Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

The Australian arid zone supports a diverse lizard and snake fauna that is significantly more species-rich than that of other arid regions in the world (Pianka 1986; Morton and James 1988; Powney et al. 2010). The precise origins of this fauna are, however, still the subject of research and some dispute. There is clear evidence for a Gondwanan origin for many components of the Australian fauna, e.g., monotremes, marsupials, passerine birds, parrots, ratites and chelid turtles, but the case for lizards and snakes is less clear. The arid zone was also not always arid, and it is only since the early Miocene that it started to take its present form (Martin 2006). A case for the Gondwanan origin of agamid lizards, with a 150-million-year divergence date between Southeast Asian and Australasian agamids based on mitochondrial DNA, was made by Schulte et al. (2003). The study was criticised, however, for methodological flaws, and a reanalysis of their data, along with the addition of nuclear (c-mos) sequences, pushed the date of this implied separation back to approximately 30 MYA (Hugall and Lee 2004). This would make an Asian origin for the Australian agamids by rafting and island hopping plausible, as the Australasian tectonic plate approached the Asian continent. Amongst the Australian agamids, two species stand out as being basal to the radiation and much older than other groups: the chameleon dragon, Chelosonia brunnea, and the mountain devil, Moloch horridus (Hugall et al. 2008), with the latter invading the developing arid zone long before any other agamid. The current consensus is in favour of an Asian origin for the agamids, dating to the early Miocene, which was followed by an extensive adaptive radiation of the group in the arid zone. There is one group of lizards, however, where a case for Gondwanan origin is compelling: the geckos and pygopodids (Oliver and Sanders 2009; Oliver and Bauer 2011). Molecular phylogenies indicate that at least five Australian lineages of diplodactyloid gecko, with wide distributions in the arid zone, are as old or older than elapid snakes and agamids. The snake-like pygopodids appear to have radiated before Australia was occupied by snakes, and recent research suggest that the Australian elapids have evolved from primitive sea kraits (Laticaudia) that may have swum the sea gap from Asia and come ashore in a snake-deficient continent (Scanlon and Lee 2004; Underwood 1957).

A strong case for an Asian, rather than Gondwanan origin, can also be made for the varanid lizards (Fuller et al. 1998). Molecular analyses reveal the existence of three monophyletic clades: an African, Indo-Asian and an Indo-Australian clade. Within the Indo-Australia clade, the endemic dwarf monitors (Odatria) form a clade sister to the large Australian monitors (the gouldii group) (Ast 2001). These studies thus indicate that the squamate fauna of the Australian arid zone is composed of both old Gondwanan (geckos and pygopodids) and more recent Asian elements (agamid lizards, skinks, elapid snakes and varanids).

Despite this wide diversity displayed in their phylogenetic origins, all of the reptiles inhabiting the Australian arid zone must contend with the same environmental contingencies that characterise arid environments worldwide, viz.:

  1. (a)

    High day temperatures through summer that must be tolerated, regulated, or avoided

  2. (b)

    A deficiency of available free water that renders difficult

  3. (c)

    The regulation of water and electrolyte balance

  4. (d)

    Seasonal deficiencies of nitrogen (protein), phosphorus and energy (carbohydrate) essential for growth and reproduction

Temperature Regulation

Following the pioneering study of Cowles and Bogert on North American desert lizards (Cowles and Bogert 1944), early publications on Australian lizards in the arid zone focused on thermoregulation. The concept of ‘ectotherms’ with an ‘eccritic’ or ‘preferred body temperature’ (PBT), resulting from active choice on the part of the animal in a thermal gradient, was established, and early surveys compared mean body temperatures of lizards active in the field (MBT) with their PBT determined in the laboratory (Licht et al. 1966; Heatwole 1970; Bradshaw and Main 1968). Although originally thought to be an unvarying and conservative feature of the thermal biology of different reptilian taxa, detailed field studies showed that the MBT actually represents a compromise between physiological preferences and ecological opportunities (Avery 1982; Cogger 1974). Various studies aimed at measuring the accuracy of thermoregulation of arid-zone species have shown, however, that this can be quite precise. A single specimen of Australia’s largest terrestrial reptile, the perentie, Varanus giganteus, which is widespread in the arid zone, was radio-tracked for a period of 7 days by King et al. (1989) on Barrow Island. The authors found that its body temperature during the day was relatively constant, averaging 35.8 °C when active, similar to that of other varanid species (Pianka 1986). The body temperature did not fall below 30 °C at night, however, due to the thermal buffering of its large body mass (8.8 kg).

