate uric acid in their chromatophores, connective tissue and liver during the dry season when metabolic rate is much reduced (Geise and Linsenmeir 1986). The small phyllomedusine P. hypochondrialis has about the same resistance to EWL as its larger congeners such as P. sauvagei, but excretes less of its nitrogen waste as uric acid (Shoemaker et al. 1975), perhaps also because of a lesser adaptive advantage for uricotely in small frogs.

                   Physiological Adaptations to Extreme Environments

Fig. 4.9 Relationship between cutaneous resistance to EWL (R; s/cm), Skin surface temperature (Tskin, white bars; °C) is maintained lower than the critical thermal maximum (CTmax, black bars; °C) by various Litoria treefrogs at high ambient temperature. For L. caerulea and L. rothii, Tskin approaches or exceeds CTmax if their normally-low evaporative water loss (grey bars) is not increased by enhanced evaporation. The cutaneous resistance values (sec/cm) are indicated, to show the reduction in R that maintains Tskin below the CTmax for these species. Data from Tracy et al. (2008).

(p.306) Waterproof skin reduces EWL but presumably also reduces the diffusional exchange of other gasses, including O2 and CO2. In Hyperolius viridiflavus, there is a significant continuous loss of CO2 across the skin in the wet season, when EWL is high, but virtually no cutaneous CO2 loss in the dry season when there is a marked increase in CO2 excretion via discontinuous respiration (Giese and Linsenmier1986b). When cutaneous EWL is low during the dry season, the proportion of respiratory water loss is high, at about 15% of total EWL. The cutaneous gas exchange of Chiromantis xerampelina and Phyllomedusa sauvagei is reduced concomitant to their low cutaneous EWL, being similar to the low cutaneous gas exchange of reptiles (Stinner and Shoemaker 1987). Interestingly, the blood gas status (blood pCO2, HCO - 3 and pH) of these frogs is similar to that of “normal” non-waterproof frogs, suggesting that they hyperventilate to avoid hypercapnia in the absence of significant cutaneous CO2 loss.

Being waterproof and relieved of excessive EWL has apparently not relieved the selective pressures on the role of bladder water as a water store that is so important for non-waterproof frogs. The bladder store of dilute urine is used to preserve normal hydration levels during dehydration. For example, the waterproof Phyllomedusa has a bladder volume of about 20–30% of body mass (Shoemaker and Bickler 1979) and a low urine concentration. Consequently, bladder urine serves as a water store that buffers EWL for Phyllomedusa (Shoemaker and Bickler 1979) just as it does for more typical anurans, (Shoemaker et al. 1992). The tolerance to water loss of waterproof frogs is high (e.g. up to 50% for Hyperolius viridiflavus; Giese and Linsenmeir 1986b).

4.3 Hibernation Adaptations

Hibernation is winter dormancy. Two different environmental stresses, a lack of food and low ambient temperatures, characterize the conditions amphibians must contend with in order to survive the winter dormancy. Food is generally not available during winter, so metabolism can only be supported by energy reserves. The advantage for ectotherms of experiencing low ambient temperatures during winter is that body temperature, and hence rate of metabolism, will be low. Consequently, the rate of depletion of energy reserves will be slow and the probability of death by starvation is diminished at these low temperatures. The obvious danger of hibernating at low temperature for an amphibian is the potential for the environmental temperature to get low enough to freeze and kill it. This danger is exacerbated when a species distribution encompasses higher latitudes and/or higher elevations where freezing temperatures are more likely.

The environmental conditions experienced by an amphibian in a temperate zone winter differ between aquatic and terrestrial environments. Ice-covered water mechanically prevents aerial breathing, hence gas exchange can only occur across (p.307) the skin and/or gills. Aquatic environments have a much lower O2 capacitance than terrestrial environments (Chapter 1). When water is covered by ice there is a much slower rate of O2 exchange with the environment, so aerobic metabolic activity of organisms living in the water and substrate are much more likely to deplete the O2and create hypoxic (low P O 2 ) conditions. Hypoxia is much more likely to develop in shallow bodies of water and when snow is deep, shading the water and preventing photosynthetic generation of O2. Consequently, physiological and behavioral mechanisms involved in dealing with hypoxia are important for winter survival in aquatic environments. A variety of data suggest that winterkill resulting from hypoxia is an important source of mortality for aquatic amphibians (Bannikov 1948; Manion and Cory 1952; Bradford 1983; Licht 1991; Pasanen and Sorjonen 1994). Tadpoles are more tolerant of hypoxia than adults, and have a lower critical p O 2 (Bradford 1983). This differential hypoxia tolerance is probably the result of higher surface area/mass (gills and skin) and a lower hemoglobin P50 for tadpoles compared to submerged metamorphs and adults. Submergence in fresh water also drives diffusional water uptake and ion loss across the integument during the winter, potentially creating ionic/osmotic problems. An advantage of an aquatic environment is that the amphibian is very unlikely to be exposed to water temperatures low enough to cause death by freezing. An aquatic environment will also have less variation in temperature than terrestrial environments at the same mean temperature due to the high specific heat capacity of water (Chapter 1), leading to lower rates of metabolism, lower rates of depletion of energy reserves and an extension of the possible length of time hibernation is possible (see below).

An amphibian hibernating in a terrestrial environment has the advantage of remaining normoxic, hence aerobic metabolism will not be compromised. However the probability of both experiencing a greater variation in ambient temperature and being exposed to subzero temperatures is greater in a terrestrial winter environment. The increased temperature variation has the effect of raising the average rate of metabolism and decreasing the length of time that the amphibian can survive on its chemical energy reserves. The other potential difficulty with a terrestrial environment is that that ambient temperatures can drop significantly below 0°C. This increases the risk of freezing. Dehydration is yet another physiological problem for amphibians over-wintering in a terrestrial environment.

The higher M O 2 of a frog at a variable temperature derives from the Q10 relationship of metabolism and temperature. Take a hypothetical situation where one frog experiences a constant daily 5°C environmental temperature, whereas another spends half the day at 0°C and the other half at 10°C. The mean temperature of 5°C is the same for both frogs. If the Q10 is 2, and the metabolic rate at 0°C is 2 (arbitrary units) then the metabolic rate at 10°C will be 4 (4 = 2(21)). The average metabolic rate will be 3 ((4 + 2)/2) for the variable frog and lower at 2.82 for the constant 5°C frog (2.82 = 2 (20.5)). Because the relationship of metabolism to temperature is usually exponential, metabolism will be greater with increased (p.308) variation despite having the same mean temperature. This effect increases as the Q10 for metabolism increases. The preceding example used a typical Q10 of 2, but it has been demonstrated to be 4.3 at these low temperatures (Tattersall and Boutilier 1997)

Temperate zone amphibians hibernate in three major environmental situations: aquatic, subterranean terrestrial, and shallow terrestrial under leaf duffand bark. The choice of hibernation sites varies interspecifically and has a strong behavioral component (Licht 1991). Metamorphs can also vary from adults in their choice of hibernation site (Laitinen and Pasanen 1998; Lamoureux and Madison 1999). The choice of over-wintering sites is not absolute for the adults of any species. Site preferences in the lab or field border on 70–90% for one site versus the other (Licht 1991; Pasanen and Sorjonen 1994; Laitinen and Pasanen 1998; Lamoureux and Madison 1999; Holenweg and Reyer 2000). This lack of absolute intraspecific fidelity for an aquatic or terrestrial over-wintering site means that during any given winter there will be individuals of the same species in both terrestrial and aquatic over-wintering sites. This may represent an evolutionary bet-hedging strategy. The intraspecific proportion of individuals choosing more terrestrial sites can also be related to age of frogs, with new metamorphs choosing aquatic sites and adults choosing terrestrial sites (Laitinen and Pasanen 1998; Lamoureux and Madison 1999). Small frogs have a higher surface area/ volume ratio, which would enhance cutaneous gas exchange in an aquatic environment but enhance the rate of evaporative water loss in a terrestrial environment. Both consequences would favor aquatic hibernation sites for juveniles. It is not clear whether this aquatic–terrestrial shift represents an ontogenetic shift of site preference or an age-related distribution of over-wintering sites as a consequence of higher rates of mortality for individuals selecting aquatic sites. Use of subterranean sites requires the ability to dig or the presence of fissures in the soil, so small size or well-developed metatarsal tubercles would be anatomical correlates of this pattern. The advantages of being able to penetrate deep in the soil are that temperature variation is lower, temperature is less likely to fall below freezing and there is a lower likelihood of dehydrational stress. This minimizes the possibility of freezing mortality, but since the soil temperature is higher, energy reserves will be expended faster. This pattern is characteristic of toads and many salamanders. Some frogs (tree frogs and some Rana) hibernate predominantly under bark or at the interface of soil and leaf litter. This entails a greater risk of both freezing and dehydration. Some species that occupy superficial terrestrial sites have some capacity to tolerate freezing of extracellular fluids (see below). The aquatic over-wintering environment is the preference of most frogs and obviously aquatic salamanders. This has the advantage of a low likelihood of freezing, a more predictable rate of expenditure of energy reserves via temperature selection, balanced by the risks of the development of hypoxia, hyperhydration and the loss of electrolytes.

(p.309) This begs the question, is an aquatic or terrestrial environment the better situation for an amphibian to survive the winter? The answer, of course, is it depends. From a distributional perspective it appears that selection for terrestrial hibernation sites (both superficial and subterranean) characterizes the most extreme high latitudinal distributions for anurans. The most northerly distributed anurans tend to over-winter in superficial terrestrial locations, exactly the sites most likely to freeze. What might be the selective advantages of these over-wintering sites besides escaping the potential threat of hypoxia? Terrestrial sites allow greater adult mobility during more equitable weather in late fall and early winter. This gives the individuals more latitude in finding appropriate thermal and hydric microenvironments. Terrestrial sites will be the first hibernation sites to thaw, hence breeding activity can occur early in the spring. Given the short growing season at high latitudes, this gives a temporal advantage to the larvae of these species.

4.3.1 Aquatic Hibernation and Hypoxia

There is strong evidence that winter can serve as the “hammer of death” for aquatic hibernators. Winterkill due to hypoxia is well known for fish (Greenbank 1945; Barica and Mathias 1979) and populations of R. temporaria (Bannikov 1948; Pasanen and Sorjonen 1994), R. muscosa (Bradford 1983), and R. pipiens (Manion and Cory 1952; Licht 1991). These data clearly establish the strong selective pressure operating for hypoxia tolerance and the importance of understanding how amphibians cope with hypoxic, aquatic conditions over winter.

