M. W. 1993. Density-related mortality in Cameraria hamadryadella (Lepidoptera: Gracillariidae) at epidemic and endemic densities. -Oikos 66: 515-525.To determine the role of natural enemies and host-plant quality in the population dynamics of Cameraria hamadryadella. we examined survival patterns and causes of mortality of a single population of C. hamadryadella during the epidemic vcars of 1982-1984 and the endemic years of 198,S-1990. C hamadryadella is a leat-mining microlepidopteran whose primary host is white oak, Quercus alba. Populations of C. hamadryadella range in density over four orders of magnitude from endemic densities of 0.004 individuals per leaf to outbreak densities of 22.79 per leaf.In an epidemic year (1984). 53.5% of the individuals studied died because of "other causes" which we contend is a combination of host resistance and intraspecific competition. Parasitism and predatit)n accounted for 21.2"o and 24.7"o ot' total mortality, respectively, and only 0.6"o of the cohort survived. Survival was signifieantly greater in the endemic year (1990). and death by "other causes" was also 29.2% lower in the endemic year. The reduction in mortality due to other causes in the endemic year is due to lower rates of starvation, and to reduced intraspecific competion. However, mortality from starvation, parasitism, and predation did not differ between epidemic and endemic years indicating a lack of temporal densitydependence. During the epidemic, only mortality caused by host and intraspecific competition was spatially positively density-dependent on the scale of individual leaves. The overall impact of parasitoids was density-independent, and the effect of predators was signifieantly inversely density-dependent on the leaf scale. Our results suggest that natural enemies are not responsible for the decline in abundance of C. hamadryadella observed between 1984 and 1990. However, starvation possibly arising from host resistance and. secondarily, intraspecific competition could account for the observed population crash. Exactly what ecological processes cause populations of the forces that operate to "regulate" populations, and herbivorous animals to grow so explosively as to reach hence give rise to outbreaks and populations declines, epidemic proportions, and what forces act subsequently into those that arise either higher or lower in the food to cause such populations to crash to endemic abun-chain, respectively. Of course natural enemies and hostdance levels? Two major classes of hypotheses have quality could affect insect populations dynamics interevolved to explain the dynamical behavior of such "out-actively (Hilborn and Stearns 1982). However, we must break" populations: hypotheses concerning natural ene-draw distinctions between these two classes of hypothmies. and hypotheses concerning host-plant quality eses in order to formulate critical tests and develop (Barbosa and Shultz 1987), These hypotheses separate mechanistic explanations for population outbreaks.