9572 - Laserfiche WebLink

Home Browse Search

and biological factors at intra-annual and annual time scales. Distributions of hatching dates in 1991 and 1992 indicated that larvae in cohorts that hatched early survived poorly to fall. Growth rate comparisons suggested that the few early hatched fish that survived were afast-growing subset of the fish present in the same cohort in summer. I attributed this to size-selective predation mortality by introduced fishes. Larvae most likely to survive to fall were hatched late and grew relatively slowly. Slow growth rates and high survival that are incongruous with patterns for early-hatched larvae were probably due to environmental factors and to natural mortality of large predaceous red shiners Cyprinella lutrensis in mid- to late-summer. An independent individual-based computer simulation model which had agape-limited red shiners as predators and Colorado squawfish larvae as prey produced similar size-selective patterns. Results of model simulations also suggested that fish with moderate growth rates were more than twice as likely to survive as fish with low-growth rates, underscoring the biological significance of growth. Alternative hypotheses to explain recruitment patterns such as starvation and competition were not proximate explanations for the size-selective patterns observed. A physical process, stochastic flooding reduced growth rates of Colorado squawfish and combined with size-selective predation to cause very low recruitment in the lower Green River in 1992. Otherwise, recruitment was unaffected by discharge and temperature regimes in the summers of 1991 and 1992. Linear plateau regression models predicted no negative effect of mean July-August discharge level on annual abundance of Colorado squawfish juveniles except at relatively high discharge. Low abundance of juvenile Colorado squawfish in 1991 'and 1992 when size-selective patterns were evident suggested that predation may regulate recruitment in some years. vi