Premium
Correlation Between Habitat Structure and a Rocky Reef Fish Assemblage in the Southwest Mediterranean
Author(s) -
Charton J. A. Garcia,
Ruzafa A. Pérez
Publication year - 1998
Publication title -
marine ecology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.668
H-Index - 58
eISSN - 1439-0485
pISSN - 0173-9565
DOI - 10.1111/j.1439-0485.1998.tb00457.x
Subject(s) - assemblage (archaeology) , habitat , geography , fishery , mediterranean climate , reef , mediterranean sea , ecology , fish <actinopterygii> , coral reef fish , oceanography , geology , biology
. We examined the correlation between the variables describing a visually censused reef fish assemblage and several habitat structure descriptors both vertical complexity and horizontal heterogeneity in the southwest Mediterranean. We used multiple regression analyses in the framework of generalized linear models GLM, and a redundancy analysis RDA evaluated through a Monte Carlo permutation test with a 1% level of significance. A significant positive relationship was observed between habitat complexity as measured by the number of rocky boulders of several sizes, the number of species 250m ‐3 P < 0.05 and the mean total abundance P < 0.001. Species diversity decreases with the increment of large boulder numbers. although the correlation is weak adj. r 2 = 0.27, P < 0.05. The canonical ordination has discerned only the effect of depth on fish species composition, unrelated to the possible influence of habitat structure. Whether the observed response of the fish assemblage to changes of habitat structure is a consequence of the lack of available space promoting competition for space as a potential driving force for structuring Mediterranean rocky reef fish assemblages, or whether it is due to a diversification of resources following the increase of habitat complexity, cannot be decided with an observational approach such as that used here. Two immediate implications of the present study are the adequacy of estimating habitat complexity by counting size‐classified rocky boulders rather than measuring rugosity perhaps a scale‐related effect, and the fact that habitat heterogeneity has no apparent influence on parameters of the fish assemblage; the latter may reflect an inadequate sampling size 50 times 5 m transects to quantify the effect of this aspect of habitat structure on species composition. Some implications for applied issues and novel methodological approaches to the problem are suggested.