Thermoregulatory precision in monitor lizards was studied by Christian and Weavers (1996), who developed an index of thermal exploitation (Ex), based on an earlier study on North American reptiles (Hertz et al. 1993). The Ex parameter is a ratio calculated by dividing the time that a lizard spends within its set-point range by the time available for the animal to exploit this temperature range and describes the thermoregulatory characteristics of ectotherms in a heterogeneous thermal environment. In their study of three species of varanids, Ex varied from 1.0 (perfect regulation) to negative (no thermoregulation) with the most active species, Varanus panoptes, being the ‘better’ regulator over all seasons. Varanus gouldii, a species whose range also extends into the arid zone, was a perfect regulator in the wet season of the year, but Ex became negative in the dry season, when all activity ceased and they remained in their burrows (Christian and Weavers 1996). A more recent attempt to correlate active body temperatures and microhabitat occupation in central Australian agamids is that of Melville and Schulte II (2001), with thermal factors playing a considerable role.

A critical question is whether arid-zone reptiles are ever forced to endure body temperatures significantly above their PBT, and, therefore, likely to induce stress responses leading to decreased fitness (Bradshaw 1997, 2017). An early analysis with a range of agamids of the genus Ctenophorus (then Amphibolurus) established that increasing aridity of the habitat was associated with MBTs significantly above the PBT, and this was attributed to the greater amount of time more arid-living species were forced to spend in high-temperature-avoidance behaviour patterns (Bradshaw 1988). This repetitive exposure to higher than ‘preferred’ body temperatures entrained an increased thermal resistance, measured as the critical thermal maximum (CTMax), which was lost after acclimation to laboratory conditions. Greer (1980, 1990) found that the CTMax of arid-living species of scincid lizards was also higher than that of mesic species, assuming these differences to be genetic, but the effect of acclimation was not investigated. Today, one also needs to consider the possibility of epigenetic effects (Hoppeler 2015). When considered in relation to the CTMax for each species, however, the highest body temperatures recorded in the field are invariably some 5–6 °C below the CTMax, and lizards (both in Australia and in North America) thus experience wide safety margins in the field (Huey 1982; Bradshaw 1988). Many reptiles commence ‘panting’ with a gaping mouth when apparently suffering from heat stress, and the panting threshold (PT) is positively correlated with the CTMax in 14 species of North American lizards (Whitfield and Livezey 1973). Panting thresholds for the few species of Australian lizards that have been studied are, however, well above the maximum body temperature ever recorded in the field (Stebbins and Barwick 1968; Heatwole 1976). Dehydration and increases in plasma osmolality, however, have a significant impact on the PT of the North American desert iguana (Dupré and Crawford 1985a) and also impact on its thermoregulation (Dupré and Crawford 1985b). Hypernatraemia (see glossary) also significantly depresses the PBT of the rock-living agamid, Ctenophorus ornatus, operating through the pituitary peptide hormone, arginine vasotocin (AVT) (Bradshaw et al. 2007).

So far as I am aware, there has been no study to date on the thermoregulation of any snake inhabiting the Australian arid zone. It is thus not possible to reach any conclusions regarding their thermoregulatory abilities and potential exposure to thermal stress. An early paper documented PBTs in a number of Australian snakes in the laboratory, ranging from 29.6 to 34.5 °C (Lillywhite 1980). Only one of these, the Western brown snake Pseudonaja nuchalis, is common in the arid zone, and its PBT was 34.0 ± 1.2 °C. Suggestions in the literature that the PBT of snakes is generically invariant (Rosen 1990) have been contested by careful studies involving telemetry. A 4-year study, for example, of the Australian blacksnake, Pseudechis porphyriacus (not an arid-zone species, however) showed excellent thermoregulatory capacities, maintaining a body temperature between 28 and 31 °C over a variety of seasons (Shine 1987). Available CTMax data for North American snakes also show a 6.8 °C differential when compared with maximum-recorded body temperatures in the field (Bradshaw 1988; Huey 1982), suggesting that thermal stress is never a problem. The most notable feature about snakes in the arid zone is their behavioural use of nocturnality as a means of avoiding high day temperatures (Greer 2000). Dehydration and hypernatraemia also depress the PBT by some 6–7 °C of the Western tiger snake, Notechis scutatus (also not an arid-zone species) operating also through AVT and lowering rates of evaporative water loss (Ladyman and Bradshaw 2003; Ladyman et al. 2003, 2006). There are some chelid species that occur in the arid zone (e.g. the flat-shelled turtle, Chelodina steindachneri), but nothing is known of their ecophysiology or how they survive for long periods in dry river beds (Kuchling 1999).

