<<

Bennett, A. F. 1997. and the of physiological characters, pp. 3-16. In: Handbook of , Sect. 13: . W. H. Dantzler, ed. Oxford Univ. Press, New York. 1. Adaptation and the evolution of physiological characters

Department of and Evolutionary , University of California, ALBERT F. BENNETT 1 Irvine, California

among the biological sciences (for example, behavioral CHAPTER CONTENTS [I241). The Many meanings of "Adaptationn In general, comparative physiologists have been Criticisms of Adaptive Interpretations much more successful in, and have devoted much more Alternatives to Adaptive Explanations to, pursuing the former rather than the latter Historical inheritance goal (37). Most of this Handbook is devoted to an Developmentai pattern and constraint Physical and biomechanical correlation examination of mechanism-how various physiologi- Phenotypic size correlation cal systems in various animals. Such compara- Genetic correlations tive studies are usually interpreted within a specific Chance fixation evolutionary context, that of adaptation. That is, or- Studying the Evolution of Physiological Characters ganisms are asserted to be designed in the ways they Macroevolutionary studies Microevolutionary studies are and to function in the ways they do because of Incorporating an Evolutionary Perspective into Physiological Studies which results in evolutionary change. The principal textbooks in the field (for example, refs. 33, 52, 102, 115) make explicit reference in their titles to the importance of adaptation to comparative COMPARATIVE PHYSIOLOGISTS HAVE TWO GOALS. The physiology, as did the last comparative section of this first is to explain mechanism, the study of how - Handbook (32). Adaptive evolutionary explanations isms are built functionally, "how animals work" (113). are woven throughout the fabric of the science. Studies of this type are undertaken at a variety of However, the assumption that adaptation is the sole organizational levels: organismal, organ system, organ, explanation for natural diversity in living systems has , , or molecular and biochemical. Ideally, a been called into question (54, 80, 81, 82, 132, 133) study will span several of these, integrating information and has become a point of contention among many from lower to higher levels and illuminating mecha- organismal and evolutionary . The rote appli- nisms from higher to lower. The second goal is to cation of adaptive explanations for all physiological explain the diversity of functional systems found phenomena has effectively uncoupled comparative among . In contrast to "how" questions, this (ecological) physiology from . goal attempts to answer "why" questions: Why is an Evolutionary biologists generally ignore and are igno- built the way it is and why does it function rant of comparative physiology. Their courses do not in a particular way and not in another? How and why discuss the evolution of physiological characters, and do physiological properties change through time? major textbooks in evolution (for example, ref. 43) The first goal entails equilibrium (in Lauder's sense contain virtually no reference to physiological phenom- [74])studies-that is, those which examine the present- ena. Comparative physiology should be a field central day associations among characters and do not under- to evolutionary biology, given the importance of the take causal (in the sense of historical) explanations. In functional characters that it studies and its avowed contrast, the second goal requires transformational or evolutionary orientation. If comparative physiology is evolutionary, explanations and hence must deal, at to make a serious contribution to evolutionary biologj., least implicitly, with the ecology, , and history it must rethink its historical commitment to adapta- of the organism and its . For both goals, a tionism. primary reference point is the function of an organism An increasing number of studies have tested alterna- In its . These goals are not unique tives to adaptive explanations and examined the evolu- to comparative physiology; they are broadly shared tion of functional and structural characters. A diversity 4 HANDBOOK OF PHYSIOLOGY-COMPARATIVE PHYSIOLOGY of approaches is now available for the study of the be the state of having become familiar with surround- evolution of physiological traits and mechanisms. ings. This usage is colloquial, as in permitting an ani- Comparative physiologists are now almost uniquely mal to adapt to experimental conditions. It usually positioned to utilize these approaches to understand refers to placing an organism in an apparatus or envi- how important functional characters have changed and ronment prior to beginning experimental measure- may change through time. The types of organismal ments to minimize fright or activity. (4) Adaptation can character that interest comparative (ecological) physi- be a functional change in an organism after exposure ologists (for example, energy exchange, thermoregula- to new conditions or a new environment. Organisms tion, and locomotor performance) are precisely those frequently reorganize their physiological systems, par- that everyone expects will be of importance to ticularly their rate processes, in response to changed and evolutionary success. Such studies must test (and conditions. The most familiar of these responses is not assume) the role of adaptation in shaping those temperature adaptation (acclimation or acclimatiza- characters. tion), but the term is also used to describe alteration of The purpose of this chapter is to examine what we muscle structure and function to altered stress patterns, know about the evolution of physiological systems and changes in osmotic and ionic regulation in different the role of adaptation in shaping them. First, the several environmental media, etc. The term applies to such meanings of the term "adaptation" for physiologists changes in both laboratory and natural conditions. are explored and their relevance to character evolution The documentation and physiological basis of such discussed. Then, alternatives to, or constraints on, ad- plasticity is a particularly popular area of physiological aptation are examined. Methods for the study of the study (61, 100). (5) Adaptation can be the process of evolution of physiological characters in both the field improvement of fitness in a population of organisms and the laboratory are then discussed, as well as their in response to natural selection. This sense refers to importance to future directions in comparative physi- the Darwinian evolutionary process, and it is in this ology. sense that most evolutionary biologists use the term. It assumes that individuals with characteristics favored in their particular environment will reproduce more THE MANY MEANINGS OF "ADAPTATION" and that the favored characters will increase in fre- quency over generations. The literature in evolutionary Physiologists and other organismal biologists employ biology dealing with the intricacies of this concept is the term "adaptation" to describe an especially wide large (readers are referred to the following books and variety of different biological responses, only some of articles for a more comprehensive discussion of evolu- which actually refer to evolutionary phenomena. Many tionary adaptation: 7, 16-18, 81, 87, 89, 94, 104, of these uses are not widely shared with, or understood 122, 132). (6) Adaptation can involve a character that by, other biologists, a situation that easily may lead to originated due to natural selection. This usage refers confusion and misunderstanding. This section clarifies to a structure or function that is the product of the the multitude of uses and definitions of this word. preceding process. In this sense, a nasal salt gland A physiologist or organismal might well may be an adaptation resulting from the process of use adaptation in any of the following senses: (1) adaptation (definition 5) to a saline environment. (7) Adaptation can be the alteration of a sensory or ner- Adaptation can involve a preexisting character that is vous response under constant . The rate of beneficial to an organism in a particular environment firing of sensory cells or neurons usually changes or circumstance. The distinction between this and the acutely with the onset or offset of a stimulus. If that previous definition is whether the character has been a stimulus is maintained, the response may either de- target for selection in the current environment or is a crease, as in pressure sensation, or increase over time, feature that is adventitiously useful. Preadaptation, as in dark accommodation. Both are termed adapta- protoadaptation (44), and (55) connote an- tion. (2) Adaptation can be a syndrome of physiological cestral features previously acquired that find a new responses to environmental stress. Specifically, this us- utility in new circumstances. (8) Adaptation can in- age refers to the General Adaptation Syndrome (116) volve a generally and phylogenetically widespread in vertebrates, in which a variety of stressors may structure or function necessary for, or promoting, sur- activate release of adrenal hormones with widespread vival, a feature of many living systems, sometimes physiological effects, including raising pressure, crucial for existence but not evolved in reference to stimulating glucose mobilization, increasing heart rate, particular environmental circumstances in extant popu- etc. If prolonged, the response may result in adverse lations. For example, the use of DNA as a genetic physiological conditions and . (3) Adaptation can material may be considered an adaptation for repro- CHAPTER 1: 5 duction or mitochondria as for aerobic res- adaptation may not necessarily predispose organisms