The physiological problems that develop during hypoxic exposure essentially ramify from a decreased diffusional flux of O2 influx across the skin, the only available exchange surface. Therefore the overall O2 flux is decreased during submergence. This can compromise aerobic metabolism under hypoxic conditions, and decrease the rate of ATP production, as can be seen in Fig. 4.10.

Amphibians that over-winter in aquatic environments have behavioral choices in both the temperature and P O 2 of the water they occupy at any time. The likelihood of hypoxia developing is less likely near the surface, and in moving water, whether from seeps or streams, since convective movement of non-hypoxic fresh water will replenish the O2. Adult green frogs (Rana clamitans) leave ponds in late fall to occupy sites with flowing water (stream or seep sites) that will remain oxygenated through the winter (Lamoureux and Madison 1999). Recent metamorphs of R. clamitans remain in ponds to over-winter. Moving to flowing water during winter also seems to be a strategy of R. muscosa to prevent hypoxic stress (Matthews and Pope 1999). Bullfrogs (R. catesbeiana) select inlets of underground tributaries (Stinner et al. 1994). The European common frog (R. temporaria) when confronted with hypoxic conditions in a thermal gradient, shift preferred temperature from 6.8°C in normoxia, to 1.9°C in hypoxia (Tattersall and Boutilier (p.310) 1997). This reduces metabolism by 80% and postpones the shift from aerobic to anaerobic metabolism. The European common frog (R. temporaria) also selects a higher P O 2 in an oxygen gradient when submerged (Tattersall and Boutilier 1999). These results imply that submerged amphibians can behaviorally exploit heterogeneous oxygen environments, as well as lowering metabolism (demand for O2), by behaviorally selecting lower temperatures in hypoxic environments. Behavior is the first key to surviving the winter for submerged amphibians, and from a physiological perspective this requires coordinated movement at low temperatures. Basically, the central nervous system and skeletal muscles have to work at low temperatures in a winter pond for the animal to exploit the thermal and oxygen heterogeneity of its environment.

                   Physiological Adaptations to Extreme Environments

Fig. 4.10 A summary of the negative physiological consequences of hypoxic submergence in amphibians (see text).

The physiology of aquatic hibernation is equally fascinating. The cardiorespiratory implications of aquatic submergence need to be considered first in order to set the stage for understanding the tissue P O 2 conditions. First, skin is the only gas exchange surface, and the aquatic environment boundary layers can significantly reduce the cutaneous O2 conductance in still water. Normally the skin only receives about 10–20% of the pulmocutaneous blood flow and its venous drainage (the blood with the highest P O 2 ) returns and is mixed with systemic venous drainage. Since there is no pulmonary gas exchange, mixing within the heart essentially becomes irrelevant. Complete mixing of right and left side blood occurs in Bufo marinus at low temperature and resulting low heart rate (Hedrick et al. 1999). Systemic and cutaneous P O 2 s are basically equal. Cutaneous blood flow with low (p.311) P O 2 would maximize cutaneous O2 diffusional uptake. This would be especially important in hypoxic aquatic situations where the blood P O 2 has to be below the water P O 2 for O2 to diffuse into the animal. Consequently, the maximal P O 2 of the systemic blood will always be lower than the P O 2 of the aquatic environment. We might expect blood P O 2 to be low and cutaneous blood flow to be high during aquatic submergence to increase the rate of uptake from the water. The problem is that the blood supplied to the skin is the same P O 2 as systemic blood; a low blood P O 2 and high flow helps on the uptake side but compromises the systemic O2delivery side. Consequently, a left to right shunt is not necessarily favorable. We saw in Chapter 2 that systemic arterial P O 2 s are normally in a range of 8.7–12 kPa, leading to hemoglobin saturations of 85–95%. Measurements of blood P O 2 from normoxic submerged frogs range from 0.4 to 1.2 kPa (Ultsch et al. 2004; Stewart et al. 2004). Given normal blood O2 loading curves this compromises O2 carrying capacity, that is 5–10% hemoglobin saturations at these P O 2 s. Couple this to low blood flow rates resulting from the low heart rates at low temperature, and systemic O2 transport is severely limited. Submergence consequently severely limits aerobic metabolic capacity, even in normoxic waters.

As we will see below, aerobic metabolism is very low during hibernation, but how much aerobic scope is there during submergence? We can estimate this from the data of Ultsch et al. (2004) who determined the critical P O 2 (Pc) for resting metabolism of submerged R. pipiens. The V O 2 at Pc was 0.05 nmol/g sec, the P aO 2 at this point was 1.7 kPa and using the limiting mitochondrial P O 2 of 0.1 kPa (St-Pierre et al. 2000), the tissue diffusing capacity equals V O 2 / Δ P O 2 or 0.05/(0.17−0.1) = 0.7 nmol/sec g kPa within the error of our determination in Chapter 2 of 0.9 nmol/sec g kPa. Consequently, at rest in normoxic water the factorial scope for further metabolic activity is dictated by the difference in P aO 2 in fully saturated water and at Pc divided by the P aO 2 at Pc ((0.4 − 0.17)/0.17 = 1.4, data taken from Ultsch et al. 2004). When this is compared to factorial increments of 6–10 for amphibians with all modes for respiratory exchange available, it becomes clear that hibernating R. pipiens are operating close to the limits of aerobic metabolic capacity at rest during hypoxic exposure. The only way to provide a measure of metabolic capacity from both an absolute and factorial perspective is to depress V O 2 , which is what we will delineate in the sections below.

The metabolic demand for O2 decreases during hibernation in two ways, a Q10 driven decrease in metabolism and an hypoxic induction of a hypometabolic state. The Q10 driven decrease in metabolism is the result of the processes delineated in Fig. 4.11.

There are two things to note in this scheme: there will be a reduction in metabolism and this reduction is matched by compromised O2 flux in terms of decreased blood flow rates, decreased O2 loading of hemoglobin and decreased respiratory O2 conductance. Consequently, there is concordance between the decreased rate of aerobic ATP synthesis and metabolic ATP utilization still maintaining (p.312) a homeostatic cellular energy balance. There is little capacity to increase aerobic ATP generation.

                   Physiological Adaptations to Extreme Environments

Fig. 4.11 A summary of the physiological consequences of hibernation at low temperature for amphibians (see text).

The second component of the physiology of aquatic hibernation is the induction of a hypometabolic state, below that normally associated with the Q10 effect on metabolism (Donohoe et al. 1998). Metabolic depression will increase the length of time energy reserves will last as well as provide some hypoxic protection. Standard V O 2 declined by 40% after 3 months of submergence in 3°C aerated water compared to standard V O 2 prior to submersion for R. temporaria at 3°C (Donohoe et al. 1998). This hypometabolism was intensified with a 75% reduction from standard V O 2 if the frogs were submerged in progressively hypoxic water from 20 to 4.5 kPa (Donohoe and Boutilier 1998). The 40 and 75% reductions in aerobic metabolism increases the length of time energy reserves will last by two and four fold. The substrates utilized to drive V O 2 during a period of submergence were a combination of fat and glucose at the start, to only glucose at the end, in both normoxic and hypoxic (4.5 kPa) submergence. The source of the glucose at the beginning of hibernation was liver glycogen stores, and both liver and skeletal muscle glycogen stores at the end of 90 days submergence (Donohoe and Boutilier 1998; Donohoe et al. 1998). Anaerobic metabolism was used in the first month of hypoxic submergence based on an increase in lactate concentration (p.313) (Donohoe and Boutilier 1998). The use of anaerobic metabolism during the first month of hypoxic submergence increased the rate of glycogen utilization in skeletal muscle, cardiac muscle and liver compared to normoxic submergence. The use of anaerobic metabolism with its accumulation of lactate and depletion of glycogen has been demonstrated in other amphibians exposed to anoxia (Rose et al. 1965; Christiansen and Penney 1973). Intracellular levels of ATP, phosphocreatine, AMP, and ADP did not change over time in either normoxic or hypoxic submergence, indicating that normal homeostatic intracellular energy status was maintained. The ability to reduce aerobic metabolic needs, and the utilization of anaerobic metabolism, at the same time systemic O2 transport is compromised, allowed the energy status of the cells to be maintained.

How is this hypometabolism achieved? To understand hypometabolism one has to understand a chemical energy budget at a cellular level—essentially, how is the ATP partitioned at rest and what avenues of O2 utilization can be minimized and still maintain cellular integrity? Unfortunately, there are insufficient data to discuss this solely for amphibians, though there is an extensive understanding for mammals (see Rolfe and Brown 1997; Section 4.5). Consequently we will assume the avenues of ATP expenditure are apportioned similarly between mammals and amphibians to start the discussion. The actual partitioning varies with organ; we will use skeletal muscle cells since they represent about 40% of the body mass in amphibians (Putnam 1979), and their V O 2 conforms to tissue P O 2 (Boutilier 2001). Consequently as the P O 2 declines during hypoxia their metabolic rate will decline, and this would make them likely candidates to contribute to the hypometabolic state. About 86% of the total O2 consumed is by the mitochondria, the other 14% of O2 being utilized by a variety of peroxisomal oxidases associated with fatty acid oxidation, endoplasmic reticulum oxidases, and some involved with the production of superoxide and hydrogen peroxide (the free radical problem). Of the mitochondrial O2 consumption in skeletal muscle cells at rest, up to 50% of the resting O2 consumption can be associated with a “proton leak” from the intermembrane space across the inner mitochondrial membrane back to the mitochondrial matrix (Rolfe and Brand 1996). The other 50% of mitochondrial O2consumption is coupled to the production of ATP. At rest in skeletal muscle, about 40% of the ATP generated is utilized by the Na-K-ATPases, Ca-ATPases and myosin ATPases, which maintain the concentration differences of these ions between the intracellular and extracellular spaces. The remaining 60% is used in the synthesis of protein, mRNA, glycogen, urea and everything else. Obviously, during skeletal muscle activity the fraction utilized by the ATPases increases dramatically to accommodate the action potentials, sarcoplasmic reticulum Ca fluxes which initiate the interaction of actin–myosin and the cycling of actin–myosin cross bridges during shortening.

