Chronic browsing by an introduced mammalian herbivore in a tropical island alters species composition and functional traits of forest understory plant communities

Mammalian herbivores have large-scale impacts on vegetation, altering structure and species composition, especially in tropical grasslands and savannas. However, there is limited understanding of the potential impacts of mammalian herbivores in tropical wet forests, where they are typically less abundant. We investigated the effects of an introduced mammalian herbivore chital (Axis axis) on vegetation structure, composition and leaf functional traits of tropical evergreen forests of the Andaman Islands, India. Across seven islands, representing a gradient of herbivore densities, increasing chital presence was associated with decreased understory richness, understory density and adult tree richness, but was not related to adult tree density or size class distributions. We also found a significant decrease in community-level leaf palatability traits (specific leaf area decreased and leaf thickness increased) with increasing chital habitat use. This community-level shift in leaf trait values was better explained by intra-specific variation in leaf traits across islands rather than changes in species composition. In summary, we show persistent long-term impacts of an introduced mammalian herbivore on understory tropical tree communities although there is little impact on adult tree communities. Our results also show that functional traits of species can be altered in response to novel herbivory, even at herbivore densities where there are no detectable impacts on adult forest structure or composition. Such altered functional traits may potentially alter ecosystem functioning in these forests even without changes in vegetation structure.This dataset reports vegetation community metrics from a few islands in the Andaman Islands, India. It has three kinds of data, with different types of collection. 1. Adult tree inventory: Collected across 100 m x 5 m belt transects with random start points. All trees >10 cm gbh measured. 2. Understory inventory: Collected within 3 uniformly spaced 5 m x 5 m quadrats along the adult tree transect. All understory plants > 10 cm height considered. 3. Community functional traits: Collected from leaves of 20 most dominant plants in the total understory community across each island. Leaves collected from healthy plants across the area sampled in the island. 4. Herbivore habitat use intensity: Dung pellet group counts of chital (Axis axis) at the transect level