|
December 2002 FLOW-SEDIMENT EFFECTS ON RIVERINE FISH
<br />4.5
<br />N 4.0
<br />3.5
<br />>. 3.0
<br />2.5
<br />co 2.0
<br />r 1.5
<br />C 1.0
<br />5 0.5
<br />0
<br />• Run 0 a
<br />0 Riffle 0
<br />°o U900
<br />b 0
<br />O •
<br />•O % '
<br />O• O 4 00 O •
<br />• O o 040
<br />0 O•
<br />•
<br />0
<br />9
<br />8
<br />0
<br />co
<br />E 7
<br />0
<br />6
<br />U)
<br />5
<br />C
<br />4
<br />3
<br />0.5 1.0 1.5 2.0 2.5 3.0 3.5
<br />In(Chlorophyll a)
<br />b
<br />
<br />CU
<br />E
<br />0
<br />s
<br />N
<br />U_
<br />N
<br />d
<br />c
<br />v3
<br />2
<br />1
<br />3
<br />0 0
<br />c
<br />c
<br />E -1
<br />0
<br />Y
<br />a -2
<br />c
<br />1731
<br />Y :J v i v ?
<br />In(Measured fish biomass)
<br />
<br />
<br /> J
<br />
<br />0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 3 4 5 6 7 8 9
<br />In(Chlorophyll a) In(Fish biomass)
<br />FIG. 5. Interrelationships among biological parameters. Solid fines represent regression lines in (a), (b), and (d), and
<br />1:1 line in (c). (a) Relationship between ln(chlorophyll a biomass) and ln(invertebrate biomass); (b) relationship between
<br />ln(chlorophyll a biomass) and ln(fish biomass); (c) predicted and measured ln(fish biomass [as catch rate]), regression based
<br />on ln(chlorophyll a biomass) and ln(invertebrate biomass); (d) relationship between ln(pikeminnow density) and ln(fish
<br />biomass [as catch rate]).
<br />brate biomass (Fig. 5c; F,,3, 17.75, P < 0.00001, r2
<br />= 0.54). These regressions improved substantially
<br />when two outlying points were removed (r2 = 0.66,
<br />0.72, and 0.74, respectively). The multiple regressions
<br />were conducted by sample reach using mean values
<br />over the three sample periods. The weighted means
<br />(weighted by the fractions of riffle and run area in each
<br />reach) of chlorophyll a and invertebrate biomass were
<br />regressed against fish numbers and biomass. Highly
<br />significant regressions also were found for numbers of
<br />all fish (F2,33 = 25.8, P < 0.00001, r2 =0.61) and for
<br />numbers of soft-rayed, native fish (F2,33 = 24.1, P <
<br />0.00001, r2 = 0.59) when regressed against chlorophyll
<br />a and invertebrate biomass. Both regressions improved
<br />when the two outlying points were removed (r2 = 0.81
<br />and 0.79, respectively). A less significant relationship
<br />(F233 = 5.6, P = 0.008, rz = 0.25) also was found
<br />between the numbers of forage-sized fish (100-300
<br />mm) and chlorophyll a and invertebrate biomass. Re-
<br />moving the same outlying points resulted in an im-
<br />proved relationship (r2 = 0.45).
<br />Body condition (K) was averaged by strata over all
<br />sample periods and compared with averages of chlo-
<br />rophyll a and invertebrate biomass. K. of both bluehead
<br />sucker and flannelmouth sucker was significantly cor-
<br />related with chlorophyll a (r2 = 0.77 and 0.70, re-
<br />spectively; P = 0.0015, n = 11) and invertebrate bio-
<br />mass (r2 = 0.58 and 0.67, respectively; P = 0.007, n
<br />= 11). Similar relationships were found for roundtail
<br />chub Kn, but these were not significant (chlorophyll a,
<br />P = 0.24, rz = 0.42; invertebrate biomass, P = 0.19,
<br />n = 5, r2 = 0.49), likely 'reflecting that only five strata
<br />could be included in the analysis (too few roundtail
<br />chub downstream of stratum 7).
<br />Fish biomass-adult pikeminnow densities relation-
<br />ship.-Significant relationships were found between
<br />densities (individuals/km) of adult pikeminnow and
<br />catch rates of other fish that might serve as forage.
<br />Catch rates of other fish within sample reaches were
<br />averaged by stratum and again averaged over the three
<br />sample periods. Significant regression relationships
<br />were found between density of adult pikeminnow and
<br />number of soft-rayed, native fish (F,,, = 27.6, P =
<br />0.001, rz = 0.80) and number of forage-sized (100-
<br />300 mm) soft-rayed native fish (F,,, = 12.8, P = 0.009,
<br />r2 = 0.65). Density of adult pikeminnow was also high-
<br />ly correlated with biomass of all fish (Fig. 5d; F,,, _
<br />22.4, P = 0.002, r2 = 0.76).
<br />To determine whether the availability of physical
<br />habitat might help explain the observed variation in
<br />pikeminnow densities, regressions were conducted us-
<br />ing the areas of riffles, runs, and combined other non-
<br />riffle/run habitats present in each stratum. Of these,
<br />only riffles had a significant relationship with densities
|