EXTRACTING INFORMATION FROM ECOLOGICAL DATA CONTAINING HIGH SPATIAL AND TEMPORAL VARIABILITY: BENTHIC MICROFLORAL PRODUCTION

Publisher Website
Google Scholar
Abstract
About 1,000 sediment cores from a salt water lagoon in Ventura County, CA were incubated between September 6 and October 6, 1981. The data were analyzed using stepwise regression, time series, multichannel information, and entropy data analysis. Stepwise regression analyses were not informative in explaining changes in productivity because the correlations between productivity and individual environmental variables (chlorophyll a, incident radiation, water temperature, mean tidal range, hours of subaerial exposure, benthic community respiration, pheophytin a/chlorophyll a, and initial dissolved oxygen concentration) varied through time. Time series and multichannel information analysis were used to uncover dynamic behavior in the data and to optimize sampling strategy. The information analysis indicated that, for a long range study, considerably more information could be obtained by concentrating samples into a 14-day interval within each month, rather than sampling throughout the month. Entropy data analysis indicated that productivity was controlled by environmental variables acting in combinations; that is, the microflora respond to the environment as a whole, and no single variable limits productivity. The study emphasizes the applicability of entropy data analysis lo ecological data, because of its ability to model data containing sporadic changes, spikes, and dynamic relationships.

This publication has 23 references indexed in Scilit: