Student projects in Costa Rica – part 2: From poison dart frogs to cannibalistic tadpoles at La Selva Biological Station
The January-term course in Costa Rica, Introduction to Field Ecology course, is over. I’m sure everyone is happy to be home, but I hope the memories of the course will stay with people for a long time. I’m finally getting my life in order again, and I thought I would share some information about our final stop on the course – La Selva Biological Station – and post abstracts from the amazing projects that folks did there.
La Selva was a great place for our course to visit. La Selva itself is fairly small (about 1600 hectares) but it is connected to the massive 36,000-hectare Braulio Carrillo National Park. The large combined area and the elevational gradient from lowlands to the 3500 meter peaks in Braulio Carrillo make this area a unique conservation area on the Caribbean slope. The flora and fauna are very diverse: there are over 700 species are trees at La Selva, a lot of beautiful snakes such as the eyelash viper (that many of us saw) and the dangerous and aggressive fer-de-lance (that one of us stepped on!), thousands of arthropod species such as the army ant Eciton burchelli (video), charismatic amphibians like the strawberry poison dart frog (see below), and more than 400 species of birds (representing almost half of Costa Rica’s bird species) (here’s a gratuitous baby peccary video). There’s also extensive infrastructure at La Selva. There is also modern lab space, comfortable living quarters, a cafeteria, and an extensive trail system. Some of the trails through the forest are even paved. La Selva is really wet (it gets about 4 meters of rain a year) and a lot of foot traffic on paths would quickly degrade the surrounding area. Paved paths also make it easier for researchers to travel to their study sites; the station even rents bikes that you can ride along the paths (here’s a video from some of our group). The diversity and infrastructure is part of the reason that La Selva is one of the most important places for tropical research in the world: since the station was established by the Organization for Tropical Studies in 1968, there have been over 1600 scientific papers published based on research at the site.
Our group contributed to this research output with 4 great projects. Here are their abstracts:
Factors that influence territorial calls in the strawberry poison dart frog, Oophaga pumilio
Katelyn Bojan, Maya Peters, Danny Oseid
There has recently been growing concern regarding the potential effects of global climate change on the diversity and population density of amphibians in the tropics (Whitfield et al. 2007). Consequently, it is important to understand the behavior and lifestyle of tropical amphibians for conservation purposes. One amphibian species that is especially important is the charismatic Oophaga pumilio, the strawberry poison dart frog (video). This organism communicates with potential mates and competitors in form of bright coloration and territorial calls. Here, we investigated the factors that influence these signals by addressing: 1) the effect of call intensity and length on the magnitude of neighbor response and 2) the traits of frogs that determine response intensity. It is unknown how these frogs respond to varying calls and how they allocate their energy towards the expensive signals of bright coloration and territorial calls. There are two alternative hypotheses concerning allocation of energy towards these signals. The tradeoff hypothesis suggests that brighter, bigger frogs will call less, whereas the handicap hypothesis postulates that the highest quality individuals will have the best genes, and therefore will produce both signals equally well. We predicted that 1) longer and louder calls would signify a greater relative threat and evoke greater response calls and 2) signaling intensity would fit either the tradeoff or handicap hypotheses. We found frogs were found in disturbed areas and classified them by size (small, medium, or large) and brightness by a scale. We then played vocal recordings of Oophaga pumilio near identified frogs at half or full volume and for a length of 20 or 60 seconds, and recorded time to respond and call duration. We found that louder and longer recordings both evoked a delayed response (F3,56 =6.17 , p= 0.0160; F3,56 = 3.80 , p= 0.0562) and a significantly longer response call (F3,56 =4.24, p= 0.0440; F3,56 = 3.83, p= 0.0551). These results suggest that louder, longer calls signal closer, more aggressive opponents, and frogs recognize these as a greater threat. In addition, larger (F2,26 =4.28 , p= 0.0246), and brighter (F3,26 = 2.94 , p= 0.0520) frogs generally produced longer response calls. This result is consistent with the handicap hypothesis (but not the tradeoff hypothesis), suggesting that the highest quality individuals in the population are both brightly colored and have a good territorial chirp. Overall, this study demonstrates the complexity of Oophaga pumilio signaling behavior and indicates that this organism is an appropriate model system for studying behavioral ecology. Based on these findings, we hope that continued effort will be taken to conserve this species in the long term.
Whitfield SM et al. (2007) Amphibian and reptile declines over 35 years at La Selva, Costa Rica. Proceedings of the National Academy of Sciences, USA 104: 8352–8356.
