8142 - Laserfiche WebLink

Home Browse Search

I I I I I I I I I I I I I I I I I I I f.....l.........'........ .. ,- t ,. ~,. -- 41 used to test for homogeneity of slopes. Contrasts of least-squares means were used to compare treatment effects where interaction terms were not significant. When the interaction term was significant, least-square means cannot be used since the covariate behaves differently between treatments. In such a case a multiple comparisons procedure was used to compare means by date. Least squares regression and ANCOV A also were used to examine relations among larval size, larval age, and otolith size. m. Results Both lapilli and sagittae are present at hatching. Asterisci did not appear until 14 to 15 d after hatching at any feeding level and were markedly smaller in size at time of appearance in low-feed and starved larvae. Diameters were not measured, but asterisci were similar to lapilli in size about 2 weeks after appearing in Ad libitum larvae. Both sagittae and lapilli were similar in size (30 to 40 .um; Figure 7a) at hatching. Lapilli remained almost round and grew slowly while sagittae elongated and increased in size at a faster rate, their length almost twice exceeding lapillus diameter 3 weeks after hatching (Figure 7b). The rapid increase in size made sagittal otoliths the most desirable for increment counting. Increments counts on sagittae and lapilli, however, were consistently similar when lapilli were needed in a few cases where saggital otoliths were not clear. Intercepts of least-squares regressions for number of increments against known age were not significantly different from zero for any treatment. The first increment thus forms the day of hatching (Table 3, Figures 8-10). Razorback sucker larvae do not begin exogenous feeding until 5 to 7 d of age (Papoulias and Minckley 1990). Although not '-