The O2 consumption of skeletal muscle cells associated with the “proton leak” is of particular interest during hibernation since it is a futile cycle from (p.314) the perspective of generating useful energy for the cell in the form of ATP. This proton leak proceeds in the same direction as the flux through the mitochondrial F1F0-ATPase but is uncoupled from the synthesis of ATP (Rolfe and Brand 1996). This futile cycle of proton pumping into the intermembrane space and diffusion back to the matrix and not through the F1F0-ATPase creates a large inefficiency of the coupling between O2 consumption and ATP synthesis. As Boutilier (2001) correctly cautioned, the production of 38 ATP per glucose is only true when mitochondria are operating maximally at P O 2 levels greater than 20 kPa, in the presence of saturating substrate and ADP concentrations (state 3 mitochondria). The reality is that these special conditions are very far from normal resting values of P O 2 , ADP and probably substrate concentrations in skeletal muscle (Chapter 2). In the frog, Neobatrachus kunapalari, skeletal muscle represents 35% of its body mass and accounts for 60% of its standard V O 2 and 85% of its minimal V O 2 (Flanigan and Guppy 1997). Obviously, if 50% of this V O 2 was involved with the ATP futile proton leak, it would be hard to sustain the rest of cell functions that require ATP, let alone maintain chemical energy reserves to extend survival time. Is the proton leak decreased during hypoxic exposure? The answer is yes, by reducing the rate of state 4 mitochondrial respiration. This is achieved by decreasing the rate of substrate oxidation by the electron transport system, which decreases the mitochondrial membrane potential but the inner mitochondrial membrane proton conductance does not change (St-Pierre et al. 2000). The F1F0-ATPase, normally the ATP synthase, will run backwards consuming ATP, acting as an ATPase to pump H+ from the matrix to the intermembrane mitochondrial space if O2 is in limited supply, exactly the mitochondrial situation during hypoxia. This backward flux would tend to reestablish the membrane potential or proton motive force in the mitochondria at the expense of consuming potentially limiting intracellular ATP concentrations (St-Pierre et al. 2000b; Boutilier and St-Pierre 2000). During hypoxia the activity of the mitochondrial F1F0-ATPase from amphibian skeletal muscle is inhibited. The inhibition of the F1F0-ATPase also prevents its action as an ATPase during anoxic conditions. This allows the use of ATP generated by the mitochondria and anaerobic metabolism for other cellular functions in lieu of maintaining the membrane potential across the inner mitochondrial membrane. Inhibition of the F1F0-ATPase also decreases the proton leak by decreasing the membrane potential which drives the proton leak (St-Pierre et al. 2000a, 2000b; Boutilier and St-Pierre 2002). Membrane proton conductance does not appear to change with hypometabolism.

Amphibian mitochondria change upon hypoxic exposure. The hypoxia-induced hypometabolism in amphibians is reflected as reductions in the rates of state 3 (maximal) and state 4 (resting) rates of V O 2 of isolated mitochondria (St-Pierre et al. 2000c). This is a consequence of a reduction in the activity of the electron transport chain (St-Pierre et al.2000a; St-Pierre and Boutilier 2001). This reduction in mitochondrial resting and maximal rates of V O 2 would lead to (p.315) a decreased rate of aerobic ATP production by the mitochondria. This change in V O 2 did not occur in mitochondria isolated from frogs submerged in normoxic water. Hypoxia also induced a greater O2 affinity of the mitochondria (St-Pierre et al.2000c). This allows mitochondrial function at lower cellular P O 2 s, exactly the conditions associated with hypoxia. The stresses and rates of hypoxic hypometabolism are reflected in the intrinsic aerobic characteristics of the mitochondria, a reduced rate of metabolism and ability to function at reduced P O 2 .

Hypoxic submergence also decreases the enzyme activity in both the tricarboxylic acid (TCA) cycle and glycolysis (St-Pierre and Boutilier 2001). The activity of citrate synthase, a key regulatory enzyme of the TCA cycle, is decreased following hypoxic submergence. So not only are the activities of mitochondrial enzymes of the electron transport change decreased (St-Pierre et al. 2000), but also a regulatory enzyme of TCA cycle which supplies substrate to the electron transport chain has decreased activity with hypoxia. This will slow the rate at which ATP is generated aerobically. The activity of lactate dehydrogenase was also significantly lower in frog skeletal muscle during hypoxic submergence (St-Pierre and Boutilier 2001), which will reduce the rate of ATP synthesis via anaerobic metabolism and will limit the synthesis of its by-product lactic acid. The decreased lactate dehydrogenase activity would favor oxidative utilization of the substrate pyruvate where the yield of ATP/glucose molecule is obviously greater. The increased efficiency would slow the rate of depletion of glycogen stores and limit the accumulation of a metabolic acid. Amphibians have limited buffering capacity to deal with metabolic acids in comparison to reptilian hypoxic hibernators (Stewart et al. 2004).

Another source of metabolic savings during aquatic hibernation is a decrease in the ATP spent maintaining the concentration differences of Na, K and Ca between the intracellular and extracellular compartments. The electrical potential across the cell membrane is a consequence of the diffusional efflux of K from cells and a smaller influx of Na. The diffusional flux, creating the membrane potential, is based on both the conductance for these ions and the concentration differences for these ions across the cell membrane. The maintenance of homeostatic sodium and potassium concentrations is dependent upon the Na-K-ATPase and the energy required is provided by oxidative metabolism (Sick and Kreisman 1981). Ions would normally not move through the non-ionic lipid bilayer that makes up the cell membrane because they are charged. The conductance, or permeability, for the ions is determined by protein channels specific for each ion. The concentration differences are established and maintained by the Na-K ATPase for Na and K, and by Ca-ATPase for Ca. If the diffusional flux of these ions is decreased, a decreased expenditure of ATP will be needed to move the ions back against their concentration differences and metabolism can be reduced. There are two potential avenues for decreasing the diffusional flux of these ions during aquatic hibernation and a developing hypoxia—decreasing the concentration differences and/or decreasing the conductance. There is strong evidence that (p.316) both mechanisms are involved with the development of hypometabolic states as organs are exposed to hypoxic conditions. The first is decreasing the concentration difference of Na and K. It is clear from early studies on amphibians that simple exposure of amphibians to low temperature led to a decrease in extracellular Na concentration, principally the result of osmotic water uptake that was not excreted by the kidneys (Jorgensen 1950; Jorgensen et al. 1978; Miller et al. 1968). It was also established that lowered temperature also leads to the loss of K from skeletal muscle (Miller et al. 1968), but the metabolic significance of these effects was not discussed. The concentration difference across the cell membrane for both Na and K will be decreased, which will decrease the rate of diffusional flux of each of these ions, decreasing the need for ATP expenditure to pump the diffusing ions back against their concentration difference. More detailed, recent work found essentially the same result for Na and K in skeletal muscle during both normoxic and hypoxic (8 kPa) submersion (Donohoe et al. 2000). Interestingly, cardiac muscle maintained its intracellular K concentration, implying that maintenance of the K concentration difference might be important for muscle that still must maintain the same activity to convectively move the blood and its limited O2 content. The other potential mechanism to decrease the normal diffusional flux of these ions is to decrease the number of ion channels present or open in the membrane for these ions. This is referred to as channel suppression or channel arrest. Donohoe and his colleagues present evidence that both K and Na conductance as well as Na-K-ATPase activity is decreased following submersion in hypoxic conditions (Donohoe et al. 2000). It seems clear from these data that a contributing element of the hypometabolism of skeletal muscle associated with normoxic and hypoxic submergence results from a decreased expenditure of ATP via the Na-K-ATPases responsible for maintaining the ion concentration differences between the extracellular and intracellular compartments of skeletal muscles.

A summary diagram of the array of cellular and biochemical changes associated with hypoxic submergence in amphibian skeletal muscle that are involved in both the hypometabolic response and allow extended life at low temperature is presented in Fig. 4.12.

The effects of hypoxia and anoxia on organ system function extend beyond relating them to environmental physiology but also their use as experimental models for understanding stroke and ischemia in organ function. Obviously, during a stroke the brain becomes hypoxic, just as regions of cardiac muscle becomes hypoxic/ anoxic following coronary artery blockage. Both processes are major sources of mortality for humans, and the organ stress and mechanism of injury are the same for frogs exposed to hypoxic submergence. The advantage that ectothermic organ system models present to biomedical researchers is that their tissues rate of metabolism is very much lower, hence the processes of dysfunction are spread over a longer period of time. This allows a more detailed analysis of the (p.317) mechanisms involved in compensation. It is referred to as “slow death”. The two most popular models are amphibian brains and their hearts. These experiments are generally run at room temperature so their application to understanding the processes associated with hypoxic submergence at 3°C is problematic.

                   Physiological Adaptations to Extreme Environments

Fig. 4.12 A summary of the cellular and biochemical adaptations of amphibian skeletal muscle with hypoxic submergence at low temperature (see text).

The amphibian heart has a remarkable capacity to maintain function and recover from hypoxic exposure. Isolated cardiomyoctes at 3°C from hypoxic acclimatized frogs have a much lower ATP turnover when exposed to hypoxia, indicative of a profound, yet reversible, metabolic depression compared to cardiomyoctes from control frogs (Currie and Boutilier 2001). Isolated toad hearts, perfused at 21°C with hypoxic buffer (<2 kPa) for 30 min at physiological filling and diastolic afterload pressures, ceased generating physiological outputs, but upon reoygenation 50–90% of their function returned (McKean et al. 1997). Interestingly, when exposure to hypoxia occurred under conditions of lowered preload and afterload the hearts continued to maintain a cardiac output and actually showed enhanced outputs upon reoxygenation. The data indicate that toad hearts will continue to maintain cardiac output under conditions of lowered demand and they recover from hypoxia quite well (McKean et al. 1997). The arterial pressure of bullfrogs declines with temperature (Rocha and Branco 1997, (p.318) 1998). Consequently, it may be reasonable to infer the normal cardiovascular conditions for hypoxic submergence would entail a lowered diastolic afterload on the heart, exactly the conditions the in vitro heart maintained cardiac output. Just as we saw with skeletal muscle, hypometabolism is a key to balancing reduced O2availability to support some cardiac output during hypoxic exposure. Future work looking at the regulation of preload and afterload during hypoxic submergence at low temperature would certainly prove interesting as they relate directly to controlling the work, hence metabolic need, of the heart.

