The Evolution of Endothermy and Its Diversity in Mammals and Birds

Author(s): Gordon C. Grigg, Lyn A. Beard, and Michael L. Augee

Source: Physiological and Biochemical Zoology, Vol. 77, No. 6, Sixth International Congress of
Comparative Physiology and Biochemistry Symposium Papers: Evolution and Advantages of
Endothermy (November/December 2004), pp. 982-997
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/10.1086/425188 .
Accessed: 24/03/2013 11:26

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to
Physiological and Biochemical Zoology.

http://www.jstor.org

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
982

The Evolution of Endothermy and Its Diversity in

Mammals and Birds

Gordon C. Grigg1 thermy, including the capacity for homeothermic endothermy

Lyn A. Beard1 during incubation, but are very relaxed in their thermoregu-
Michael L. Augee2 latory precision and minimise energetic costs by using ecto-
1
School of Integrative Biology (Zoology), University of thermy facultatively when entering short- or long-term torpor.
Queensland, Brisbane, Queensland 4072, Australia; 2Linnean They also have a substantial layer of internal dorsal insulation.
Society of New South Wales, P.O. Box 82, Kingsford, New We favor theories about the evolution of endothermy that in-
South Wales 2032, Australia voke direct selection for the benefits conferred by warmth, such
as expanding daily activity into the night, higher capacities for
Accepted 6/30/04 sustained activity, higher digestion rates, climatic range expan-
sion, and, not unrelated, control over incubation temperature
and the benefits for parental care. We present an indicative,
stepwise schema in which observed patterns of body temper-
ABSTRACT
ature are a consequence of selection pressures, the underlying
Many elements of mammalian and avian thermoregulatory mechanisms, and energy optimization, and in which homeo-
mechanisms are present in reptiles, and the changes involved thermy results when it is energetically desirable rather than as
in the transition to endothermy are more quantitative than the logical endpoint.
qualitative. Drawing on our experience with reptiles and echid-
nas, we comment on that transition and on current theories
about how it occurred. The theories divide into two categories,
depending on whether selection pressures operated directly or
Introduction
indirectly on mechanisms producing heat. Both categories of
theories focus on explaining the evolution of homeothermic Endothermy has arisen in insects, fishes, reptiles, birds, and
endothermy but ignore heterothermy. However, noting that mammals, as well as in some plants. However the evolution of
hibernation and torpor are almost certainly plesiomorphic endothermy in mammals has attracted the most debate and
(pancestral, primitive), and that heterothermy is very common attention, and it is regarded as one of the most significant
among endotherms, we propose that homeothermic endo- transitions in vertebrate evolution. We bring two perspectives
thermy evolved via heterothermy, with the earliest protoendo- to the debate that are different from published views of other
therms being facultatively endothermic and retaining their ec- authors; a focus on the relevance of thermoregulatory mech-
tothermic capacity for “constitutional eurythermy.” Thus, anisms in reptiles, many of which foreshadow those used by
unlike current models for the evolution of endothermy that mammals and birds, and a consideration of the implications
assume that hibernation and torpor are specialisations arising of the view that mammalian and avian torpors are probably
from homeothermic ancestry, and therefore irrelevant, we con- plesiomorphic (see review by Grigg [2004]). Although we will
sider that they are central. We note the sophistication of ther- concentrate on mammalian endothermy, many of our consid-
moregulatory behavior and control in reptiles, including precise erations are relevant to birds also.
control over conductance, and argue that brooding endothermy What endotherms have is the capacity for substantial heat
seen in some otherwise ectothermic Boidae suggests an incip- production as a by-product of cellular work, either with or
ient capacity for facultative endothermy in reptiles. We suggest without muscular contractions, and its retention or loss gov-
that the earliest insulation in protoendotherms may have been erned by mechanisms that regulate thermal conductance. The
internal, arising from redistribution of the fat bodies that are heat-production, heat-loss equation is such that endotherms
typical of reptiles. We note that short-beaked echidnas provide have the capacity to display body temperature patterns that are
a useful living model of what an (advanced) protoendotherm conspicuously different from those within the capacities of
may have been like. Echidnas have the advantages of endo- ectotherms.
Current ideas about the evolution of endothermy focus on
the evolution of homeothermic endothermy. This is not sur-
Physiological and Biochemical Zoology 77(6):982–997. 2004. 䉷 2004 by The prising because it is usually assumed that endotherms are typ-
University of Chicago. All rights reserved. 1522-2152/2004/7706-3079$15.00 ically homeothermic, with a stable, high body temperature.

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
 Evolution of Endothermy and Its Diversity 983

However, homeothermy may not be characteristic of even the product, higher resting heat production and the capacity for
majority of endotherms. Substantial daily and seasonal Tb var- homeothermic endothermy.
iations occur in very many mammals and birds (reviews by The difficulty of establishing that a higher active metabolic
McKechnie and Lovegrove [2002] and Geiser [2004]) and, as rate implies a high resting rate as well has been a problem with
data accumulate from studies of free-ranging species, more and this model (see discussion by Hayes and Garland [1995]), and
more are found to show heterothermic patterns. Even some this may be why Ruben (1995) included the possibility of “in-
primates are heterothermic (e.g., Madagascan lemurs; Daus- ertial homeothermy” (sensu McNab 1978), in which homeo-
mann et al. 2000). Not only is heterothermy common, but thermy comes as a correlate of higher thermal inertia at large
there is good evidence to suggest that hibernation is probably body size.
plesiomorphic (Augee and Gooden 1992) and that hibernation Our aim is not necessarily to try to discriminate between
and short-term (daily) torpor are likely to be homologous these two groups of models but to use our own research on
(Grigg 2004), implying that torpor in the broader sense is ple- reptiles and echidnas to better define the appropriate questions,
siomorphic. This has implications for thinking about the evo- broaden the discussion, and further develop some ideas we
lution of endothermy (Grigg and Beard 2000; Grigg 2004). In presented previously (Grigg and Beard 2000).
this article we will review the evidence available from among Who were the earliest protoendotherms? It is thought that
reptiles, show that short-beaked echidnas have many of the mammals arose late in the Triassic, from the cynodont therapsid
attributes that might be expected in a putative protoendotherm reptiles about 200 mya (Rowe 1992). Ruben and Jones (2000)
and draw the threads together to propose a logical stepwise noted that because endothermy occurs in all three classes of
scenario for the evolution of endothermy in a way that incor- extant mammals, by parsimony their last common Mesozoic
porates the implications of torpor being plesiomorphic. The ancestor, about 160 mya, must also have been endothermic,
scenario also includes a new idea about the vexed question of with a high resting metabolic rate and hair. However, endo-
the evolution of insulation. We think our scenario is broadly thermy almost certainly arose even earlier than the evolution
compatible with much of what has been written by others about of mammals. Hillenius (1994) suggested that the presence of
probable selection pressures but extends and expands it. respiratory turbinals in the nasal passages provides a useful
marker for the possession of endothermy, implying that “mam-
Current Ideas about the Evolution of Endothermy malian” endothermy first arose in the Permian therapsid rep-
tiles, about 250 mya in the late Permian. Thus, the first “mam-
Ruben (1995, 1996), Hayes and Garland (1995), and Farmer malian” protoendotherms were likely to be the dog- or
(2000) all provide excellent and comprehensive reviews of the bear-sized therocephalians. These animals are thought to have
literature on this topic. Numerous models have been proposed, been active carnivores, 20–100 kg, living in tropical and sub-
and they fall broadly into two categories. One category includes tropical climates, essentially reptilian in most features and ap-
models that assume direct selection for the benefits of high Tb parently lacking hair or any external insulation (Ruben and
and high resting metabolic rate, such as the thermal niche Jones 2000). These authors also considered that these animals
expansion model (Bakken and Gates 1975; Crompton et al. may have been inertial homeotherms, but, as we will discuss
1978), the homeothermy and increased metabolic efficiency later, that idea can be dismissed.
model (Heinrich 1977; Avery 1979), the endothermy via inertial
homeothermy model (McNab 1978), and the parental care The Relevance of Torpor Being Plesiomorphic
model (Farmer 2000). Implicit in all these models is the notion
that warmth allows a higher resting metabolic rate and, there- Torpor, hibernation, and estivation in mammals are all quali-
fore, has a high selective value. These ideas are sometimes re- tative descriptions of deviations from what has come to be
ferred to collectively as the thermoregulatory hypotheses. regarded as “normal” mammalian homeothermy. Lyman and
The second broad category includes models that consider Blinks (1959) considered hibernation to be functionally ad-
the benefits of warmth per se to be insufficient to warrant the vanced and specialized, and Wang (1989, p. 392) wrote that
costs, and that the high metabolic rates typical of endotherms “the wide occurrence of torpor in species from at least six
must have arisen as a consequence of selection for other ad- mammalian and eight avian orders suggests that it is polyphy-
vantages such as postural changes that increase exercise (Heath letic in origin.” This remains the prevailing view. Ruben (1995)
1968; Bakker 1971), increased brain size (Hulbert 1980), or accepted Wang’s (1989) judgment and dismissed the relevance
higher aerobic capacity (Bennett and Ruben 1979; Bennett of torpor/hibernation to considerations about the evolution of
1991; Ruben 1995). The aerobic capacity model has gained endothermy.
prominence in recent years. According to this model, selection However, this dismissal may not be justified. A more par-
for increased aerobic capacity to support a higher level of sus- simonious interpretation of the taxonomic distribution and
tained activity has led consequentially to higher resting meta- other aspects of both short- and long-term torpor, reviewed
bolic rates and, apparently as a more or less accidental by- by Grigg and Beard (2000) and Grigg (2004), leads to the