piration. to greater competitive fitness. For example, bacteria Given this diversity of definitions, physiologists acclimated to a lower temperature (32°C) outcompete should be particularly careful to clarify the sense in the same bacterial acclimated to a higher temperature which they are using the word "adaptation." It is (42°C) in both 32" and 42°C environments (79). The important to recognize that all of these are correct assumption that phenotypic adaptations are necessarily definitions in the sense that these are ways in which beneficial to the individuals in which they occur (29, the word is actually used. We might regard some of 63, 106) is unwarranted and susceptible to the same these uses, particularly definitions 3 and 8, as unfortu- criticisms brought against the assumption of evolution- nate but not incorrect. An evolutionary biologist might ary adaptation (79; see later under CRITICISMS OF insist that only definitions 5 and 6 are appropriate (and ADAPTIVE INTERPRETATIONS). these are the senses in which it is generally used in this A problem with the plethora of ways in which "ad- chapter). Such insistence, however, will not change aptation" is used by physiologists and other organismal common usage. biologists is the blurring of distinctions between them This set of definitions divides into two distinct cate- and the assumption that all are related or equivalent. gories, phenotypic and evolutionary adaptation: the Adaptation of a sensory cell (definition 1) or of an first four refer to phenotypic alterations in individual organ system (definition 4) may or may not be evolu- organisms and the second four to multigenerational tionary adaptations according to definition 6 or 7. The changes in of organisms. (This is the dis- distinctions among these categories should be kept tinction made previously by Bock and von Wahlert clearly in , and one should not be assumed to [17] between physiological and evolutionary adapta- imply the other. Investigations of phenotypic adapta- tion.) The first set encompasses changes in organisms tion (definitions 1-4) should study both their occur- ranging from very rapid to long-term developmental rence and their mechanistic basis (the most usual type events that occupy the entire lifetime of an individual. of study, for example, refs. 61,100) and their provision In view of the very different of organismal and of benefit to the organism (less commonly examined population processes, it is quite possible that pheno- [67]). Interpretations of evolutionary adaptation of typic adaptation may have little in common with evolu- physiological and morphological characters and their tionary changes and processes, even in response to study are the topic of the remainder of this chapter. similar environmental perturbations. For instance, mechanisms of temperature adaptation-in definitions 4 and 6-are usually quite distinct, the former involv- CRITICISMS OF ADAPTIVE INTERPRETATIONS ing alterations in concentration and the latter protein composition (61). Evolutionary adaptation Adaptation as the explanation for a biological phenom- clearly involves a genetic process. Phenotypic adapta- enon is easier to invoke than to demonstrate. For many tion may also have a genetic basis (20, 109, 120), for characters, a satisfactory investigation into evolution- example, differential gene activation (4, 71). ary development may be difficult or impossible. The Is there a unifying theme in this diversity of defini- response to this difficulty has often been to claim, tions? The only common feature I see is change through rather than to investigate, adaptation. A mechanistic time or biological change in response to environmental study of a morphological or physiological character change. Temporal variation is the element that distin- frequently is concluded with the assertion of a plausible guishes adaptive from more static biological processes. means through which natural selection might have One might be tempted to add "beneficial" to these operated to produce the evolution of the character, common definitions, implying that these changes im- generally by contending that the character benefitted prove the functioning and/or fitness of the organisms survivorship and . Indeed, the American in a changing environment. That is probably often true Journal of Physiology: Regulatory, Integrative and for definitions 1-4 and is certainly true for definitions Comparative Physiology specifically encourages such 5-8. However, the mere fact of phenotypic alteration speculation (30). The problem with such assertions is in a new environment cannot be taken as evidence of that they often amount to dogmatic appeals to a partic- benefit. The fact that a lobster turns bright red in ular evolutionary mechanism rather than a testable boiling water does not imply that becoming red is a scientific hypothesis. Any such interpretation is only positive reaction to minimize heat stress. Such effects one of a multitude of different possible adaptive scenar- may be direct responses to the environmental change ios, let alone nonadaptive explanations, for the posses- and may be neutral or even detrimental in their impact sion of a particular trait by a particular organism (80). on competitive fitness (62, 67, 79, 95). Phenotypic This tendency to claim adaptation as the universal 6 HANDBOOK OF PHYSIOLOGY-COMPARATIVE PHYSIOLOGY explanation for biological diversity has evoked critical tion that adaptive explanations are only rote applica- attack, not from opponents of evolutionary theory tions of articles of faith undercuts the idea of a support- but from evolutionary biologists themselves. George ive linkage between comparative and evolutionary Williams (132) maintained that "evolutionary adapta- biology. If any character could unfailingly be explained tion is a special and onerous concept that should not by adaptation, it could equally well be explained by be used unnecessarily" and discussed several alternative some other cause, such as special creation. This is explanations for the possession of biological traits. uncomfortably close to the pre-Darwinian situation, in Lewontin (81) and Gould and Lewontin (54) explicitly which biologists regarded the uniqueness and perfec- challenged the uncritical application of adaptation as tion of organismal design as evidence of an omnipotent a dogmatic assertion. In the latter article, they attacked creator. The suggestion that organismal biologists sim- what they characterized as the "adaptationist pro- ply exchanged one religious faith for another under- gram." They elevated the latter to a philosophical mines their evolutionary credentials, or at least suggests approach to biology and an agenda for its study and that their approach and thinking have not been funda- proceeded to criticize it as failing to consider alternative mentally altered by evolutionary thought. To be uncon- explanations for the evolutionary development of bio- nected to evolution is to be unconnected to the central logical characters. biological principle. Particular criticism (54, 83) was directed to the asser- Third, a particularly disturbing aspect of the critique tion of optimality. Optimality approaches have played was that such assertions of adaptation are not only a prominent role in comparative biology-for example, unjustified but also unscientific. In other words, if in ecological energetics (125) and respiratory physiol- adaptive statements are not presented as hypotheses ogy (symmorphosis; 123, 131). Gould and Lewontin susceptible to falsification, the enterprise is not science. (54) maintained that the adaptationist program seeks What could be a more damning charge to a ? to demonstrate an optimal design of each organismal Little wonder then that these critiques, challenging character. Any failure to find an optimal design in a fundamental values and interpretations of comparative particular trait is explained as a necessary compromise biology, were taken so seriously. with other traits, resulting in an organism that is de- Finally, the criticisms were delivered partly as lam- signed as well as possible. Again, the failure to test poons, characterizing the approach as no more than a these assertions was criticized. The entire approach series of "just so" stories that might have been invented was characterized (or caricatured) as "panglossian," a by Rudyard Kipling. Mockery perturbs beyond mere reference from to Dr. Pangloss, who believed criticism. that everything happened for the best in this best of all The response of the to this criticism has possible worlds. (Quoting to lampoon adapta- been mixed. Some (for example refs. 1, 23, 88, 102, tionism is not original to Gould and Lewontin, but 103, 121, 133), with varying levels of success, continue dates back to the turn of the century [6]). to grapple with the concept of adaptation to make it These critiques were widely influential and caused more palatable to organismal and comparative biolo- considerable consternation and controversy within the gists. Some have simply ignored the controversy and comparative and organismal communities. Why was continue to operate within an adaptationist frame- this issue considered so important and troubling to work. Others (for example, 19, 38, 59) have accepted comparative biologists? Other criticisms of organismal the essentials of the criticism and have attempted more biology, such as the irrelevance of integrative or sys- direct analyses of the adaptive process. These investiga- temic studies in comparison to molecular studies, have tions examine adaptation as a hypothesis rather than not perturbed the community in similar ways. There as an assertion (37) and undertake comparative and are several reasons for this concern. experimental