PARENT SESSION
Oral Session 16: Statistics and Biometrics.
Presiding: E Garton and T Simons
Monday, August 2, 8:00 AM to 11:30 AM, Meeting Room D 139.

Wavelet analysis as an ecological monitoring tool.

Anthony, Jill^*,1, ¹ Oregon State University, Corvallis, Oregon, USA

ABSTRACT- Long-term, large-scale ecological monitoring programs are increasingly important in conservation as the questions of sustainability become more complex. The advanced statistics of pattern analysis detect ecological pattern across multiple scales, characterize change in pattern over time, measure current levels of variability, and identify changes in variables that require action. Additionally, effective ecological monitoring must involve sampling designs sufficiently detailed to detect ecologically significant patterns at multiple scales, yet logistically tractable and resource-efficient for sustained use. Pattern detection and the quantification of complex ecological data increase our comprehension of the reciprocal relationship between pattern and process across scales. Integrating individual features at one scale as a texture at multiple scales translates ecological patterns as a multi-dimensional volume. Moreover, visualizing complex relationships as a space-time volume captures a closer approximation of the way ecological relationships exist in nature: nonlinear, multi-scalar, uncertain. Wavelet analysis is a sophisticated tool with adequate flexibility to investigate the reciprocal relationship between pattern and process. Whether to use one- or two-dimensional wavelet analyses depends on the available data, question of interest, and specific parameters of the study design. We explore the advantages of wavelet analysis as an ecological monitoring tool relative to several other time series and spatial statistics used in landscape ecology (e.g., Fourier analysis, autocorrelation). With examples from the Breeding Bird Survey, we demonstrate the effectiveness of wavelet analysis to optimize density, extent, and sampling resolution in multi-scalar ecological sampling and monitoring programs.

Key words: pattern, avian, scale