Preparing seeds for the experiments.

Anthropogenic pressures can have significant effects on local plant and animal communities. A large proportion of medium and large-sized frugivore populations, which are main seed dispersers in many communities, are declining to extinction globally as a result of habitat fragmentation and deforestation. Such decline is often associated with an increase in the abundance of certain rodents, a phenomenon recently defined as “rodentation” by Dirzo et al. (2013). Rodentation can have significant effects on the diversity and structure of plant communities, and the patterns of forest regeneration by altering seed dispersal patterns and plant establishment. For instance, intense rodentation may result in an increase in seed predation and facilitation of species invasion. However, studies of such impacts on natural communities are still lacking. This study will assess the impacts of non-native rodents in Ranomafana rainforests on seed dispersal patterns of native fruit species. Since species invasions are a challenging conservation issue worldwide, understanding processes and impacts of biological invasions is critical and can provide us better insights for forest management and control of invasive species.

Seed-dispersal mutualisms are crucial in forest regeneration; thus, knowledge of their importance to the organization of natural communities is critical in predicting the impacts of their decline on plant communities. This project aims in understanding the impacts of non-native rodent invasions on lemur-plant mutualism. My previous study showed that lemurs play critical roles as primary seed dispersers and have important consequences on the demography of their feeding plants and on forest-gap regeneration. However, in recent field expeditions, I have found almost complete post-dispersal seed predation and removal regardless of the micro-habitats where seeds were dispersed. Unfortunately, we do not know what species remove the seeds and whether such removal is beneficial for the plants.

Rodents can act as seed predators in many communities, leading to limitations of plant establishment. However, they may also assist in effective seed dispersal through secondary removal away from parent trees and/or into more suitable environments. In a secondary forest in Ranomafana, an increased invasion of the non-native rat Rattus rattus was reported with an increase in the occurrence of non-native guava trees from forest edge to interior habitats. Such invasion may disrupt seed dispersal mutualisms leading to negative impacts on forest regeneration patterns.

With a combination of observation data and field experiments, this project will specifically determine:

(a) which rodent species (whether native or not) removes seeds from their primary dispersal location,

(b) how rodents might change the spatial patterns of seed dispersal in term of dispersal-distance from con-specifics and dispersal microhabitats,

(c) how invasion of non-native rodents may lead to change in forest regeneration patterns.