studies on the of func- First, these critiques centered on the philosophical tional characters. The approaches of some of these and interpretive bases of the field, not on its choice of studies in evolutionary physiology are presented below, subject matter. Such fundamental criticisms cannot be but before that, alternative hypotheses to adaptive ex- dismissed as mere subjective prejudice. planations are discussed. Second, the critiques exposed the rather tenuous connections between comparative biology as com- monly practiced and evolution. Most comparative biol- ALTERNATIVES TO ADAPTIVE EXPLANATIONS ogists believe in the importance of evolution as the grand unifying concept in biology and think that their There are many possible explanations for the posses- work both is in accord with it and provides support sion of a particular trait or character by an organism. for it as a science. Gould and Lewontin's (54) imputa- Adaptation in the sense of definition 5 above is only CHAPTER 1: EVOLUTIONARY PHYSIOLOGY 7 one of them. Several categories of alterative explana- have a new utility and benefit not present ancestrally, in tion, many of them cited by Gould and Lewontin (54), which case it is referred to as a preadaptation or are discussed below. All involve explanations for the exaptation (55; see ref. 12 for an example). correlation of phenotypic characters within individuals Analysis of a character within a phylogenetic frame- or between characters and environmental factors. Inter- work can aid in the determination of its historical and pretation of correlation in comparative studies is al- evolutionary development (19, 21,22, 49, 59, 64, 75). ways difficult (52), particularly in regard to unravelling For example, the pattern of concordance between the the basis of the evolution of physiological characters. occupation of a specific environment and the appear- Some of these explanations are subject to experimental ance of the character in a group of related taxa can testing and thus permit an investigation of alterative help determine whether the character might be an hypotheses to adaptation. Rejection, or at least consid- evolutionary response to the current environment of eration, of such alternative hypotheses is an important the organism. By way of example, a phylogenetic study step in the investigation of the evolution of a character of in Australian scincid lizards (66) and its adaptive basis (37). found that the evolution of heat tolerance was coinci- An initial difficulty, however, lies in the stipulation dent with the occupation of environments. This of exactly what constitutes a "character" (or "trait" coincidence argues against high thermotolerance being or "feature") to be investigated. Any organism can a preexisting (ancestral) condition in this group and potentially be atomized into a series of thousands of supports, but does not confirm, its adaptive nature. different traits, each requiring analysis and explana- To consider another example, desert-dwelling ring- tion. Is, for instance, the shape of every individual scale tails (Bassariscus astutus, a small carnivore in the rac- on a an individual character of importance and is coon family Procyonidae) have a low basal metabolic it shaped separately by selection or another process? rate and a high heat tolerance in comparison to most Or is there a more broadly encompassing character of other (26). Low basal metabolic rates are interest, of which the shape of a set of scales, or of all frequently found among desert-dwelling of scales, on the fish is only a part? Because allometric, and and are generally thought to be phenotypic, and genetic correlation (see later under adaptations to environments of high temperature and Phenotypic Size Correlation and under Genetic Corre- low (5). Is this condition in ringtails a lation) are so widespread among different features of specific adaptation that permitted this population to an organism, the latter explanation is more likely. occupy a desert environment? A phylogenetically based However, identification of these broader characters is examination (26, 48) found that the divergence in difficult (101). Even the characteristic nature of single metabolic rates between desert and montane popula- and obvious morphological structures may be debat- tions of ringtail is statistically much greater than that able. Gould and Lewontin (54), for instance, asked elsewhere in the procyonid lineage. This divergence is whether the can legitimately be considered consistent with an adaptive explanation that lower a character which would require an explanation (adap- metabolic rate was evolved in concert with the occupa- tive or not). They maintained that it is simply the tion of a hot and arid environment. product of differential evolutionary development of Analysis of character evolution within a phylogenetic two facial growth fields, requiring no further justifica- context may also permit determination of the ancestral tion or explanation. state and the directionality of evolution within any Possible explanations, in addition to adaptation, for descendent group. In addition, rates of evolution of the possession of a character by an organism include physiological characters may be estimated if phyloge- the following six. netic divergence times are known. (The following refer- ences are suggested for a discussion of the theory and methods of such analyses: 7,19,39,40,46,50,59, 64, I. Historical Inheritance 75, 86; potential interpretive problems of historically An organism may possess a character simply because based adaptive studies are discussed in refs. 42, 76, it was present in its ancestors. It may have no relation and 80.) to evolution in the current environment of the popula- tion. The trait may even be detrimental, if only in requiring energy for its formation and maintenance. It 2. Developmental Pattern and Constraint may be in the process of being eliminated, as is presum- Change, and hence evolution, typically occurs in bio- ably the case for vestigial organs (for example, non- logical systems by modification of preexisting struc- functional in cave animals or pelvic girdles in tures. Novelties, including adaptations, must therefore cetaceans). The trait may be neutral. The trait may even arise sometime during ontogenetic development as an 8 HANDBOOK OF PHYSIOLOGY-COMPARATIVE PHYSIOLOGY alternation of some developmental sequence. (An im- ture of an organism may be best explained in terms of portant exception to this generalization is the endosym- its development and not in terms of adult or current biotic incorporation of exogenous biological material.) utility or design. Such familiar evolutionary patterns as and pedomorphosis appear to result from differential rates 3. Physical and Biomechanical Correlation of maturation of different developmental fields or structures (53). (The role of development in initiating Any structure will have an array of attendant properties and constraining evolutionary novelty is extensively incidental to its design. Those that are crucial to its discussed in refs. 53, 58, 127, 132.) The necessity of function will be taken into account in its construction, operating on pre-existing structures and functions acts but unimportant factors will not be subjected to a as a brake on potential evolutionary change. Improved designer's scrutiny. Properties such as density, thermal function might be theoretically obtainable but would coefficient of expansion, or melting point may be im- require such fundamental reworking of the develop- portant design criteria in some circumstances and irrel- mental scheme as to be practically impossible. evant in others. Nevertheless, any structural material Ontogenetic development is a complicated series of will possess a discrete value for the entire suite of interactions among genes, gene products, and morpho- physical properties. logical structures. Timing of the interactions among Biological structures are no different. The designing these elements can have a crucial influence on ultimate agent may have been selection, but selection will have shape and function. The problem for evolutionary operated only on features of functional importance. change is to perturb the developmental program Other aspects of the structure,. no matter how striking enough to produce a specific improvement but not so they may appear, may be completely incidental. Con- much as to interfere with other orderly interactions sider, for instance, vertebrate blood. Its most apparent necessary to produce a functioning organism. Thus, feature is its red color and its change in color with adaptations frequently involve alternations rather late oxygenation state, but these are undoubtedly incidental in the developmental scheme, such as rates of matura- properties arising from the use of iron in the oxygen- tion of specific organs or tissues. Alterations early binding structure of hemoglobin. It is doubtful that in development may so fundamentally affect diverse the changing color of blood, convenient as it is for structures that development is terminated or the re- physiologists, was an important or selected factor in sulting organism is not viable or competitive. Thus, the early evolution of vertebrate hemoglobin as an early development often remains relatively conservative oxygen carrier. Other reasons for blood color need not in comparison to later development. be invoked, nor is it even a separate character that Characters arising early in the developmental plan must be explained. From this point of view, no further may thus be retained in adult organisms without appar- explanation, adaptive or otherwise, is required to ac- ent present utility. Some of these, such as the navel (or count for the presence of