Molecular evidence for pika diet composition in the Columbia River gorge

Background/Question/Methods

Climate change has already profoundly affected species distributions. Habitat specialists appear to be particularly sensitive to climate change, but plastic behaviors, such as incorporating novel food items into the diet, may allow specialists to tolerate new resource landscapes. The American Pika (Ochotona princeps) is a habitat specialist in the high mountains of western North America at elevations of 2500m and higher.  Resurveys have shown local extinctions of pikas in the Great Basin in response to climate change. However, pikas survive near sea level in the Columbia River Gorge in an unusual temperate rainforest climate. Gorge pikas exhibit dietary plasticity unusual for this species because they consume large amounts of moss, which covers their habitat. Over two years, we observed pika foraging behavior and collected fecal samples from marked animals at two sites where over half of the diet is composed of mosses. We sought to expand these observations with molecular evidence of the diet, reconstructed from fecal samples. We included fecal samples from marked animals, whose feeding habits we observed in the field. We extracted DNA from pika feces and amplified the plastid psbA-trnH spacer with PCR. PCR products were then gel-purified and sequenced.  Plant species were identified using BLAST.

Results/Conclusions

We successfully recovered sequences of 10 plant species from pika feces, including several species that are not consumed by pikas in typical habitat. These include ferns (Polypodium spp.), Douglas Fir (Pseudotsuga menziesii), and a liverwort (Ptilidium pulcherrium). In fact, typical food resources for pikas (grasses and forbs) comprised less than 30% of both sequences and diet observations in the field. However, we did not detect any mosses with this technique. Potential explanations for the disparities between sequence data and field observations include PCR inhibitors in fecal samples, poor quality DNA, or improper collection and storage of fecal samples. In addition, although the primers and PCR conditions that we used were validated to identify bryophytes in the diet of barnacle geese, they may still underestimate the moss species consumed by pikas in the Columbia River Gorge. Several techniques may improve the results of future studies, including testing primer specificity or optimizing sample collection and extraction protocols. Our results highlight the need to validate molecular reconstructions of diet with field observations before analyzing diet across seasons or habitats.