The utility of visual estimation of cover for rapid assessment of graminoid abundance in forest and grassland habitats in studies of animal foraging

Abstract

2 3 2 4Running head: Rapid assessment of graminoid abundance by visual estimation 2 5 2 6 2 The utility of visual estimation of cover for rapid assessment of graminoid abundance in 2 7 forest and grassland habitats in studies of animal foraging 2 8 2 9 Abstract: 3 0Questions: To assess the feasibility of using visually-estimated cover in rapid assessment of 3 1 herbivore food species abundance in the grass layer, we asked the following questions: 1) 3 2What is the relationship between total graminoid cover and biomass in forests, and does 3 3 height improve the prediction of biomass from cover? 2) How does total cover relate to 3 4 biomass in a grassland habitat? 3) How does elephant food species graminoid cover relate to 3 5 individual species biomass? 4) How well does species diversity of forest understorey grass 3 6 layer, calculated from cover data, mirror that calculated from biomass data? Location: 3 7 Nagarahole National Park, India. Methods: We estimated the abundance of graminoids 3 8 through visual estimation of cover and weighted harvested biomass in forest and grassland 3 9plots. In forests, two estimates of total graminoid abundance, total graminoid cover and sum 4 0 of species covers, were used. In the grassland, only total graminoid abundance was measured. 1We examined the relationship between cover estimates and biomass, and the additional utility 4 2 of height in predicting biomass, using multiple regressions and AIC-based model selection. 3We also assessed similarity in cover-and biomass-based Simpson's and Shannon-Wiener 4 4 diversity indices using regressions. Results: Graminoid cover explained a large portion of 4 5 variation in total graminoid biomass in forest and grassland habitats. The sum of species 4 6covers was better than total cover in estimating total graminoid biomass in the forest. The 4 7 benefit of including height to estimate total biomass was moderate in forests but substantial in 4 8 grassland. Cover estimates were good proxies of food species biomass, and the addition of 4 9 height did not yield better models for most species. Species diversity indices calculated from 5 0 cover largely matched those based on biomass. Conclusions: Visual estimation of species 5 1cover is a good alternative to biomass harvesting for rapid assessment of abundance of 5 2 graminoids consumed by generalist herbivores, like elephants. 5 3