The amphibian brain also can function and recover following over 30 h of anoxia at 5°C (Hermes-Lima and Storey 1996). The amphibian brain is dependent upon the Na-K-ATPase to maintain the gradients of Na and K by oxidative metabolism (Sick and Kreisman 1981). The amphibian brain can withstand exposure to anoxia at 25°C for 3–5 h (Lutz and Reiners 1997; Knickerbocker and Lutz 2001) before significant quantities of K starts to escape and Ca starts to enter. It appears that even with significant Ca entry, over 70% of brain cells survived when measured 12–16 h post oxygenation (Hedrick et al. 2005). This suggests that the mammalian brain model, which ascribes the increase in intracellular Ca as activating irreversible proteolytic activity, may not apply to amphibians. The mechanism for these differences is unclear. Anoxia decreases cellular ATP levels and induces a loss of intracellular K (Knickerbocker and Lutz 2001). This is followed by a sharp increase in K loss with the release of the neurotransmitters dopamine, aspartate, taurine and GABA, and a decrease in glutamate (Lutz and Reiners 1997; Milton et al. 2003). The decrease in glutamate with anoxia is completely the opposite of the mammalian brain response to anoxia and has been implicated in the greater hypoxia tolerance of frog brains in comparison to mammalian brains (Lutz and Reiners 1997). The increase in the excitotoxic transmitter dopamine separates the frog brain from the more anoxia-tolerant turtle brain (Milton et al. 2003). Increased perfusion of the frog brain during anoxia, in lieu of other organs, does not seem to contribute to the maintenance of central nervous system function during anoxia (Soderstrom-Lauritzen et al. 2001). The amphibian brain demonstrates a unique pattern of responses in comparison to more anoxia tolerant reptiles and less anoxia tolerant mammals. The brain is quite anoxia tolerant and seems to recover from short-term anoxia quite well. There are no data on functional correlates of nervous system function during and following hypoxia.

A problem that has not been addressed in aquatic hibernators under hypoxic, low temperature conditions is how kidneys continue to excrete water. Obviously, osmotic water entry will continue to occur under aquatic hypoxic conditions. The maintenance of kidney function is critical to prevent ion loss while maintaining water excretion, or the osmotic concentration will continue to become hypo-osmotic. The maintenance of filtration pressures and ion reabsorption under these hypoxic conditions is critical to maintaining a homeostatic osmotic internal environment and is essentially unstudied.

(p.319) 4.3.2 Shallow Terrestrial Hibernation and Freezing

If an amphibian over-winters in a terrestrial environment, the threat of hypoxia is gone, but the threat of freezing and dehydration “loom” on the osmotically hostile and more thermally extreme terrestrial “horizon”. Does winter also operate as a selective force for shallow terrestrial hibernators like hypoxic winterkill in aquatic ones? It is harder to visually discover evidence for death of amphibians burrowed under duff than in ponds. The only way to document the potential importance of temperature and dehydration as a source of mortality for terrestrial hibernators is via long-term studies of marked amphibian populations. By measuring winter survivorship over multiple years and then correlating rates of mortality to the severity of winter conditions, one can demonstrate the importance of winter weather to survivorship. A marvelous study of survivorship of both Rana esculenta and R. lessonae from Switzerland populations over 6 years provides unique insights into the evolutionary biology of terrestrial hibernation (Anholt et al. 2003). Both species are shallow terrestrial hibernators (Berger 1982). Over the 6-year study, winter survivorship varied from 10 to 80% for both species. The first principle component, accounting for about 60% of the variation in survivorship, included winter minimum temperature, average low temperature and its standard deviation. Other measures of average high temperature, average mean temperatures and their standard deviation loaded in a negative direction. A composite principal component measure of weather variables found that 90% of the variation in over-winter survival was accounted for in both species. Survival was lowest in winters with low minimum temperatures that were coupled to high and variable maximum temperatures (Anholt et al. 2003). The lowest survival was the winter with the lowest minimum low temperature (−12.9°C), implicating freezing as a selective factor. Why should high and variable temperatures during winter also negatively impact survival? There are a variety of potential explanations. High and variable body temperatures will increase metabolic expenditure via a Q10 mechanism, which will increase the rate of expenditure of energetic reserves (see above). High and variable air temperatures will encourage movement as both species have been described as moving in the field even when the temperature is close to 0°C (Holenweg and Reyer 2000). Warm temperature may encourage frogs to change hibernation sites and become trapped by the subsequent colder weather in sites that are both less insulated and more dehydration prone. The data do indicate that minimum temperature and energy reserves may both be significant sources of variation in over winter survival in shallow terrestrial environments. It is clear that winter weather conditions can be a strong selective element in the biology of terrestrial hibernators.


The physiological problems of shallow terrestrial hibernation principally derive from the stresses brought on by freezing. Data for an exposed site with little snow (p.320) cover to insulate the soil from the extremes in low air temperature common in the continental environment of South Dakota are given in Fig. 4.13 (from Swanson and Graves 1995) and provides an example of the potential for freezing. Average temperature for many months is below freezing at shallow depths, but is ameliorated by soil depth. The minimum temperatures at most depths down to a meter could lead to a freezing event for an amphibian in any winter month; again the severity decreases with depth. These data emphasize that environmental conditions leading to freezing of amphibians are a real threat. The addition of significant snow accumulation can create a microenvironment much closer to 0°C by minimizing convective exchange with air temperatures lower than the snow and creating a microenvironment with more infrared radiative input than the winter sky.

The physical variables involved in freezing derive from the interaction of water and solute. As solute is added to a solution there is an increase in the osmotic concentration, a freezing point depression, a lowering of the vapor pressure and an increase in the boiling point (these are the colligative properties of a solution). Consequently, the temperature at which fluid will freeze is related to its osmotic concentration. A 1 osmolal solution will freeze at −1.86°C compared to 0°C for pure water. Given that amphibian body fluids are 0.25 osmolal, we would expect them to freeze at 0.25 (−1.86) = −0.47°C. If an amphibian were to sequester additional solutes it would freeze at a lower temperature than −0.47°C. Solute accumulation provides some resistance to freezing but this is not significant when put in the perspective of the lowered ambient air temperatures possible during winter.

The problems of freezing are multifaceted. Water in solution will preferentially freeze first. This will concentrate the unfrozen solutes as water is removed from the liquid state by freezing and create a hyperosmotic stress for cells adjacent to the ice as the solutes are left behind. These cells will lose water and shrink, leading to high ionic strength inside the cells. As water is frozen there is a hypovolemic stress, since there is less liquid fluid to pump through the circulatory system as well as in (p.321) increase in hematocrit and blood viscosity, as the lymphatic fluid and plasma form ice crystals. This is reminiscent of what we delineated for dehydrational stress, with the major difference being that only a very low metabolism to support life is needed at these low temperatures. A third problem is mechanical/ structural damage associated with formation of the ice crystals. If the ice forms inside cells, then cell membranes can be ruptured and the differences in concentration of solutes from the intracellular and extracellular spaces will be lost. If the ice crystals form in extracellular spaces then they can mechanically rupture the vasculature and organs. Upon thawing, the extracellular space will now be hypo-osmotic to the cells. This can also lead to cell rupture as a consequence of the swelling resulting from water entering from the diluted extracellular fluid resulting from ice melting. Couple these issues to evaporative water loss in the terrestrial environment and you also introduce the additional physiological problems of dehydration that we discussed in Chapter 3. Imagine a partially frozen amphibian that is beginning to freezer burn as a consequence of the evaporative water loss, and it can be easily seen why freezing has been long thought to be a terminal process in vertebrates.

                   Physiological Adaptations to Extreme Environments

Fig. 4.13 A summary of (A) the mean and (B) minimum soil temperatures at various depths during the Fall and Winter in South Dakota (Swanson and Graves 1995).

Schmid (1982) delineated an amazing capacity of certain amphibians to tolerate freezing. Three species of terrestrial, hibernating northerly latitude frogs (Rana sylvatica, Hyla versicolor and H. crucifer) were capable of recovering from being frozen for 5–7 days at −6°C, environmental conditions that led to the death of two other northerly latitude species (R. pipiens, R. septentrionalis) that normally over-winter in water. He found that at least one species (H. versicolor) accumulated high concentrations of glycerol and inferred that this contributed to freeze tolerance (Schmid 1982). This observation set in motion a variety of subsequent studies concerning the physiology of and mechanisms involved in freeze tolerance in amphibians. Not only is the information of interest to environmental physiologists, but it may also delineate mechanisms that could be important in the cryopreservation of vertebrate organs for future use, a process of significant medical interest but currently not possible.

The pattern of an amphibian’s body temperature during a freezing event (Fig. 4.14) can be described as an initial decrease in body temperature, the rate of which is determined by a combination of the local thermal environmental, the surface area and body mass of the animal. The nominal freezing should take place at the temperature equivalent to the osmotic concentration of the animal. In this example, and for amphibians in general, the actual freezing occurs at a lower temperature than is predicated from the colligative properties of its body fluids. This additional decrease in temperature before freezing occurs is known as supercooling. Eventually some of the animal’s body water begins to freeze, and with this phase transition from liquid to solid, heat energy is released. This temperature is known as the crystallization temperature (when ice begins to crystallize). The heat energy released raises the body temperature of the animal, and this temperature increase is identified as the start of the exotherm. As more and more (p.322) water is frozen there is continued release of heat and a continued elevated body temperature. Eventually, an equilibrium temperature is achieved when the water has stopped freezing. The integral of the exotherm represents a rough estimate of the total amount of body water that has frozen.

                   Physiological Adaptations to Extreme Environments

Fig. 4.14 A schematic of an amphibian’s body temperature during a freezing episode; see text for explanation of the terms.

The cryopreservation literature gives some insight into the environmental and physical variables that increase cell survival from a freezing episode. As we saw above, osmotic volume stresses are placed on the cell both during the freezing phase (decrease in cell volume) and the thawing phase (increase in cell volume). This translates into mechanical stresses which can rupture the cell membrane (Pegg and Diaper 1988; Muldrew and McGann 1994). During freezing the concentration of the solutes increases as water is removed to form ice. The stress to red cell function was inferred to be disruption of protein structure and function as a consequence of the increased ionic strength of the concentrating solutes (Lovelock 1953a). The addition of glycerol ameliorated the effect of freezing damage, and the mechanism for this protection was interpreted as a colligative dilution of the ions (Lovelock 1953b). An alternative mechanism was proposed (Mazur et al. 1981) in which aqueous intercellular channels grow narrower as a consequence of being frozen, leaving less space for the cells and subjecting them to shear and compression forces. This is reflected functionally as the greater the fraction of unfrozen water the greater is cell survival (Mazur et al. 1981). Recent evidence seems to indicate that osmotic injury leads to cell membrane rupture which allows propagation of extracellular ice crystals to propagate into the intracellular space (Muldrew and McGann 1990, 1994). What is clear from empirical evidence is that freezing of the intracellular space compromises cell survival. The key to enhancing survival is to have freezing occur at a rate that allows the intracellular and extracellular environments to achieve an osmotic equilibrium as water is being removed during the formation of ice in the extracellular fluid. This obviously depends on the rate of water and solute flux across the cell membrane. If the rate of cooling is too fast, then osmotic equilibrium will not be achieved, the likelihood of membrane (p.323) damage increases, and the probability of intracellular ice formation increases. At any non-equilibrium freezing rate, the degree of supercooling increases the probability of intracellular ice formation. Consequently, extensive supercooling then freezing increases the risk of intracellular ice formation and has a negative effect on cell survival.

