7195 - Laserfiche WebLink

Home Browse Search

74 NESLER ET AL. reflect the temporal variation in number of drifting larvae. In contrast, seine sampling throughout a 75-km river reach over the 5-6 d required to float the reach by raft could easily miss this discrete group. In 1980-1982, seine sampling may have missed the concentration of larvae in the river. In 1983-1984, 'the seining catch per effort for larval Colorado squawfish was greater than in 1980- 1982 (Haynes et al. 1985), and some portion of the peak production of larvae was sampled. How- ever, it remains uncertain whether one-time seine sampling throughout a river reach can adequately quahtify the transient pulse of drifting Colorado squawfish larvae in the Yampa River. It is pre- sumed that sampling of drifting larvae and estima- tion of spawning dates were more representative in 1983-1986, when a combination of sampling techniques or drift-netting alone were used, than in 1980-1982, when seine sampling alone was used. Thus, the low catches of Colorado squaw- fish larvae in 1980-1982 might not indicate poor reproductive success. In using the predictive age-growth equations, we assumed that growth and incubation time for wild Colorado squawfish stocks in the Yampa River were similar to those of hatchery stocks raised in Willow Beach National Fish Hatchery, Arizona. It is probable that there is greater year- to-year variability in both these variables among wild than among hatchery progeny. Another as- sumption represented by the polynomial age- growth equation (Figure 2) is that all larvae are approximately 6.7 mm total length at hatching. Size at hatching within a year class of wild prog- eny (as well as within hatchery stocks) is certainly variable. Hamman (1981) reported that hatching sizes of Colorado squawfish larvae varied from 6.0 to 7.5 mm total length; differences between the smallest and largest larvae were 0.5 to 1.0 mm for a given batch of eggs. Size differences of this range result in 3- to 5.5-d differences in posthatch- ing age estimated with the polynomial equation. As pointed out earlier, the incubation time's re- ported by Hamman (1981) show a 2.5-d range, or a 1-1.5-d variation about the mean incubation time of 5 d used here. These factors may apply to a few, some, or all of the larvae in a particular year. On the positive side, mean water temperatures at the Yampa Canyon spawning ground ranged from 22 to 23.5°C in 1981-1984 (Tyus et al. 1987). This compares favorably to the 20-24°C water temperatures present at the hatchery during the Hamman (1981) study upon which the larval age- growth equations were based. Also, Hamman (1981) reported that cold river water was added periodically to the recirculated-water raceways holding the Colorado squawfish larvae, lowering the water temperature from 23-24°C to 17-18°C. Whether intentional or not, this action was prob- ably aclose simulation of water temperature fluctuations that occur in the Yampa River. Be- cause growth data used in the predictive equa- tions were averaged over all rearing conditions reported in Hamman (1981), the age-growth equa- tions probably provide a fair approximation of growth conditions for wild Colorado squawfish larvae. Colorado squawfish collected by drift net in Yampa Canyon were predominately protolarvae (7.3-9.0 mm total length) whose nutritional needs were still supplied from a yolk sac. Thus, varying food resources in the wild for drifting Colorado squawfish larvae would be of limited concern with regard to the validity of the age-growth equations. However, the effect of a variable environment upon the growth of actively feeding mesolarval Colorado squawfish presents another unknown variable. For example, the comparison of the 1984 spawning peak with the flow regime suggested that the flow spike occurred 1 d later than the start of peak spawning activity. This incongruity in alignment may be a function of the predominant influence of seine-collected larvae, which pro- vided the basis of the 1984 spawning peak on July 8-12. Colorado squawfish larvae collected by seines were typically larger than those captured by drift nets in 1983 and 1984 (Table 1). It is possible that some of the seine-collected larvae were actively selecting habitats with low-velocity water that was warmer than in the main channel, and grew faster than would be predicted. Earlier spawning dates would be calculated from these larger larvae than actually occurred. Valdez et al. (1985) showed that, for native cyprinids in the Colorado River, smaller larvae usually appeared in drift nets and larger larvae were found along shorelines sampled by seines. Extrapolation of flow patterns from USGS gages distant from the Colorado squawfish spawn- ing area suggested another source of variation. Comparison of the Maybell and Deerlodge hydro- graphs demonstrated very similar patterns, but differences in the timing of peak spike flows between the two gage sites suggested that a spike peak arrives at primary spawning area in Yampa Canyon 1 d after it passes the Deerlodge gage. Of greater importance to the assumption of similar