8173

• UCREFRP Catalog Number: 8173
• Author: Gido, K. B. and D. L. Propst.
• Title: Habitat Use and Association of Native and Nonnative Fish in the San Juan River, New Mexico and Utah.
• USFW Year: n.d.
• USFW - Doc Type: \
• Copyright Material: NO

occurred. A chi-square test statistic was used to determine if
this proportion was greater or less than expected. The expected
frequency was simply the proportion of the total seine hauls in

which the nonnative species/age-class was captured. For example,

if juvenile Q. lutrensis was present in 50% of the total seine

hauls, it would be expected to occur in 50% of the seine hauls

where native ~. latipinnis occurred. A Bonferroni corrected

alpha level was used because of the large number of comparisons

(n=99) .

Spearman's rank correlations were calculated to determine if

the abundance of species pairs was positively or negatively

correlated across habitats. Because a large number of habitats

had only one of the species/age-class present we only considered

those habitats where both were present (Ludwig and Renyolds

1988). Thus, for many of the less-common species there were low

sample sizes (<5 comparisons); these were omitted from the

analysis.

Variation in habitat availability and habitat sampled among

seasons and sample sites was examined using a Kruskal-Wallis

test. If significant differences were found, pair-wise

comparisons by sites were tested using Wilcoxon sign ranks test

and P-value was adjusted using a Bonferroni correction (Sokal and

Rohlf 1995). Seasons were divided into the four calendar seasons

of the year. Additionally, differences in habitat variables were

tested between habitat available and that sampled using Wilcoxon

sign ranks test. All calculations were computed using SPSS v7.0

9