                   Physiological Adaptations to Extreme Environments

Fig. 4.15 The crystallization temperature when ice forms for a variety of amphibians. The filled squares are species with well developed freeze tolerance, the unfilled squares are species that will survive a short freezing exposure and the triangles are species that have not been demonstrated to withstand any freezing. Data from Schmid 1982; MacArthur and Dandy 1982; Storey and Storey 1986; Layne and Lee 1989; Constanzo et al. 1992; Layne 1995; Swanson and Graves 1995; Churchill and Storey 1996; Swanson et al. 1996; Irwin et al. 1999; Croes and Thomas 2000; Voituron et al. 2003; Voituron et al. 2005.


The extent of supercooling and the tissue freezing (crystallization) temperature of amphibians varies from −1 to −4.3°C. It does not appear to differ between freeze tolerant and intolerant species (Fig. 4.15). The degree of supercooling is consistent with other vertebrates but less than for most insects (Lee and Constanzo 1998). Consequently, the extent of supercooling is modest and conveys limited protection to low environmental temperatures, but does not appear to contribute to the ability to tolerate freezing in amphibians.

The variation observed in crystallization temperature within and between species probably results from the experimental conditions under which the crystallization temperature was determined. Water will supercool but once an ice crystal begins to form it serves as a nucleation site to propagate further ice crystal formation. The initial formation of an ice crystal can result from spontaneous aggregation of water molecules or be initiated by another substance (Lee and Constanzo 1998). (p.324) Amphibian plasma has agents that serve as ice-nucleating agents, which prevent extensive supercooling (Wolanczyk et al. 1990). The problem is that the crystallization temperature of the plasma and its ice-nucleating agents is around −8°C, obviously much lower than the experimentally determined TCof all amphibians to date (Fig. 4.15). What then serves as the nucleator for ice crystal formation and prevents extensive supercooling down to the temperature for plasma ice-nucleation? There are a couple of nucleation site possibilities, one internal the other external. The intestinal tract and skin of amphibians contains bacteria, which nucleate ice crystal formation in supercooled water at −3°C (Lee et al. 1995). These are the two organs that supercool the least, and are closest to the whole animal crystallization temperature in comparison to the other organs of the body (Layne 1995). A possible external nucleation site is ice in the local environment with which the skin of the frog is in contact. To prevent dehydration, over-wintering amphibians will select moist substrates. Water in the interstices of soil and leaves has a lower osmotic concentration than an amphibian’s body fluids, and hence freezes first. As frost forms on the surface of the hibernation site, these ice crystals could serve as nucleators across the permeable amphibian skin. External ice will trigger freezing in frogs at temperatures of −0.5 to −0.8°C, which is warmer than their normal supercooling temperature of −1.8°C (Layne 1991; Constanzo et al. 1999). Crystallization temperatures are consistently higher when determined on a wet substrate in comparison to a dry substrate (Swanson et al. 1996). It seems clear that the environmental conditions normally associated with a hibernation site can significantly raise the crystallization temperature; therefore the determination of crystallization temperature in a dry laboratory setting will have limited ecological relevance to an animal in a normally moist hibernation site.

The combination of a supercooling episode followed by a freezing episode can be lethal for amphibians. The western chorus frog (Pseudacris triseriata) shows 100% survival from freezing following supercooling down to −2°C, 60% survival from −2 to −3°C, 25% survival from −3 to −4°C and no survival if the freezing followed −4 to −6°C supercooling (Swanson et al. 1996). Clearly, natural selection should operate to favor behavioral selection of environments and/ or physiological mechanisms that would raise crystallization temperatures and minimize the degree of supercooling for amphibians that tolerate freezing of their body fluids. The permeable integument of amphibians clearly is an advantage from the perspective of allowing external ice to serve as a nucleator and preventing extensive supercooling (Constanzo et al. 1999).

Pattern and Magnitude of Frozen Body Water

The sequence of ice propagation throughout the body has been followed during freezing and thawing with proton magnetic resonance imaging (MRI) of Rana sylvatica (Rubinsky et al. 1994). The first region to freeze was adjacent to the (p.325) region of skin contacting the chamber, consistent with environmental nucleation. The progression of ice crystal formation was from the exterior to the interior, at first in the subcutaneous lymphatic sacs and pleuroperitoneal cavity. The extracellular fluids (lymph and interstitial) were the first to freeze, and the last to freeze were the major vessels and liver. Thawing was a more uniform process with the core organs melting first (Rubinsky et al. 1994).

The pattern of organ dehydration during freezing for two freeze-tolerant species and four species with low freeze tolerance (Fig. 4.16), shows that the freeze tolerant species have greater water loss from the organs than freeze-intolerant species. This would suggest that the freeze tolerant species are more likely to equilibrate with the hyperosmolality induced by ice formation than the intolerant species. This loss of water and concentration of solute would decrease the likelihood of intracellular ice formation.

The amount of total body water that freezes seems to be an important aspect of surviving freezing. The fraction of body water frozen can be measured by calorimetry. The ice content is influenced by the plasma osmolality (Layne 1999), although the influence of osmolality is less than would be predicted from the colligative effects alone (Claussen and Constanzo 1990). Freeze tolerance in the three water frog species of the Rana esculenta complex is related to the ice content with freezing; above a threshold fraction of frozen body water (50–58%) there was no survival of freezing (Voituron et al. 2003, 2005). This is consistent of low (p.326) fractions of frozen water (35–46%) in the treefrogs Hyla versicolor and H. crucifer (Schmid 1982; Storey and Storey 1985). However, the model study frog for freeze tolerance, R. sylvatica, seems to be able to tolerate a greater fraction (65%) of its body water being frozen (Layne and Lee 1987; Constanzo et al. 1993).

                   Physiological Adaptations to Extreme Environments

Fig. 4.16 The pattern of tissue water content upon freezing in two freeze-tolerant species (line connected) and four freeze-intolerant species: (A) muscle; (B) liver; (C) heart; (D) brain. Data are from Constanzo et al. 1993; Lee et al. 1992; Layne and Jones 2001; Irwin and Lee 2003; Voituron et al. 2003; Voituron et al. 2005.


A cryoprotectant is a chemical that enhances physiological function and/or increases survival during freezing. Cryoprotectants are usually divided into two broad categories, colligative and membrane stabilizers (see Storey 1997). Colligative cryoprotectants are those that reach high enough concentrations that they can significantly depress the freezing point and reduce the amount of ice formed during freezing. Membrane stabilizing cryoprotectants are generally in lower concentrations and prevent liquid–gel transitions of the cell membranes. The three chemicals invoked as being cryoprotectants in amphibians are glucose, glycerol and urea. All are important colligative cryoprotectants. Their role and the role of other possible membrane stabilizing cryoprotectants is essentially unstudied for amphibians.

Glucose is the most studied of amphibian cryoprotectants, especially for R. sylvatica. The freeze-tolerant R. sylvatica has six times the liver glycogen and 12 times the glycogen phosphylase activity compared to the non-freeze tolerant R. pipiens (Mommsen and Storey 1992). The synthesis of glucose has both a short and long-term regulatory component initiated at the start of freezing. At the initiation of ice formation an inactive form of the enzyme responsible for breaking down glycogen to form glucose-1-phosphate, glycogen phosphorylase a, is activated (Storey et al. 1996). At the same time there is a longer-term increased synthesis of glycogen phosphorylase enzymes (Storey and Storey 1988). This conversion to the active form (minutes) and upregulation of synthesis of phosphorylase (hours) leads to a sustained increase in the formation of glucose-1-phosphate. The glucose-1- phosphate is converted to glucose-6-phosphate, which has the potential to move down a variety of metabolic pathways. Inhibition of the glycolytic pathway is correlated with the inhibition of 6-phosphofructo-1-kinase and 6-phosphofructo-2-kinase (Storey 1987; Vasquez-Illanes and Storey 1993). There is also a decrease in activity of the enzyme glycogen synthetase, responsible for synthesis of glycogen from glucose (Russel and Storey 1995). This glucose mobilization process is not initiated until ice begins to form, so it is not activated in an anticipatory fashion, based on maintenance at low temperature (Storey and Storey 1985; Storey and Storey 1986a, b). How the ice starts this regulatory cascade is not understood but it seems to be mediated via the sympathetic nervous system since it can be stopped with β-blockers (Storey and Storey 1996). The mediation via the sympathetic nervous system is consistent with the more general amphibian stress response mediated by catecholamines released by the sympathetic nervous system. The glucose mobilized in the liver is distributed to the other organs of the body via the (p.327) circulation (Storey 1987; Storey and Storey 1986). To escape the liver cells and enter the cells of other organs it has to traverse the cell membranes of the organs. This is achieved by an increase in the number and/or activity of membrane glucose transporters in the freeze tolerant R. sylvatica compared to the less tolerant R. pipiens (King et al. 1993, 1995). A summary of the various physiological and biochemical events associated with glucose mobilization during a freezing episode in R. sylvatica are delineated in Fig. 4.17.

The potential advantages of glucose as a “cryoprotectant” are numerous. Glycogen (a glucose polymer) is an important storage form of chemical energy, hence glucose is readily mobilized requiring no anabolic reactions. Glucose is the principal substrate that can be used to supply anaerobic metabolism in (p.328) amphibians. As we will see below with the arrest of circulation upon freezing, the only way that O2 can move in a frozen frog is by diffusion, hence aerobic metabolism is essentially stopped and cellular ATP needs have to be supplied anaerobically (Layne et al. 1989). Glucose will raise the osmotic concentration of the cell, limiting the decrease in cell volume associated with the removal of water from the intracellular fluid with ice formation. Finally there is some evidence that glucose can serve to stabilize cell and organ function during the increase in osmotic concentration associated with freezing. There is an inverse relationship between the severity of freezing injury and the quantity of exogenous glucose added in rapidly cooled frogs, a process where endogenous glucose mobilization is compromised (Constanzo et al. 1991). Exogenously administered glucose did not influence the degree of tissue water lost during freezing but did reduce the body ice content and red cell injury (Constanzo et al. 1993) as well as the length of time frogs could survive freezing (Layne et al. 1998). The magnitude of the hyperglycemia upon freezing in ranid frogs is correlated with the tolerance of freezing; R. sylvatica shows the greatest response and other ranids accumulate less than 20 mmol/l in their plasma. Exogenously administered glucose did not make a normally freeze intolerant species, Bufo paracnemius, cryoresistant (Steiner et al. 2000). It is clear that glucose is acting as a cryoprotectant from both a metabolic and colligative perspective.