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
984 G. C. Grigg, L. A. Beard, and M. L. Augee

alternate conclusion that daily torpor and hibernation are likely Barnes 1999), keyed tightly to seasonal events and enabling
to be plesiomorphic and, furthermore, are supported by similar survival through Arctic winters, are specializations from more
mechanisms expressed to different extents. Augee and Gooden facultative styles of heterothermy seen in their relatives still
(1992) proposed that the ability to hibernate, occurring as it living in less severe conditions (Grigg 2004). Echidnas provide
does in such a wide diversity of mammalian groups and across another example (see later), with both facultative and obligate
all three orders of extant mammals, argues for a plesiomorphic hibernation being shown within the one species, depending on
origin, not a polyphyletic one. Quite independently, Malan where they live (Grigg and Beard 2000; Kuchel 2003). In mild
(1996, p. 2) noted similarities in the Tb patterns expressed by climates where food is available all year, echidnas hibernate
hibernators to those seen seasonally in many reptiles and came facultatively, perhaps depending on overall energy status. In
to the same conclusion: that “we should now consider the areas snow clad in winter, where both harsh conditions and
phylogeny of hibernation, not as the repetitive and independent food shortage prevail, they match the pattern seen in classical
occurrence of a secondary adaptation in various phyla, but hibernators, and all individuals hibernate every year.
simply as a recurring expression of primitive traits.” If the physiological and behavioral capacities for short-term
The idea is not new; Cade (1964) interpreted the occurrence torpor and hibernation are expressions of persistent/reemergent
of torpor/hibernation in five out of 14–15 distinct lineages capacities reflecting reptilian ancestry, and because the ther-
within the Rodentia as evidence of intermediate stages in the moregulatory systems present in reptiles foreshadow so strongly
“history of rodent evolution of homeothermic mechanisms and those of mammals and birds, it seems reasonable to suggest
the associated loss of tolerance for deep hypothermia.” He also that studies of torpor and hibernation should be a central theme
drew particular attention to the significance of torpidating/ in studies of the evolution of endothermy. Indeed, the entry
hibernating mammals’ tolerance for low Tb, which Eisentraut into daily torpor or hibernation may involve a reversal of some
(1960) had called “constitutional eurythermy.” More recently, of the same physiological mechanisms that accompanied the
Lovegrove et al. (1999) interpreted the finding of torpor in evolution of endothermy from ectothermy (Grigg and Beard
elephant shrews (Macroscelidae) as support for a plesiomorphic 2000).
origin of heterothermy.
Furthermore, while there is strong evidence that endothermy Evidence from Living Reptiles
must have arisen independently in birds and mammals, the
mechanisms underlying torpor/hibernation in both groups If endothermy arose in a lineage of therapsid reptiles, clues to
could well be monophyletic in origin. Malan (1996) certainly its origin may survive among extant reptiles. Few, if any, reptiles
saw these attributes in both mammals and birds as stretching are simply passive thermoconformers. There are behavioural
back to a common ancestry in the reptiles. With torpor/hi- and physiological mechanisms that cause many reptiles to have
bernation being found in 11 out of 26 orders of birds, and still a high (and often relatively stable) body temperature during
counting, including the ancient mousebirds (Coliidae; Mc- daytime hours (see review chapters in Gans and Pough 1982),
Kechnie and Lovegrove 2002), parsimony again suggests ple- and these have been selected for, presumably, because of the
siomorphy, as argued by these authors. As in mammals, avian benefits accruing from being warm. It could be argued that
evidence continues to accumulate; Lane et al. (2004) concluded there are few associated costs when the heat source is external,
that heterothermy within the Caprimulgiformes is plesiomor- but this overlooks the potential high cost of predation exposure
phic. It seems likely that the physiological mechanisms that in a cool, basking lizard. A really impressive example of the
support torpor in birds and mammals share the same ancient benefits accruing from warmth is provided by Australia’s fresh-
origins and, therefore, the evolution of their endothermy may water crocodile, Crocodylus johnstonii, in which social hierarchy
be more similar than is usually acknowledged. and the competition for basking sites dictates which individuals
The plesiomorphic model makes it easy to see how attributes in the population are thermoregulators rather than thermo-
shown by a heterotherm in a milder climate could be selected conformers, with undoubted advantages in growth and repro-
for and enable range expansion into more severe climates, with ductive success (Seebacher and Grigg 1997).
attendant specializations to fine-tune behaviors, tie them to Modern reptiles provide a rich source of information relevant
relevant day-length cues, and even to change membrane struc- to the evolution of endothermy, but, before exploring that, we
ture, which may be necessary if drops to very low Tb will be will examine an idea that has been influential even to the extent
encountered (Aloia et al. 1986). That is, we interpret the pat- of possible inclusion in variants of the aerobic capacity hy-
terns shown by the “classic” hibernators not as specializations pothesis (Ruben 1995; Ruben and Jones 2000), the hypothesis
from homeothermic endothermy but as specializations at one that endothermy arose via passive homeothermy. (Note that
end of a whole spectrum of examples with attributes deriving the term “homeothermy” is used in older literature and even
from ancestral heterothermic capabilities differentially ex- some modern literature as if it were synonymous with what
pressed (Grigg 2004). The complex physiology and behaviour would now be termed “homeothermic endothermy.”)
in arctic ground squirrels, for example (Barnes 1989; Boyer and Endothermy did not evolve via “inertial homeothermy.”

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
 Evolution of Endothermy and Its Diversity 985

Table 1: Comparison of patterns, mechanisms, and regulation between reptilian ectotherms and their descendants, the
endothermic mammals and birds
Reptilian Ectotherm Mammalian Endotherm Avian Endotherm
Pattern of Tb Daily and seasonal fluctuations, Commonly high and stable, com- Mostly high and stable, com-
Tb may be high and stable monly heterothermic, modal monly heterothermic, modal
during activity periods. Size re- values differ between three values higher than in the three
lated. Facultative endothermy classes. Many show torpor mammalian classes. Many
in brooding pythons. and/or hibernation. show torpor, few show
hibernation.
Mechanisms Primarily behavioral, heat pro- Behavior still important, heat Behavior still important, heat
duction from RMR low, little production from RMR high production from RMR high
heat from leaky membranes, (leakier membranes), physio- (leakier membranes), physio-
physiological mechanisms in- logical mechanisms include logical mechanisms include
clude panting, changes in pe- changes in peripheral blood changes in peripheral blood
ripheral blood flow. Muscle flow, panting, sweating, insula- flow, panting (gular flutter),
thermogenesis in brooding py- tion. Locomotion. Shivering. insulation. Locomotion. Shiv-
thons foreshadows some at- Regulatory NST from un- ering. Regulatory NST, possibly
tributes of endotherms, includ- known source in marsupials from sarcoplasmic reticulum.
ing the capacity to increase (skeletal muscle?) and possibly
metabolic rate when ambient all mammals. BAT a special-
temperature falls. ized tissue forming a thermo-
genic organ in many eutherian
mammals, particularly euthe-
rian hibernators and young.
Regulation Role of hypothalamus in regulat- Role of hypothalamus in regulat- Role of hypothalamus in regulat-
ing behavior and physiology ing behavior and physiology ing behavior and physiology
relevant to thermoregulation. relevant to thermoregulation. relevant to thermoregulation.

MacNab (1978) proposed ingeniously that endothermy evolved in modal daily Tb between winter and summer was similar in
via inertial (passive) homeothermy, gained through large body both larger and smaller individuals, 4⬚–5⬚C. Even a 1-ton croc-
size and maintained by selection for the advantages accom- odile, seeking sun or shade, water or land, and doing so in a
panying homeothermy as body sizes decreased. It is therefore way that minimizes change in Tb, is not thermostable across
relevant to explore extant species to see how large a reptile seasons. Inertial homeothermy at a seasonal scale could occur
would have to be to show passive homeothermy across both in only very large reptiles, much larger than 1 ton, at any
daily and seasonal timescales. Data from a study of daily and latitude. Furthermore, the acquisition of inertial homeothermy
seasonal changes in Tb in estuarine crocodiles, Crocodylus po- does not necessarily imply the acquisition of warmth as well.
rosus, in a tropical habitat and over a size range from 32 to Admittedly, Tb in reptiles large enough to be passively homeo-
1,000 kg (Grigg et al. 1998; Seebacher et al. 1999) show clearly thermic does increase with an increase in body size (Seebacher
that dog- or bear-sized protoendotherms in a tropical or sub- et al. 1999), but the crocodilian data show that the combination
tropical climate could not have been inertial homeotherms. of passive homeothermy and substantially above-ambient Tb
Movements to and from the water and between sun and shade could not occur in reptiles in the size range of the Permian
minimized daily changes in Tb. Tb cycled daily through about theriocephalians (i.e., less than 100 kg). The evidence suggests
7⬚C in the smallest animals (32–42 kg) decreasing to 2⬚–3⬚C that McNab’s idea that endothermy evolved via passive homeo-
in the largest animals (600–1,000 kg). So the largest individuals thermy can be disregarded.
were essentially passive homeotherms at the timescale of a single A comparison between reptiles and the higher vertebrate en-
day and, if thermostability within 2⬚C is accepted as homeo- dotherms. The extent to which reptilian thermoregulatory sys-
thermy, this pattern was observed in animals larger than about tems foreshadow those in mammals and birds is striking; most
500 kg. That is, inertial homeothermy over the course of even of the differences are more quantitative than qualitative (Table
a single day is found in only quite large individuals, even in a 1).
tropical environment. What was interesting, however, and very Reptiles have sophisticated control over thermal conductance.
relevant to this discussion, is that at a seasonal scale the range Reptiles can effect considerable control over thermal conduc-

