Factors affecting site productivity of a Costa Rican secondary rain forestin relation to Vochysia ferruginea, a commercially valuable canopy tree species

We estimated site productivity and its variation in a 28-year-old secondary rain forest in Florencia, San Carlos, Costa Rica, in relation to the domin ant tree species, the commercially important Vochysia ferruginea. Thirty-si x sample plots (20x20 m(2)) were installed in such a way as to cover the wh ole range of substrate variation at the site. The stand was assumed to be e ven-aged and of dominant height, estimated as the mean total height of the four tallest Vochysia trees in each plot, used as a measure of site product ivity. Intra-plot coefficients of variation of total height of these four t rees were <20%, indicating homogeneous conditions within plots in terms of site productivity and phytosanitary state. Of the dominant trees, however 8 % were found to have partially shaded crowns, which could lead to a subesti mation of dominant height. For this reason, estimations of dominant height using two of the four trees per plot (these having fully illuminated crowns ) were also developed. Dominant height of Vochysia was significantly correl ated with mean plot canopy height and basal area of all species, and with i ts own basal area. Soil variables significantly correlated with dominant he ight were, at soil depth 0-12 cm, clay(%), organic matter(%), P, Fe (all ne gatively correlated) and Cu (positively correlated). Adjusted multiple regr ession models of dominant height on soil variables were fitted by stepwise regression (Forward method). For four trees per plot, the best fit was obta ined using P, Mn, percent organic matter and percent clay as independent va riables (R-2=79.5%, S.D. 1.8 m). For two trees with fully illuminated crown s, best fit was obtained from a model using the same variables except P. R- 2 was lower for this latter model (74.4%), but the S.D. was also lower (1.2 m). Dominant height appears to be a potentially useful indicator of site p roductivity in neotropical secondary forests of similar ages, although it a ppears important, in mixed forests, to use crown variables to aid in the de finition of dominant trees. There may be an important effect of site qualit y on production of timber in such forests, and such effects should be taken into account in the development of models of stand dynamics, growth and yi eld. (C) 1999 Elsevier Science B.V. All rights reserved.