| Summary: | Thesis (Ph.D.)--Memorial University of Newfoundland, 2001. Biology Includes bibliographical references The early life history (ELH) traits of fish relate directly to critical aspects of survival, development and growth that are reflected in aquaculture recruitment and the quality of progeny. The ELH traits of the eggs and larvae of yellowtail flounder, Pleuronectes ferrugineus (Storer), were investigated to determine which ELH traits relate to commonly used egg and larval quality criteria and how these traits could be used to predict recruitment in mass rearing aquaculture. -- Chapter 1 shows that inducing females to spawn with GnRHa produces higher mortality during embryogenesis compared with spontaneously ovulating females. However, the variability in mortality remains relatively constant. In addition, temperature affects time to hatch, but not egg mortality, which was determined through comparing the dry mass of filtrate to the dry mass of known volumes of eggs. -- Chapter 2 shows that egg size was more variable than carbon and nitrogen content and therefore, less effective as an egg quality criterion. It is hypothesised that variation in egg size is greatly affected by hydration processes that are under environmental influence. -- Chapter 3 shows that abnormal cleavage patterns during the 4-8 cell stage of embryogenesis significantly decreased hatching success, but did not produce any curved larvae. This suggests that hatching success may be underestimated and that visual egg quality determinations based on viability criteria may be less effective than previously thought. -- Chapter 4 shows that multivariate analyses of five larval and three egg ELH traits had significantly higher predictive capabilities than any single trait. However, because of its relationship to other ELH traits, larval standard length was found to be the best single, quantitative metric of egg batch quality. Nonetheless, a trained assessment of larval quality, based on visual observations, is nearly as effective. Males were typically found to contribute less than 5% to the variability in ELH traits (range 0 - 15%) while females generally contributed over 30% (range 6% - 83%). However, relative contributions varied according to whether the analysis was performed on individual or mean levels of data and (or) at what time larval measurements were made. This result suggests that, with regard to ultimate larval success, the maternal contribution decreases and the paternal contribution may increase through ontogeny. -- Comparative statistical analyses shows that multivariate treatment is superior to univariate treatment of data at the same level. Furthermore, the method of aggregating data. e.g. mean-level or population-level groupings, may significantly change results. This scale result is due in part to significant changes in sample sizes as data are aggregated. Model I and Model II (reduced major axis) regressions are compared and were found to be different when ordinate data are highly variable at a given abscissa value.
|
|---|