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
986 G. C. Grigg, L. A. Beard, and M. L. Augee

tance through changes in peripheral blood flow. The hysteresis of limited endogenous heat production not associated with
between rates of heating and cooling was described initially by muscular work, the sort of difference upon which, presumably,
Bartholomew and Tucker (1963) in a laboratory study of natural selection can operate. Leakiness in membranes, al-
bearded dragons, Pogona (pAmphibolorus) barbata. Its occur- though less than in mammals (Else and Hulbert 1987), is a
rence and relevance to reptiles under natural conditions was feature of reptilian tissues also, so it is possible to imagine
confirmed in a field study by Grigg and Seebacher (1999) in selection pressures operating to increase this to the extent of
the same species. Increased blood flow allows a cool reptile to providing substantial heat as a by-product. The temperature
gain heat rapidly while basking; after basking, decreased flow increases in I. iguana were only small, 1⬚–2⬚C, and oxygen
leads to decreased conductance and allows a warm reptile to consumption was not measured simultaneously, but the results
maintain its heat load for longer in a cooling environment. are provocative.
Excitingly, the fieldwork showed that this species also has the Also provocative, Rismiller and McKelvey (2000) described
capacity to reduce peripheral blood flow and avoid further patterns of Tb in a varanid lizard in winter that appear strik-
warming under high operative temperatures, that is, a “reverse ingly similar to the spontaneous arousals that characterize
hysteresis.” These carefully regulated changes in conductance hibernating mammals. Will study of reptiles shed light on the
foreshadow mechanisms seen in both mammals and birds. enigmatic but characteristic periodic arousals that charac-
Shivering thermogenesis in brooding pythons, and its integra- terize hibernation in eutherian, marsupial, and monotreme
tion with basking. Shivering thermogenesis by brooding female hibernators?
Boidae is usually dismissed as a curious adaptation in a few Selection for warmth rather than increased aerobic capacity?
very specialized reptiles. However, the shivering thermogenesis The “thermoregulatory hypothesis” assumes that there was pos-
in brooding pythons (Hutchison et al. 1966; Harlow and Grigg itive selection for increased resting metabolic rate in the evo-
1984) shows that reptiles have more of the capacities of en- lution of the higher body temperatures, presumably within the
dotherms than is generally recognized and, therefore, is very therapsid-cynodont lineage in the case of mammals, because
relevant to discussions about the evolution of endothermy. of the advantages accrued by warmth. As discussed, reptiles
These snakes are able to shiver and raise the temperature of show a number of mechanisms that operate to collect, retain,
their eggs above a threshold allowing development, as well as and even, in a couple of instances, generate warmth quite apart
increasing the rate of development. They are, therefore, fac- from increased aerobic capacity. Much of the heat production
ultative endotherms. Very significantly, and just like mammals associated with high resting metabolic rate in endotherms
and birds, they respond to a lowered ambient temperature by comes from otherwise futile work, and that by itself might be
increasing their metabolic rate by increased shivering, thus offered as an argument in favor of the thermoregulatory
maintaining Tb. hypothesis.
Furthermore, their integration of shivering with typical ec- The “protoendotherms” may have been facultative endotherms.
tothermic basking behaviour is particularly interesting. When Many elements of endothermy are present and functional in
the snake in Harlow and Grigg’s (1984) study on Morelia spilota living reptiles: excellent peripheral vascular control, a sophis-
had an ambient heat source available a short distance away, ticated regulatory control of thermoregulation, examples of fac-
she would leave the eggs each morning to bask and, when ultative endothermy by shivering thermogenesis, and an ex-
warmed to 32⬚–33⬚C, return and coil once more around the ample of small circadian increases in Tb not associated with
eggs, shivering from then on as required to maintain Tb. This activity. Given this, we propose that the evolution of endo-
is very reminiscent of the way ectothermy and endothermy are thermy is likely to have occurred via stages in which selective
integrated in an energetically sensible way in some mammalian pressures favored the enhancement of elements present within
endotherms, for example, in the torpidating carnivorous mar- reptiles and that the early “protoendotherms” were facultative
supial, Pseudantechinus macdonnellensis, which uses the morn- rather than obligate endotherms, like the only known extant
ing sun to rewarm following an overnight torpor (Geiser et al. reptilian endotherms.
2002), and in two species of mouse lemur, Microcebus, in which
the early morning Tb rises passively with ambient temperature Echidnas and the Protoendotherm
until the start of the next activity session (Schmid 1996). In
general, though often overlooked, behaviour is a very important Having identified characteristics of reptiles that seem to be
element in thermoregulation of all mammals and birds, just as foreshadowing endothermy, can we find extant mammals or
it is in reptiles. birds that seem to echo strongly the features likely to be found
Some additional observations from reptiles. Tosini and Men- in a putative protoendotherm? Among the extant mammals,
aker (1995) described physiologically generated circadian short-beaked echidnas seem to have many of them. Martin
rhythms in Tb in iguanas, Iguana iguana, kept at constant tem- (1903) wrote about echidnas in the context of the development
perature, in which activity cycles were not the cause of the of “homeothermism.” After a comprehensive study of captive
increases in Tb. This seems to be an example among reptiles echidnas in outdoor pens in Melbourne, Australia, and based

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
 Evolution of Endothermy and Its Diversity 987

on their heterothermy, low Tb, reliance on muscular work for

heat, and apparent lack of NST, Augee (1978, p. 118) wrote,
“unless fossil evidence is obtained to the contrary, the best
interpretation of monotremes in regard to endothermy/ho-
meothermy is that they represent an early stage in its evolution.”
Both of these studies predated the availability of data from
radiotelemetry of echidnas free-ranging in their natural habitat
and, therefore, the realization that echidnas are hibernators
(Grigg et al. 1989) and also torpidators (Grigg et al. 1992a).
These studies, combined with data from other workers, provide
a much better understanding of echidna thermal relations, and,
therefore, a reappraisal of their relevance to ideas about the
evolution of endothermy is timely.
Short-beaked echidnas match in many respects the attributes
that one might design in a putative protoendotherm. They
occur all over Australia, from the tropics to cold temperate
regions, above the snowline, and by the coast. This distribution
implies very substantial adaptability to different thermal re-
gimes, and their thermal relations have now been studied in
detail across a wide variety of habitats and climates (Grigg et
al. 1989, 1992a; Beard et al. 1992; Nicol and Andersen 1993,
2000; Rismiller and McKelvey 1996; Grigg and Beard 2000;
Brice et al. 2002a, 2002b; Kuchel 2003).
A number of generalizations emerge that we will illustrate
using published information supplemented with previously un-
published data from our study site near Stanthorpe in SE
Queensland, a comparatively mild, subtemperate climate:
1. Although the modal Tb of echidnas is close to 32⬚C (the
same as the other monotremes; Dawson et al. 1978; Grigg et
al. 1989, 1992a, 1992b, 2003), they are relaxed about main-
taining a stable Tb, with typical daily cycles of 2⬚–5⬚C (Fig. 1a),
sometimes larger (Fig. 1b), that correlate with their daily activity
cycle. Daily maxima and minima may not correlate with am-
bient temperatures (Grigg et al. 1992a; Kuchel 2003). A fac- Figure 1. Two examples of daily cycles in body temperature in a male
ultatively endothermic protoendotherm might have been sim- echidna (95) at our southeast Queensland study site, which has a warm
ilarly relaxed about regulating Tb and might have had a lower temperate climate. a, July 4–14, 2001 (southern winter). b, November
modal Tb than is typical of most mammals or of birds. 13–23, 2001 (southern summer).
2. Behaviour is important in echidna thermal relations, and
shifts in the time of day at which echidnas are active contribute et al. 1992a). A protoendotherm would presumably also have
significantly to the similarity of daily patterns seen in markedly had to rely on behaviour to assist in its thermoregulation.
different climates. Thus they are more diurnal in cooler climates 3. An early endotherm might have been expected to make
and more nocturnal in warmer climates and often shift their considerable use of basking. Basking has rarely been observed
active period seasonally. For example, in the hot, semiarid con- in short-beaked echidnas in the wild, perhaps because they seem
tinental climate of central Queensland, they may be active at to spend all their waking hours foraging actively. Echidnas in
any time of the day during winter and spring (unless hiber- captivity and in the wild occasionally do lie in full sun, giving
nating) but are strictly nocturnal in summer, seeking shelter in every impression of basking. Tb in a 3.3-kg individual rose from
caves and burrows to avoid the heat (Brice et al. 2002a). Avoid- 28.0⬚ to 30.8⬚C in about 90 min, during which she shifted
ing heat is important, for they lack sweat glands and presumably position occasionally as if to remain in the sun. How much of
rely on the heat sink provided by ambient conditions to counter this increase can be attributed to insolation is, however, un-
heat production associated with exercise. On unusually warm known. Manning and Grigg (1997), working on freshwater
days in the Australian Alps, where they are normally active turtles, have shown that observations on apparent basking be-
during the day, we recorded them losing heat quickly after haviour can be very misleading, and more work is needed to
seeking shade or a cool stream when Tb reached 34⬚C (Grigg establish how much echidnas make use of basking.

