Maximum likelihood estimation of ‘unobservable’ growth and development rates using scale data: Application to carp aquaculture in India

The present study on the carp species Labeo rohita in Karnataka State, south India, shows that it is possible to obtain quantitative estimates of pre-reproductive growth rate, age at first spawning and present age by appropriate interpretation of scale features such as the number of growth checks, circuli counts and spacing between circuli. The unobserved life history variables and their causal relationships with the observed scale features are specified a priori and estimated in a confirmatory factor analysis model. The estimates of pre-reproductive growth rate, age at first spawning and present age can be used to guide selection programmes and to compare populations in different aquaculture regimes. As a preliminary indication of the usefulness of this technique in studying domestication, it is noted that the fish chosen as breeders are not the fastest growing fish in the particular fish farms studied.