Multi-scale analysis of habitat association in a guild of blennioid fishes

The degree to which reef fish are associated with particular reef characteristics has been the subject of much debate. It is increasingly clear that the strength of the relationship between reef fish and their habitat may be dependent on the scales at which the reef habitat is categorised. Consequen...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Marine ecology. Progress series (Halstenbek) 1995, Vol.125 (1/3), p.31-43
1. Verfasser: Syms, Craig
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The degree to which reef fish are associated with particular reef characteristics has been the subject of much debate. It is increasingly clear that the strength of the relationship between reef fish and their habitat may be dependent on the scales at which the reef habitat is categorised. Consequently, scale must be explicitly incorporated into any investigation of fish and habitat association. I addressed the problem of scale by examining changes in the composition of a guild of blennioid fishes (comprising 13 species in the families Tripterygiidae and Blenniidae) relative to the scale at which their habitat was defined. Correspondence Analysis was used to display differences in guild structure. At large, geographical scales, characteristic blennioid assemblages could be detected. Changes in guild structure were due partly to differences in numerical dominance of a set of generalist species and,to a lesser extent, species composition. At brand scales, the blennioid assemblage displayed speciesspecific depth patterns and association with macroalgal cover. A core group of species was found at all depths, while others were restricted in depth and biogenic habitat type. The degree of shelter provided by topographic features characterised the blennioid assemblage at fine scales, and habitat specialisation became apparent at this scale. The patterns detected in this survey indicate scales at which questions about processes generating these patterns may be profitably addressed. I conclude that the incorporation of more than 1 scale in ecological studies is useful for reducing ambiguity and for generating a future research program of hypotheses.
ISSN:0171-8630
1616-1599
DOI:10.3354/meps125031