red blood in vertebrates and at least the umbilical cord associated with it), may blue blood in molluscs. Nor is it of adaptive importance have had considerable utility during development, Oth- that bones are white or that bile is green. It is not an ers, such as nipples in male mammals, may not. Other adaptive feature that bones break when put under structural features may be rationalized only in terms too much stress. These are correlated properties that of the operation of a developmental program, even at require no further justification or explanation. the expense of functional capacity in adults. Consider, for instance, the orientation of rods and cones in the 4. Phenotypic Size Correlation vertebrate retina. These light-sensitive cells are pointed toward the pigmented layer and sclera of the and Various traits, be they structural or functional, may be inverted with respect to incident illumination coming extensively correlated within an organism. For in- through the pupil and lens. Consequently, light must stance, individuals having the largest heart size may pass through several layers of neurons before reaching also have the largest wings. This correlation may be a the photocells, which must then send information back reflection of common genetic influence on these fea- through those neurons, which then pass it back through tures (see later under Genetic Correlation, below) and/ the photocell layer in the optic nerve. No engineer or a phenotypic association between characters (101). would have designed such a photosensitive circuit, and If the characters are phenotypically correlated, selec- its explanation is not to be found in any adaptive tion on one will affect the other and a priori it may be design. Rather, its orientation is best understood in impossible to tell which is an adaptation and which is reference to the pattern of ontogenetic development of merely being carried along in the course of evolution. the eye arising from the brain. Thus, a structural fea- If the characters are functionally coupled, as might be CHAPTER 1: EVOLUTIONARY PHYSIOLOGY 9 the case for heart size and large wings if selection influence numerous physiological, morphological, and favored endurance, both might be adaptive characters. behavioral characters. Second, a genetic association However, it is also possible that only one might be of among traits may result from linkage disequilibrium, importance and the covariation of the second might in which alleles at two loci are statistically associated falsely be assumed to be evidence of a functional associ- and do not recombine randomly. The values of traits ation between the traits. resulting from each locus thus tend to be associated The most common source of phenotypic correlation with each other. Genetic correlations produced by ei- is covariation of characters with size. Nearly all struc- ther of these mechanisms have similar consequences. If tural and functional characters vary with body size (24, selection favors one of the correlated characters, the 114). Bigger organisms generally have bigger organs so second may also increase in the population, even if it that a correlation between heart and wing size may is neutral or disadvantageous. In the latter case, the reflect nothing more than body size heterogeneity balance of selection coefficients will determine the di- within the population. Further, differential reproduc- rectionality and rapidity of evolution. (For discussions tion by individuals with large hearts andlor wings of the effect of genetic correlation on the evolution of might also be misinterpreted as being due to either of physiological traits, see refs. 3, 101). these factors. In fact, selection may have been on The problem with genetic correlation as a confound- body size itself or on performance related to body size ing factor in evolutionary analysis is that breeding independent of, or even in spite of, heart and wing size. experiments are necessary to know whether traits are Clearly size-related phenotypic correlation may genetically correlated. Comparative phenotypic analy- geatly complicate any analysis of the functional or sis alone will not permit determination of whether a evolutionary importance of a single character. There- trait stands by itself or is genetically correlated with fore, attempts are commonly made to eliminate size other factors (2, 40, 69, 101). For many organisms of effects. Frequently, however, the relationship between interest to physiologists, breeding experiments are very a character and body size is not straightforward. De- difficult or even impossible. Consequently, much of parture from geometric similarity, in which linear ele- what we know about genetic correlations comes from ments scale as the cube root of volume, is termed experiments on laboratory-maintained organisms, such and is common in biological systems (24, as mice or fruit flies. The results of these studies should 96, 105, 114). The structural and functional bases of give physiologists cause for serious concern about the the allometric scaling of most organismal characters, genetic independence of the characters they study. such as the scaling of basal metabolic rate to the three- As an example, consider the extensive and well- quarter power of mass, are poorly understood and controlled experiments of Rose, Service, and their co- sometimes contentiously debated (60). Attempts to workers (reviewed in refs. 57, 107) on history eliminate the influence of size by dividing by mass (for selection in Drosophila melanogaster. In the labora- example, mass-specific metabolic rate), or any other tory, selection for both early- and late-life reproduction allometric exponent assumed a priori, may fail to re- was imposed on replicated populations. As anticipated, move the effect of size completely. This failure may the pattern of reproduction evolved in response to this iead to incorrect functional or evolutionary conclusions selection, as did , later-reproducing flies living because of remaining, though weaker, size correlations. longer. In addition, however, a suite of diverse physio- One method for reducing the influence of size and logical characters also evolved in these experiments, other size-related characters is analysis of size-corrected including flight ability, , tolerance residuals (8, 45, 59). A residual analysis may help to ethanol vapor, and starvation resistance. Breeding eliminate body size as an explanation for the posses- experiments demonstrated that some of these physio- sion, or significance, of a character but does not elimi- logical traits are genetically correlated with age-specific nate the more general problem of disentangling other reproduction, the factor on which selection was im- phenotypic correlations from evolutionary analyses. posed. In another set of experiments (63), selection on physiological characters in Drosophila led to associated changes in life-history characters. 5. Genetic Correlation Given these underlying genetic associations, it would Two dissimilar phenotypic characters may share a com- not be correct to abstract one of these traits in a natural mon genetic basis. This genetic correlation can be population and conclude that selection had operated generated by two different mechanisms. The first and adaptively in its formation. If, for example, one de- most common is , in which a single gene tected a low rate, or great tolerance, of water loss in a affects more than one trait. For instance, an allele that desert population of Drosophila, the temptation might increases testosterone might be expected to be irresistible to conclude that selection had operated 10 HANDBOOK OF PHYSIOLOGY-COMPARATIVE PHYSlOLOGY on that character. In fact, however, a reproductive or spectives. For example, physiological studies have tra- other life-history character may have been the target ditionally emphasized central tendency (for example, of selection and desiccation resistance may have been averages or regression values) and to look upon indi- only a genetically correlated response. vidual variation as an unfortunate and suspect compli- cating factor (8). From an evolutionary perspective, however, individual variation is a prime precondition 6. Chance Fixation for the operation of selection and possible evolution of In considering evolution of a physiological character, a trait within a population. Individual variation from we generally think of selection on standing genetic this viewpoint is something of interest in its own right. variability in a large, outbreeding population. Under New questions then become apparent in regard to this such conditions, it is highly unlikely that the frequency variability: What are its correlated properties among of deleterious characters will increase or become fixed individuals? Is the trait inherited? Is it an important in the population. Natural populations, how- trait? For instance, in regard to maximal locomotor ever, may be neither large nor randomly breeding. If speed, what, physiologically and morphologically, effective population size is small (for example, under makes a fast individual fast? Is being fast an inherited loo), an allele may increase and even become fixed trait? Is it important to be fast? Similar questions could in the population by chance alone. This problem is be asked about urine-concentrating ability, digestive exacerbated by fluctuations in these already low popu- efficiency, maximal oxygen consumption, etc. These lation sizes, which may occasionally drop to very few questions arise directly from an interest in studying the breeding individuals (a bottleneck). Probability of ran- evolution of organismal characters. dom fixation of an allele and its attendant traits in- Here I consider two different