Monday, August 4, 2008 - 1:30 PM

COS 2-1: What factors explain distribution patterns of herbivores along an altitudinal gradient? A tri-trophic community approach

Genoveva Rodríguez-Castañeda, Tulane University, Lee A. Dyer, University of Nevada, and Grant L. Gentry, Samford University.

Background/Question/Methods

Many studies have focused on the distribution of adult insects to characterize diversity patterns across latitude and altitude. This study examined how host specific interactions between Piper (Piperaceae) host plant species and specialized geometrid caterpillars, Eois spp., change across different habitats along an altitudinal gradient. First we asked whether the patterns of density and diversity of Eois are due to: a) increased richness of Eois per Piper species, b) higher average host specificity, or c) higher host diversity. To determine which aspects of the tri-trophic community explained the distribution of Eois species richness, we utilized multiple regression to examine the interactions between peaks of herbivore diversity and local community metrics. Eighty-five 10 m diameter understory plots were created along an altitudinal gradient (400-3200m) in the North-Eastern Andes within the Sucumbios and Napo provinces of Ecuador. All Piper plants present in plots were collected, measured, and weighed. Eois caterpillars were collected and reared at the Yanayacu biological station (2070m elevation). Mutualist ants were collected and associated predator and herbivore species were counted.

Results/Conclusions

A total of 9075 trophic links between enemies, Eois, and Piper were counted from 2000-2007; this included 104 morphospecies of Eois feeding on 40 species of Piper. We found that altitude and availability of resources affected Eois density and diversity in the following ways: 1) Eois species density peaks at middle elevations (1900-2100m), 2) Eois, a genus specialist has higher average host specificity (i.e. higher proportion of monophagous species) at middle elevations, 3) Eois density was highly correlated with host plant biomass, and 4) Eois diversity is not correlated with higher host diversity. These results contrast with what has been found by other tropical ecologists. Notable top-down patterns existed as well. Densities of foraging ants decreased with increasing elevation, while Piper ant mutualists at first increased with elevation (400-1200m) and then decreased sharply (1300-2600m). Parasitism rates of Eois were higher in the lowlands and predation rates of Eois at first decreased with elevation (0-1500 m) and then increased again (1500-2600m). These patterns indicate that a combination of local processes such as high Piper species biomass, absence of mutualistic ants and low parasitism rates are the best predictors of high peaks in Eois diversity across an altitudinal gradient. This supports the general hypothesis that local factors are more important in determining local herbivore species richness than historical or regional plant traits.