                   Physiological Adaptations to Extreme Environments

Fig. 4.17 A summary of the various physiological and biochemical events associated with glucose mobilization during a freezing event in Rana sylvatica (see text).

Glycerol is another putative cryoprotectant produced by amphibians (Schmid 1982). The concentration of glycerol has been reported to increase in a variety of amphibians, but to the greatest degree and only significant colligative way in the gray treefrog, Hyla versicolor (Schmid 1982; Storey 1984; Storey and Storey 1985, 1986, 1987; Layne 1999; Irwin et al. 1999; Croes and Thomas 2000; Layne and Jones 2001). Glycerol concentrations of about 0.3 M have been measured for H. versicolor in both fall and late winter (Schmid 1982). The concentration increase upon freezing seems to be better developed in adult compared to immature H. versicolor (Storey and Storey 1985). The pattern of increase and removal suggests that it may be less mobile than glucose. Some data to suggest it is constitutively synthesized, as well as being stimulated during freezing (Schmid 1982; Irwin and Lee 2003; Zimmerman et al. 2007). The kidneys increase their fractional reabsorption of glycerol with cold acclimation (Zimmerman et al. 2007). There do not appear to be differences in glycerol synthetic capacity in colder climates for H. versicolor (Irwin and Lee 2003).

Urea has recently been described as a cryoprotectant for amphibians (Constanzo and Lee 2005). Amphibians commonly use urea to balance their internal osmotic concentration with that of the environment see above). In this manner water can be osmotically taken up from the environment. The freeze-tolerant wood frog (Rana sylvatica) enclosed in outdoor enclosures accumulated up to 65 mmol/l urea in the fall, and accumulated up to 90 mmol/l in dry laboratory conditions (Constanzo and Lee 2005). During freezing, the concentrations of urea and glucose were (p.329) similar, even though glucose was synthesized during the freezing, and urea was synthesized before the freeze. This suggested that the cryoprotectant properties of urea should be explored. Urea was comparable to glycerol, and as good or better than glucose, in preventing freeze–thaw damage to erythrocytes (Constanzo and Lee 2005). Urea is suited for a role in cryoprotection as it is the principal nitrogenous waste of terrestrial amphibians, equilibrates quickly between the intracellular and extracellular spaces, and is broadly used as an organic osmolyte during osmotic stress. The physiological role of urea as a solute to ameliorate osmotic stress in terrestrial hibernation sites has already been demonstrated in most amphibians studied to date. Given urea’s osmotic role for amphibians, it is preadaptated to function as a colligative cryoprotectant. Urea may also be a contributor to metabolic depression (Constanzo and Lee 2005). The concentrations of urea, glycerol and glucose reported in the plasma of freeze-tolerant amphibians, R. sylvatica and H. versicolor, subjected to freezing during fall and winter are summarized in Fig. 4.18. The concentration of the principal cryoprotectant for each species is quite variable, probably reflective of synthetic rate variation, rate of freezing, glycogen storage and time of sampling.

Physiology of Freezing

The process of freezing has important implications for organ system function. The cardiovascular system plays a key role in the distribution of cryoprotectant synthesized and released from the liver in conjunction with its convective gas transport role. As we saw, with ice formation water is removed, thereby increasing the osmotic concentration of the remaining unfrozen extracellular fluid and creating a hyperosmotic environment for the cells. As ice forms in the extracellular space there is less lymph and plasma to fill the circulatory system, creating a hypovolemic stress on the cardiovascular system, which would decrease cardiac output via a Frank–Starling mechanism. As plasma volume is decreased, the red cell fraction (hematocrit) will increase, raising the blood viscosity. Combined with (p.330) the increase in viscosity generated by a decreased temperature and the presence of ice, the hyperviscous stress is magnified; the end result of these perturbations is a decrease in blood flow. Since glucose is not produced until ice has started to form, the cardiovascular system has to function under these hyperosmotic, hypovolemic and hyperviscous stresses. Couple this to a low heart rate of four beats per minute and the delivery of cryoprotectant to peripheral tissues will be a slow process. The release of the heat energy of fusion raising the body temperature as ice forms (exotherm) actually may assist the process of cryprotectant distribution via a Q10contribution to an increase in heart rate to 8 beats per minute, doubling the blood flow as the body temperature increases from its supercooled point (Layne et al. 1989). Also contributing to the increased heart rate will be sympathetic stimulation. The synthesis of glucose is mediated via sympathetic stimulation (Storey and Storey 1996). Sympathetic stimulation would also increase heart rate, increase the strength of cardiac contraction and mobilize blood via venoconstriction. As freezing reaches its equilibrium point the heart stops beating (Layne et al. 1989). Cardiac contractility is the first physiological function to return upon thawing (Layne and First 1991) and is the least affected organ system by freeze duration (Layne et al. 1998). Interestingly, isolated frozen ventricular strips incubated in glucose from the freeze-tolerant R. sylvatica rapidly regain contractility after thawing (Canty et al. 1986). Ventricular strips from the mildly freeze-tolerant frog R. pipiens not only do no regain contractility after freezing but lose contractility when exposed to glucose equivalent in concentration to cryprotectant concentrations (Canty et al. 1986). The mechanism for this difference is currently not understood, but it suggests that the ability to function at low temperature and recover function upon thawing is more complicated than simply the synthesis and accumulation of cryoprotectant.

                   Physiological Adaptations to Extreme Environments

Fig. 4.18 The plasma concentration of glucose, glycogen and urea in R. sylvatica and H. versicolor during a freezing event. Data from summary of Constanzo and Lee 2005; Layne and Jones 2001; Storey and Storey 1985; Layne 1999.

There are lower survival rates with rapid freezing in comparison to slow freezing (Constanzo et al. 1991). One of the dangers of a rapid freezing event is that it may not allow enough time for the synthesis and distribution of cryoprotectant at these low temperatures (Constanzo et al. 1992). Given a natural hibernation site with reasonable leaf litter and snow cover, the rate of cooling during a freezing event will probably be slow.

The metabolic effect of the freezing-induced circulatory arrest is that the cells have to operate anaerobically. Two things ameliorate the dangers of being forced to operate solely anaerobically: (1) metabolic rates are extremely low at these subzero temperatures; and (2) the principal substrate for anaerobic metabolism, glucose, has been distributed as a cryoprotectant so glucose exists in very high intracellular concentrations. Not all tissues accumulate glucose to the same degree: the organs accumulating the most are liver, heart, brain, kidney and leg muscles accumulate the least (Storey 1987). The energy charge within cells goes down during freezing and lactate begins to accumulate most dramatically in brain, heart and kidney (Storey and Storey 1984, 1985; Storey 1987). There is some evidence for alanine (p.331) accumulation in leg muscles and skin, indicating that anaerobic breakdown of muscle amino acids (aspartate, glutamate and glutamine) during freezing may be more important than glucose breakdown in skeletal muscle during a freezing event (Storey 1987; Hemmings and Storey 2001).

The freeze-tolerant frog R. sylvatica has different responses of its neural and muscular function in comparison to other non-freeze tolerant amphibians. In particular, the structure and function of the Ca-ATPase from isolated skeletal muscle sarcoplasmic reticulum is temperature insensitive and has seven unique amino acid substitutions in comparison with a freeze intolerant congener (Dode et al. 2001; Hemmings and Storey 2001). This would allow normal rates of calcium cycling, and presumably muscle function at very low temperatures. The nervous system of R. sylvatica was also found to be less likely to show large, prolonged depolarization with accompanying uncontrolled nervous system activity in response to cold exposure when compared to freeze intolerant species (Dalo et al. 1995). The mechanisms of these adaptations remain to be fully delineated. It does suggest a suite of adaptations that would allow coordinated normal activity at low temperatures for R. sylvatica. It normally takes about 2 days post-thaw for R. sylvatica to return to normal behaviors and demonstrate normal behaviors (Constanzo et al. 1997; Layne et al 1998).

There has been much discussion of the similarities of freezing and dehydration tolerance. The corollaries are obvious in terms of hyperosmotic, hypovolemic and hyperviscous stresses, compromising cardiovascular function. Couple this to a stimulation of glucose mobilization that differs only in the magnitude of the increase in glucose concentration between dehydration and freeze stress, and the conclusion that freeze tolerance is an extension of adaptations involved in dehydration tolerance appears inevitable. Unfortunately, there are some strong distinctions in the behavior and osmotic response of organs between the two stresses. First, there are distinct differences in both the proteins synthesized and the rate of protein synthesis in response to either freezing or dehydration (Storey et al. 1997). Dehydration is a stress that can occur at any time of the year and is very rapid (hours to days). The key to escaping the environmental conditions leading to the negative water balance requires movement on the animal’s part to find water. Amphibians increase their activity when they start to dehydrate, presumably trying to find water (Heatwole and Newby 1972; Putnam and Hillman 1977). This requires a mobilization of metabolic capacity, not the shutdown of metabolism one sees with surviving a freezing event during winter. Freeze tolerance is part of an environmentally predictable, long-term low energy and low temperature winter adaptation, rather than an acute episode that can occur in any season. The overall osmotic response of organs graphically demonstrates this dichotomy. If we compare the magnitude of water loss from muscle, heart, liver and brain during a freezing event and a dehydrational event we see completely different responses in the behavior of these organs. Freezing leads to substantial water losses from heart (p.332) liver and brain, but little lost from the skeletal muscle (Fig. 4.19). This decreases the likelihood of intracellular freezing occurring in the vital organs. Dehydration on the other hand results in a large amount of water being lost from the skeletal muscle (the largest organ system) and the least from the crucial organs (brain, liver and cardiac muscle) involved in a coordinated metabolic response to move and find water (Fig. 4.19). It is important to recognize not only the similarities but also the fundamental physiological differences between these stresses. As Swanson et al. (1996) point out, the anuran families with the greatest dehydration tolerance (Pelobatidae and Bufonidae) are not tolerant of freezing at all.

                   Physiological Adaptations to Extreme Environments

Fig. 4.19 (A) The amount of organ water lost in R. sylvatica and H. versicolor during a freezing event compared to freeze intolerant frogs and (B) a comparison of R. sylvatica to data from seven species of amphibians at their critical activity point following dehydration (data taken from Fig. 3.17).