sorts of studies evolu- creases greatly in such situations. Further, any propen- tionary physiology. The first concerns conclusions sity within such small populations .toward nonrandom about comparisons of characters among different pop- mating or an unequal sex ratio of breeding individuals ulations, species, genera, or higher taxa, examining the within the population further reduces the effective pop- endpoints of different phylogenetic lineages. This is the ulation size and increases the likelihood of . more familiar kind of comparative study. The second In general, we do not know effective population size type investigates the process of evolution, the origin in natural demes of different organisms or the impor- and operation of selection and other processes on char- tance of drift in influencing the distribution of alleles acters, within a single population. The former may be and traits within them. Drift, as opposed to adaptation, considered macroevolutionary and the latter, micro- is therefore another potential source of variation in evolutionary. The former concentrates more on pattern natural populations. Experimental evolutionary studies analysis, the latter on analysis of adaptive process. The attempt to eliminate drift effects by maintaining large former, usually of necessity, involves relatively few population sizes and replicated experimental and con- individuals (for example, a dozen) in each ; the trol lines (28, 78, 108). latter requires observations on hundreds of individuals or more. (For discussions of interrelationships between Potential explanations for the presence of a.trait in intra- and interpopulation or -species studies, see refs. a population of organisms are many. Adaptation is but 34, 47, 49, 111.) one of these. How then does one sort through these possibilities and determine the evolutionary source and adaptive importance of a particular feature of interest? Macroevolutionary Studies Comparative macroevolutionary studies attempt to ex- plain the differential distribution of characters among STUDYING THE EVOLUTION OF PHYSIOLOGICAL different biological groups. The first, and probably CHARACTERS most valuable, step in the evolutionary analysis of such data is the abandonment of the preconception that As Bock (16) has pointed out, studying the evolution any differences that may be observed are necessarily of functional traits is not a simple undertaking. It is, adaptive. Adaptation should be considered as only one however, one that has considerable reward because it of a series of competing hypotheses to be investigated fulfills the second goal of organismal and comparative (37, 80). The alternative hypotheses most easily investi- biology-to explain how organisms came to be the gated are historical inheritance (explanation 1; see way they are, not just how they work in the here earlier under ALTERNATIVES TO ADAPTIVE EXPLANA- and now. Examining physiological processes from an TIONS) and phenotypic correlation (explanation 4). De- evolutionary viewpoint can give us valuable new per- velopmental constraint, explanation 2, is a special case CHAPTER 1: EVOLUTIONARY PHYSIOLOGY 11 of 1, and rejection of the latter would also all of them may have to be considered as being an reject the former. Biomechanical correlation, explana- alternative target of adaptation. If body sizes are un- tion 3, is largely a problem of too narrow a definition equal among the groups compared, an analysis of of a character; a proper search for correlated pheno- phylogenetically independent contrasts of the variable typic traits would assist in evaluation of this alterna- of interest on body size should be undertaken (see tive. The investigation of genetic correlation, explana- methods in refs. 43, 72). A promising experimental tion 5, and chance fixation, explanation 6, while highly approach for directly testing the importance of body desirable, are usually impractical or impossible in or- size and its influence on numerous different traits is ganisms most commonly the subject of such studies. allometric engineering (117, 118). Body size is manipu- Crucial to the evaluation of the historical hypothesis lated by such techniques as yolk removal or hormonal is a knowledge of the phylogenetic relationships of the supplementation, permitting the direct measurement of groups of interest (21, 49, 51, 59, 64, 65). These the effect of body size itself, uncomplicated by diverse relationships will determine which groups should be correlations inherent in intergroup comparative part of the study, and thus this information must studies. precede data collection. For example, to study a pre- In comparative macroevolutionary studies of this sumptive adaptation (definition 6) to desert environ- sort, adaptation as a hypothesis for the existence of a ments, different subgroups (for example, populations character cannot be tested directly. However, examina- or species), some of which live in and some of tion and rejection of these alternative hypotheses per- which do not, must be included. Further, the character mits one to have more confidence that adaptation must be measured in a closely related group or groups might be the real explanation for the character. Consid- (outgroups, sister taxa) to provide an indication of eration and rejection of such alternatives is very much dire~tionalit~and ancestral state of the character. Thus, in the spirit of Williams' (132) assertion that adapta- a minimum of three groups, and preferably many more, tion "should be used only as a last resort. It should is required in any such analysis (48). Assistance from not be invoked when less onerous principles . . . are a systematist familiar with the group in question is sufficient for a complete explanation." Since such com- usually highly valuable in determining these phyloge- parative studies are the only means we have to examine netic relationships. the evolutionary history of natural living systems (ma- An important complicating factor in the analysis of nipulative experiments being impossible), it is well such data is the lack of statistical independence of worth being as careful as possible in their analysis and different subgroups in a study, since they are, to differ- interpretation. ent degrees, all related to each other (39,59). (Excellent and extensive discussions of the philosophy and meth- Microevolutionary Studies odology of this approach can be found in the references cited earlier under Historical Inheritance section.) Some In contrast to the analysis of the historical patterns interesting and instructive examples of its application resulting from evolution described in the previous sec- are studies of thermoregulation in scombroid tion, microevolutionary studies concentrate on analyz- (14), locomotor performance in anoline lizards (81, ing the current processes of evolution. They concen- 85), and rates of and water loss in geckos trate on a single population or series of populations (31). If a character can be shown by this analysis not and analyze the possibility and the actual operation of to have been present in the ancestor, the historical selection on biological characters. They show what is inheritance alternative can be rejected. possible in the evolution of physiological characters- Analysis of phenotypic correlation, explanation 3, for example, which traits will respond to selection, requires an examination of the statistical associations which traits are correlated with each other, and how of the character with other traits. Body size is the most rapidly and extensively adaptive evolution may pro- likely confounding correlate, so the best starting place ceed. (The approach described here for microevolution- is to determine whether body size varies among the ary analyses of physiological characters is taken from groups compared. If body size does not differ among refs. 2, 10, 15, 16, 18, 73, which provide more de- groups, size and other factors correlated with it are tailed discussions.) unlikely to complicate the analysis. Correlations with For a biological character to be able to respond to other traits, after correction for phylogenetic associa- selection, it must be both variable and heritable (35). tion (50, 86), may then be sought directly with such That is, within a population, individuals must differ techniques as principal component analysis. Failure to from each other with respect to the character, and find correlated traits weakens the phenotypic correla- some of this variation must have a genetic bases: off- tion hypothesis. If correlated traits are found, any or spring must resemble their parents in regard to the 12 HANDBOOK OF PHYSIOLOGY-COMPARATIVE PHYSIOLOGY trait. Most physiological characters are quantitative tion, see ref. 117) of the effects of such factors as (having many character states or being continuously size and other size-correlated variables on the trait distributed). These traits may be under the influence of is advisable. many different genes and are consequently amenable to After demonstration of the correlation of the trait analysis through the techniques of with differential survival (and perhaps identification of (36). One of the first steps of a microevolutionary the portion of the life cycle affected), the use of the study might be the demonstration of variability and trait under natural conditions-that is, its ecological in the character of interest. The character importance-should be examined. Why is it that hav- cannot respond evolutionarily to selection if it is not ing a particular condition of the character is beneficial? both variable and heritable. However, most physiologi- Arnold (2) advocated dividing such studies into two cal traits that have been examined apparently do pos- more manageable portions, termed the performance sess these properties (8, 49, 93). For many physiologi- and fitness gradients. In the former, the effect of a cal traits, such as rate processes, interindividual character on a performance trait (for example, escape variability is quite high, with coefficients of variation ability) is examined in the laboratory; in the latter, the of around 20%-30%. Given this relatively great vari- performance trait is analyzed in its natural context. ability, physiologists have also sought, usually success- Many physiological characters (for example, rate pro- fully, to demonstrate that these traits are measurably cesses) may already be regarded as complex perfor- repeatable through time for an individual (8, 68). In mance traits in this context. Demonstration of the addition, these traits have moderate, but nonetheless ecological effect of the performance trait-for example, highly significant, levels of heritability (8, 49, 93, 119). exactly how it aids in predator escape, food acquisition, Such investigations frequently require that the organ- mate attraction, starvation resistance, etc.