How to write better thesis

Abstract

We assessed the effects of controlled selective logging on stand-level acorn production and early tree recruitment in a tropical montane oak-bamboo forest in Costa Rica (2600-2800 m elevation). For this, we took advantage of a mast fruiting event during year 2000 in a population (N = 870 individuals >10 cm dbh) of one of the dominant canopy tree species, Quercus costaricensis. The forest was selectively logged for timber between 1991 and 1992 under two harvesting intensities (20 and 30% basal area removal of trees >10 cm, dbh). We found that more sound acorns fell per unit area in plots subjected to the highest harvesting intensity than in adjacent, unlogged plots (180,000 and 100,000 total acorns/ha, respectively); acorn production in the lightest harvesting intensity plots was marginally higher than in unlogged plots. We suggest that this result is largely due to the presence of small (10-30 cm dbh) trees that commenced their reproduction only in the logged plots. Our results also suggest that logging may have enhanced stand-level acorn production in Q. costaricensis by augmenting crown illumination levels in retained individuals: within plot types, the crowns of fruiting individuals were better illuminated than those not in fruit; among plot types, and for a given size class, trees in the unlogged plots had their crowns less well-illuminated than those in the logged plots. Because more sound acorns fell per unit area in the more intensively-logged plots, vertebrate predators apparently became satiated and left proportionally more acorns to germinate on the forest floor with respect to unlogged plots. Seedling densities from the 2000 acorn cohort mirrored the observed differences in acorn production and acorn survival between the more intensively-logged and unlogged plots. Acorn germination and seedling establishment success, however, did not vary among these plot types. Our results point out to the potentially beneficial role of controlled logging on stand regeneration in this montane forest, being detectable even one decade after its implementation. (C) 2002 Elsevier Science B.V. All rights reserved