Agricultural production has expanded rapidly in tropical regions, transforming the landscape and fragmenting tropical forests. However, little is known about the long-term effects of agricultural production practices (i.e., management intensification) on population connectivity and dispersal. Our study integrates genetic and landscape data to examine landscape factors influencing the connectivity of a tropical small mammal. Specifically, we investigated landscape factors driving the population structure of Heteromys goldmani, a common forest rodent, in a coffee agroecosystem. We evaluated 5 different landscape variables; (1) tree cover, (2) slope, (3) elevation, (4) distance-to-stream and (5) dispersal barriers (streams), to measure their influence in driving H. goldmanigenetic patterns in the coffee agroecosystem.
Results/Conclusions
Our results show that H. goldmani dispersal is limited in the coffee landscape, characterized by discrete genetic populations with limited gene flow. We found that all landscape variables, except streams, have a significant correlation with the observed genetic structure. These results suggest that streams do not represent a barrier to dispersal of H. goldmani in our study site. On the other hand, distance-to-stream was the variable with the strongest correlation with genetic structure. Areas close to streams in these coffee farms are steep, hard to access, and tend to be unmanaged. Our results suggest that these areas are serving as habitat or corridors, facilitating dispersal, and allowing H. goldmani to survive within the coffee farms.