Fig. 1 Maps of (a) B06-1 (plot 1 burned in 2006) and (b) B06-2 (plot 2 burned in 2006)

Fig. 2 Univariate spatial pattern analysis of pre- (a, b) and post-fire (inset figures in a, b) in B06-1 and B06-2, respectively Fire-enhanced dead trees are also included in spatial pattern analysis prefire. Capital letters are the codes of null hypothesis models (Table 1). “+” and numbers followed indicate significant aggregation at the corresponding range of scales. Black and grey solid lines indicate g(r) and their corresponding lower and upper 99% confidence limits constructed by 99 Monte Carlo simulations, respectively

Fig. 3 Univariate spatial pattern analysis of fire-enhanced dead trees in B06-1 (a), B06-2 (b), respectively a) and b) and its inset figures are double-cluster pattern analysis of fire-enhanced dead trees at smaller and larger scales, respectively. Capital letters are codes of null hypothesis models (Table 1). “+” and numbers followed stand significant aggregation and the radii of the corresponding clusters, respectively. Black and grey solid lines stand g(r) and their corresponding lower and upper 99% confidence limits constructed by 99 Monte Carlo simulations, respectively

Fig. 4 Bivariate spatial pattern analysis of adults (pattern 1) and saplings (pattern 2) in B06-1 (a) and B06-2 (b), respectively Adults are survivals not less than 10 cm in DBH, and saplings survivals not more than 10 cm in DBH and not more than 1.3 m in height in B06-1 and B06-2, respectively. Capital letters are the codes of null hypothesis models (Table 1). “+” and “-” and numbers followed stand significant aggregation and repulsion at the corresponding range of scales, respectively. Black and grey solid lines stand g12(r) function and their corresponding lower and upper 99% confidence limits constructed by 99 Monte Carlo simulations, respectively

Fig. 5 Bivariate spatial pattern analysis of survivals (pattern 1) and fire-enhanced dead trees (pattern 2) in B06-1 (a) and B06-2 (b) and its inset figures, respectively Capital letters are the codes of null hypothesis models (Table 1). “-” and numbers followed stand significant repulsion at the corresponding range of scales. Black and grey solid lines stand function values and their corresponding lower and upper 99% confidence limits constructed by 99 Monte Carlo simulations