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
988 G. C. Grigg, L. A. Beard, and M. L. Augee

4. Echidnas have good capacity for changes in thermal con-

ductance. By analogy with reptiles, and mammals and birds,
the putative protoendotherm could be expected to have good
control over thermal conductance. Dawson et al. (1978, p. 100)
described a fourfold increase in conductance in Zaglossus (long-
beaked echidna) between Ta 15⬚C and Ta 30⬚C, which they
attributed to a “considerable increase in peripheral blood flow
to facilitate heat dissipation.” Schmidt-Nielsen et al. (1966)
described similar changes in thermal conductance in Tachy-
glossus. The changes in conductance can be ascribed to changes
in peripheral blood flow. In both echidnas there is a substantial
muscle layer (panniculus carnosus) external to the fat layer (pan-
niculus adiposus) in which the spines are deeply embedded and
that is well supplied with blood. This dorsal fat layer is ideally
placed to act as effective insulation. However, it can be bypassed
by blood flowing peripherally either during basking for heat
gain or when heat loss is required. Echidnas in cooler climates
are conspicuously more hairy than in warm climates, where
they have almost no hair between the spines.
5. Echidnas lack sweat glands and do not pant (Augee 1976)
and on the basis of this and early experimental work (Robinson
1954, cited in Griffiths 1968) echidnas have come to be regarded
as intolerant of heat (Griffiths 1978). However, Brice et al.
(2002a) have shown in fieldwork that they rest in surprisingly
warm daytime shelters and maintain Tb below Ta. Further work
on this is in progress in our laboratory. Do they reduce met-
abolic rate and show patterns reminiscent of the “reverse hys-
teresis” discovered in Pogona, as might be expected in a proto-
endotherm?
6. Short-beaked echidnas also show both short-term and
long-term hypothermias that fit the patterns usually identified
as short-term torpor and hibernation, respectively. Of partic-
ular interest in the context of the present topic, echidnas often
and at any time of the year show large, short-term drops in Figure 2. Two examples at our southeast Queensland study site of
Tb, sometimes in a daily cycle (Fig. 1b) and sometimes the departures from the normal daily pattern of body temperature caused
result of a day in which the activity period was foregone (Fig. by the echidna foregoing its normal daily foraging activity. By analogy
2). Such large departures from “normal” Tb are usually iden- with common practice, such departures might be said to be indicative
tified as torpor events, even if no information is recorded about of daily torpor. a, Echidna 117, male, December 26, 2000–January 5,
2001. b, Echidna 93, female, October 22–December 1, 2000.
the capacity for locomotor coordination (see below).
Short-beaked echidna long-term torpors may last for weeks
or months, interrupted by periodic arousals to normal body winter, were broken by periodic arousals and qualified as hi-
temperatures (Fig. 3), matching the pattern of so-called classical bernation. She could not find a distinction on the basis of
hibernation thought originally to occur only in Eutheria. Thus, metabolic rate either. Grigg (2004) took up this question further
short-beaked echidnas were the first mammals in which both and concluded that daily torpor and hibernation are likely to
torpor and hibernation were described (Grigg et al. 1992a), a be supported by similar mechanisms. That is, they share a high
phenomenon now known also in the edible dormouse, Glis glis degree of homology. Such flexibility in the adoption of ecto-
(Wilz and Heldmaier 2000). The occurrence of both daily tor- thermy or endothermy could be expected in an early, facultative
por and hibernation in the same species begs questions about endotherm.
the extent to which these patterns are supported by different 7. Likewise logical in a protoendotherm, and particularly
mechanisms or simply expressions of the same mechanism ex- significant for our present topic, echidnas show hibernation in
pressed to different extents. Kuchel (2003) found that it was surprisingly mild climates. The biggest differences in Tb patterns
impossible to make distinctions in echidnas; 1-, 2-, and 3-d found so far between studies at different locations is that echid-
torpors merged indistinguishably into longer periods that, in nas in warmer climates spend less time in hibernation than

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
 Evolution of Endothermy and Its Diversity 989

season varied considerably between individuals—approxi-

mately 1–4 mo. Data from other sites are less numerous, mak-
ing generalizations more difficult. In the hot, semiarid conti-
nental climate of Idalia National Park in central Queensland,
all three individuals from which data loggers were recovered
showed hibernation for about 3 mo, and all showed torpor
events (Brice et al. 2002a and personal communication).
That echidnas often hibernate in very mild climates where
food shortages are unlikely, and display such a diversity of
hibernation patterns, has prompted considerable interest (Nicol
and Andersen 1996, 2000; Grigg and Beard 2000; Kuchel 2003).
We (Grigg and Beard 2000) concluded that the purpose of
hibernation in benign climates is to achieve an energetic ad-
vantage. Animals well prepared for the spring breeding season
might be “putting themselves on ice” for the winter. This pre-
dicts that in a climate where hibernation is facultative, animals
with good body condition in autumn would hibernate, while
animals in lesser condition would choose to continue to forage
through the winter. Kuchel (2003) explored this proposition
and found somewhat equivocal results. She did, however, con-
firm that hibernation represents a saving on the year’s energy
expenditure though, as would be expected, less than in hiber-
nators in colder climates.
8. Despite being conspicuously heterothermic, echidnas do
have the capacity for homeothermic endothermy, as clearly
shown during incubation. Just before laying their single egg,
during incubation, and for a few days thereafter, a total of 2–
3 wk, they have the capacity to regulate Tb very tightly (Beard
and Grigg 2000; see also Fig. 3a). The exact source or sources
of heat during this time remain unresolved.
9. Another aspect of short-beaked echidna thermal relations
is that they have the clear capacity for apparently normal ac-
tivity at very low body temperatures in comparison to other
Figure 3. Two examples of patterns of hibernation shown at our warm mammals. Although this has received only passing mention in
temperate, southeast Queensland study site. a, Echidna 90, female, previous accounts, it is very important in the present context.
April 2000–February 2002. Note the conspicuous period of “homeo- Echidnas can be foraging at the lowest Tb of their daily cycle,
thermy” in late September, from before laying the single egg, during
presumably shortly after emerging from the resting site. The
incubation, and for a short period thereafter, showing a capacity to
regulate Tb very tightly. b, Echidna 115, female, nonbreeding that year, lowest Tb at which Kuchel (2003) sighted an animal active and
March 2000–March 2001. foraging was 20.6⬚C. It is apparent that they have the capacity
to be active over a Tb range of more than 15⬚C. This capacity
for constitutional eurythermy is characteristic of reptiles as well
those in colder climates, and in the warmer areas hibernation and would be expected in an early, facultative endotherm, and
seems to be facultative rather than obligate. All individuals in its loss in homeothermic endotherms (that show constitutional
the Tasmanian and Australian Alps studies hibernated, every stenothermy) was identified by Cade (1964) as an important
year, for periods up to about 7 mo. Males entered earlier and correlate of the evolution of ectothermy to “homeothermy”
aroused earlier than females at both locations. In these colder (phomeothermic endothermy). It could be argued that the
climates, some (younger?) individuals remain in hibernation term “torpor” should not be used unless Tb falls below the
until well into the spring, missing the breeding season. In con- point where coordinated activity ceases. In echidnas, this would
trast, hibernation is facultative in the warm, temperate, con- be below 20⬚C (Kuchel 2003). At one level, this throws doubt
tinental climate of our study site near Stanthorpe, Queensland. on whether defining torpor according to the extent of the drop
Here, Kuchel (2003) found that hibernation occurred in nine in Tb (see Barclay et al. 2001) is sufficient, as has been discussed
out of 15 echidna-years of data, males being less likely to hi- by McKechnie and Lovegrove (2002) and Grigg (2004). At
bernate than females, and that the length of the hibernation another level, however, it is worth remembering that the terms

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
990 G. C. Grigg, L. A. Beard, and M. L. Augee