4.3.3 Deep Terrestrial Hibernation

The key to being a good deep, terrestrial hibernator is an ability stay below the frostline. The groups of amphibians that over-winter in this environment are the toads and many salamanders, which hibernate in deep hibernacula. There is strong evidence that toads will actively change depth in response to soil temperature changes (Breckenridge and Tester 1961; Tester and Breckenridge 1964a, b; (p.333) van Gelder et al. 1986), though not all stay below the frost line (Breckenridge and Tester 1961). Toads burrowed deeper to stay below the freezing point in a severe winter, while the toads were relatively inactive as the freezing point stayed well above their initial burrow depths in a milder winter. Salamanders will also respond to decreasing soil temperature by moving deeper (Vernberg 1953). There is some evidence that over-winter mortality is correlated with the depth of the soil frost line (Tester and Breckenridge 1964b; Kelleher and Tester 1969; Petranka 1979). Juvenile toads appear to be more susceptible to winter mortality, and this difference has been attributed to their hibernating at shallower depths, based on a reduced ability to dig compared to adults (Kelleher and Tester 1969). Populations of salamanders appear more vulnerable to winterkill in areas where their possible depth of soil penetration is limited during severe weather (Petranka 1979). Annual variation in the depth of the soil frost line is inversely related to the depth of snow cover (Tester and Breckenridge 1964b). Presumably the hypometabolic characteristics associated with aestivators are important for deep terrestrial hibernators as well (see above). Again, the environmental conditions of winter can act as a strong selective agent in deep terrestrial hibernators, just as for aquatic and shallow terrestrial hibernators.

4.4 Metabolic Depression

Metabolic depression is an adaptation whereby animals can survive with a metabolic rate below their normal minimal value (Hochachka and Guppy 1987; Storey 1988; Storey and Storey 1990; Guppy and Withers 1999; Withers and Cooper 2008). In ectotherms such as amphibians, metabolic depression is survival below the standard metabolic rate (SMR), which is the minimum resting rate of metabolism at any particular temperature (see Section 3.3). Metabolic depression occurs during periods of restricted resources, for example survival of short- or long-term anoxia (see Section 4.3), during extended periods of starvation such as hibernation (see Section 4.3) or aestivation (see Section 4.2). The adaptive significance of metabolic depression is to prolong survival until appropriate environmental conditions are restored. Metabolic depression has been reported in virtually all animal phyla, and the extent of depression can vary from moderate at about 60–80% of SMR, to commonly 20–25% of SMR, and in extreme examples to <1% of SMR (Guppy and Withers 1999). Metabolic depression in amphibians is generally to 20–25% of SMR.

Amphibians occur in a surprisingly wide range of habitats given their thin, permeable skin, high rates of evaporative water loss, poor tolerance of high temperatures, and susceptibility to freezing. Behavioral avoidance rather than physiological adaptation is often the first line of defense for amphibians against adverse environmental conditions. Dormancy, or inactivity, is a common response (p.334) of amphibians to cold or desiccating conditions. It is sometimes termed hibernation or brumation in winter cold, and aestivation or retraherence in summer heat and dryness. Behavioral avoidance in particularly harsh environments is associated with long periods of inactivity interspersed with short periods of intense feeding and reproduction. For example, the annual temperature cycle of subpolar environments can restrict amphibians to a short spring/ summer activity period. In desert environments, activity may be restricted to short periods after rainfall, which in extreme cases is less frequent than every 12 months.

Metabolic depression to below standard metabolic rate is an important physiological adaptation for survival of such extended periods of inactivity, particularly in desert environments where the inactivity period can extend for a year or more, and ambient temperature is high so standard metabolic rate remains high. The adaptive significance of metabolic depression is self-evident. For every decrement in metabolic rate there is an equivalent (reciprocal) increment in survival time based on endogenous energy stores; a reduction to 90% of SMR results in a 111% survival time; reduction to 50% a 200% survival time; 20% of SMR a 500% survival time.

4.4.1 Aestivation and Hibernation

Many arid-zone amphibians burrow underground or seek refuge under bark or in tree hollows to escape the temperature and humidity extremes of an arid environment (Section 4.2.3). They remain dormant underground for extended periods of months to even years. The aestivation period is a well-defined physiological state of inactivity, especially for cocoon-forming species but also for non-cocooning species. For example, in natural conditions the water-holding frog Cyclorana platycephala can apparently survive cocooned underground for over 5 years if above-ground conditions are unfavorable, although an annual activity after heavy cyclonic rains is more typical (van Beurden 1980). For non-cocooning fossorial frogs such as Scaphiopus and Notaden, the period of survival underground is more likely less than annual, and even extended seasons of adverse environmental conditions might be interspersed with short periods of activity during favorable surface conditions (e.g. Notaden; Thompson et al. 2005). In the laboratory, C. platycephala can survive for over 18 months (Cartledge 2007), and 3 to 12 months is routine for many cocooning frogs and salamanders. Non-cocooning frogs survive in appropriate burrowed conditions for 3 to 6 months, for example Scaphiopus (Seymour 1973) and Notaden (McMaster 2007).

The steady-state metabolic rate of amphibians during aestivation is generally about 20–25% of the normal SMR (Withers 1993; Guppy and Withers 1999; Fig. 4.20). This level of depression is similar to that of aestivating lungfish (Protopterus aethiopicus; Delaney et al. 1974; Fishman et al. 1986), marbled swamp eel (Synbranchus marmoratus; Rosa 1977; Bicudo and Johansen 1979) and (p.335) salamander fish (Lepidogalaxias salamandroides; Pusey 1989), and some reptiles and invertebrates such as molluscs (Guppy and Withers 1999).

                   Physiological Adaptations to Extreme Environments

Fig. 4.20 Summary of resting metabolic rate (SMR; large black circles) and depressed metabolic rate during aestivation (large gray circles) for amphibians, with SMR for anuran amphibians (small open circles). Data from: Seymour (1973), Loveridge and Withers (1981), McClanahan et al. (1983), Etheridge (1990), Gatten et al. (1992), Withers (1993), Flanigan and Guppy (1997), Fuery et al. (1998), Flanigan et al. (1991), Withers and Thompson (2000), McMaster (2007).

The metabolic rate of amphibians declines relatively rapidly from SMR to a depressed level, generally taking 2–4 weeks from the onset of aestivation to plateau at about 20–25% of SMR (Fig. 4.21). After 4 weeks or so, the metabolic rate is relatively constant for weeks to months. This pattern of metabolic decline during aestivation is very similar to that of lungfish (Delaney et al. 1974; Fishman et al. 1986) and the marbled swamp eel (Rosa 1977; Bicudo and Johansen 1979). In lungfish, the decline in metabolic rate with onset of aestivation is like that for the onset of starvation, so induction of metabolic depression seems to involve the same cues and mechanisms. In amphibians, starvation also depresses metabolism but apparently not so dramatically; starved Scaphiopus have about 50% of the normal metabolic rate (Seymour 1973a).

The adaptive value of metabolic depression during aestivation is to extend the survival time using endogenous energy stores. The primary energy store and hence metabolic fuel during aestivation is lipid, as indicated by the respiratory exchange ratio and depletion of body fat reserves during aestivation (Jones 1980a; Loveridge and Withers 1981; McClanahan 1967; Seymour 1973; van Beurden 1980). Lipid is stored primarily in the abdominal fat bodies, and sometimes axillary and inguinal fat pads, as well as in the liver and other organs. The fat bodies can comprise 3–5% of the body mass prior to aestivation, and are considerably reduced after aestivation. For the water-holding frog Cyclorana platycephala, both small and large individuals have about 5% fat body mass, which would sustain depressed metabolic rate for over a year (van Beurden 1980; Cartledge 2007). (p.336) Loveridge and Withers (1981) estimated that 3% body fat stores would suffice for metabolic needs over 260 days of aestivation for the large (about 500 g) Pyxicephalus adspersus. For the smaller Scaphiopus couchii, lipid stores account for about 72% of their aestivation energy requirements and protein for about 23% (McClanahan 1967; Jones 1980). Large aestivating frogs are expected to have a longer survival time than smaller individuals/ species because of their lower mass-specific metabolic rate. van Beurden (1980) modeled energy store utilization of Cyclorana platycephala over extensive periods of aestivation. Some carbohydrate (e.g. liver glycogen) is used at the start of aestivation but the primary energy source during aestivation is the fat body, with more protein use towards the end of prolonged aestivation when the fat stores approach exhaustion. Survival is longer for large C. platycephala because of their lower mass-specific metabolic rate but equivalent mass-specific fat body stores (see Fig. 4.21).

                   Physiological Adaptations to Extreme Environments

Fig. 4.21 Decline in metabolic rate from 100% of SMR during entry into metabolic depression for a salamander (a, Siren intermedia, Etheridge 1986) and anurans (b, Neobatrachus centralis, Withers 1993; c, Scaphiopus couchii, Seymour 1973; d, Pyxicephalus adspersus, Loveridge and Withers 1981; e, Cyclorana maini, Withers 1993; Lepidobatrachus llanensis; f, McClanahan et al. 1983). Approximate survival time based on 5% initial body mass as abdominal fat bodies and calculated from average metabolic rate during metabolic depression for these amphibians (assuming 38 kJ/g, 20.1 ml/kJ).

Lipid metabolism during aestivation presumably follows the standard β-oxidation pathway of vertebrate cellular metabolism which requires additional input of C to prevent the depletion of TCA cycle intermediates. This additional C is generally provided by metabolism of pyruvate or phosphoenolpyruvate to oxaloacetate (Pinder et al. 1992). Although metabolism of lipid stores is often viewed as a concomitant “water store” because of the high metabolic water yield from fat metabolism (e.g. 146 moles H2O from metabolism of 1 mole of palmitoyl-CoA; Pinder et al. 1992), the total water yield of fat metabolism is less than (p.337) that of carbohydrate or protein metabolism because of its higher “bound” water content (see Chapter 2).

Protein metabolism provides some energy during aestivation, particularly towards the end of a long aestivation period (e.g. van Beurden 1980; Jones 1980) and results in the accumulation of often substantial concentrations of urea (see Section 4.2) particularly for cocoon-forming species that are sealed from the environment by their cocoon (except for gas exchange through their nares). Protein metabolism can even be accelerated during aestivation as the external soil water potential declines (McClanahan et al. 1972). During extended aestivation, urea accumulation can reflect 17–32% metabolism of body protein in Scaphiopus (McClanahan 1967; Jones 1980). This urea is apparently not accompanied by accumulation of counteracting solutes (Withers and Guppy 1996; see Section 4.2.2) and acts as a balancing osmolyte and ameliorates the adverse effects of high ionic concentration on enzyme function (Grundy and Storey 1994; Cowan and Storey 2002).