Water Turnover and Osmoregulation

Reptiles possess a metanephric kidney but lack the countercurrent multiplier mechanism of birds and mammals needed to elaborate an hyperosmotic urine (Dantzler and Bradshaw 2009; O’Shea et al. 1993). This renders them particularly vulnerable to significant perturbations of their milieu intérieur should they be faced with high electrolyte intakes or lack sufficient free water needed to excrete dietary salts (Cooper 2017). Some species possess cephalic salt-secreting glands that help in the maintenance of osmotic homeostasis (e.g. most varanids and many scincid lizards (Bradshaw 1986)), but they are absent in Australian agamids and terrestrial snakes (Saint Girons and Bradshaw 1987; Saint Girons et al. 1981).

The increased availability of tritiated water and liquid scintillation counters in the 1970s and ‘80s led to the first measurements of rates of water turnover of free-living lizards in North America, followed soon after by those of Australian arid-zone species. When compared with rates in tropical and subtropical species, those in arid-zone species were significantly lower, reflecting their overall enhanced water economy (Nagy 1982). Water turnover increases virtually linearly with body mass, and Nagy’s regression for arid-zone species of mL.d−1 = 20.5 kg0.91 was updated to 32.2 kg0.98 by Withers and Bradshaw (1995) and is close to the allometric relationship for free-ranging varanids with mL.d−1 = 38 kg1.19 (Bradshaw 1997). In all three cases, the exponent of the equation does not differ significantly from 1.0 (Fig. 1).

Fig. 1
figure 1

Allometric relationship between body mass and field metabolic rate in kJ.d−1 (FMR, squares) and water turnover rate in mL.d−1 (circles) for semiarid- and arid-zone lizards. (Adapted from Withers and Bradshaw, 1995)

Full size image

Interspecific regression equations such as these, which are based on average values for a wide range of species, can, however, mask more subtle differences between species that may only be apparent when intraspecific regression equations are compared. A comparative study of the relationship between body mass and rate of evaporative water loss in several skinks, geckos and an agamid lizard found large and significant differences in the values of ‘a’ and ‘b’ in the allometric relationship y = ax b where y is water loss and x is body mass in grammes (Fig. 2). These data from an early preliminary study suggest that there are substantial differences in the value of the constant ‘a’, which represents the rate of water loss of a 1 g (or 1 kg) animal depending on the scale. The gecko Gehyra variegata has a very high ‘a’ value of 8.71 and thus will have a high rate of water loss when small, but the low value of the exponent ‘b’ of 0.26 means that its rate of evaporative water loss will decrease, relative to its body mass, as it grows in size. The barking gecko, Underwoodisaurus milii, on the other hand, with b = 0.97, does not benefit from any relative increase in its water economy as it grows. The agamid, Ctenophorus (formerly Amphibolurus) ornatus, is the ‘best’ adapted in terms of water economy, with a = 3.16 and b = 0.55, and it is evident that lizards such as Ctenotus labillardieri and Diplodactylus vittatus must, if they are to survive, restrict their movements to moist habitats because of their coupled high ‘a’ and ‘b’ values and potential for desiccation.

Fig. 2
figure 2

A plot of mean values for ‘a’ and ‘b’ from the respective allometric equations (y = ax b) describing the relationship between body mass (x) and rate of evaporative water loss (y) for six species of Australian lizards from three families. (Adapted from Bradshaw, 1986, and references therein. NB Ctenotus lesueuri = C. australis)

Full size image

In the absence of a renal concentrating system, increases in water economy in reptiles can only be achieved by either reducing intake of water or by limiting losses. Under extreme conditions of salt loading, some lizards close down kidney function, cease filtering and become temporarily anuric (Bradshaw 1997), but reduction in rates of respiratory and cutaneous water loss underpins their enhanced water economy (Lillywhite 2006; Zucher 1980).