“torpor” and “hibernation,” like “poikilothermy” and “homeo- possibly the oldest?) is from muscular work, either as a con-
thermy,” are descriptive of patterns, not mechanisms. They are sequence of locomotor activity or from shivering. There is little
useful because of the descriptive information they convey but work on heat production as a by-product of activity in reptiles.
cannot be defined tightly because of the diversity of patterns What might be the significance of muscular heat production
to which they are applied. When we have sufficient under- in the evolution of endothermy in birds and mammals?
standing of the mechanisms at work that explain the observed 1. Locomotor activity. The daily increases in Tb in both short-
patterns, those terms, like poikilothermy and homeothermy (in beaked (detailed analysis by Kuchel 2003; Fig. 1) and long-
the old sense), will be replaced. beaked (Grigg et al. 2003) echidnas coincide with their daily
In short, echidnas are facultative ectotherms that, although foraging activities. Indeed, Augee (1978) considered that the
having resting metabolic rates much higher than reptiles, shiv- main source of heat in short-beaked echidnas was their mus-
ering thermogenesis, a conspicuous heat input from locomotor cular activity. The picture cannot be so simple, however, because
activity, and the capacity for homeothermic endothermy, also otherwise the daily Tb minima, at the end of rest periods, would
show a strikingly heterothermic pattern of Tb most of the time. depend upon ambient temperatures, and this is not the case
They also show striking abandonments of their normal daily (Grigg et al. 1992a; Kuchel 2003). It is comparatively easy to
pattern, in the form of periods of daily torpor at any time of imagine that the benefits of warmth gained from muscular work
the year and periods of classical winter hibernation of widely could be recognised in an evolutionary sense and lead to direct
varying duration that are longest in cooler climates. We do not selection for less visible and probably less expensive mecha-
consider that the short-beaked echidna displays a primitive or nisms of heat production that do not involve muscular work.
inadequate monotreme grade of thermoregulatory capacity. Perhaps the evolution of higher resting metabolic rates in mam-
Rather, within its lineage, and like many marsupials and euthe- mals and birds, via leaky membranes, came as the evolution
rians, it has retained and refined the capacity to tolerate low of a replacement for heat produced by muscular work. The
body temperatures and, for reasons dictated primarily by energy strong expression of heat produced by muscular work on the
management, can enter a diversity of torpor/hibernation states daily Tb cycle of echidnas may provide an example of a situation
that may not be very different physiologically from each other. that could have been present among protoendotherms.
We think, therefore, that in having many attributes that one 2. Shivering thermogenesis. Shivering thermogenesis occurs
might expect to find in an advanced protoendotherm, short- in birds and mammals (and boid snakes). It occurs in all three
beaked echidnas provide a useful, living model. classes of mammals, including our model protoendotherm, the
short-beaked echidna. It is a significant contributor to the in-
Physiological Steps in the Evolution of Endothermy creased oxygen consumption seen at ambient temperatures be-
low the thermoneutral zone. In birds and those mammals that
Any model for the evolution of endothermy must accommo- lack BAT, many authors would assume that all of that increased
date realistic interpretations of each of the elements of differ- oxygen consumption is attributable to shivering, but this may
ence between reptilian ectothermy and mammalian or avian not be so (see point 5 below).
endothermy, as well as elements carried over. The elements of 3. High resting metabolic rates. Resting metabolic rates are,
difference relate particularly to the various sources of heat pro- by definition, measured within the thermoneutral zone, and
duction and the acquisition and control of lowered thermal the heat produced as a consequence of that metabolism, fed
conductance. into a heat balance equation, will account for the maintenance
of a high and stable Tb. Importantly, because it suggests that
heat from this source is a continuing background production
Increased Biochemical Heat Production
rather than a regulated source, this heat production is an in-
We identify five endogenous sources of heat in mammals; mus- trinsic property of the tissues themselves for, when the meta-
cular work associated with locomotion and other activity, oth- bolic rate of excised tissues (liver, heart, brain) is determined,
erwise futile muscular work (shivering), heat produced from that of mammalian tissue exceeds that of reptilian tissue by a
whatever comprises basal metabolic rate (including futile work factor of approximately 5 (Else and Hulbert 1987). Whole-
associated with maintaining gradients across leaky membranes), organism metabolic rates reflect a similar difference between
heat produced by the uncoupling of oxidative phosphorylation mammals and reptiles. Further, these and other authors have
using the uncoupling protein UCP1 in brown adipose tissue explored the source of the measured difference and found
(BAT; review by Cannon and Nedergaard 2004; usually referred higher mitochondrial densities and increased leakiness in mam-
to as regulatory nonshivering thermogenesis, NST) and, quite malian membranes, in both visceral tissues and in skeletal mus-
possibly, an additional source of regulatory, controllable NST cle. Heat is produced as a consequence of the additional ionic
located in skeletal muscle, similar to that described in birds pumping necessary to maintain or reestablish concentration
(see review by Bicudo et al. [2002]). gradients (Else et al. 1996, 2004; Rolfe and Brand 1997; Rolfe
The most obvious source of increased heat production (and et al. 1999).

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
 Evolution of Endothermy and Its Diversity 991

This source of higher “background” or resting heat produc- Bicudo et al. [2002]). Ye et al. (1995, 1996), aware of the re-
tion is a major difference between the endotherms and the ported lack of BAT in marsupials, explored other possible
reptiles. Shivering (as in pythons) and muscular work (e.g., sources of NST in the macropodid marsupial Tasmanian bet-
foraging echidnas) can provide substantial periods of regulated tong, B. gaimardi. Working on an isolated, perfused hindleg
high Tb. However, understanding the evolution of endothermy preparation, they found that catecholamines modified skeletal
depends on understanding the evolution of the capacity for muscle oxygen consumption and lactate and glycerol produc-
high rates of metabolic heat production in animals at rest. This tion and concluded that the skeletal muscle vascular bed made
is particularly interesting in the context of the aerobic capacity a significant contribution to whole-body thermogenesis of bet-
hypothesis, in which it is envisaged that higher resting meta- tongs. Further work on bettongs has reinforced this hypothesis.
bolic rates evolved only as a correlate of selection for higher, Bettongs are homeothermic endotherms that lack both BAT
aerobically driven activity levels. Hayes and Garland (1995) and UCP1 (Rose et al. 1999; Kabat et al. 2003a) but that show
have drawn attention to the difficulty of validating this exper- norepinephrine-stimulated NST (Rose et al. 1998, 1999). Fur-
imentally. It seems unlikely but is difficult to prove that oth- thermore, B. gaimardi does have UCP2 and UCP3 (Kabat et
erwise futile biological work would evolve as an automatic cor- al. 2003a) and, as pointed out by these authors, although these
relate of higher aerobic capacity. In other words, heat are identified as uncoupling proteins mainly by their structural
production by futile work across leaky membranes seems to be similarity to UCP1, rat white adipose tissue is known to up-
a specific mechanism for producing heat, which could be se- regulate UCP2 in response to cold stress (Loncar 1991). Both
lected for directly, as was, presumably, thermogenic BAT in the UCP2 and UCP3 are known from both skeletal muscle and
eutherian mammals. white adipose tissue in eutherians as well, including those lack-
4. Brown adipose tissue (BAT) and nonshivering thermogenesis. ing BAT, at least as adults.
Note that BAT is not a prerequisite for endothermy nor, indeed, Comprising such a large proportion of total body tissue,
diagnostic of mammals, as was asserted by Cannon et al. (2000, skeletal muscle is a very logical potential source of regulatory
p. 387), who wrote that “our mammalian prerogative of being NST, and a catecholamine control over the production of me-
able to produce extra metabolic heat without muscular activity tabolites with the thermogenic potential of glycerol and lactate,
(i.e., nonshivering thermogenesis), is anatomically localized to metabolized to lipids and carbohydrate in the liver (Ye et al.
a unique mammalian tissue, brown adipose tissue, and, mo- 1995), provides an attractive hypothesis. Alternatively, the
lecularly, to the unique mammalian protein UCP1 (thermo- source could be in the sarcoplasmic reticulum, as in birds (Bi-
genin), uniquely found in the brown adipose tissue.” This is cudo et al. 2002). Either way, there is a need for an explanation
an important point because BAT seems not to be present in for what remains a mystery at present: the mechanism that
monotremes (Augee 1969, 1978), the metatheria (Hayward and provides controllable, regulatory NST in mammals that lack
Lisson 1992; Kabat et al. 2003a, 2003b) although controversially brown fat. It seems likely that a widespread, ancient, and con-
reported in a single dasyurid (Hope et al. 1997). BAT is also trollable mechanism of regulatory NST will be found in the
not commonly present in eutherian mammals (Rothwell and skeletal muscle of mammals. If it turns out to be the same
Stock 1985) and has not been found in birds (Brigham and mechanism in both birds and mammals, that could imply that
Trayhurn 1994). Nevertherless, Rose et al. (1999) and Kabat et endothermy in these two groups stems from a common origin
al (2003a, 2003b) have clearly demonstrated that norepineph- (Grigg 2004).
rine-stimulated NST occurs in cold-acclimated Tasmanian bet- Whether or not echidnas have the capacity for regulatory
tongs, Bettongia gaimardi, and Tasmanian devils, Sarcophilus NST is unknown at present. That they do is suggested by ob-
harrisii, both of which are homeothermic endotherms that lack servations that daily Tb minima can be independent of ambient
both BAT and UCP1. temperature (Grigg 1992a; Kuchel 2003). On the other hand,
This is not to detract from the importance of BAT as a source Augee (1978) injected an echidna with the neuromuscular-
of heat in the many species of Eutheria in which it does occur, junction-blocking drug Flaxedil and found that it was unable
and particularly those eutherian hibernators that rely on its to maintain Tb. He concluded that echidnas are dependent on
capacities during rewarming. However, arousals from hiber- heat produced by locomotor activity and shivering. However,
nation in marsupial hibernators and in the short-beaked echidnas in captivity cannot be relied on to maintain Tb when
echidna are equally impressive events and not at all dependent Ta falls (Kuchel 2003), so further work is needed.
on BAT. It seems more likely that BAT and the thermogenic
organs it comprises are a specialization that occurred within
Insulation (Peripheral Changes in Blood Flow, Redistributed
the eutherian. It is clear that the step from ectotherm to en-
Body Fat, Hair, Feathers)
dotherm did not depend on the evolution of BAT.
5. Another heat source? There may be another, as yet not There has been a great deal of discussion about which came
understood source of regulatory NST in mammals, as has ap- first, insulation or increased heat production. One argument is
parently been found in the skeletal muscle of birds (review by that insulation must have come first, because otherwise any

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
992 G. C. Grigg, L. A. Beard, and M. L. Augee

