Statistical evaluation of methodologies for genetic strain evaluation in small to medium sized experimental facilities
Date
1994
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Dalhousie University, Halifax, NS, CA
Abstract
A major goal of aquaculture genetics is to provide improved fish that will eventually benefit not only the private industry but also the small fish farmers. It is essential to develop strain testing methods that have sufficient statistical power to detect differences in strains. True differences that are not detected can mean millions of dollars of opportunity loss for the aquaculture industry. In this study, several strain testing procedures were evaluated in the context of their statistical power to detect economically important strain differences.
Two strains of Nile tilapia (Oreochromis niloticus) were reared under various experimental comparison procedures. The strains were size-matched (graded) and grown separately, together, and grown separately but with an internal reference fish (red tilapia) in each replicate cage or tank. Fish of mixed-sizes and ages (ungraded) were simultaneously grown separately, together, and separately but with an internal reference fish. On-farm, strain testing was also done in four rice-fish farms.
The power to detect true differences was low when strains were grown together and were of mixed sizes. Use of an internal reference fish was inadequate to remove environmental sources of variation when fish were of mixed sizes. Initial size differences resulted in apparent growth depensation under experimental conditions and growth compensation in rice-fish farms. The size-dependency of growth rates in fish may have important implications in strain testing and selection programmes in aquaculture. Large differences between two growth means are not always due to genetic variance, which is what is sought for strain testing or selection. Environmental variance (initial size differences) plays a major role in observed differences in growth rates of fish.
Size-grading or having almost the same common starting size among genotypes before strain testing may help minimize environmentally induced variation like initial size differences. This procedure is more powerful than mixed size rearing at detecting true differences.
Description
item.page.type
Thesis
item.page.format
Text
Keywords
AQUATIC FAUNA, MARINE ENVIRONMENT, GENETIC ENGINEERING, GENETIC IMPROVEMENT, ON-FARM RESEARCH, STATISTICAL DATA, ECONOMIC IMPLICATIONS, TESTING