Littoral macroinvertebrate communities: spatial scale and ecological relationships

SUMMARY 1. Spatial correlations between ecological patterns and processes are thought to be scale‐dependent, yet surprisingly few studies have evaluated the correspondence between different levels of spatial scale and ecosystem structure and function. 2. We evaluated the strength of relationships be...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Freshwater biology 2002-10, Vol.47 (10), p.1840-1854
Hauptverfasser: Johnson, R. K., Goedkoop, W.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:SUMMARY 1. Spatial correlations between ecological patterns and processes are thought to be scale‐dependent, yet surprisingly few studies have evaluated the correspondence between different levels of spatial scale and ecosystem structure and function. 2. We evaluated the strength of relationships between the benthic macroinvertebrate communities of stony littoral habitats and levels of ecological scale and geographical position, using partial constrained ordination. Our hypothesis was that correlation strength would be inversely related to ecological scale, i.e. habitat > ecosystem > riparian > catchment > ecoregion. 3. The effect of habitat was greater than that of other levels of spatial scale: 23% of the variance in taxonomic composition and 11% of that in functional composition was explained by habitat variables alone. However, greater spatial scales were also important. For example, the combined influence of riparian, catchment and ecoregion classification accounted for 24% (taxonomic) and 11% (functional) of the explained variance. 4. Relationships between organisms and scale variables were, however, non‐linear and a substantial amount of the functional variance was hidden in joint effects. These findings were not unexpected, and presumably indicate a close interdependence between local and regional‐scale variables.
ISSN:0046-5070
1365-2427
DOI:10.1046/j.1365-2427.2002.00932.x