On the calculation of birth rates and death rates in fluctuating populations with continuous recruitment.

A two-instar and a three-instar model for the calculation of birth and death rates is proposed which both use only a minimum of assumptions. It is shown that the proposed approach is a generalization of the methods of Edmondson (1960, 1968), Caswell (1972), and Paloheimo (1974). The validity of the two methods is compared with two-instar and three-instar models according to Edmondson, Caswell and Paloheimo. Since it is impossible to use field data for the comparison, a set of 27 computer simulations was run. From these simulations the "true" and estimated values were obtained. The comparisons showed that it is better to use the new "two-instar" model in an environment with no instarspecific mortality. But it is better to use "three-instar" models when strong instarspecific selection prevails. Among the "three-instar" models the best results are obtained by the method of Paloheimo when predicting birth rates of young and of adults. It proved that the bias produced by the several methods is mainly influenced by the amount of deviation from the situation of steady state.