increase in heat production could not lead to higher Tb, and The Evolutionary Pathways
so the benefit could not be selected for. Alternatively, it is argued
that insulation could not have come first because an ectotherm In this review we have proposed that the first endotherms were
with insulation would be compromised in its capacity to raise facultatively endothermic and that short-beaked echidnas pro-
Tb in the ectothermic way—by basking. Cowles’s (1958) famous vide a useful model for a putative protoendotherm, albeit an
report of an experiment with an insulating mink jacket seems advanced one because it is a facultative ectotherm rather than
to make this point well. However, such arguments refer to a facultative endotherm. As different mammalian lineages
evolved to fill different niches, basic facultative endothermy
external insulation, fur or feathers, and may not apply to in-
was modified in different ways, not only leading to obligate
ternal insulation such as layers of fat. We note that an internal
endothermy in many cases but also, commonly, being retained
fat layer could serve quite well for insulation and, by way of
to various extents by modern heterotherms.
variable blood flow in the skin, could either be effective in
Benefits of having a high Tb must include thermal niche
providing insulation or be bypassed, depending on need, by expansion and generally increased metabolic efficiency and
opening up peripheral blood flow. We have discussed above higher aerobic capacity, perhaps particularly for reproduction.
the capacity of reptiles for sophisticated control over peripheral Brooding pythons present a good example where the mecha-
blood flow in response to prevailing need, and changes to blood nism for producing warmth has apparently been the result of
flow could bypass an internal insulating fatty layer, or reinstate direct selection, the benefits being accelerated development of
it, in a putative protoendotherm. the embryos and an expanded geographic distribution, despite
There is a logical step that can be made from the reptiles in the energetic cost to the brooding female. It is interesting that
this case. Reptiles are typically capital rather than income breed- energy optimization apparently favors resumption of an ec-
ers (in the sense of Bonnet et al. 1998) and store fat in con- tothermic lifestyle once brooding is complete. One of the dom-
spicuous, lobed fat bodies within the peritoneal cavity in an- inant issues accounting for the type of heterothermy/homeo-
ticipation of reproduction. A redistribution of the fat bodies thermy a mammal or bird displays must be dictated by the
could provide effective, and bypassable, insulation. management of energy rather than temperature regulation per
Echidnas are also capital breeders to a substantial extent. se. The work of Schmid (1996), Ortmann et al. (1996), and
Dausmann et al. (2000) provides excellent examples of the way
Body mass may commonly change by 20%–30% seasonally with
in which the pattern of Tb expresses the interplay between
fat buildup during the summer and loss during hibernation
energy availability, ambient temperatures, facultative ecto-
and, particularly, in the posthibernation reproductive season
thermy, and energy savings. Other factors will also be impor-
(Grigg et al. 1992a). They store a substantial amount of this
tant; for example, hibernation in echidnas may provide security
fat subcutaneously, up to 2–3 cm in thickness, where it can from predation for a significant part of the year (Grigg and
function for insulation. Despite this thick insulatory layer, Daw- Beard 2000), and body size will be a very influential factor on
son et al. (1978) and Schmidt-Nielsen et al. (1966) described energy management and, thus, on the expressed pattern of Tb
large changes in thermal conductance in both Zaglossus and (Geiser 2004). An additional aspect of energy management may
Tachyglossus that can, presumably, be ascribed to changes in be an increased life span. Hibernation in mild climates may
peripheral blood flow. stretch a life span over more breeding seasons, thus increasing
Thus, the situation in echidnas suggests that a protoendo- potential lifetime reproductive output (Grigg and Beard 2000).
therm lacking external insulation but having a layer of sub- However, because of the high selective value of functions and
cutaneous body fat and reptile-like capacities for controlling structures that have a direct impact on reproduction, regardless
peripheral blood flow could, unlike Cowles’s jacketed lizards, of cost (peacock feathers and champagne being good examples),
bask effectively and function ectothermically but also have a we agree with Farmer (2000) and Koteja (2000) that repro-
much reduced thermal conductance more typical of a mammal ductive outcomes will have been of primary importance. It is
noteworthy that the only time we see homeothermy in echidnas
or bird. Heat produced by either incremental (and selected for)
is while they are incubating an egg and shortly afterward. Also,
increases in resting metabolic rate or specific thermogenic
as pointed out by Farmer (2000) and Koteja (2000), parental
mechanisms in skeletal muscle or elsewhere could have been
care requires sustained vigorous activity. The energetic require-
retained by an internal layer of body fat. That is to say, mech-
ments of development (egg production, incubation/gestation),
anisms for heat production could have preceded the evolution nutrition (lactation/egg provisioning and parental foraging),
of external insulatory devices. This is important because it offers and protection of young could be primary selective forces for
a solution to the difficulty referred to by McNab (1978, p. 19) body temperatures affording higher metabolic rates in both
“that small endotherms cannot be directly derived from small birds and mammals.
ectotherms because of the requirement of the simultaneous Table 2 is an attempt to portray schematically how, by step-
change of thermal conductance and the rate of metabolism.” wise acquisition of physiological and behavioral mechanisms,

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
 Evolution of Endothermy and Its Diversity 993

Table 2: An example of stepwise acquisition of or change in emphasis on physiological and behavioral mechanisms by which
the diverse patterns shown by endotherms may have arisen from ectothermy
Mechanism Pattern
1 Control over emergence, sun basking, control over shut- Morning warm-up, warm and stable Tb for much of the
tling, control over peripheral blood flow, shelter seek- day, cooling to ambient temperature overnight and on
ing at day’s end, or tolerance for wide range in T, little days when emergence does not occur (e.g., poor
heat produced by maintenance of gradients across leaky weather, seasonal retreat). In many, including nocturnal
cell membranes. (Lizard-like, 20 g to ca. 20 kg.) reptiles, “constitutional eurythermy.”
2a As 1, plus internal insulation, shelter-seeking late in eve- As above, plus Tb maintained warm well into the evening,
ning when Tb falls below chosen level. (No known active season extended.
model?)
2b As 1, plus large body size. (Crocodiles, to 1,000 kg.) Daily cycles in Tb decrease as size increases, but without
endothermic heat production; even the largest extant
reptiles show marked seasonal Tb cycles and are not in-
ertial homeotherms.
2c As 1, plus facultative thermogenesis by muscular contrac- Tb raised substantially by either “shivering” or basking, to
tions (locomotion or shivering), behavior interleaved warm eggs. When not brooding, python resumes ecto-
with basking if opportunity presents, increase in oxygen thermic behavior and pattern.
consumption if Ta falls. (Boid pythons.)
3 As 1, and 2a and/or 2c, plus modest increase in resting Tb able to be maintained overnight, but lowered Tb often
MR from thermogenesis in leakier cell membranes, dictated by energetic considerations. Active season ex-
perhaps some regulatory NST, maybe no need to seek tended, more energy for reproductive activities.
shelter during the night in warm seasons. (No known
model? Early protoendotherm.)
4a Cool climate: As 1–3, plus significant resting MR (and Tb warm and stable throughout day, regular/occasional
regulatory NST?); no need to seek shelter in active sea- torpor, active season extended further, facultative win-
son, energetics favors seeking a retreat in cool season ter hibernation.
(facultative hibernation/torpor). (Echidna-like, except
above snowline; cynodonts?)
4b Warm climate: As 1–3, 4a. Energetic balance favors year As 4a except climate too warm for torpor/hibernation to
round activity (no hibernation/torpor). (Echidna in be effective for energy savings.
warm climate? Cynodonts?)
4c Cold climate: As 1–3, 4a. Energetics demands seeking a As 4a except Tb warm and stable Tb throughout day and
retreat in winter (obligate hibernation). (“Classic” hi- night during the active season, entry to prolonged
bernators, echidnas above snowline, small to medium hibernation in winter to cope with cold/food
eutherian and marsupial torpidators/hibernators) unavailability.
5 Various climates: As 1–3, 4a–4c plus significant regulatory Tb essentially stable day and night and throughout the
NST heat production; energetic balance mitigates year.
against abandonment of endothermy at any time of the
year and favours homeothermic endothermy (body size
very relevant). “Constitutional eurythermy” lost. (All
the mammals and birds that are regarded as “typical
endotherms.”)
Note. The putative mechanisms (shown in the “Mechanism” column) are cumulative, underpinning the observed patterns of Tb (shown in the “Pattern”
column). Extant or hypothetical animal models are shown in the left column. Mechanisms 2a, 2b, and 2c and 4a, 4b, and 4c represent alternates. Note that the
table is meant to be indicative rather than comprehensive, particularly to illustrate an approach in which the acquisition of a particular mechanism leads to a
change in the pattern exhibited and how different sets of different circumstances may lead to different outcomes (modified from Grigg and Beard 2000).

the diverse patterns shown by endotherms could have arisen some may show torpor and/or hibernation. We have argued
stepwise from the basic ectothermic reptilian template. The that hibernation by echidnas in mild climates may be a strategy
model incorporates a stage or stages of facultative endothermy for economizing on energy requirements and expenditure by
in which heat-production mechanisms are selected to augment adopting periodically a cool/cold Tb, even though there is no
the heat gained by insulation and implemented selectively ac- immediate or looming food shortage (Grigg and Beard 2000).
cording to need. Facultative endotherms may thus have the In this context, echidnas and many other heterothermic mam-
energetic benefits of both ectothermy and endothermy, and mals and birds are facultative ectotherms. Referring to echidnas,

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
994 G. C. Grigg, L. A. Beard, and M. L. Augee

