8161 - Laserfiche WebLink

Home Browse Search

, - E E ......... 1000 900 800 700 600 500 400 300 200 1100 1000 900 800 700 600 500 400 300 200 o .c: ..... 0) c: Q) CO ..... ~ 693 GROWTH AND SURVIVAL OF COLORADO SQUAWFISH Il!I!!fllllmlffi~lII~ IIIl 1l11111!lll 11111 A lllllllllll!!! III ~ Illlllll~ I Jlllllll!111 l B , 5 10 15 20 25 30 35 40 45 50 55 Age (years) FIGURE 4.-Lengths from Monte Carlo simulations using (A) size-specific growth rates as shown in Tables I and 2 and (B) constant growth rate for fish 550 mm TL and longer (see Table 2). Bars represent lengths of 95% of fish of each age-group. Differences in size distribution were, however, present between fish (~550 mm) caught in tram- mel nets and by electrofishing. The only data set with comparable periods and reach of capture for both methods was in 1994 in a section (rk 246- 275) of the upper reach. Captures by electrofishing (N = 13) included significantly (P = 0.016) more large fish than by trammel netting (N = 21), a bias consistent with observations of others (see Reyn- olds 1983). Although we could not test if distri- butions of fish caught by trammel nets were rep- resentative of the population, we assumed they were because fish were confined and actively trapped, thereby reducing or eliminating possibil- ities for size selectivity (i.e., differential trap shy- ness, escapement ability, or susceptibility to elec- tric fields). Survival estimates were therefore cal- culated for fish 550 mm and longer from trammel- net data only. Estimates.-For 1991-1994 data combined, suitable survival rates varied from 0.83 to 0.87 (P 50 40 A 30 20 ",,11111111111 (i) 10 L- as Q) ~ Q) 0) <( o 50 B 40 30 20 ",,11111111111 10 o 400 500 600 700 800 900 1000 Total length (mm) FIGURE 5.-Simulated range of estimated ages for Colorado squawfish of given length. Simulations used (A) size-specific growth rates as shown in Table 2 or (B) an assumed constant growth rate for fish 550 mm TL and longer. Bars in both graphs represent ages of 95% of fish of each length group. < 0.05). Only a narrow range of estimates was not significantly different from the measured distri- bution, even with P < 0.001 (Figure 6). The best fit for the measured distribution was for a survival rate of 0.85 (Figure 7). Similar but broader ranges of survival were estimated for individual years, largely a reslllt of smaller sample sizes. All years combined produced a range of suitable survival rates that narrowed to 0.82-0.87 (Table 3). For a growing or declining population with a stable age distribution, the effect on survival rate would be a change of = 1.0% for each 1.0% in population increase or decrease (Table 4). Estimated survival rates were the same when growth rates for fish 550 mm and longer used in calculation of stable length distributions varied by fish size (as in Table 2) or when assumed to be constant (as in Figure 4b). Twenty simulations of stable age distributions that used different random-number sequences pro- duced identical suitable survival rates and nearly identical K-S d-value statistics.