During the six-week OTS Tropical Biology course in Costa Rica, students participated in three different faculty-led projects and worked in groups (or alone) to design two independent projects. The independent projects that I worked on with some friends were in La Selva and Las Cruces. Here are some highlights of those mini-projects.
1. Microhabitat preference of leaf litter herpetofauna at La Selva
Objective: To determine if herpetofaunal diversity was associated with certain microhabitats and what microclimate conditions might explain patterns of diversity.
Methodology: My friend Peter and I had 4 sites, with three 3 x 3 meter quadrats of each microhabitat (fallen log, tree buttress and leaf litter). Two of the sites were close to a river, and had had mud deposited over the leaf litter in many of the quadrats due to flooding at La Selva. Data was collected for this project over 3 days – on the first day, we set up the quadrats and took environmental measurements (canopy cover, leaf litter layers, mud depth, temperature, etc.), and then on the next two days we surveyed for reptiles and amphibians. Troubles encountered included a bullet ant nest, army ants marching through several of our sites, and torrential rains. My favorite memory of possibly the whole trip is when it was raining cats and dogs, we were at the most muddy site, over our ankles in mud, and trying to look for lizards and frogs in a quadrat that had become a mud pit.
The Results: 36 hours after initiating the experiment, we started analyzing data. On the surveys we encountered 8 different species – 4 frogs and 4 lizards. The most frequently encountered species was Craugastor bransfordii. Generalized linear models revealed that only site had any significant interaction with species richness, with richness being higher at two of the sites. One of the sites had significantly fewer leaf litter layers, and mud depth varied by site, but, contrary to our expectations, there were no differences in any of the environmental variables between microhabitats. This may have been because we measured variables for the quadrat, when we should have focused exclusively on the specific microhabitat in question. We also should have included randomized control quadrats.
The moral of the study: There is a lot of variability in leaf litter systems, so more replication is better, and it is important to take into account variation at multiple scales (i.e. variation between sites and between quadrats within a site, temporal variation, etc.).
2. Stream frog and lizard community variation across a land use gradient at Las Cruces
Objective: To investigate how stream amphibian and reptile community composition, species richness and abundance of a common stream lizard (Anolis aquaticus) varies across three different land use areas.
Methodology: My friends Thais, Peter, and I set up six stream transects along one creek – one transect in a garden area, two in selectively-logged primary forest, one in abandoned pasture and two in transitional areas. We did nocturnal surveys on 4 nights, and on the 2nd night, started marking the common stream anole (Anolis aquaticus) with nail polish so that we could get a more accurate estimate of their abundance. During the day, we collected some environmental measurements from the transects. The main obstacles that we encountered during this study were illness and exhaustion, as it was nearing the end of the six-week course and some of us were running on coffee, tequila and fumes.
The Results: We encountered 10 species in our surveys. Anolis aquaticus was the most common species, and the only one found on all six transects. We observed that some A. aquaticus were found sleeping in male-female pairs, but the one pair that we marked on our transect separated after we marked them. The transects in more disturbed habitats (garden and abandoned pasture) had the lowest number of leafy branches crossing the stream, the highest observed herpetofaunal species richness, and were more similar in community composition to each other than to adjacent transects. Anolis aquaticus abundance did not vary significantly by site but was higher in primary forest and transitional transects. However, even on our last night of sampling, we were still catching a lot of unmarked individuals, suggesting that our estimates of A. aquaticus abundance were not terribly accurate.
The moral of the study: Species richness of stream reptiles and amphibians was highest in the more disturbed areas, suggesting that disturbance is not always bad for herpetofaunal communities. However this study was only conducted in a single stream during the rainy season, so we can not extrapolate our results to other habitats, seasons or locations.