Basal area growth rates of five major species in a Pinus-Cunninghamia forest in eastern China affected by asymmetric competition and spatial autocorrelation

We analyzed basal area (BA) growth using growth data obtained from permanent plots over 4 years for five major tree species in Anhui Province, eastern China. The studied species were dominant conifers (Pinus massoniana and Cunninghamia lanceolata) and co-dominant broad-leaved species (Castanopsis eyrei, Castanopsis sclerophylla, and Loropetalum chinense). A hierarchical Bayesian approach was used to estimate species-specific parameters and to quantify a spatially autocorrelated random effect. We selected a model that included only the following relevant predictor variables: initial size, asymmetric competition, spatially autocorrelated random effect, and random effect associated with plots. For all species analyzed, the model accounted for significant proportions of the variation (R (2) = 70-98 %) in BA growth rates. The initial slope of the relationship between BA growth rate and the initial BA tended to be steeper for P. massoniana than for C. lanceolata. The BA growth rate increased from an initial low value and then leveled off, with a lower maximum BA growth rate for C. lanceolata than for P. massoniana. The BA growth rate of P. massoniana was significantly affected by asymmetric competition with neighbors. The results of our analyses were used to predict to what extent thinning neighboring trees at different intensities would reduce competition impacts on BA growth of P. massoniana and C. lanceolata. Our results also helped to clarify the ecological characteristics of the species analyzed, as well as the spatial distribution of unknown factors influencing tree growth.