PARENT SESSION
Oral Session #84: Statistical Ecology.
Presiding: L. Subhash
Thursday, August 8. 9:00 AM to 11:30 AM. Mesquite Room, Radisson.
Predicting optimum sampling strategies to capture species-environment interactions.
PIERCE, KENNETH*,1, URBAN, DEAN1, 1 Duke University, Durham, NC
ABSTRACT- We developed a model to predict the effectiveness of geostatistical sampling plans designed to capture multiple interacting spatial phenomena. While it is widely regarded that issues of scale are important to study ecological systems, it is often acknowledged after the fact. We implemented two studies designed to capture the interaction between seed source and environment as they affect woody seedling spatial patterns. The studies were conducted in Sequoia National Park, one was on a 1.25 ha stem map and the second was across the surrounding 60 km2 watershed. For big long-lived sessile species, studies at too fine a scale approximate a uniform gradient. At too wide a scale, communities have little overlap. In the smaller study, 80% of the species were represented on >90% of the sample plots and seedling patterns appeared random in space. At the wider scales, communities were obviously separated by the complex elevation gradient. While species are obviously separated it is difficult to discern why or where environment is affecting species distributions. Our model examined the relationship between seedrain, local trees and the environmental gradient. Results indicate, based on the environmental gradients and dispersal kernels of species at our filed site, that an intermediate extent would have been most effective at capturing both environmental and biotic interactions. Models of this sort could greatly assist in planning field sampling and monitoring protocols.
KEY WORDS: sampling design, plant-environment interactions, mantel test, gradient analysis