Department of Biology
My research to date has concerned primarily the fields of evolutionary and comparative physiology. Much of this work has focused on interactions between organisms and their environmentswhole-organism , with a particular emphasis on performance traitsaspects of the ; such traits have a particular relevance to evolutionary patterns, since they represent phenotype acted on directly by selectionphysiological . My interests lie in asking two sorts of evolutionary questions about performance traitsthe actual : what sorts of patterns exist in populations or species for selection to act on, and what are results of selection over many generations?
Physiological ecologists have often assumed that, for many species, locomotor ability is important in the outcome of predator-prey interactions, and, as a corollary, that natural selection will tend to favor the evolution of faster individuals. Relatively few studies have actually addressed this topic, however, and the expectations are well worth testing in a natural system since selection and evolution do not always work in the expected way.
Populations of guppies (Poecilia reticulata) in Trinidadian streams provide an ideal system to test the evolutionary impact of predation on escape performance. Within most streams, one can find guppy populations exposed to both high levels of predation (areas where predatory species of fish are present) and low levels of predation (where those species are absent). Along with fellow post-doc Shyril OSteen and my graduate advisor Albert Bennett, I have been making replicate comparisons of escape performance between these two sorts of populations. To examine overall escape performance, Dr. OSteen and I conducted arena trials in which natural predators attacked small schools containing guppies from both high- and low-predation populations; we then used the relative number of survivors from each population type as an index of escape performance. In our tests of wild-caught fish, we found that guppies from high-predation areas were much more effective at avoiding predatory attacks than guppies from low predation areas. Similar tests of these populations using lab-reared descendents of wild-caught fish (thus fish that had never actually encountered a predator) revealed a smaller but still significant difference in performance, suggesting that some of the difference in escape ability among populations has a genetic basis.
In a second set of studies, I have been investigating one of the possible bases for these observed differences in overall escape performance. Using a high-speed video system, I am examining the "rapid-start" (or Mauthner-cell initiated) escape response, which quickly turns and accelerates a fish as the initial response to a striking predator. The results so far indicate that, for wild-caught fish, guppies from high-predation areas are faster than guppies from low predation areas. Interestingly, however, in this case the difference does not persist in lab-reared fish, suggesting that differences in wild-caught fish do not have a genetic basis, and hence that rapid-start performance is not undergoing evolutionary change in these populations.
Possible mechanisms by which endothermy may have evolved in mammals and birds have been a subject of considerable interest. Although many proposed hypotheses are by necessity purely speculative, others are more amenable to testing through the use of extant species. I am currently working with Dr. Bennett and James Hicks on a study designed to test one model for the evolution of endothermy, known as the thermoregulatory hypothesis. This hypothesis argues that incremental increases in metabolic rate could have provided a fitness advantage to animals, despite increased energetic costs, through the benefits provided by increases in body temperature. A major assumption of this model is that moderate increases in metabolic rate will necessarily allow animals with reptilian physiology to increase their body temperatures. We tested this assumption directly by taking advantage of the four- to five-fold increase in metabolic rate seen in many lizards after consumption of a large meal, and found that higher metabolic rates did not promote either endothermy or homeothermy in animals as large as one kilogram. These results suggest that endothermy may have evolved through an alternative pathway.
In addition to my work on vertebrates, I have also been investigating model systems more suitable for studying long-term evolution under controlled conditions. These studies, funded in part by an NSF postdoctoral fellowship, use laboratory strains of Escherichia coli adapted to different temperatures and pH values to examine the generality of stress adaptations.
In brief, these experiments make use of lines of E. coli founded by Al Bennett and Richard Lenski and propagated for 2000 generations, with replicate lines evolving under one of four conditions: 42°C, 37°C, 32°C, or daily alternation between 32°C and 42°C. The work I am undertaking includes: i) looking at shifts in the tolerance of different lines to life-threateningly high temperatures; ii) looking for changes in growth rate across temperatures in the different lines; iii) testing the degree to which the alternating 32/42°C cycle has produced thermal generalists, relative to the three specialist groups; and iv) determining whether adaptation to temperature stress enhances or diminishes resistance to other environmental stresses. Much of this work is ongoing.
My dissertation work focuses on physiological performance in lizards of the genus Cnemidophorus. This genus is unusual in that it contains a large number of asexual species, all of which had their origin in relatively rare hybridization events occurring between sexual species in the genus. These hybrids produced subsequent generations by the process of parthenogenesis, resulting in strictly clonal, all-female lineages that have shown considerable ecological success.
[More information on asexual Cnemidophorus]
I used this genus to test two different hypotheses that have been proposed concerning asexual animals. The first attempts to explain the high incidence of hybrid origins in asexual animals by suggesting that such hybrids gain an advantage from the high levels of heterozygosity typically present in their genomes. The second is that low genetic variability in asexual populations leads to reduced phenotypic variation relative to sexual species. In addition, I looked at sexual dimorphism within sexual Cnemidophorus species.
These hypotheses were addressed using five whole-animal traits, which reflect physiological limits on an organism's interactions with its environment. Three of these were locomotor traits, each of which reflects some aspect of physiological capacity: burst running speed (reflecting maximal muscular and biomechanical performance), treadmill endurance (a measure of aerobic capacity) and maximal exertion (a measure of anaerobic capacity). I also measured standard metabolic rate and evaporative water loss rate. All of these traits are either known or believed to be important determinants of fitness in desert reptile species like Cnemidophorus.
My work revealed that, in terms of physiological performance, asexual species of Cnemidophorus appear to be at a disadvantage relative to females in related sexual species. Asexuals tend to exhibit inferior trait values on average, and they also exhibit less phenotypic variation than their sexual counterparts. Because males tend to exhibit higher trait values than females within sexual species, these differences are exacerbated if males are included when comparing sexual and asexual species.
[More details on these Cnemidophorus studies]