8172

• UCREFRP Catalog Number: 8172
• Author: Jacobi, G. Z., J. E. Sublette, S. J. Herrmann, D. E. Cowley and M. D. Hatch.
• Title: Final Report
• USFW Year: 1998.
• USFW - Doc Type: Investigations of an Index of Biotic Integrity in New Mexico.
• Copyright Material: NO

disagreement distances
using Ward's method.'
ethod.
Fish Species Associations
between species Species-Environment RehWons
Two cluster analyses were conducted to
illuminate associations of fish species.
First, cluster analysis on presence/absence
data was made using Jaccard dissimilarity
distances - between species -and -Ward's
method. This analysis provided
assemblages of fish species in the upper Rio
Grande.
Jaccard's similarity coefficient (Sneath and
Sokal, 1973; Bridge, 1993) minimizes
errors when a large number of negative
results occur, e.g., when two species are
found only at a few of the sites, a large
number of the pairwise data are negative
(both absences). Inclusion of the negatives
gives a similarity that is too large and
conversely, a distance that is too small.
The Jaccard dissimilarity distance was
obtained as the converse of the Jac card
similarity coefficient (Bridge, 1993). The
Jaccard dissimilarity provided better
resolution than the percent disagreement
distance for upper Rio Grande fish species
because a number of fish species were
found at only one or a few sites. A
FORTRAN Program (Appendix 3) was
written and used to obtain the Jacquard
dissimilarities.
Cluster analysis was performed using
relative abundance data. In this analysis,
Euclidean distances and Ward's method
were used. This analysis, in contrast to the
analysis of presence/absence data, does not
provide a true picture of the assemblages of
fish species. Rather, it provides groupings
of abundant and less abundant species.
Species-environment relations were
examined in several ways. First, for fish
data, canonical correspondence analysis
(CCA) (ter Braak, 1985, 1986, .1987a,b;
Taylor et al., 1993) and detrended CCA
(DCCA) (ter Braak and Barendregt, 1986;
ter Braak and Looman, 1986; ter Braak and
Prentice, 1988) was ..performed .. using
CANOCO (ter Bmak, 1987-1992).
Canonical correspondence analysis is an
iterative combination of ordination and
multiple regression that attempts to relate
species distribution and abundance to
environmental variables. Detrended CCA
takes the results from CCA and rescales the
variation in such a way that the biplot can
provide a finer scale view of
species-environment relations. The end
result of CCA and DCCA is a biplot
(Gabriel, 1971, 1981) which allows
simultaneous viewing of axes of
environmental variation and ordination of
species with respect to those axes of
environmental variation. Prior to analysis
by CCA and DCCA, fish species relative
abundance data were ranked within sites by
assigning a value of one to the least
abundant species up to a value equal to the
number of species found at the site for the
most abundant species. Twenty-one of the
22 environmental variables were included in
CCA and DCCA; the categorical variable
representing land use designation was
omitted.
Second, the relationship between elevation
and the number of fish species was
investigated with linear regression.
Comparable regressions were performed for
elevation and the number of chironomid
species elevation and the number of benthic
macroinvertebrate taxa. Third, the
congruence of fish faunal regions,

7