Fruit Eating Fish: Pattern recognition versus initial quality assessment in Brycon guatemalensis
Evan Nolander, Tyler Abrahamson, Mark Painter, Leah Ruhland
When foraging, animals need to make a decision on when food is worth attempting to acquire. Food quality can often be assessed directly by sampling. However, sampling opportunities may be limited when competition for resources is intense; how such competition affects assessment strategies is poorly understood. Here we examined assessment behavior in Brycon guatemalensis, a frugivorous river fish that groups in large schools underneath fruit trees. We predicted that B. guatamalensis would rely upon a system of pattern recognition for areas with elevated amounts of high quality food, rather than individual assessments of each piece of falling fruit when deciding whether or not to attempt to consume it. We conducted 3 tests to evaluate B. guatamalensis foraging behavior; we replicated each tests at 3 sites. First, we assessed how food quality affects foraging by adding either “high quality” banana pulp or “low quality” banana peel at 20-sec intervals in distinct locations. We found that significantly more fish responded to the high quality food (mean=8.19) than to the low quality food (mean=2.00, t=10.671, df=126, p<0.001), and that high quality food was consumed in less time (mean=3.29 seconds) than was low quality food (mean=13.40 seconds, t=8.883, df=126, p<0.001). In a second test, we examined the timing of food assessment by adding food with an alternating pattern of high and low quality at 20-sec intervals in distinct locations. We again found that significantly more fish responded to the high quality food (mean=5.27) than to low quality food (mean=2.95, t=4.909, df=131, p<0.001), and that high quality food was consumed in less time (mean=2.66 sec) than low quality food (mean=12.45, t=-8.128, df=131, p<0.001). Results from the alternating and non-alternating tests were compared using a two-way ANOVA, and the differences between number of fish responding to both high quality food and low quality food were found to be significantly different (F(1,257)=26.92, p<0.001). No significant difference was found between time pre-consumption (F(1,257)=0.03, p=0.8520). A third test was run to assess if more high quality food or a higher frequency of food regardless of quality affected foraging. This was done by adding high quality food at 30-sec intervals in one area while adding food with a pattern of high quality- low- low- low every 15 seconds to another distinct area. We found that fish responded with higher numbers in the area receiving a higher amount of high quality food (mean=7.60) than the high quality food in the higher frequency test (mean=5.35, t=2.417, df=63, p=00185). However, the high quality food in the high frequency test was consumed faster (mean=0.45 sec) than that of the low frequency test (mean=3.57 sec, t=3.312, df=63, p<0.001). Low quality food was responded to by less fish (mean=4.16) and consumed more slowly (mean= 8.81 sec). Contrary to our prediction, we found that food was both individually tested for quality as well as pursued based on pattern recognition of high quality food. Overall, it can be concluded that B. guatamalensis react to food hitting the surface in a manner of pattern recognition but use quality assessment after food enters their mouth.
Using the Island Biogeography model to determine island characteristics in tropical microhabitats
Jordan Goetting, Amy Niemela, Tom Langer
Boulders in streams and fallen logs provide a unique place for tropical plants to grow, making these surfaces important contributors to species diversity. The Equilibrium Model of Island Biogeography (EMIB) is a well-established model that predicts the species richness on islands based on rates of immigration and extinction. These rates are predicted to be affected by the size of an island and its proximity to source populations (the “mainland”). In the rainforest within La Selva Biological Station the boulders and logs are distinct habitats that could act as islands within their greater environment. By accepting the EMIB as a working model of the forest, we were able to use it as a tool to determine whether or not these two microhabitats could be defined as islands. We hypothesized that species diversity on both fallen logs of a forest floor and boulders within streams would show trends predicted by the EMIB: species diversity should increase with island size and decrease with distance from source populations. We measured the surface area of ten “islands” and the distance from each island to the nearest island at each of three different sites per microhabitat. We also measured the number of plant species on each island. As predicted, there was a significant positive correlation between island size and species diversity for both microhabitats (fallen logs: R² = 0.4643 p<0.0001, boulders: R² = 0.4041 p=0.0003). However, we found that there was no significant relationship between species diversity and distance from other islands for either microhabitat (fallen logs: R2=0.0611 p=0.1879, boulders: R2=0.0440 p=0.2840). Therefore these microhabitats follow the EMIB based on the island size but not on connectedness. This implies that colonizers are not coming from the nearest island that we measured distance to, but rather they are coming from an alternate “mainland” source. We conclude that boulders and logs within a tropical ecosystem function as ‘islands’ in certain ways to increase species diversity. This is important to take into account when clearing forests or damming streams.
Tadpoles with a side of fava beans: Cannibalistic behavior of Bufo marinus tadpoles at the La Selva Biological Station, Costa Rica
Cory Birkestrand, Liz Chambers, Braedon Wieseler, Matt Scott
The prevalence of cannibalism in wild populations of aquatic animals is often underestimated. Cannibalism may be advantageous due to the nutritional benefit of consuming animal tissue, especially where food is scarce or less nutritious. A potential consequence for cannibalistic behavior is the potential disease transmission from the deceased organism to the consumer. However, there are costs associated with cannibalism due to the challenge of fighting an equal and the potential for disease transmission from a conspecific carcass. Here, we study behavior associated with cannibalism in tadpoles of the cane toad Bufo marinus. We studied an aggregation of tadpoles on the banks of the Rio Sarapiqui in La Selva Biological Station (watch their cool movements here). Rapid changes in the level of the river caused some of the tadpoles to become trapped within small tide pools that eventually lead to their death. After the water rose, the surviving tadpoles would consume carcasses of the dead B. marinus tadpoles. Here we tested whether the tadpoles were more reluctant to consume a tadpole carcass or the pulp of Passiflora fruit from a nearby tree. We conducted feeding trials in low (<50 tadpoles/m2), medium (50-250 tadpoles/m2), and high (>250 tadpoles/m2) density areas in natural habitat. We measured initial discovery time (the time it took for the first tadpole to begin feeding on an item) and the difference in time between the fourth discovery and fifth discovery (= “5th discovery time”). We predicted that tadpoles would initially take longer to begin feeding on a carcass but subsequent discovery would be faster for carcass feeders due to the potential costs associated with feeding on a conspecific. We found that, at all densities, tadpoles discovered fruit faster than carcasses (p<0.0001). However, the difference between the length of the 5th discovery time and that of the initial discovery time was much larger from carcasses than for fruit. These results suggest that tadpoles were more reluctant to feed on carcasses than on fruit, but once carcasses were fed upon by one tadpole reluctance in other tadpoles disappeared. In an isolated pool, we tested (video) whether tadpoles would become acclimated to cannibalistic feeding by examining whether discovery time would decrease on successive feeding trials. Our results showed that feeding on the carcass occurred more readily after previous exposure, suggesting that tadpoles that have previously taken part in cannibalistic activities are more likely to feed readily on a carcass in the future. Together, our results suggest that B. marinus tadpoles respond to the costs and benefits associated with cannibalism.