Analyzing the impact of neighbors’ identity and growth rate on individual tree mortality

Growth, mortality, and regeneration determine the development and yield of forest stands, with mortality representing the largest knowledge gap and challenge from measurement to modeling. Available models estimate the competition-related mortality of trees mostly via competition indices, which quantify the individual spatial constriction at the beginning of the respective growth period. Here we present a new approach that uses the species ensemble and relative growth rate of neighbors to estimate mortality in mixed stands. In contrast to the individual spatial constriction quantified by the competition indices, the relative growth rate better represents how the respective tree and its neighbors acquire the resources that are not sufficient for survival of all, incorporating a temporal component to the competition. The ability of traditional competition indices and the new index to predict mortality was evaluated over data from 69 permanent plots located in Germany and Spain, involving such species as Norway spruce ( Picea abies (L.) H. Karst. ), European beech ( Fagus sylvatica L. ), sessile oak ( Quercus petraea (Matt.) Liebl. ), stone pine ( Pinus pinea L ), resin pine ( Pinus pinaster Ait ), holm oak ( Quercus ilex L ). and junipers ( Juniperus thurifera L or J. oxycedrus L). Repeated measurements (two to five per plot) at varying intervals between 3 and 11 years (mean 6.3 years) resulted in a total of 197 plot x inventory observations. Our results evidence that joint consideration of species identity, traditional competition indices and relative growth rate results in more accurate predictions of individual tree mortality. However, the sign of the effect depends on the characteristics of the species involved. Incorporating observed or predicted past growth rates into forest models and simulators may improve the design of management practices and development of silvicultural prescriptions, especially for mixed stands. • Neighbors’ specific identity influences probability for individual tree mortality. • The sign of this effect is not constant, depending on species and types of mixture. • The past relative growth rate of a tree is a main indicator of future mortality. • Incorporating new growth rate index improves accuracy of mortality models.