Bocas del Toro Research Station

Research Projects

J.P. Lawrence

JP Lawrence

Figure 1: A sampling of the different color morphs of Oophaga pumilio found throughout the Bocas del Toro archipelago, Panama.
PhD Student
Biology Department, University of Mississippi
STRI A. Stanley Rand Short Term Research Fellow (2013)

Correlations of Coloration and Toxicity to Diet Specialization in Dendrobatid Frogs Using Illumina Techniques

The Strawberry Poison Frog (Oophaga pumilio = Dendrobates pumilio) has brought researchers from around the world to Bocas del Toro. This is because in this region, this species shows an impressive radiation of color. On each island where these frogs live, a unique morph resides, from bright orange seen on Solarte to dull greenish-brown on Pastores. The entire visible spectrum can be seen in these frogs (Figure 1). This pattern also extends to the surrounding mainland areas. Further, as the Bocas del Toro archipelago started forming within the last 9,000 years, this radiation occurred extremely rapidly. Such radiation of unique and impressive colors on a relative small scale brings up the question of “why?”

Previous research has covered everything from sexual selection to predation to toxicity in an effort to understand what evolutionary processes could promote such polytypism. This is particularly intriguing because polytypism in aposematic species (species that display warning coloration to advertise unpalatability) is counter-intuitive. As coloration is thought to be a signal to predators, such signals should facilitate predator learning. One method to facilitate predator learning is to have consistency in signal so that when predators sample one prey item, they will learn to avoid all. Having variation in signal can weaken the ability to learn avoidance.

Recent research conducted by Martine Maan and Molly Cummings at STRI have found that these colors are honest indicators of toxicity to bird predators meaning color is a predictor for toxicity (i.e., brighter is more toxic). This could mean that toxicity is driving the evolution of color, through predation. Interestingly, Poison Frogs in the family Dendrobatidae derive their alkaloid toxins from dietary sources such as ants, mites, millipedes, and beetles. Ralph Saporito has shown that there are differences among populations in toxicity, which would suggest that different populations of frogs are eating different prey items. If all of this is true, then it is possible that the diet of the frogs could be driving the evolution of the color of the frogs.

JP Lawrence

Figure 2: The other dendrobatid frogs found throughout the Bocas del Toro archipelago. From left to right: Allobates talamancae, Andinobates claudiae, Silverstoneia flotator, Phyllobates lugubris, and Dendrobates auratus.

While O. pumilio remains the most studied dendrobatid in Bocas del Toro, Bocas is home to other species of dendrobatids including Dendrobates auratus, Phyllobates lugubris, Andinobates claudiae, Allobates talamancae, and Silverstoneia flotator (Figure 2). The last two species being in a more ancestral clade of dendrobatids that is non-toxic. All five species are sympatric with O. pumilio and all are monotypic in color form (not to be confused with a monotypic genus). Once again, if colors are honest indicators of toxicity, this lack of variation in color would suggest lack of variation in toxicity. It also would suggest, then, that diet among separate populations should not be variable as compared to the diet of O. pumilio populations. As both A. talamancae and S. flotator are non-toxic, these predictions may not apply to them. However, since they are non-toxic, it has been suggested that they are generalist species, eating a far greater variety of invertebrates as compared to their toxic relatives. But no study to date has examined toxic and non-toxic species living in sympatry (and thus having access to the same invertebrate prey community). Thus the objectives of this study are:

  1. Determine toxin profile and invertebrate prey for all six species of dendrobatids for all sampled populations in the Bocas del Toro archipelago;
  2. Compare diet of polytypic O. pumilio populations to sympatric, monotypic D. auratus, P. lugubris, and A. claudiae populations;
  3. Compare diet variability among toxic and non-toxic dendrobatids living in sympatry.

Traditional methods of determining diet have involved manually sorting through gut contents to morphologically identify invertebrate prey. This has two distinct disadvantages: it is time consuming and partially digested prey cannot be identified reliably. Thus, I plan to extract DNA from the complete stomach contents of frogs and sequence them using high-throughput next generation technology (Illumina). This will amplify the COI mitochondrial barcoding locus which will allow for the identification of invertebrate prey. While partially digested prey may not be identifiable using morphological techniques, it should still have DNA for barcoding, thus allowing for more reliable identification. Toxin profiles will be determined by using both GC-MS (Gas Chromatography-Mass Spectrometry) and LC-MS (Liquid Chromatography-Mass Spectrometry) to determine the complete alkaloid toxin profile for the different populations and species.

To date, I have collected from 13 uniquely colored O. pumilio populations. The other five species are not evenly spread throughout the archipelago like O. pumilio is, so I do not have sympatric representatives for all species for all locations. I do, however, have multiple populations represented for the other five species. I am presently working on analyzing the data and determining my findings.

Back to Current Research