Seed dispersal and seed predation by animals are key natural processes that affect forest regeneration and the evolution of plant diversity. The role of rodents in this regard is controversial, having been argued as good for trees because they disperse and bury seeds, bad for trees because rodents eat seeds, or conditionally mutualistic depending on local food or predators. This relationship remains uncertain because few studies have been able to follow seeds to determine their ultimate fate. This study uses motion-sensitive radio transmitters to track the fate of rodent-dispersed seeds. The hypothesis tested is that rodents are conditional mutualists of trees: they are effective dispersers only where seeds are at low density AND rodents have a high likelihood of dying, thus leaving their buried seeds to germinate.
The factors affecting whether rodents are dispersers or predators of seeds are of major importance for forest management, where rodents still tend to be viewed as pests that should be killed. Our study will provide a framework for understanding the role of rodents, and of how disturbances such as fragmentation and hunting affect forest regeneration by modifying predator and rodent abundances. Emphasis on scientifically interesting research with charismatic rainforest animals and important conservation implications will put the study in a strong position for continued involvement in educational programs.
Seed dispersal and seed predation by animals are key natural processes that affect forest regeneration and the evolution of plant diversity. The role of rodents in this regard is controversial: 1) having been argued as good for trees because they disperse and bury seeds, 2) bad for trees because rodents eat seeds, or 3) conditionally mutualistic depending on local food or predators. This relationship remains uncertain because few studies have been able to follow seeds to determine their ultimate fate. We used motion-sensitive radio transmitters and camera traps to track the fate of rodent-dispersed seeds. We were able to track seed movement over larger areas and for longer time periods than any other study. One of our key discoveries was that many seeds recovered from rodent caches were not immediately eaten, but moved and reburied by thieving rodents. The degree to which rodents stole caches from other rodents was confirmed through extensive camera surveillance of cached seeds in combination with a population of marked agoutis. This continued shuffling of buried seeds resulted in longer dispersal distances for the seeds, which is good for survival of the lineage of the parent tree. We also used camera traps to monitor mammal communities, including predators and prey, and plotted this variation in relation to maps of nut-tree abundance. This project was conducted in collaboration with 15 graduate students, who worked with us to write 17 theses, as well as 5 undergraduates and 2 post docs. In addition to peer-reviewed papers, we share information about our research via a blog (http://Agoutienterprise.wordpress.com). Our work has been included in a major High School Biology Text Book (Miller and Levine) and in a new documentary by the Smithsonian Channel. Additionally, our database of camera trap images formed the base for a new website sharing camera trap images from around the world with the public. Since its launch in Feb 2011 this site (Smithsonian Wild: http://siwild.si.edu) has attracted >1 million unique views.
