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

an orthogonal rotation2 using the varimaX
criterion (Harman, 1§76, pp, 290-299) to
improve their interpretability. Standardized
scores (mean=0, variance=l) for each site
were obtained for each of the six
varimax-rotated factors. Thus, the original
22 environmental variables were reduced to
fewer multivariate factors having different
environmental variables determining their
extent- and- direction- in- multivariate°-space:
Because ordination is a qualitative statistical
procedure, i.e., there are no formal
hypothesis tests performed, transformation
of data is unnecessary, other than use of the
correlation matrix as described above, and
no transformations were made prior to
analysis.
Using scores on the rotated environmental
factors as data, cluster analysis was
performed to ehicidate affinities of different
sites in the upper Rio Grande drainage.
Euclidean distances between sites were
obtained and a cluster "tree" was
constructed using Ward's clustering
algorithm (Ward, 1963). All cluster
analyses were performed using routines in
STATISTICA for Windows (StatSoft,
1994b). The utility of cluster analysis to
define groupings of sites with similar
characteristics and of Ward's method has
been demonstrated (e.g., Cowley, 1979;
Hawkes et al., 1986; Corkum, 1989;
Conquest et al., 1994). Following
elucidation of clusters (aquatic ecoregions)
of upper Rio Grande sites based on the
multivariate environmental factors,
discriminant analysis (SAS, Institute, Inc.,
1989a) was performed to obtain a post hoc
assessment of which multivariate
environmental factors were most important
in separating aquatic ecoregions (Huang and
Ferng, 1990; Conquest et al., 1994).
Three-dimensional plots of site scores on
the multivariate environmental factors were
obtained using STATISTICA (StatSoft,
Inc., 1994a). These plots provided a means
for determining which sites were extreme
for the multivariate environmental factors.
"They -also-provided -a' post hoc' explanation
of why certain environmental variables were
important sources of variation in the
principal components analysis.
Utility of Aquatic Ecoregions as Opposed
to Hydrologic Units
An important consideration is whether
aquatic ecoregions capture relevant-
quantities of variation in chemical and/or
physical properties of the environment. If
aquatic ecoregions fail to stratify the
landscape, some other scheme for grouping
sites might be preferred. To answer this
question, a series of fixed effects analyses
of variance (ANOVA; GLM Procedure,
SAS Institute, Inc., 1989b) were carried out
to elucidate differences between aquatic
ecoregions for the 22 chemical, physical,
and climatic environmental parameters. A
companion set of ANOVAs were performed
to assess differences between Hydrologic
Units in the upper Rio Grande drainage
(USGS, 1976).
Site Similarity Based on Invertebrate Taxa
(Chironomid and Bendde
Macroinvertebrate Faunal Regions)
Cluster analysis was performed to assess
similarity of upper Rio Grande sites based
on the presence/absence of chironomid
2 After rotation, the principal components are referred to as factors.

5