-may be isms be brought into the laboratory to heed, or at quite difficult to accomplish. Focal observations on least to deliver offspring, so that patterns of inheritance individual organisms in nature (97) can establish pat- can be analyzed and determined. Measurements of the terns of or organismal interaction (for exam- character in large numbers of individuals are required. ple, what predators are present and how organisms If the potential for response to selection can be avoid becoming prey). These may be used to provide demonstrated, the effect of the character on survival a credible rationale for the ecological operation of the or reproduction within the natural population can be character. However, the stochastic nature of selection examined (see ref. 35 for a discussion of methods may require similar observations on hundreds of indi- of demonstrating natural selection). The character in viduals to demonstrate the effect of a known character question nay be an organismal trait, such as locomotor state (10). In addition, the. presence of the observer performance (for example, burst speed in lower verte- will have an unknown, and perhaps unwelcome, effect brates [lo, 70, 90, 128]), a particular physiological or on the results of the study. morphological character (for example, bill shape in Such microevolutionary studies are large and com- [56] or scalation pattern in lizards [41]), or a plex but not impossible. Their feasibility depends major gene product (for example, lactic dehydrogenase greatly on the choice of character and study organism. in fish [98,99], phosphoglucose isomerase in Ideally, the organism should live in a discrete, easily [130], or leucine aminopeptidase in molluscs [72]). A observed population of manageable size. It should pro- group of individuals of measured character state within duce groups of offspring of known parentage, which the population is followed through time to determine can be compared to each other. Organisms should whether those with low, intermediate, or high values be readily collected, measured, marked, and returned of the character are more or less likely to survive and without damage to their natal population and recol- reproduce in the natural environment (see refs. 35, 73, lected again later. The character should have temporal 110 for a discussion of methods). Care must be taken stability and a connection to a performance trait of to include all ages as selection might be particularly presumptive importance in nature. These desirable fea- intense on newborns or some other group and might tures restrict the field of potential study organisms but go undetected if some ages were excluded from exami- do not preclude all interesting ones. nation (112). This approach is basically correlative It should be borne in mind that only very rarely is and subject to the problems pointed out earlier under any study complete, including mechanistic analyses Phenotypic Size Correlation. If differential survival in done in the laboratory. Some steps are missing in even organisms with high values of a particular trait can be the best and most complete investigations. Getting demonstrated, it is still possible that selection may be partial answers may be better than waiting forever to operating on a correlated trait and not on the one in discover the perfect system (37, 129). question (91). Hence, analysis (and possibly manipula- In the search for experimental systems for the study CHAPTER 1 : EVOLUTIONARY PHYSIOLOGY 13 of evolutionary physiology, some investigators have organismal function and design has too long dominated turned to a variety of laboratory-cultured organisms, physiological thinking. It was an easy afterthought to such as fruit flies and bacteria. These organisms offer justify results in terms that seemed possible in view of several advantages. They can be maintained in discrete the current environmental circumstances of an organ- populations of large size, minimizing chance fixation ism. Its axiomatic application disinclined physiologists while allowing the continuous generation of to the investigation of evolutionary questions. Testable new by and recombination. alternatives to adaptation exist, and their consideration small organisms with very rapid generation times may greatly enriches our appreciation of the evolutionary be used, permitting observation of the time course as process. as trajectory of evolutionary change. In some Over the past half-century, comparative studies have organisms, direct measurement of competitive (Dar- emphasized the equilibrium (74) interaction of organ- winian) fitness is possible SO that magnitude and rate of ism and environment. Comparative (ecological) physi- can be determined. Further, the replicated ology rose to prominence as an independent field, with nature of the experiment permits an assessment of many attendant successes (see ref. 9), operating largely the repeatability, or inevitability, of any evolutionary within an equilibrium frame of reference. However, pattern and a statistical analysis of its significance the understanding of the evolution of physiological (126). systems did not keep pace with the expanding knowl- This approach has been termed "natural selection in edge of their mechanistic bases and ecological corre- the laboratory" (108) and involves the imposition of a lates. The of adaptation often precluded novel environment on replicated experimental popula- critical evolutionary investigations. Critiques of this tions, while maintaining control populations for com- adaptationist view, although unsettling, freed compara- parison. It is distinct from "artificial seiection" (as in tive biology to begin investigations of alternative expla- animal or plant breeding) in which the experimenter nations for the evolution of organismal characters. It chooses particular characters and permits breeding on is now both possible and highly desirable to incorpo- their basis alone. In the former approach, only the rate an evolutionary component into comparative environment is stipulated by the investigator and selec- physiological studies (10, 21, 38). Comparative (eco- tion works on any characters it can. Examples include logical) physiology (along with its companion field, experimental physiological studies on the evolution of functional ) is in a very strong position to temperature adaptation in fruit flies (25, 63, 69) and undertake such studies. The field has many traditional bacteria (11, 13, 77, 92) and ethanol resistance in fruit strengths, including the detailed understanding of func- flies (27). These laboratory-based studies of microevo- tional mechanisms, an excellent technological base, a lutionary physiology are meant to supplement, not strong experimental ethic, and a long-standing empha- supplant, research on natural populations. The two sis on the importance of the integration of characters approaches are mutually supportive and informative at the organismal level. Evolutionary biology has much and will lead to more rapid progress in our understand- to gain from work on the evolution of important ing of the evolution of physiological characters than organismal and physiological characters and their will either by itself. mechanistic bases. Evolutionary physiology is emerging as an important new discipline, to the mutual benefit of both comparative physiology and evolutionary biology. INCORPORATING AN EVOLUTIONARY PERSPECTIVE INTO PHYSIOLOGICAL STUDIES I thank the following individuals for their very perceptive and Organisms did not acquire their present forms and helpful comments on the manuscript: T. Bradley, A. Cullum, T. functional capacities by special creation. These traits Garland, R. Huey, R. Lenski, A. Leroi, J. McLister, S. O'Steen, M. Rose, and T. Watkins. Support was provided by National Sci- evolved from characters preexisting in their ancestors. ence Foundation grants IBN-9118346, -9208662, -9420155, and They were shaped sometimes by adaptation, sometimes -9507416. by other forces; sometimes they were constrained from evolutionary change. In addition to trying to under- stand how organisms are built and work today, com- REFERENCES parative biologists must attempt to understand how 1. Alexander, R. McN. Apparent adaptation and actual perfor- today's organisms became the way they are by studying mance. In: Evolutionary Biology, edited by M. K. Hecht, B. how organismal properties and physiological mecha- Wallace, and R. J. MacIntyre. New York: Plenum, 1991, vol. 25, p. 357-373. nisms have changed through time. 2. Arnold, S. J. Morphology, performance and fitness. Am. Zool. Adaptation as the sole explanation for all aspects of 23: 347-361,1983. 14 HANDBOOK OF PHYSIOLOGY-COMPARATIVE PHYSIOLOGY

