Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Discharge variability and the development of predictive models relating stream fish assemblage structure to habitat in northeastern Australia

Discharge variability and the development of predictive models relating stream fish assemblage... Abstract – We developed classification/multiple discriminant analysis models to predict fish assemblage structure and tested whether the predictive power of these models varied with discharge variability. Models developed for assemblages characterized by the density of component species for two rivers with low discharge variability had better predictive power than did models developed for two rivers of higher variability. Similar distinction between rivers of differing flow variability was not evident for models based on assemblages characterized by the presence or absence of component species. Factors such as the within‐river level of beta diversity, location of study sites relative to the river mouth and the degree of covariation in species' occurrence appeared important determinants of predictive power in these models. Randomization tests (Mantel tests) were used to determine the degree of association between site by site association matrices generated for fish assemblage structure (both density and presence/absence) and habitat structure (catchment, physical, microhabitat or a combination). This approach revealed that in most cases, catchment‐related variables explained almost as much of the variation in assemblage structure as variables related to in‐stream habitat structure and that greater association was detected for comparisons based on presence/absence rather than density data. The addition of in‐stream habitat variables to catchment‐related variables usually resulted in explaining the greatest amount of variation. These data suggest that most of the structure observed in the fish assemblages of the study rivers was a result of the effect of regional or catchment factors in determining which species were present at an individual site and that local factors were then important in determining the abundance of the component species. It is at this level that the effects of regional differences in discharge variability were expressed. Although significantly different from random for all comparisons, Mantel's tests revealed that a substantial amount of variation in the fish assemblage data sets could not be explained by the abiotic (habitat) data sets. It is suggested that the assemblages in question did not represent unit discrete assemblages but were composed of species varying along individual environmental gradients. Predictive models may be better achieved by modelling the distribution and abundance of individual species rather than assemblages. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ecology of Freshwater Fish Wiley

Discharge variability and the development of predictive models relating stream fish assemblage structure to habitat in northeastern Australia

Download PDF
21 pages

Loading next page...
 
/lp/wiley/discharge-variability-and-the-development-of-predictive-models-9WANvYLjMa

References (43)

Publisher
Wiley
Copyright
Copyright © 2000 Wiley Subscription Services, Inc., A Wiley Company
ISSN
0906-6691
eISSN
1600-0633
DOI
10.1034/j.1600-0633.2000.90105.x
Publisher site
See Article on Publisher Site

Abstract

Abstract – We developed classification/multiple discriminant analysis models to predict fish assemblage structure and tested whether the predictive power of these models varied with discharge variability. Models developed for assemblages characterized by the density of component species for two rivers with low discharge variability had better predictive power than did models developed for two rivers of higher variability. Similar distinction between rivers of differing flow variability was not evident for models based on assemblages characterized by the presence or absence of component species. Factors such as the within‐river level of beta diversity, location of study sites relative to the river mouth and the degree of covariation in species' occurrence appeared important determinants of predictive power in these models. Randomization tests (Mantel tests) were used to determine the degree of association between site by site association matrices generated for fish assemblage structure (both density and presence/absence) and habitat structure (catchment, physical, microhabitat or a combination). This approach revealed that in most cases, catchment‐related variables explained almost as much of the variation in assemblage structure as variables related to in‐stream habitat structure and that greater association was detected for comparisons based on presence/absence rather than density data. The addition of in‐stream habitat variables to catchment‐related variables usually resulted in explaining the greatest amount of variation. These data suggest that most of the structure observed in the fish assemblages of the study rivers was a result of the effect of regional or catchment factors in determining which species were present at an individual site and that local factors were then important in determining the abundance of the component species. It is at this level that the effects of regional differences in discharge variability were expressed. Although significantly different from random for all comparisons, Mantel's tests revealed that a substantial amount of variation in the fish assemblage data sets could not be explained by the abiotic (habitat) data sets. It is suggested that the assemblages in question did not represent unit discrete assemblages but were composed of species varying along individual environmental gradients. Predictive models may be better achieved by modelling the distribution and abundance of individual species rather than assemblages.

Journal

Ecology of Freshwater Fish – Wiley

Published: Jun 1, 2000

There are no references for this article.