Statistical analyses
All data met the assumption of normality and homogeneity of variance, and were analyzed by using SPSS 21.0 (IBM, Chicago, IL, USA).
General linear model - univariate analysis was performed to assess the effect of target species, soil nutrition condition, and plantation condition on the biomass and root: shoot ratio, with all factors as fixed factors. In addition, Duncan’s post hoc analysis was conducted to evaluate the differences of the mean value among these treatments.
General linear model - univariate analysis was used to estimate how target species, soil nutrition condition, and neighbor species influenced the intensity and importance of competition, with these factors as fixed factors. Moreover, the magnitude of effects (ω 2) of individual factor or the interaction of two or three factors were calculated by dividing each variance component to the total response variable (Graham et al. 2001). In addition, the directionality of the NImpA was assessed using one-sample t test, where a value > 0 indicates a significant facilitation effect of the neighbor on the target species, suggesting the trend of co-existence. While a value < 0 indicates a significant inhibition effect, suggesting the trend of competitive exclusion. Furthermore, we estimated the difference ofNImpA on the same target species between high and low nutrition treatments by independent sample t test.
In order to analyze the change of competitive hierarchy, the effects of target species, soil nutrition and their interaction on the competition intensity were analyzed for each neighbor species by general linear model - univariate analysis, with target species and soil N condition as fixed factors; and Duncan’s post hoc analysis was conducted to evaluate the differences of the mean value among these four treatments. The significant interaction between target species and soil nutrition condition would suggest the competition change between S. grandisand S. krylovii .
The effect of target species, plantation condition on the nutrition-addition effect was estimated by general linear model - univariate analysis, with target species and plantation condition as fixed factors. When significant interaction effect of target species and plantation condition was found, a simple-effect analysis was used to evaluate the differences among levels of one factor under a certain condition of another factor. Moreover, the difference of the nutrition-addition with zero was assessed using one-sample ttest.