we envisage that facultative endothermy/ectothermy may have Augee M.L. and B.A. Gooden. 1992. Monotreme hibernation—
been important stages in the evolution of widespread homeo- some afterthoughts. Pp. 174–176 in M.L. Augee, ed. Platypus
thermy endothermy. In some mammals in particular this ca- and Echidnas. Royal Zoological Society of New South Wales,
pacity for ectothermy has become more finely tuned and ex- Sydney.
aggerated to cope with habitats that became thoroughly hostile Avery R.A. 1979. Lizards: A Study of Thermoregulation. Insti-
in the winter, resulting in obligate hibernators, such as arctic tute of Biology’s Studies in Biology 109. University Park
ground squirrels. In many other mammals, however, and in Press, Baltimore.
most of the birds, energy-optimization-related selection pres- Bakken G.S. and D.M. Gates. 1975. Heat transfer analysis of
sures, often dictated by the energetic costs of reproduction, animals: some implications for field ecology, physiology and
apparently favored abandonment of the capacity for short- or evolution. Pp. 255–290 in D.M. Gates and R.B.Schmerl, eds.
long-term torpors, the loss of the capacity for constitutional Perspectives in Biophysical Ecology. Springer, New York.
eurythermy, and the evolution of a pattern of homeothermic Bakker R.T. 1971. Dinosaur physiology and the origin of mam-
endothermy. mals. Evolution 25:636–658.
We think this is a parsimonious model. It recognizes that Barclay R.M.R., C.L. Lausen, and L. Hollis. 2001. What’s hot
mammalian and avian endotherms display very diverse thermal and what’s not: defining torpor in free-ranging birds and
patterns. It describes the evolution of this diversity, recognizes mammals. Can J Zool 79:1885–1890.
that the evolution of endothermy and the evolution of homeo- Barnes B.M. 1989. Freeze avoidance in a mammal: body tem-
thermy are not the same thing, and proposes that the second peratures below 0⬚C in an arctic hibernator. Science 244:
step, where it occurs, follows the first. It explains how avian 1593–1595.
and mammalian endothermy, in all of their diversities, could Bartholomew G.A. and V.A. Tucker. 1963. Control of changes
have arisen from reptilian ectothermy with most of the incre- in body temperature, metabolism, and circulation in the
ments being quantitative rather than qualitative. It acknowl- Agamid lizard, Amphibolorus barbatus. Physiol Zool 36:199–
edges that the step to endothermy may have been made a 218.
number of times, and it could accommodate a step back to Beard L.A. and G.C. Grigg. 2000. Reproduction in the short-
full-time ectothermy, as proposed for crocodiles (Seymour et beaked echidna, Tachyglossus aculeatus: field observations at
al. 2004). It is a model that accommodates diversity and that
an elevated site in south-east Queensland. Proc Linn Soc N
recognizes that there need not be only one specific, detailed
S W 122:89–99.
model, but that different expressions of extent may have been
Beard L.A., G.C. Grigg, and M.L. Augee. 1992. Reproduction
achieved by different selection pressures in different groups in
by echidnas in a cold climate. Pp. 93–100 in M.L. Augee,
response to different circumstances, and at different times.
ed. Platypus and Echidnas. Royal Zoological Society of New
South Wales, Sydney.
Acknowledgments Bennett A.F. 1991. The evolution of activity capacity. J Exp Biol
160:1–23.
We are grateful to Louise Kuchel, Peter Brice, and David Ellis Bennett A.F. and J.A. Ruben. 1979. Endothermy and activity
for helpful discussions. Our research has been supported at in vertebrates. Science 206:649–654.
various times by the Australian Research Council and from Bicudo J.E.P.W., A.C. Bianco, and C.R. Vianna. 2002. Adaptive
University of Queensland Research Grants. thermogenesis in hummingbirds. J Exp Biol 205:2267–2273.
Bonnet X., D. Bradshaw, and R. Shine. 1998. Capital versus
Literature Cited income breeding: an ectothermic perspective. Oikos 83:333–
342.
Aloia R.C., M.L. Augee, G.R. Orr, and J.K. Raison. 1986. A Boyer B.B. and B.M. Barnes. 1999. Molecular and metabolic
critical role for membranes in hibernation. Pp. 19–26 in H.C. aspects of mammalian hibernation. BioScience 49:713–724.
Heller, X.J. Musacchia, and L.C.H. Wang, eds. Living in the Brice P.H., G.C. Grigg, L.A. Beard, and J.A. Donovan. 2002a.
Cold: Physiological and Biochemical Adaptations. Elsevier, Heat tolerance of short-beaked echidnas (Tachyglossus acu-
New York. leatus) in the field. J Therm Biol 27:449–457.
Augee M.L. 1969. Temperature Regulation and Adrenal Func- ———. 2002b. Patterns of activity and inactivity in echidnas
tion in the Echidna. PhD thesis. Monash University. (Tachyglossus aculeatus) free-ranging in a hot dry climate:
———. 1976. Heat tolerance of monotremes. J Therm Biol 1: correlates with ambient temperature, time of day and season.
181–184. Aust J Zool 50:461–475.
———. 1978. Monotremes and the evolution of homeothermy. Brigham M.R. and P. Trayhurn. 1994. Brown fat in birds? a test
Pp. 111–120 in M.L. Augee, ed. Monotreme Biology. Royal for the mammalian BAT-specific mitochondrial uncoupling
Zoological Society of New South Wales, Sydney. protein in common poorwills. Condor 96:208–211.

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
 Evolution of Endothermy and Its Diversity 995

Cade T.J. 1964. The evolution of torpidity in rodents. Ann Acad bernation Symposium. Biological Papers of the University
Sci Fenn Ser A IV Biol 71:77–112. of Alaska 27. University of Alaska, Fairbanks.
Cannon B., V. Golozoubova, A. Matthias, K.E. Ohlsson, A. Grigg G.C., M.L. Augee, and L.A. Beard. 1992a. Thermal re-
Jacobsson, and J. Nedergaard. 2000. Is there life in the cold lations of free-living echidnas during activity and in hiber-
without UCP1? uncoupling proteins and thermoregulatory nation in a cold climate. Pp. 160–173 in M.L. Augee, ed.
thermogenesis. Pp. 387–400 in G. Heldmaier and M. Kling- Platypus and Echidnas. Royal Zoological Society of New
enspor, eds. Life in the Cold. Springer, Berlin. South Wales, Sydney.
Cannon B. and J. Nedergaard. 2004. Brown adipose tissue: Grigg G.C. and L.A. Beard. 2000. Hibernation by echidnas in
function and physiological significance. Physiol Rev 84:277– mild climates: hints about the evolution of endothermy? Pp.
359. 5–20 in G. Heldmaier and M. Klingenspor, eds. Life in the
Cowles R.B. 1958. Possible origin of dermal temperature reg- Cold. Springer, Berlin.
ulation. Evolution 12:347–357. Grigg G.C., L.A. Beard, and M.L. Augee. 1989. Hibernation in
Crompton A.W., C.R. Taylor, and J.A. Jagger. 1978. Evolution a monotreme, the echidna Tachyglossus aculeatus. Comp
of homeothermy in mammals. Nature 272:333–336. Biochem Physiol 92A:609–612.
Dausmann K.H., J.U. Ganzhorn, and G. Heldmaier. 2000. Body Grigg G.C., L.A. Beard, J.A. Barnes, L.I. Perry, G.J. Fry, and M.
temperature and metabolic rate of a hibernating primate in Hawkins. 2003. Body temperature in captive long-beaked
Madagascar: preliminary results from a field study. Pp. 41– echidnas (Zaglossus bartoni). Comp Biochem Physiol 136A:
48 in G. Heldmaier and M. Klingenspor, eds. Life in the 911–916.
Cold. Springer, Berlin. Grigg G.C., L.A. Beard, T.R. Grant, and M.L. Augee. 1992b.
Dawson T.J., D. Fanning, and T.J. Bergin. 1978. Metabolism Body temperature and diurnal activity patterns in the plat-
and temperature regulation in the New Guines monotreme ypus, Ornithorhynchus anatinus, during winter. Aust J Zool
Zaglossus bruijni. Aust Zool 20:99–104. 40:135–142.
Eisentraut M. 1960. Heat regulation in primitive mammals and Grigg G.C. and F. Seebacher. 1999. Field test of a paradigm:
tropical species. Bull Mus Comp Zool 124:31–43. hysteresis of heart rate in thermoregulation by a free-ranging
lizard, Pogona barbata. Proc R Soc Lond Ser B Biol Sci 266:
Else P.L. and A.J. Hulbert. 1987. Evolution of mammalian en-
1–7.
dothermic metabolism: “leaky” membranes as a source of
Grigg G.C., F. Seebacher, L.A. Beard, and D. Morris. 1998.
heat. Am J Physiol 253:R1–R7.
Thermal relations of large crocodiles, Crocodylus porosus,
Else P.L., N. Turner, and A.J. Hulbert. 2004. The evolution of
free-ranging in a naturalistic situation. Proc R Soc Lond Ser
endothermy: role for membranes and molecular activity.
B Biol Sci 265:1793–1799.
Physiol Biochem Zool 77:950–958.
Harlow P. and G.C. Grigg. 1984. Shivering thermogenesis in
Else P.L., D.J. Windmill, and V. Markus. 1996. Molecular ac-
the brooding diamond python, Python spilotes spilotes.
tivity of sodium pumps in endotherms and ectotherms. Am
Copeia 1984:959–965.
J Physiol 271:R1287–R1294.
Hayes J.P. and T. Garland, Jr. 1995. The evolution of endo-
Farmer C.G. 2000. Parental care: the key to understanding en-
thermy: testing the aerobic capacity model. Evolution 49:
dothermy and other convergent features in birds and mam- 836–848.
mals. Am Nat 155:326–334. Hayward J.S. and P.A. Lisson. 1992. Evolution of brown fat: its
Gans C. and F.H. Pough. 1982. Biology of the Reptilia. Vol.12. absence in marsupials and monotremes. Can J Zool 70:171–
Physiology C: Physiological Ecology. Academic Press, New 179.
York. Heath J.E. 1968. Origins of thermoregulation. Pp. 259–278 in
Geiser F. 2004. Metabolic rate and body temperature reduction E. T. Drake, ed. Evolution and Environment. Yale University
during hibernation and daily torpor. Annu Rev Physiol 66: Press, New Haven, Conn.
239–274. Heinrich B. 1977. Why have some animals evolved to regulate
Geiser F., N. Goodship, and C.R. Pavey. 2002. Was basking a high body temperature? Am Nat 111:623–640.
important in the evolution of mammalian endothermy? Na- Hillenius W.J. 1994. Turbinates in therapsids: evidence for late
turwissenschaften 89:412–414. Permian origins of mammalian endothermy. Evolution 48:
Griffiths M. 1968. Echidnas. Pergamon, London. 207–229.
———. 1978. The Biology of Monotremes. Academic Press, Hope P.J., D. Pyle, C.B. Daniels, I. Chapman, M. Horowitz,
New York. J.E. Morley, P. Trayhurn, J. Kumaratilake, and G. Wittert.
Grigg G.C. 2004. An evolutionary framework for studies of 1997. Identification of brown fat and mechanisms for energy
hibernation and short term torpor. Pp. 1–11 in M. Barnes balance in the marsupial Sminthopsis crassicaudata. Am J
and H.V. Carey, eds. Life in the Cold: Evolution, Adaptation, Physiol 273:R161–R167.
Mechanisms, and Applications. Twelfth International Hi- Hulbert A.J. 1980. The evolution of energy metabolism in

