WSP12173

• File Number: 8240.300.02
• Description: San Juan River Recovery Implementation Program - Recovery Plans & Information
• Basin: San Juan/Dolores
• Water Division: 7
• Date: 2/1/2000
• Title: Drift of Fishes in the San Juan River - 1991-1997
• Water Supply Pro - Doc Type: Report/Study

Platania, Dudley, and Maruca. 2000. Drift of Fishes in the San Juan River 1991-1997.

FINAL

r-,
("~

Of these five predictive relationships, the most robust was the multiple logarithmic
relationship:

~
...J
--:;
1\:>

In(PF) = 1.345 + 1. 6611n(PT) 1. 12 Iln(Q)

with R' = 0.86 (n ~ II; F = 23.65; df= 2,8; P < 0.001). From the above empirical relationship, total
volume of water sampled for years/sites lacking direct measures were estimated (Table I).
Catch-per-unit-effort (CPUE) was then determined by dividing the number of drifting fish
caught by the volume of water sampled (100 m'). As there were no notable differences in catch or
catch rates among nets in any year at either site, those data were combined to yield one value for each
parameter for each set. For years and sites where CPUE data were available (Table I), catch rates for
morning and evening sets on the same day were evaluated using a paired comparisons t-test. Catch
rates were calcnlated for all species combined for this test. There was no statistically significant
difference in catch rate between morning and evening sets (t = 1.32, df= 107, P = 0.19) so the
number offish caught and volume of water sampled were combined to yield a single catch rate for
each species on a daily basis (daily CPUE).
Since daily flow could not be estimated (from models) for the five years/sites lacking flow
measurements, daily CPUE's are only available for those years/sites where flow was measured
directly. In 1993, daily CPUE values reported for the Four Comers site were from the upstream
location only; volume of water sampled was not measured at the downstream Four Comers location.
A mean catch rate was also calculated for each year/site by dividing the total catch for each species
by the total volume of water sampled (mean annual CPUE). Mean annual CPUE values were
determined for all years at both sites.

Relation of Drift CPUE to Flow Regime
We tested the hypothesis that characteristics of the flow regime were associated with the
variation observed in drift density in larval red shiner (Cyprine/la lutrensis), speckled dace
(Rhiniehthys oseulus), and channel catfish (Ietalums punctalUs). These taxa were selected becanse
they exhibited more variability in their drift thall the other species collected and the sampling period
encompassed the majority of their reproductive season. A set of descriptive hydro graph statistics was
compiled for each year at each site from USGS discharge data obtained from the Four Comers, New
Mexico and Bluff, Utah gauging stations. The statistics were calculated during the period of I
January - 31 August each year (1991-1997) and were comprised of the following seven flow
variables: maximum discharge (cubic feet per second, cfs), number of days discharge was greater
than 2,500, 5,000, 8,000, and 10,000 cfs, and number of days discharge was less than 1,000 and 500
cfs. The drift variables selected were mean annual CPUE and maximum daily CPUE. For each
species at each site, Pearson correlation coefficients (r) were calculated and evaluated between each
drift variable and each flow variable. Data from all seven years of the study were included in
analyses involving mean annual CPUE, while only tive of seven were inclnded in analyses involving
maximum daily CPUE. (Daily CPUE values were not available for 1991 and 1994 at Four Comers
or 1993 and 1994 at Mexican Hat). Analyses were performed at the 5% level of significance and
assumptions of normality were tested using normal plots and Lilliefors tests of residuals.
While the seven hydro graph variables were interrelated, they characterized different aspects
of the hydrograph (any of them may be important for reproduction of the three species examined).
Additionally, there were enough non-significant correlation coefficients among comparisolls of
hydro graph variables to warrant inclusion of the entire group. The variables selected were chosen to
determine if there was a threshold discharge or a level of flow above (or below) at which
reproduction was markedly improved. To aid in interpreting the results of the correlation analysis,

12