3. Arnold, S. J. Genetic correlation and the evolution of physiology. 24. Calder, W. A. Size, Function, and Life History. Cambridge, In: New Directions in Ecological Physiology, edited by M. E. MA: Harvard Univ. Press, 1984. Feder, A. F. Bennett, W. Burggren, and R. B. Huey. Cambridge: 25. Cavicchi, S., D. Guerra, V. Latorre, and R. B. Huey. Chromo- Cambridge Univ. Press, 1987, p. 189-215. somal analysis of heat-shock tolerance in Drosophila melanogas- 4. Baldwin, J., and P. W. Hochachka. Functional significance of ter evolving at different temperatures in the laboratory. Evolu- isoenzymes in thermal : acetylcholinesterase from tion 49: 676-684, 1995. trout brain. Biochem. 1. 116: 883-887, 1970. 26. Chevalier, C. D. Aspects of Thermoregulation and Energetics 5. Bartholomew, G. 4. Energy metabolism. In: Animal Physiology: in the Procyonidae (Mammalia: Carnivora). Irvine: Univ. of Principles and Adaptations, edited by M. S. Gordon, G. A. California, 1991 Ph.D. thesis. Bartholomew, A. D. Grinnell, C. B. Jorgensen, and F. N. White. 27. Cohan, F. M., and A. A. Hoffrnann. under New York: Macmillan, 1982, p. 46-93. uniform selection. 11. Different responses to selection for knock- 6. Bateson, W. Heredity and variation in modern lights. In: down resistance to ethanol among Drosophila melanogaster in Modern Science, edited by A. C. Seward. Cambridge: Cam- populations and their replicate lines. Genetics 144: 145-163, bridge Univ. Press, 1909, p. 85-101. 1986. 7. Baum, D. A., and A. Larson. Adaptation reviewed: a phyloge- 28. Cohan, F. M., and A. A. Hoffmann. Uniform selection as a netic methodology for studying character . Syst. diversifying force in evolution: evidence from Drosophila. Am. Zool. 40: 1-18, 1991. Nut. 134: 613-637,1989. 8. Bennett, A. F. Inter-individual variability: an underutilized re- 29. Cossins, A. R., and K. Bowler. Temperature Biology of Animals. source. In: New Directions in Ecological Physiology, edited by London: Chapman and Hall, 1987. M. E. Feder, A. F. Bennett, W. W. Burggren, and R. B. Huey. 30. Dannler, W. H. Perspectives. Am. 1. Physiol. 263 (Regulatory Cambridge: Cambridge Univ. Press, 1987, p. 147-169. Integrative Comp. Physiol. 34): Rl, 1992. 9. Bennett, A. F. The accomplishments of ecological physiology. 31. Dial, B. E., and L. L. Crismer. A phylogenetic analysis of In: New Directions in Ecological Physiology, edited by M. E. physiological-ecological character evolution in the lizard Feder, A. F. Bennett, W. W. Burggren, and R. B. Huey. Cam- Coleonyx and its implications for historical biogeographic re- bridge: Cambridge Univ. Press, 1987, p. 1-8. construction. Syst. Biol. 41: 178-195, 1992. 10. Bennett, A. F., and R. B. Huey. Studying the evolution of 32. Dill, D. B., E. F. Adolph, and C. G. Wilber, eds. Adaptation to physiological performance. Oxf. Surv. Evol. Biol. 7: 251-284, the Environment. Washington, DC: Am. Physiol. Soc., 1964. 1990. 33. Eckerr, R., D. Randall, and G. Augustine. Animal Physiology: 11. Bennett, A. F., and R. E. Lenski. Evolutionary adaptation to Mechanisms and Adaptations (3rd ed.), New York: Freeman, temperature. 11. Thermal niches of experimental lines of Esche- 1988. richia coli. Evolution 47: 1-12, 1993. 34. Emerson, S. B., and S. J. Arnold. Intra- and interspecific relations 12. Bennett, A. F., and R. E. Lenski. Evolutionary adaptation to between morphology, performance, and fitness. In: Complex temperature. V. Adaptive mechanisms and correlated responses Ornanismal Functions: Intenration and Evolution in Verte- in experimental lines of . Evolution 50: 493- bra;es, edited by D. B. ~aieand G. Roth. Chichester, UK: 503, 1996. Wiley, 1989, p. 295-314. 13. Bennett, A. F., R. E. Lenski, and J. E. Mittler. Evolutionary 35. Ender, J. A. Natural Selection in the Wild. Princeton: Princeton adaptation to temperature. I. Fitness responses of Escherichia Univ. Press, 1986. coli to changes in its thermal environment. Evolution 46: 16- 36. Falconer, D. S. Introduction to Quantitative Genetics (3rd ed), 30, 1992. London: Longman, 1989. 14. Block, B. A., J. R. Finnerty, A. F. R. Stewart, and J. Kidd. 37. Feder, M. E. The analysis of physiological diversity: the pros- Evolution of endothermy in fish: mapping physiological traits peas for pattern documentation and general questions in ecolog- on a molecular phylogeny. Science 260: 210-214, 1993. ical physiology. In: New Directions in Ecological Physiology, 15. Boake, C. R. B., ed. Quantitative Genetic Studies of Behavioral edited by M. E. Feder, A. F. Bennett, W. W. Burggren, and R. B. Evolution. Chicago: Univ. of Chicago Press, 1994. Huey. Cambridge: Cambridge Univ. Press, 1987, p. 38-75. 16. Bock, W. J. The definition and recognition of biological adapta- 38. Feder, M. E., A. F. Bennett, W. W. Burggren, and R. B. Huey. tion. Am. Zool. 20: 217-227, 1980. eds. New Directions in EcoIogical Physiology. Cambridge: Cam- 17. Bock, W. J., and G. von Wahlert. Adaptation and the form- bridge Univ. Press, 1987. hnction complex. Evolution 19: 269-299, 1965. 39. Felsenstein, J. Phylogenies and the comparative method. Am. 18. Brandon, R. N. Adaptation and Environment. Princeton: Nut. 125: 1-15, 1985. Princeton Univ. Press, 1990. 40. Felsenstein, J. Phylogenies and quantitative characters. Annu. 19. Brooks, D. R., and D. A. McLennan. Phylogeny, Ecology, and Rev. Ecol. Syst. 19: 445-471, 1988. Behavior. Chicago: Chicago Univ. of Chicago Press, 1991. 41. Fox, S. F. Natural selection on morphologcal of the 20. Bull, J. J. Evolution of phenotypic variance. Evolution 41: 303- lizard Uta stansburiana. Evolution 29: 95-107, 1975. 315, 1987. 42. Frumhoff, P. C., and H. K. Reeve. Using phylogenies to test 21. Burggren, W. W. Does comparative respiratory physiology have hypotheses of adaptation: a critique of some current proposals. a role in evolutionary biology (and vice versa)? In: Physiological Evolution 48: 172-180, 1994. Strategies for Gas Exchange and Metabolism, edited by A. J. 43. Futuyma, D. J. Evolutionary Biology (2nd ed.), Sunderland, Woakes, M. K. Grieshaber, and C. R. Bridges. Cambridge: MA: Sinauer, 1986. Cambridge Univ. Press, 1991, p. 1-13. 44. Cans, C. : An Approach to Vertebrate Biology. 22. Burggren, W. W., and W. E. Bemis. Studying physiological Philadelphia: Lippincott, 1974. evolution: paradigms and pitfalls. In: Evolutionary Innovations, 45. Garland, T., Jr. Physiological correlates of locomotory perfor- edited by M. H. Nitecki. Chicago: Univ. of Chicago Press, 1990, mance in a lizard: an allometric approach. Am. 1. Physiol. 247 p. 191-228. (Regulatory Integrative Comp. Physiol. 18): R806-R815, 23. Cain, A. J. The perfection of animals. Biol. J. Linn. Soc. 36: 3- 1984. 29, 1989. 46. Garland, T., Jr. Rate tests for phenotypic evolution using phylo- CHAPTER 1: EVOLUTIONARY PHYSIOLOGY 15

independent contrasts. Am. Nut. 140: 509-519, 69. Huey, R. B., L. Partridge, and K. Fowler. Thermal sensitivity of 1992. Drosophila melanogaster responds rapidly to laboratory natural 47. Garland, T.,Jr., and S. C. Adolph. Physiological differentiation selection. Evolution 45: 751-756, 1991. of vertebrate populations. Annu. Rev. Ecol. Syst. 22: 193- 70. Jayne, B. C., and A. F. Bennett. Selection on locomotor perfor- 228, 1991. mance capacity in a natural population of garter snakes. Evolu-

48.. . Garland, T., Jr., and S. C. Adolph. Why not to do two-species tion 44: 1204-1229, 1990. comDarative studies: limitations on inferring adaptation. Phys- 71. Johnson, T. P., and A. F. Bennett. The thermal acclimation of ,il.iool. 67: 797-828, 1994. burst escape performance in fish: an integrated study of molecu- 49. Garland, T., Jr., and P. A. Carter. Evolutionary physiology. lar and cellular physiology and organismal performance. 1. Exp. Annu. Rev. Physiol. 56: 579-621, 1994. Biol. 198: 2165-2175, 1995. 50. Garland, T., Jr., P. H. Harvey, and A. R. Ives. Procedures for the 72. Koehn, R. K., R. I. E. Newell, and F. W. Immermann. Mainte- analysis of comparative data using phyl~geneticall~independent nance of an aminopepudase by natural contrasts. Syst. Biol. 41: 18-32, 1992. selection. Proc. Natl. Acad. Sci. U.S.A. 77: 5385-5389, 1980. 51. Garland, T., Jr., R. B. Huey, and A. F. Bennett. Phylogeny and 73. Lande, R., and S. J. Arnold. The measurement of selection on thermal physiology in lizards: a reanalysis. Evolution 45: 1969- correlated characters. Evolution 37: 1210-1226, 1983. 1975, 1991. 74. Lauder, G. V. Form and function: structural analysis in evolu- 52. Gordon, M. S., G. A. Bartholomew, A. D. GrinnelI, C. B. tionary morphology. 7: 430-442, 1981. Jorgensen, and F. N. White. Animal Physiology: Principles and 75. Lauder, G. V., R. B. Huey, R. K. Monson, and R. J. Jensen. Adaptations. New York: Macmillan, 1982. and the study of organismal form and function- 53. Could, S. J. Ontogeny and Phylogeny. Cambridge, MA: Belk- advances in systematics are defining new directions for func- nap, 1977. tional morphology and comparative physiology. Bioscience 45: 54. Gould, S. J., and R. C. Lewontin. The spandrels of San Marco 696-704,1995. and the Panglossian paradigm: a critique of the adaptationist 76. Lauder, G. V., A. M. Leroi, and M. R. Rose. Adaptations and programme. Proc. R. Soc. Lond. B Biol. Sci. 205: 581-598, history. Ecol. Evol. 8: 294-297, 1993. 1979. 77. Lenski. R. E., and A. F. Bennett. Evolutionary response of 55. Gould, S. J., and E. Vrba. Exaptation-a missing term in the Escherichia coli to thermal stress. Am. Nut. 142: S47-S64, science of form. Paleobiology 8: 4-15, 1982. 1993. 