Stocking response of Eucalyptus growth depends on site water deficit across a 2100-km gradient in Brazil

Abstract

Stocking in forests strongly influence stand growth by influencing resource capture, tree survival, and competition between trees and understory vegetation. Many studies have examined these patterns for specific locations and conditions, though with limited populations of inference. The influence of stocking was examined in the TECHS experimental platform, at 16 sites, with 18 clones, across a 2100 km geographic gradient. Stocking was examined by varying spacing within columns, with constant (3-m) spacing between rows. We examined how wood production and quality related to genotypes and environmental conditions, with stocking ranging from 480 to 3300 trees ha−1. Mean annual increment (Mg ha−1 yr−1) for the entire rotation increased with stocking, asymptoting above 1500 trees ha−1. Water-stressed sites had much lower responses to stocking levels than wetter sites. The three driest sites showed only a 25% increase in MAI (7 to 9 Mg ha−1 yr−1) with stocking increases from 480 to 3300 trees ha−1, whereas the 3 most productive sites increased by 50% (20 Mg ha−1 yr−1 to 30 Mg ha−1 yr−1). A similar result occurred for the clones, where the response to the change in stocking was higher in the more productive clones. From a qualitative point of view, tree breakage declined as stocking increased, going from 0% of broken trees in the highest stocking to a maximum of 3% with the stocking of around 480 trees ha−1. Injury to trees increased with stocking, going from about 4 to 9%. Mortality showed a similar pattern, increasing from 8 to 14% with increasing stocking. The percentage of intact (surviving, undamaged) trees at the lowest stocking was 83%, reaching a peak of 94% between 700 and 1100 trees ha−1, declining again to about 85% at the densest stocking. Survival rates declined when stocking exceeded 2900 trees ha−1 on wetter sites, compared with just 900 trees ha−1 on the driest sites. Overall, an increase in stocking from 480 trees ha−1 to 3300 trees ha−1 led to 50% increase in growth. To put this in perspective, the poorest 25% of sites averaged 59% less growth than the top 25%. The poorest 25% of genotypes grew 47% less than the best 25%. The interactions of effects, such as stocking by site, had effects that were of the same magnitude as these main effects. Management decisions about optimal stocking may be substantially improved by considering the main effects and especially their interactions.