<p>Invasive exotic species pose an enormous threat to the world's biological diversity. Invasions can alter native communities, replacing local biotas with non-indigenous species introduced by humans. Exotic plant invasions can have negative effects on native flora, which can be in turn detrimental to the herbivores that depend on the vegetation. In this dissertation, I examined the association of an exotic invasive weed, Lantana camara L., with the Asian elephant (Elephas maximus), its food resources (grass and browse), habitat use and feeding behaviour in Mudumalai Tiger Reserve, southern India. Exotic plant invasions are often associated with alterations or declines in native floral species. I first examined the association of L. camara and measured environmental covariates with floral species assemblage and richness, elephant browse plants, percentage grass cover and percentage grass occupancy. A multivariate analysis revealed a significant association of L. camara with floral species assemblage and richness, some elephant browse plants and grass cover within the moist deciduous forest (MDF) and dry deciduous forest (DDF), but not in the thorn forest (TF) of Mudumalai. My results suggest that L. camara appears to be capable of altering the floral community in some habitats. These results also suggest that changes in the floral community and a reduction in grass cover due to L. camara invasion could be detrimental to elephant and other herbivores that depend on grass in this reserve. I then examined the association of L. camara with habitat use by elephant. Elephant dung density was used to assess elephant habitat use from 62 line transects, each 1-km in length. I found no evidence that L. camara was associated with elephant habitat use across habitats, although the interaction term between one habitat (DDF) and L. camara was significantly associated with elephant dung density suggesting that the effect of L. camara was different in different habitats. This indicates that L. camara is associated with elephant habitat use within certain habitats. Habitat and impact of human settlements were significantly associated with elephant habitat use across habitats within Mudumalai. In the DDF, however, only L. camara was associated with elephant habitat use. I conclude that while no significant effects of L. camara were seen across habitats, in specific habitats, negative associations of this invasive plant with elephant habitat use, possibly through the reduction of grass cover, are possible. These results indicate that L. camara appears detrimental to elephant in certain habitats and removal of L. camara in these habitats should be prioritised so as to facilitate growth of grass and native browse species, especially if elephant populations continue to expand. Lastly, I examined the association of elephant behaviour, assessed from feeding and stepping rates, with variation in L. camara invasion. Fifty-seven elephants were observed for a total of 64.3 hours using the focal-animal sampling method. Elephant were never observed to feed on L. camara, but rather fed on grass and browse that were present within and around L. camara patches. Feeding rates (number of trunksful·min⁻¹) were negatively associated with L. camara invasion. A path analysis, which assesses both direct and indirect effects of independent variables, indicated that the total effect of L. camara on feeding rates was 11% less than the direct negative association owing to a positive indirect relationship between L. camara and feeding rates through grass cover and browse density. Lantana camara was not significantly associated with variation in stepping rates (number of steps·min⁻¹). Rather, stepping rates were negatively associated with grass cover and positively associated with browse density. My results indicate that L. camara is potentially capable of changing elephant feeding rates, likely through a loss of grass areas due to L. camara invasion. Wild elephants do not eat L. camara, and this invasive plant appears to take the place of an important food source. My results indicate that managers should prioritize their focus on certain habitats to control the impact of L. camara on elephants and vegetation. However, this study was of a correlational nature based on observational data. Experimental work is therefore needed to test for causal relationships among the variables I measured, over multiple seasons and in different habitats. Experimental evidence will enhance our understanding of how invasive weeds modify floral communities, elephant habitat use and behaviour and help determine whether L. camara is a 'passenger' or 'driver' of these changes in this ecosystem.</p>
Rapid adaptation of insect herbivores to an invasive plant