PARENT SESSION
Oral Session #84: Statistical Ecology.
Presiding: L. Subhash
Thursday, August 8. 9:00 AM to 11:30 AM. Mesquite Room, Radisson.
Standardizing sampling effort on autosimilarity for comparing assemblages: effect of species-abundance and species-occurrence distributions.
Cao, Yong*,1, Hawkins, Charles1, Larsen, Phil2, Van Sickle, John2, 1 Department of Fisheries and Wildlife, Logan, UT2 Western Ecology Division, EPA-NHEERL, Corvallis, OR
ABSTRACT- Biological assemblage surveys are conducted to address a variety of questions in ecology, environmental sciences and natural resources management. How well biological samples represent the assemblages of interest affects data interpretation and conclusions. Standard samples may unequally represent the assemblages from which the samples are drawn and thus can introduce bias into assemblage comparisons. Standardizing sampling effort on autosimilarity based on the Jaccard Coefficient (AJC) may resolve this problem, but the effectiveness of this standardization relies on consistent relationships between the percentage of total species richness captured in an assemblage (%TSR) and the AJC value. However, different assemblages often exhibit different species-abundance or species-occurrence distributions. Species-abundance distributions have been extensively investigated, but little is known regarding species-occurrence distributions. In this study we use survey data and simulations to address the effect of variation in both species-abundance and species-occurrence distributions on %TSR-AJC relationships. We first examined species-occurrence distributions for several taxonomic groups at the local and regional scales. Negative exponential, logistic and geometric distributions were observed for different taxa and at different spatial scales. We then simulated a variety of assemblages based on these three types of species-occurrence distributions and two types of species-abundance distributions (logseries and lognormal). %TSR-AJC relationships were fit and compared across these simulated assemblages. Our results showed that %TSR-AJC relationships were highly robust to variation in both types of species-abundance distributions and to variation in each type of species-occurrence distributions, but it is somewhat sensitive to differences among the three types of species-occurrence distributions particularly at low to intermediate levels of AJC. However, because the three types of distributions can be easily differentiated from one another, the proper model needed to estimate %TSR can be identified, thus circumventing any major difficulty. Our results support standardizing sampling effort on AJC to improve the comparison of biological assemblages.
KEY WORDS: autosimilarity, community comparison, sampling effort, species richness