56. Grant, P. R., B. R. Grant, J. N. M. Smith, I. J. Abbott, and L. 78. Lenski, R. E., M. R. Rose, S. C. Simpson, and S. C. Tadler. K. Abbott. Darwin's finches: population variation and natural Long-term in Escherichia coli. I. Adapta- selection. Proc. Natl. Acad. Sci. U.S.A. 73: 257-261, 1976. tion and divergence during 2,000 generations. Am. Nut. 138: 57. Graves, J. L., E. C. Toolson, C. Geong, L. N. Vu, and M. R. 1315-1341, 1991. Rose. Desiccation, flight, glycogen, and postponed senescence 79. Leroi, A. M., A. F. Bennett, and R. E. Lenski. Temperature in Drosophila melanogaster. Physiol. 2001. 65: 268-286, 1992. acclimation and competitive fitness: an experimental test of the 58. Hall, B. K. Evolutionary . London: beneficial acclimation assumption. Proc. Natl. Acad. Sci. U.S.A. Chapman and Hall, 1992. 91: 1917-1921.1994. 59. Harvey, P. H., and 1M. D. Pagel. The Comparative Method in 80. Leroi, A. M., LM. R. Rose, and G. V. Lauder. What does the Evolutionary Biology. Oxford: Oxford Univ. Press, 1991. comparative method reveal about adaptation? Am. Nut. 143: 60. Heusner, A. A. Energy metabolism and body size. I. Is the 0.75 381-402,1994. mass exponent of Kleiber's equation a statistical artifact? Respir. 81. Lewontin, R. C. Adaptation. Sci. Am. 239: 156-169, 1978. Physiol. 48: 1-12, 1982. 82. Lewontin, R. C. as an adaptationist program. 61. Hochachka, P. W., and G. N. Somero. Biochemical Adaptation. Behav. Sci. 24: 5-14, 1979. Princeton: Princeton Univ. Press, 1984. 83. Lewontin, R. C. The shape of optimality. In: The Latest on the 62. Hoffmann, A. A. Acclimation: increasing survival at a cost. Best: Essays on Evolution and Optimality, edited by J. Dupre. Trends Ecol. Evol. 10: 1-2, 1995. Cambridge, MA: MIT Press, 1987, p. 151-159. 63. Hoffmann, A. A,, and P. A. Parsons. Evolutionary Genetics and 84. Losos, J. B. Ecomorphology, performance capability, and scaling Environmental Stress. New York: Oxford Univ. Press, 1991. of West Indian Anolis lizards: an evolutionary analysis. Ecol. 64. Huey, R. B. Phylogeny, history, and the comparative method. Monogr. 60: 369-388, 1990. In: New Directions in Ecological Physiology, edited by M. E. 85. Losos, J. B. The evolution of form and function: morphology Feder, A. F. Bennett, W. W. Burggren, and R. B. Huey. Cam- and locomotor performance ability in West Indian Anolis liz- bridge: Cambridge Univ. Press, 1987, p. 76-98. ards. Evolution 44: 1189-1203, 1990. 65. Huey, R. B., and A. F. Bennett. A comparative approach to field 86. Martins, E. P., and T. Garland, Jr. Phylogenetic analyses of and laboratory studies in . In: Predator- the correlated evolution of continuous characters: a simulation Prey Relationships, edited by M. E. Feder and G. Lauder. study. Evolution 45: 534-557, 1991. Chicago: Univ. of Chicago Press, 1986, p. 82-98. 87. Mayr, E. Adaptation and selection. Biol. Zentralbl. 101: 161- 66. Huey, R. B., and A. F. Bennett. Phylogenetic studies of coadapta- 174,1982. tion: preferred temperatures versus optimal performance tem- 88. Mayr, E. How to carry out the adaptationist program? Am. peratures of lizards. Evolution 41: 1098-1 115, 1987. Nut. 121: 324-334, 1983. 67. Huey, R. B., and D. Berrigan. Testing evolutionary hypotheses 89. Michod, R. E. On adaptedness and fitness and their role in of acclimation. In: Animals and Temperature: Phenotypic and evolutionary explanation. 1. Hist. Biol. 19: 289-302, 1985. Evolutionary Adaptation, edited by I. A. Johnston and A. F. 90. Miles, D. B. related differences in locomotion and Bennett. Soc. Exp. Biol. Sem. Ser. 59, 1996, p. 205-237. morphology in two populations of Urosaurus ornatus. Am. 6fL hey, R. B., A. E. Dunham, K. L. Overall, and R. A. Newman. 2001. 27: 44A, 1987. Variation in locomotor performance in demographically known 91. Mitchell-Olds, T., and R. G. Shaw. Regression analysis of natu- populations of the lizard Sceloporus merriami. Physiol. 2001. ral selection: statistical and biological interpretation. Evolution 63: 845-872, 1990. 41: 1149-1161. 1987. 16 HANDBOOK OF PHYSIOLOGY-COMPARATIVE PHYSIOLOGY

92. Mongold, J. A., 4. F. Bennett, and R. E. Lenski. Evolutionary pressures and life history trade-offs. Proc. R. Soc. Lond. B adaptation to temperature. IV. Adaptation of Escherichia coli Biol. Sci. 246: 11-17, 1991. at a niche boundary. Evolution 50: 35-43, 1996. 113. Schmidt-Nielsen, K. How Animals Work. Cambridge: Cam- 93. Mousseau. T. A., and D. A. Roff. Natural selection and the bridge Univ. Press, 1972. heritability of fitness components. Heredity 59: 181-197, 1987. 114. Schmidt-Nielsen, K. Scaling: Why is Animal Size so Important! 94. Munson, R. Biological adaptation. Phil. Sci. 38: 200-215, 1971. Cambridge: Cambridge Univ. Press, 1984. 95. Padilla, D. K., and S. C. Adolph. Plastic inducible morphologies 115. Schmidt-Nielsen, K. Animal Physiology: Adaptation and Envi- are not always adaptive: the importance of time delays in a ronment (4th ed.). Cambridge: Cambridge Univ. Press, 1990. stochastic environment. Evol. Ecol. 10: 105-117, 1996. 116. Selye, H. The Story of the Adaptation Syndrome. Montreal: 96. Peters, R. H. The Ecological Implications of Body Size. Cam- Acta, 1952. bridge: Cambridge Univ. Press, 1983. 117. Sinervo, B., P. Doughty, R. B. Huey, and K. Zamudio. Allomet- 97. Pough, F. H. Organismal performance and Darwinian fitness: ric engineering: a causal analysis of natural selecrion on off- approaches and interpretations. Physiol. Zool. 62: 199-236, spring size. Science 258: 1927-1930, 1992. 1989. 118. Sinervo, B., and R. B. Huey. Allometric engineering: an experi- 98. Powers, D. A. A multidisciplinary approach to the study of mental test of the causes of interpopulational differences in genetic variation within species. In: New Directions in Ecologi- performance. Science 248: 1106-1109, 1990. cal Physiology, edited by M. E. Feder, A. F. Bennett, W. W. 119. Sorci, G., J. G. Swallow, T. Garland. Jr., and J. Clobert. Burggren, and R. B. Huey. Cambridge: Cambridge Univ. Press, Quantitative genetics of loc~rnotorspeed and endurance in the 1987, p. 102-130. lizard Lacerta vivipara. Physiol. Zool. 68: 698-720, 1995. 99. Powers, D. A., T. Lauerman, D. Crawford, and L. DiMichele. 120. Stearns, S. C. The evolutionary significance of phenotypic Genetic mechanisms for adapting to a changing environment. plasticity. Bioscience 39: 436-445, 1989. Annu. Rev. Genet. 25: 629-659, 1991. 121. Stephens, D. W., and J. R. Krebs. Theory. Princeton: 100. Precht, H.. J. Christofersen, H. Hensel, and W. Larcher. Tem- Princeton Univ. Press, 1986, p. 206-215. perature and Life. Berlin: Springer-Verlag, 1973. 122. Stern, J. T. The meaning of "adaptation" and its relation 101. Price, T., and T. Langen. Evolution of correlated characters. to the phenomenon of natural selection. Evol. Biol. 4: 39- Trends Ecol. Evol. 7: 307-310, 1992. 66, 1970. 102. Prosser, C. L. Adaptational Biology: to organisms. 123. Taylor, C. R., and E. R. Weibel. Design of the mammalian New York: Wiley, 1986. respiratory system. 1. Problem and strategy. Respir. Physiol. 103. Quellar, D. C. The spaniels of St. Marx and the Panglossian 44: 1-10, 1981. paradox: a critique of a rhetorical programme. Q. Rev. Biol. 124. Tinbergen, N. On aims and-methods in . 2. Tierpsy- 70: 485-489, 1995. chol. 20: 410-433, 1963. 104. Reeve, H. K., and P. W. Sherman. Adaptation and the goals 125. Townsend, C. R., and P. Calow. Physiological Ecology: An of evolutionary research. Q. Rev. Biol. 68: 1-32, 1993. Evolutionary Approach to Use. Sunderland, MA: 105. Reiss, M. J. The Allometry of Growth and Reproduction. Sinauer, 1981. Cambridge: Cambridge Univ. Press, 1989. 126. Travisano, M., J. A. Mongold, A. F. Bennett, and R. E. Lenski. 106. Rome, L. C., E. D. Stevens, and H. B. John-Alder. The influence Experimental tests of the roles of adaptation, chance, and of temperature and thermal acclimation on physiological func- history in evolution. Science 267: 87-90, 1995. tion. In: Environmental Physiology of the Amphibians, edited 127. Waddington, C. H. New Patterns in Genetics and Develop- by M. E. Feder, and W. W. Burggren. Chicago: Univ. of ment. New York: Columbia Univ. Press, 1962. Chicago Press, 1992, p. 183-205. 128. Watkins, T. B. Predator-mediated selection on burst swimming 107. Rose, M. R. Genetics of increased lifespan in Drosophila. performance in tadpoles of the Pacific frog, Pseudacris Bioessays 11: 132-135, 1989. regilla. Physiol. Zool. 69: 154-167, 1996. 108. Rose, M. R., J. L. Graves, and E. W. Hutchison. The use of 129. Watt, W. B. , physioIogical ecology, and popula- selection to probe patterns of pleiotropy in fitness characters. tion genetics: the mechanistic tools of evolutionary biology. In: Life Cycles: Genetics, Evolution and Co-Ordination, Funct. Ecol. 5: 145-154, 1991. edited by F. Gilbert. New York: Springer-Verlag, 1990, p. 130. Wan, W. B., R. C. Cassin, and LM. S. Swan. Adaptation at 29-42. specific loci. Ill. Field behavior and survivorship differences 109. Scheiner, S. M., and R. F. Lyman. The genetics of phenotypic among Colias PGI are predictable from in vitro plasticity. 11. Response to selection. J. Evol. Biol. 4: 23- biochemistry. Genetics 103: 725-739, 1983. 50, 1991. 131. Weibel, E. R., C. R. Taylor, and H. Hoppeler. The concept 110. Schulter, D. Estimating the form of natural selection on a of symmorphosis: a testable hypothesis of structure function quantitative trait. Evolution 42: 849-861, 1988. relationship. Proc. Natl. Acad. Sci. U.S.A. 88: 10357-10361, 111. Schluter, D. Bringing population and phylogenetic approaches 1991. to the evolution of complex traits. In: Complex Organismal 132. Williams, G. C. Adaptation and Natural Selection. Princeton: Functions: Integration and Evolution in Vertebrates, edited by Princeton Univ. Press, 1966. D. B. Wake and G. Roth. Chichester, England: Wiley, 1989, 133. Williams, G. C. A defense of in evolutionary p. 79-95. biology. 0x6 Sum. Evol. Biol. 2: 1-27, 1985. 112. Schluter, D., T. D. Price, and L. Rowe. Conflicting selection