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
996 G. C. Grigg, L. A. Beard, and M. L. Augee

mammals. Pp. 129–139 in K. Schmidt-Nielsen, L. Bolis, and to the cold? Pp. 7–12 in F. Geiser, A.J. Hulbert, and S.C.
C.R. Taylor, eds. Comparative Physiology: Primitive Mam- Nicol, eds. Adaptations to the Cold. Tenth International Hi-
mals. Cambridge University Press, Cambridge. bernation Symposium. University of New England Press, Ar-
Hutchison V.H., H.D. Dowling, and A. Vinegar. 1966. Ther- midale, New South Wales.
moregulation in a brooding female Indian python, Python ———. 2000. Patterns of hibernation of echidnas in Tasmania.
molurus. Science 151:694–696. Pp. 21–28 in G. Heldmaier and M. Klingenspor, eds. Life in
Kabat A.P., R.W. Rose, J. Harris, and A.K. West. 2003a. Mo- the Cold. Springer, Berlin.
lecular identification of uncoupling proteins (UCP2 and Ortmann S., J. Schmid, J.U. Ganzhorn, and G. Heldmaier. 1996.
UCP3) and absence of UCP1 in the marsupial Tasmanian Body temperature and torpor in a Malagasy small primate,
bettong, Bettongia gairmardi. Comp Biochem Physiol 134B: the mouse lemur. Pp. 55–61 in F. Geiser, A.J. Hulbert, and
71–77. S.C. Nicol, eds. Adaptations to the Cold. Tenth International
Kabat A.P., R.W. Rose, and A.K. West. 2003b. Non-shivering Hibernation Symposium. University of New England Press,
thermogenesis in a carnivorous marsupial, Sarcophilus har- Armidale, New South Wales.
risii, in the absence of UCP1. J Therm Biol 28:413–420. Rismiller P.D. and M.W. McKelvey. 1996. Sex, torpor and ac-
Koteja P. 2000. Energy assimilation, parental care and the evo- tivity in temperate climate echidnas. Pp. 23–30 in F. Geiser,
lution of endothermy. Proc R Soc Lond Ser B Biol Sci 267: A.J. Hulbert, and S.C. Nicol, eds. Adaptations to the Cold.
479–484. Tenth International Hibernation Symposium. University of
Kuchel L. 2003. The Energetics and Patterns of Torpor in Free- New England Press, Armidale, New South Wales.
Ranging Echidnas (Tachyglossus aculeatus) from a Warm ———. 2000. Spontaneous arousal in reptiles? body temper-
Temperature Climate. PhD thesis. University of Queensland. ature ecology of Rosenberg’s goanna, Varanus rosenbergi. Pp.
Lane J.E., R.M. Brigham, and D.L. Swanson. 2004. Daily torpor 57–64 in G. Heldmaier and M. Klingenspor, eds. Life in the
in free-ranging whip-poor-wills (Caprimulgus vociferus). Cold. Springer, Berlin.
Physiol Biochem Zool 77:297–304. Rolfe D.F.S. and M.D. Brand. 1997. The physiological signifi-
Loncar D. 1991. Convertible adipose tissue in mice. Cell Tissue cance of mitochondrial proton leak in animal cells and tis-
Res 266:49–161. sues. Biosci Rep 17:9–16.
Lovegrove B.G., M.J. Lawes, and L. Roxburgh. 1999. Confir- Rolfe D.F.S., J.M.B. Newman, J.A. Buckingham, M.G. Clark,
mation of plesiomorphic daily torpor in mammals: the and M.D. Brand. 1999. Contribution of mitochondrial pro-
round-eared elephant shrew Macroscelides proboscideus (Ma- ton leak to respiration rate in working skeletal muscle and
croscelidae). J Comp Physiol 169:453–460. liver and to SMR. Am J Physiol 276:C692–C699.
Lyman C.P. and D.C. Blinks. 1959. The effect of temperature Rose R.W., N. Kuswanti, and E.Q. Colquhoun. 1998. The de-
on the isolated hearts of closely related hibernators and non- velopment of endothermy in the Tasmanian bettong and the
hibernators. J Cell Comp Physiol 54:53–63. effects of cold and noradrenaline. J Comp Physiol 168:359–
Malan A. 1996. The origins of hibernation: a reappraisal. Pp. 363.
1–6 in F. Geiser, A.J. Hulbert, and S.C. Nicol, eds. Adapta- Rose R.W., A.K. West, J.M. Ye, G.H. McCormick, and E.Q.
tions to the Cold. Tenth International Hibernation Sympo- Colquhoun. 1999. Non-shivering thermogenesis in the mar-
sium. University of New England Press, Armidale, New supial (Bettongia gaimardi) is not attributable to brown ad-
South Wales. ipose tissue. Physiol Biochem Zool 72:699–704.
Manning B. and G.C. Grigg. 1997. Basking behaviour is not of Rothwell N.J. and M.J. Stock. 1985. Biological distribution and
thermoregulatory significance to the “basking” freshwater significance of brown adipose tissue. Comp Biochem Physiol
turtle Emydura signata. Copeia 1997:579–584. 82A:745–751.
Martin C.J. 1903. Thermal adjustment and respiratory exchange Rowe T. 1992. Phylogenetic systematics and the early history
in monotremes and marsupials. Philos Trans R Soc Lond Ser of mammals. Pp. 129–145 in F.S. Szaly and M.C. McKenna,
B Biol Sci 195:1–37. eds. Mammalian Phylogeny. Springer, Berlin.
McKechnie A.E. and B.G. Lovegrove. 2002. Avian facultative Ruben J.A. 1995. The evolution of endothermy in mammmals
hypothermic responses: a review. Condor 104:705–724. and birds: from physiology to fossils. Annu Rev Physiol 57:
McNab B.K. 1978. The evolution of endothermy in the phy- 69–95.
logeny of mammals. Am Nat 112:1–21. ———. 1996. Evolution of endothermy in mammals, birds
Nicol S. and N.A. Andersen. 1993. The physiology of hiber- and their ancestors. Pp. 359–393 in I. Johnston and A.F.
nation in an egg-laying mammal, the echidna. Pp. 56–64 in Bennett, eds. Animals and Temperature. Cambridge Uni-
C. Carey, G.L. Florant, B.A. Wunder, and B. Horwitz, eds. versity Press, London.
Life in the Cold: Ecological, Physiological and Molecular Ruben J.A. and T.D. Jones. 2000. Selective factors associated
Mechanisms. Westview, Boulder, Colo. with the origin of fur and feathers. Am Zool 40:585–596.
———. 1996. Hibernation in the echidna: not an adaptation Schmid J. 1996. Oxygen consumption and torpor in mouse

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions
 Evolution of Endothermy and Its Diversity 997

lemurs (Microcebus murinus and M. myoxinus): preliminary Tosini G. and M. Menaker. 1995. Circadian rhythm of body
results of a study in western Madagascar. Pp. 47–54 in F. temperature in an ectotherm (Iguana iguana). J Biol
Geiser, A.J. Hulbert, and S.C. Nicol, eds. Adaptations to the Rhythms 10:248–255.
Cold. Tenth International Hibernation Symposium. Univer- Wang L.C.H. 1989. Ecological, physiological and biochemical
sity of New England Press, Armidale, New South Wales. aspects of torpor in mammals and birds. Pp. 361–401 in
Schmidt-Nielsen K., T.J. Dawson, and E.C. Crawford. 1966. L.C.H. Wang, ed. Advances in Comparative and Environ-
Temperature regulation in the echidna (Tachyglossus aculea- mental Physiology. Vol. 4. Springer, Heidelberg.
tus). J Cell Physiol 67:63–72. Wilz M. and G. Heldmaier. 2000. Comparison of hibernation,
Seebacher F. and G.C. Grigg. 1997. Patterns of body temper-
estivation and daily torpor in the edible dormouse, Glis glis.
ature in wild freshwater crocodiles, Crocodylus johnstoni:
J Comp Physiol B 170:511–521.
thermoregulation vs. conformity, seasonal acclimatisation
Ye J.M., S.J. Edwards, R.W. Rose, S. Rattigan, M.G. Clark, and
and the effect of social interactions. Copeia 1997:549–557.
E.Q. Colquhoun. 1995. Vasoconstrictors alter oxygen, lactate
Seebacher F., G.C. Grigg, and L.A. Beard. 1999. Crocodiles as
dinosaurs: behavioural thermoregulation in very large ecto- and glycerol metabolism in perfused muscle of rat-kangaroo
therms leads to high and stable body temperatures. J Exp (Bettongia gaimardi: Marsupialia). Am J Physiol 268:R1217–
Biol 202:77–86. R1223.
Seymour R.S., C.L. Bennett-Stamper, S.D. Johnston, D.R. Car- Ye J.M., S.J. Edwards, R.W. Rose, J.T. Steen, M.G. Clark, and
rier, and G.C. Grigg. 2004. Evidence for endothermic an- E.Q. Colquhoun. 1996. Alpha-adrenergic stimulation of ther-
cestors of crocodiles at the stem of archosaur evolution. Phys- mogenesis in a rat kangaroo (Marsupialia: Bettongia gai-
iol Biochem Zool 77:1051–1067. mardi). Am J Physiol 40:R586–R592.

This content downloaded from 157.89.65.129 on Sun, 24 Mar 2013 11:26:07 AM

All use subject to JSTOR Terms and Conditions