A method to determine rates and patterns of variability in ecological communities

• Published in: Oikos Vol. 91; no. 2; pp. 285 - 293
• Main Authors: Collins, Scott L., Micheli, Fiorenza, Hartt, Laura
• Format: Journal Article
• Language: English
• Published: Copenhagen Munksgaard International Publishers 01.11.2000; Munksgaard International Publishers, Ltd; Blackwell
• Subjects: Animal and plant ecology; Animal, plant and microbial ecology; Biological and medical sciences; Datasets; Experimental data; Fundamental and applied biological sciences. Psychology; General aspects; General aspects. Techniques; Linear regression; Methods and techniques (sampling, tagging, trapping, modelling...); Modeling; Ordination; Simulations; Species; Synecology; Time series; Vegetation; Biocenosis; Steppe; Old field; Analytical method; Lakes; Euclidean space; Simulation model; Measurement method; Autocorrelation; Time variation; Zooplankton
• DOI: 10.1034/j.1600-0706.2000.910209.x
• ISSN: 0030-1299

It is well known that ecological communities are spatially and temporally dynamic. Quantifying temporal variability in ecological communities is challenging, however, especially for time-series data sets of less than 40 measurement intervals. In this paper, we describe a method to quantify temporal variability in multispecies communities over time frames of 10-40 measurement intervals. Our approach is a community-level extension of autocorrelation analysis, but we use Euclidean distance to measure similarity of community samples at increasing time lags rather than the correlation coefficient. Regressing Euclidean distances versus increasing time lags yields a measure of the rate and nature of community change over time. We demonstrate the method with empirical data sets from shortgrass steppe, old-field succession and zooplankton dynamics in lakes, and we investigate properties of the analysis using simulation models. Results indicate that time-lag analysis provides a useful quantitative measurement of the rate and pattern of temporal dynamics in communities over time frames that are too short for more traditional autocorrelation approaches.