The role of the pituitary hormone arginine vasotocin (AVT) in controlling kidney function in three species of agamid lizards was investigated by Ford and Bradshaw (2006) who posited an ‘adaptationist’ scenario (sensu Gould and Lewontin (1979)), testing whether desert species would display enhanced renal responses to dehydration and salt loading when compared with mesic species. In fact, they found little support for this hypothesis, apart from a primarily glomerular rather than tubular response to salt loading in the mesic species, Pogona minor. Surprisingly, AVT levels in the desert species, Ctenophorus nuchalis, showed no significant correlation with changes in plasma osmolality, although increasing levels of AVT were associated with a marked antidiuresis that was both glomerular and tubular in nature. This study thus supports the conclusion that desert-living reptiles are ‘exapted’ rather than adapted to the exigencies characteristic of the arid zone (Bradshaw 1986, 1988).

The Western netted dragon, Ctenophorus nuchalis, is widespread in sandy areas of the arid zone in Australia and was the subject of a long-term ecophysiological and population study at Shark Bay in Western Australia (Bradshaw 1986). An unexpected discovery to emerge from mark-and-recapture data with this large (up to 60 g), fast-growing lizard is that it is an annual species, along with many other species in the genus. Following winter rains, they breed and lay eggs in spring, then progressively lose condition and die during summer such that, by autumn, few if any adults in the population survive and the young emerge to colonise what is an ‘empty’ environment (Bradshaw 1981). A detailed study of seasonal changes in body condition, with modelling of daily changes in thermoregulatory activity, established that coping with high environmental temperatures in summer precludes the lizards from feeding adequately (even though food is available) and thus leads to their demise (Bradshaw and De’ath 1991). Their gradual slide into negative water balance as summer progresses, with a corresponding catastrophic decline in field metabolic rate (FMR) and dry matter intake (DMI), measured with doubly labelled water, is shown in Fig. 3.

Fig. 3
figure 3

Seasonal changes in field metabolic rate (FMR) measured as CO2 production in mL.g−1.h−1 and estimated dry matter intake (DMI) in mg.g−1.d−1 in the Western netted dragon, Ctenophorus nuchalis, in Shark Bay, Western Australia. (Adapted from Nagy and Bradshaw, 1995)

Full size image

A closely related species, the ringtail dragon, Ctenophorus caudicinctus, is a rock-dweller and lives sympatrically with C. nuchalis in the arid Pilbara region of Western Australia. This species breeds invariably in autumn, following cyclonic rains, and a study of the two species together found that C. nuchalis deferred breeding in spring in some years when winter rains failed, survived the summer period and bred in autumn with C. caudicinctus (Bradshaw et al. 1991). This implies that the summer death seen with C. nuchalis at Shark Bay is a post-reproductive phenomenon, and both C. nuchalis and C. caudicinctus, like C. fordi (Cogger 1978), are thus semelparous species (Henle 1991; Dickman et al. 1999). The reasons why there is a breakdown in the ability to maintain homeostasis following reproduction in these lizards is not clear, although parallels with small semelparous dasyurid marsupials are obvious (Bradley 2003). Changes in circulating levels of corticosteroid-binding globulin (CBG) and free and bound concentrations of the adrenal steroid, corticosterone, have been implicated in stress reactivity studies of the North American lizard, Urosaurus ornatus (Jennings et al. 2000), but have yet to be measured in these Australian species. What has been documented in C. nuchalis is a marked change in the normal adrenal response in spring to the stress of confinement, or injections of adrenocorticotrophic hormone (ACTH), when repeated after breeding in late summer at Shark Bay, with the majority of individuals failing to respond, whilst others show an aberrant overresponse to the stressor (Bradshaw 1997, 2017) (Fig. 4).

Fig. 4
figure 4

Frequency distribution of plasma corticosteroid concentrations in the agamid lizard Ctenophorus nuchalis injected with adrenocorticotrophic hormone (ACTH) after dexamethasone blockade in spring and summer at Shark Bay in Western Australia. The aberrant, non-Gaussian response of the lizards in late summer to the ACTH challenge is evident. (Adapted from Bradshaw 1986)

Full size image

A number of studies have been conducted in the wet/dry tropics of Australia, using doubly-labelled water to document the energetics of agamid and varanids lizards which are very informative, linking activity, thermoregulation and foraging patterns. Two species of varanid lizards, for example, were studied along the South Alligator River in the Northern Territory of Australia and displayed divergent patterns of activity and resource exploitation (Christian et al. 1995). Rates of energy expenditure and water turnover were faster in Varanus gouldii than in V. panoptes, but levels were higher in both species in the wet/dry tropics than in other varanids, such as Varanus giganteus occupying the arid zone (Green et al. 1986). Varanus gouldii, for example, with an average water influx of 50.7 mL.kg−1.d−1, was turning over approximately 8% of its total body water content per day (TBW), compared with only 3.9% for V. giganteus on Barrow Island (King and Green 1999). The frilled-neck lizard, Chlamydosaurus kingii, is a large 600 g agamid that is common in the tropical north of Australia. It ceases activity in the dry season of the year, spending it perched on tree trunks. Rates of water turnover are rapid, as would be expected in the wet season with a mean annual rainfall of 1600 mm in Darwin, but water turnover rates fall to 13.6 mL.kg−1.d−1 in the early dry season which is only approximately 2% of the TBW per day; later in the dry season, this falls even further to 5.6 mL.kg−1.d−1 (2.9 mL.d−1) (Christian and Green 1994). This is, however, still substantially greater than the rate of metabolic water production for a lizard of this size (ca. 0.3 mL.d−1), showing that whilst apparently inactive, the lizards are still feeding sporadically. The only other study of an arid-living varanid is that of the small (12 g) arboreal Varanus caudolineatus in the Murchison region of Western Australia (Thompson et al. 1997). Water influx in summer averaged 31.6 mL.kg−1.d−1, and FMR was 0.46 mL CO2.g−1.h−1, which enables a calculation of the ‘Water Economy Index’ (WEI) of 0.14 (Nagy and Peterson 1988). This is considerably less than that reported for any other varanid species.

There is a single published study of energetics and water flux rates of a free-ranging gecko, the marbled velvet gecko, Oedura marmorata, in both tropical regions and in central Australia (Christian et al. 1998). As would be expected, both FMR and rates of water influx are much lower in the arid population than in the geckos from the tropical locations. The FMR of arid geckos in spring was 40.1 kJ.d−1, compared with FMRs ranging from 109.8 to 148.9 kJ.d−1 for the tropical geckos. The authors do raise the possibility that the two populations may not in fact be from the one species, and a recent paper on the systematics of the Oedura complex confirms this with the arid population currently listed as Oedura aff. Marmorata (P Doughty pers. comm.). Rates of metabolism and water flux have been reported in the northern death adder, Acanthophis praelongus, again not an arid-zone species, with a similar tropical pattern of reduced rates in the dry season compared with that in the wet (Christian et al. 2007).

figure a
figure b

Conclusion

This short review serves to highlight how little we actually know of the ecophysiology of arid-zone reptiles in Australia. Despite having a reptilian fauna that is at least three times as rich in number of species as that of other desert regions of the world, ecophysiology has been little studied, and we only have speculations and hypotheses to account for reptile abundance and diversity (Morton and James 1988; Pianka 1986, 1989, 2014; Byrne et al. 2008). Although rates of water and energy turnover of arid-zone species are slower than those occurring in tropical regions of Australia, this does not appear to be the result of any specific adaptations of the former. In fact some species, such as the semelparous lizard Ctenophorus nuchalis, typically ‘avoid’ the exigencies of the arid zone by what may be thought of as the bizarre habit of breeding and depositing eggs in spring, only to die in the oncoming summer. The population thus survives during the hottest period of the year in the egg phase, and juveniles emerge each autumn to a habitat devoid of adults. Natural selection also does not operate to enhance summer survival as the lizards are then post-reproductive. Longer-living lizards, such as the many varanids, have to survive during the summer months and, in most cases, do this by drastically reducing their activity and often retreating below ground. In common with reptiles living in other regions of the world, the reason for their abundance and diversity in desert areas is linked to their ectothermy and slow rates of resource utilisation, compared with those of birds and mammals, and better described as ‘exaptations’ (Gould and Vrba 1982), rather than adaptations (Bradshaw 1988).