Abstract
A population of Tipula sacra was studied intensively from 1969 to 1974 in a series of abandoned beaver ponds. Beavers recolonized the ponds from 1974 to 1976 and again in 1978. The basic life‐cycle is of 2 years duration, but cohort‐splitting occurs in the autumn as a result of some larvae growing faster and completing the life‐cycle in 1 year. Some periods are more favourable for growth than others and the size of the fast‐growing cohort varies from year to year. This insight allows some anomalies in the estimates of population size to be explained and life budgets have been produced for three year‐classes. The short‐fall in realization of egg potential is massive, apparently due to heavy female mortality by predation and the prevention of oviposition by inclement weather. Mortality rate is relatively low and constant from stage II larvae until pupation, but increases to adult emergence. There is a correlation between larval density and mortality rate. Certain aspects of the life‐history (e.g. high fecundity, variability in growth rate) appear to fit the species to life in a temporally unstable habitat.