Behavior is the first key to surviving winter for amphibians (Section 4.3). For example, winter-acclimated Rana temporaria decreases its preferred temperature to 6.8°C in normoxia and 1.9°C in hypoxia (Tattersall and Boutilier 1997) compared to their normal thermal preference of about 29.6°C (Strübing 1954). Although hibernating amphibians are sometimes described as “torpid” (e.g. Holzapfel 1937; Willis et al. 1956; Pinder et al. 1992), it is not easy to determine whether this is simply a cold-induced lethargy or actually an altered metabolic state i.e. metabolic depression. Field and laboratory evidence suggests that overwintering frogs remain responsive and behavior is important in avoiding stresses such as hypoxia (see Tattersall and Boutilier 1997). Nevertheless, hibernating amphibians must cope with physiological effects of cold and starvation (Section 4.3). Similar strategies are used by both aquatic and terrestrial hibernators such as accumulation of lipid energy stores prior to hibernation, reliance on aerobic lipid metabolism (as in aestivating amphibians), and maintenance of cell membrane and protein function at low temperatures (Pinder et al. 1992).

In general, metabolic rate does not appear to decrease during hibernation other than from the standard physiological Q10 effect; the Q10 for biological processes such as metabolism is generally between about 2 and 3, and increases with decreasing temperature (Herbert and Jackson 1985; Ott et al. 1980; Withers 1992; Rome et al. 1992; see Section 3.5). The metabolic savings accruing from decreased temperature (e.g. from 22 to 2°C with Q10 decreasing from 4 to 2) is about 90% (Figure 4.22) which is similar to the savings accruing from metabolic depression during aestivation (about 80%) at a constant 22°C. Further metabolic depression after the thermal Q10 effect would provide a small, additional absolute energy savings but a substantial factorial reduction in MR, and hence an increase in survival time. However, the metabolic depression from decreased Ta and a Q10effect results in >100 days survival time for even a 1-g frog, >200 days for a 10-g frog, and even longer survival times for larger frogs (Fig. 4.22). These survival (p.338) times would seem more than adequate for survival of a cold winter period (e.g. 3 months), so further reduction in energy consumption by metabolic depression would appear unnecessary (and would result in relatively small absolute energy savings). This lack of necessity to further prolong survival time during hibernation is in contrast to the necessity for metabolic depression during aestivation, since survival times of even a year or more may be routinely required at often high ambient temperatures.

                   Physiological Adaptations to Extreme Environments

Fig. 4.22 Model of relative decline in metabolic rate of an amphibian with a decrease in body temperature from 22 to 2°C (assuming Q10 decreases linearly with Ta from 4 to 2; see text). Survival time is calculated from predicted standard metabolic rate (see Chapter 1.4) assuming 5% fat body stores and 38 kJ/g fat for frogs varying in mass from 1 to 1000 g.

Nevertheless, the decrease in metabolic demand for O2 during hibernation can be augmented by metabolic depression as well as an hypoxic induction of further hypometabolism beyond the Q10 effect (Donohoe et al. 1998; Section 4.3). For example, SMR of R. temporaria declined by 40% after 3 months of submergence at 3°C compared to prior to submersion at 3°C, and this hypometabolism was intensified with a 75% reduction from SMR if frogs were submerged in water made progressively hypoxic (Donohoe and Boutilier 1998). These 40 and 75% reductions in aerobic metabolism increase survival time by two and four fold.

4.4.2 Mechanisms of Metabolic Depression

It is not immediately clear how metabolic depression to below SMR is achieved. It could be caused by a reduced ATP supply that forces ATP-consuming processes (p.339) to be reduced, or ATP-consuming processes could be turned down, causing a reduction in the demand for ATP. The mechanisms may or may not be similar for intrinsic metabolic depression during aestivation and hypoxia-induced metabolic depression. Metabolic depression that accompanies aestivation occurs in the absence of any obvious cell signals, that is there is no change in temperature, no obvious cause of cell hypoxia and no change in body fluid solute concentrations. In hibernation, the decrease in temperature reduces metabolic rate through a Q10effect, and there may or may not be additional intrinsic metabolic depression. To understand possible mechanisms of metabolic depression, we first need to understand the processes that require O2 and ATP.

The partitioning of O2 consumption and ATP expenditure varies amongst organs, but skeletal muscle is a useful model for metabolic depression since it represents about 40% of the body mass in amphibians (Putnam 1979), and its metabolic rate conforms to tissue P O 2 (Boutilier 2001). In rat skeletal muscle, mitochondria consume most (about 86%) of the total O2 used, with about 14% consumed by various peroxisomal oxidases associated with fatty acid oxidation, endoplasmic reticulum oxidases, and the production of superoxide and hydrogen peroxide free radicals (Rolfe and Brand 1996). About 52% of total O2consumption is associated with the “proton leak” across the inner mitochondrial membrane, and 34% is coupled to the production of ATP. With respect to ATP-consuming processes, about 40% of the ATP is utilized by Na-K-ATPases, Ca-ATPases and myosin ATPases to maintain ion concentration differences between the intracellular and extracellular spaces, and 60% is used in the synthesis of protein, mRNA, glycogen, urea and everything else. Skeletal muscle of resting Neobatrachus centralis has a low protein synthesis rate that is unaffected by aestivation (Fuery et al. 1998a). The liver has a different partitioning of O2 and ATP with a larger synthetic role than skeletal muscle; protein synthesis is more than ten times that of skeletal muscle and accounts for about 38% of the resting cell O2 consumption (Fuery et al. 1998a). In a whole rat, estimates suggest that SMR is partitioned into about 8% non-mitochondrial and 20% proton leak from the skeletal muscle and liver alone, with about 20% for ATPases, 20% for protein synthesis and 32% for “other” processes (Rolfe and Brand 1996; Fig. 4.23). Whether this partitioning in rats applies equally well to an ectothermic amphibian or not, it is clear that to achieve a metabolic depression to 20% of SMR there must be a substantial decrease in non-mitochondrial and mitochondrial O2 consumption, the rate of ATP production and the extent of a proton leak, and that ATPase-dependent processes for both ion gradient maintenance and biosynthesis must be impacted. Adenylate energy charge is maintained during aestivation by Neobatrachus kunapalari (Flanigan and Guppy 1997), so ATP production has not been compromised relative to demand.

There is little scope for metabolic depression of the non-mitochondrial component of cellular O2 consumption because of its small overall contribution to (p.340) cell metabolism. Nevertheless, there are decreases in lipid peroxidation and antioxidant defenses in aestivating spadefoot toads Scaphiopus couchii (Grundy and Storey 1998).

                   Physiological Adaptations to Extreme Environments

Fig. 4.23 Hypothetical partitioning of whole animal O2 consumption into non-mitochondrial respiration and mitochondrial respiration, and mitochondrial respiration into the proton leak, ATPase activity, protein synthesis and “other” processes, at rest (based on data for rats; Rohlf and Brand 1996). The change in partitioning during metabolic depression is not known, but presumably involves reduction in all of the SMR components (see text).

There are four general characteristics of metabolically-depressed cells, at the molecular level; a decrease in pH, latency of mRNA but reduced protein turnover, changed protein phosphorylation and maintenance of membrane ion gradients (Guppy and Withers 1999). Reduced intracellular pH is often associated with metabolic depression but whether it is necessary for metabolic depression is equivocal (Guppy et al. 1994). Decreased protein synthesis is at the level of translation, as mRNA levels are generally stable during metabolic depression. Liver protein synthesis varies from 12% of the O2 consumption of liver in Bufo marinus to 38% in Neobatrachus centralis (Fuery et al. 1998a, b). Liver protein synthesis and O2 consumption are reduced in vitro during aestivation by 55% and 67% respectively in N. centralis, but this only contributes about 5% of the metabolic depression of the whole frog (Fuery et al. 1998). No in vitro metabolic depression or decrease in protein synthesis was observed for skeletal muscle of N. centralis, and the low protein turnover of skeletal muscle precludes any significant contribution to overall metabolic depression. In contrast, in Neobatrachus (p.341) pelobatoides there was a 70% decrease in skeletal muscle O2 consumption during aestivation, which accounted for about 70% of the frog’s metabolic depression (Flanigan et al. 1991), and in N. wilsmorei there was a 50% reduction in skeletal muscle in vitro (Flanigan et al. 1993). A mechanism for reduced protein synthesis during aestivation might involve EIF2α regulation of initiation of translation; the amount and extent of EIF2α phosphorylation increases during aestivation in Neobatrachus sutor (Pakay et al. 2003). Protein phosphorylation has been suggested to play a role in metabolic depression by reducing enzyme activity including protein synthesis. Phosphorylation of enzymes provides an important control of intermediary metabolism during aestivation in the spadefoot toad Scaphiopus couchii (Cowan and Storey 1999). There is a major reorganization of cell metabolism, with changes in activities of key enzymes and depression of signal transducing enzymes. However, there has been no direct quantification of the role of protein phosphorylation in metabolic depression. Ion gradients are consistently maintained during metabolic depression, even though the cost of maintaining ion gradients is thought to be high. Aestivating N. wilsmorei maintain ion gradients (Flanigan et al. 1993). In aestivating Neobatrachus kunapalari, ion gradients are also maintained, as are the levels of Na-K-ATPase (Flanigan and Guppy 1997). It seems that during metabolic depression “high priority” functions such as ion gradient maintenance are sustained (perhaps at a reduced cost?) whereas “low priority” functions such as protein synthesis are turned down.

In the aestivating frog Neobatrachus kunapalari, skeletal muscle represents 35% of its body mass and seems to account for 60% of its SMR and 85% of its depressed metabolic rate (Flanigan and Guppy 1997). Obviously, if 50% of this depression was involved with the futile proton leak, it would be hard to sustain the rest of the cell functions that require ATP, let alone maintain chemical energy reserves to extend survival time. Consequently, it is not surprising that the proton leak is decreased during hypoxic exposure by reducing state 4 mitochondrial respiration (St-Pierre et al. 2000).

We have already considered mechanisms for hypoxia-induced metabolic depression during hibernation with respect to the chemical energy budget at a cellular level (Fig. 4.12). There is a reduced activity of the electron transport chain (St-Pierre et al. 2000a; St-Pierre and Boutilier 2001), hence reduced aerobic ATP production by the mitochondria. Hypoxic submergence also decreases enzyme activities in both the TCA cycle and glycolysis (St-Pierre and Boutilier 2001). The activity of citrate synthase, a key regulatory enzyme of the TCA cycle, is decreased following hypoxic submergence. The activity of lactate dehydrogenase is also significantly lower in frog skeletal muscle during hypoxic submergence (St-Pierre and Boutilier 2001). Consequently, the rate of ATP synthesis via aerobic and anaerobic metabolism is reduced during hypoxia. (p.342)