Laserfiche WebLink
hatching was completed. Surviving larvae were then counted and returned to their <br />incubation chambers. <br />Daily observations to determine age at swim bladder inflation and first feeding were <br />made with a dissecting microscope. Age at swim bladder inflation and first feeding in each <br />treatment was taken as the age at first observed occurrence in an individual larva in the <br />treatments. Newly hatched brine shrimp nauplii Anemia sue. were presented to larvae twice <br />daily ad libitum and presence or absence of food in the gut was documented 1 h later. <br />Larvae were maintained in experimental treatments for 7 d after hatching and then counted <br />and measured. Larvae were maintained in the fifth incubation chamber for all treatments <br />until first feeding was observed. All larvae were examined daily for obvious abnormalities <br />(i.e., lack of eyes, spinal deformities, swim bladder anomalies). <br />Percent hatch, percent survival of larvae to d 7 (of embryos that hatched), and percent of <br />larvae with abnormalities (of embryos that hatched), were calculated from the four <br />replicates in each treatment. Percentages were transformed (aresine and square-root) to <br />stabilize statistical variances and normalize the data. Main effects of temperature at three <br />levels (18, 22, 26°C) and constant versus fluctuating temperature regimes and their <br />interaction (temperature x regime) were analyzed with fixed-effects two-factor analysis of <br />variance (ANOVA). The Tukey test for multiple comparisons was used if the overall F-test <br />was significant (P < 0.05). Statistical analyses were conducted with the SAS statistical <br />package (SAS Institute 1988). The data for the 30°C constant temperature treatment were <br />excluded from statistical analyses because all larvae were deformed. <br />7 <br />