Spatial eigenfunction analyses in stream networks: do watercourse and overland distances produce different results?

Abstract

SUMMARY1. The use of spatial variables is a common procedure in ecological studies. The techniqueis based on the definition of a connectivity⁄distance matrix that conceptually defines thedispersal of organisms. The shortest distance between two points is a straight line. Despitethe fact that a straight line may not represent the easiest dispersal path for many kinds oforganisms, straight-line distances are often used to detect patterns. We argue that othertypes of connectivity⁄distance matrices will better represent dispersal paths, such as thewatercourse distance for aquatic organisms (e.g. fish, shrimps).2. We used empirical and simulated community data to evaluate the usefulness of spatialvariables generated from watercourse and overland (straight-line) distances.3. Spatial variables based on watercourse distances captured patterns that straight-linedistances did not, and provided better representations of the spatial patterns generated bydispersal along a dendritic network. Keywords: community, dispersal, overland, stream networks, watercourse