Some members of the Servedio Lab: Maria Servedio, postdoc Rob Lachlan, visiting student Machteld Verzijden
♦ Research
♦ People
♦ Publications
♦ CV
♦ Contact & Links
♦ Programs
♦ Photos
The primary research projects in my lab span topics from evolutionary genetics to behavioral ecology. Prior, current and future projects of mine focus on theoretical studies of speciation but include work on mate choice, aposematic coloration, brood parasitism, systematic methodology, and species delimitation. My main goal is to use mathematical models to integrate rigorous evolutionary theory with hypotheses explaining behavioral and ecological patterns and phenomena.
Speciation and reinforcement Mate choice
Warning Coloration and mimicry Additional interests
All species concepts that address populations in physical contact require mechanisms of isolation to prevent the fusion of gene pools. These mechanisms can occur at several stages during the process of interbreeding. Individuals from different populations can mate, but their hybrid offspring can have low fitness (postzygotic incompatibilities). Alternatively, mating could occur, but there could be incompatibilities, such as between sperm and eggs, that are expressed before hybrid zygotes are even produced (postmating, prezygotic incompatibilities). Both of these categories are forms of postmating isolation between species. Finally, individuals from the different populations can be prevented from mating in the first place (premating isolation). One of the most interesting ways that this can happen, in my opinion, is through mate choice; females can evolve to choose not to mate with males of a different population. Most of my work on speciation concentrates on the evolution of premating isolation through mate choice.
When mechanisms of postmating isolation are present in a population, it becomes adaptive for premating isolation to evolve, to prevent the wastage of gametes in unproductive matings or unfit hybrids. This adaptive evolution of premating isolation is called reinforcement (in its broad sense). Using mathematical models, I have been making predictions about what biological factors will tend to make reinforcement more or less likely to occur.
In one project, for example, I used population genetic models to show that reinforcement between two panmictic populations occurs less often as directional asymmetry in gene flow increases (Servedio and Kirkpatrick 1997). Reinforcement will therefore be especially rare under the pattern of one-way migration from a founder population into a peripheral isolate. This implies that it may be difficult for peripheral isolates to gain complete reproductive isolation.
Populations under vicariance vs. peripheral isolation.
A combination of sensory bias and ecological character displacement can cause direct selection on preferences to mimic the pattern of reproductive character displacement.
It has been traditional for studies of reinforcement to require that hybrids have low fitness. In another project, I explore other forces that may promote the evolution of premating isolation (Servedio 2001). Direct selection, for example, would be placed on female preferences if mating choices alter the reproductive output of females. I found that during the process of preference divergence, direct selection on preferences can drastically overwhelm the force caused by selection against hybrids. Direct selection on preferences can also mimic the pattern of reproductive character displacement, generally assumed to be the signature of reinforcement. Direct selection should therefore be carefully eliminated as a cause of preference divergence by empiricists before reinforcement is assumed to have occurred.
In this same paper I also explore another source of selection on preferences besides low fitness of hybrids, namely postmating, prezygotic incompatibilities. These occur when females have reduced fertility or suffer higher mortality at the time of mating if they mate with the wrong type of male. I show that postmating, prezygotic incompatibilities can be a more effective than the low fitness of hybrids as a cause of premating divergence. Postmating, prezygotic isolating mechanisms may therefore prove to be an important force occurring in conjunction with traditional reinforcement, or may explain patterns of preference divergence when hybrids have high fitness.
More recently I have collaborated with Glenn-Peter Sætre (Servedio and Sætre, 2003)* on a project that builds on recent empirical findings that genes controlling post- and prezygotic isolation are often linked on the sex chromosomes. We use population genetic models to demonstrate that such linkage results in a positive feedback loop in which both post- and prezygotic isolation are strengthened, and that this effect is exaggerated by sex linkage.
Other projects on speciation and reinforcement are described in these publications, and are ongoing in the lab.
Much of the work in speciation described above involves mechanisms of sexual selection and mate choice. I also have a general interest in mate choice and sexual selection, and am continuing work in this area.
One example of a past project concerns the evolution of mate choice copying. Mate choice copying occurs when females in lekking or promiscuous species are more likely to mate with a male they observe to be chosen by other females. With Mark Kirkpatrick (Servedio and Kirkpatrick, 1996) I demonstrated that contrary to prior conjectures, mate choice copying can evolve even when it is not costly for a female to gather information and make choices about mates. Since such costs have rarely been found, this implies that mate choice copying may be less taxonomically limited than previously thought.
Warning coloration and mimicry:
Another topic in behavioral ecology that I am interested in is the evolution of interspecific signaling involved in warning coloration and mimicry. Below I describe two such projects, one concentrating on the evolution of warning coloration in general, and the second concerned with the mimetic signaling involved in brood parasitism.
In Servedio (2000a) I use dynamic models to explore the effects of predator learning, forgetting, and recognition errors on the evolution of warning coloration in distasteful prey species. I show that predator learning under a taste aversion model, with one-trial learning and a constant probability of attacking in error, promotes the evolution of warning coloration by the fixation of very bright mutations or of successive mutations each which cause small increases in a prey’s conspicuousness. In contrast, learning under a classical conditioning model, with gradual learning and forgetting, makes it very difficult for aposematic coloration to become established unless the frequency of conspicuous morphs can cross an often high threshold through chance factors alone. The way in which predators learn and forget can therefore have a profound effect on both whether warning coloration will evolve and the manner in which such evolution might proceed.
In another project, the evolution of mimicry and discrimination is explored in the context of brood parasitism. Brood parasitism presents an evolutionary paradox. Although it seems that hosts should evolve to consistently reject the mimetic eggs of cuckoos, they fail to do so, even when the genetic variation for rejection is clearly present. With Russ Lande (Servedio and Lande, 2003)* I built a quantitative genetic model to understand the evolutionary constraints acting on host discrimination ability and host and cuckoo egg characters. We found a stable equilibrium for coexistence of the host and cuckoo where there is cuckoo egg mimicry, evolutionary displacement of the host egg away from the cuckoo egg phenotype, and host discrimination against unusual eggs. Both host discrimination and host egg displacement are fairly weak at the equilibrium. Cuckoo egg mimicry, although imperfect, usually evolves more extensively and quickly than the responses of the host. Our model therefore provides evidence for both the evolutionary equilibrium and evolutionary lag hypotheses of host acceptance of parasitic eggs.
The equilibrium value (follow trajectories) for host egg size (zh), cuckoo egg size (zc), and host discrimination (b). The arrow shows the optimal egg size values under natural selection alone.
I have also collaborated with John Wiens in research projects on methods of phylogeny reconstruction (Wiens and Servedio, 1997,1998) and on species delimitation. In one paper (Wiens and Servedio, 2000), for example, we propose a new statistical method that would allow researchers collecting morphological data to apply the use of frequency cut-offs in the presence or absence of character states in species delimitation studies.
*This work was funded by the National Science Foundation grant DEB 0234849. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the view of the National Science Foundation.
Home People Publications CV Opportunities Contact & Links Programs Photos