Caigual Stream, Trinidad
Focal Streams
We created a new series of introduction experiments in 2008-9. We initiated four replicate introduction experiments in which we transplanted guppies from a single HP locality to four previously guppy-free localities. These experiments recreate the scenario of HP guppies invading and adapting to a predator-free habitat. HP guppies initially invade at low densities, so they have abundant food resources and select the highest quality prey, leading to dramatic population growth. From here, an LP phenotype may evolve as a direct response to reduced mortality and/or as an indirect response to increased population density and reduced resource availability.
|
We overlay these introductions with an experimental manipulation. The four focal streams consist of two pairs. One member of each pair has an intact forest canopy and one has a thinned canopy. We thinned the canopy to manipulate resource availability independently of predation; thinning increases light penetration and primary production. We have since confirmed that thinning has had this effect in three ways. Primary productivity and the biomass of invertebrates is higher under the thinned canopies. Guppy growth rates, asymptotic body sizes and body condition are also consistently higher under the thinned canopies. If risk of mortality is the sole cause of guppy evolution, then thinning should have no effect. If, instead there is an interaction between mortality risk and resource availability, the course of evolution should differ between intact and thinned canopies and selection should fluctuate with resource availability. Our first two years of data show evolution of male coloration, morphology and age of maturation and fluctuating selection and trade-offs between viability and fertility selection.
|
Our collaborators have generated a time-series description of the ecosystem in the guppy introduction site and in a control section of stream upstream from where the guppies were introduced, beginning one year before the guppy introduction. Thus far, we have found that guppies significantly deplete the abundance of invertebrates in the introduction site.
Other collaborators have performed a time series, mark-recapture study of the Rivulus populations in the introduction and control sites. Matt Walsh, a recent PhD student, established that guppies cause Rivulus life histories to evolve. His results yield the prediction that they do so via their indirect effect on Rivulus abundance. In past introduction sites (experiments that are now >25 years old), we found that the presence of guppies was associated with reduced Rivulus abundance and increased individual growth rates. Matt’s results imply that Rivulus that co-occur with guppies have adapted to increased resource availability. In our focal stream experiments, we have documented a significant reduction in Rivulus abundance and shift in their size distribution in response to the guppy introduction after only two years.
|
Our experiment was designed to generate resources for the future study of the genetics of adaptation. Guppies were collected as juveniles from a single locality where they live with a diversity of predators (high predation environments). They were reared to maturity in single sex groups, then mated before being introduced. All introduced fish were individually marked. Scales were kept from each of them to provide a source of DNA. All four populations are intensively censused once a month so that we can follow individual growth and movements. All new recruits are individually marked and scales are collected. All individuals are photographed every time they are caught. The photographs provide a source of data for characterizing growth rate, shape and male coloration. We have genotyped all individuals at 12 hypervariable, tetranucleotide microsatellite loci and have been able to reconstruct the pedigrees of four replicate, evolving populations. We archive around 80% of the DNA from each scale sample.
|
Guanapo River, Trinidad