RESUMEN
Species composition and diameter class structure were investigated in 1.5 hm2 (100 m x 150 n) permanent plot in Picea koraiensis and Abies nephrolepis forest of Xiaoxing' an Mountains. The spatial distribution pattern and spatial association of main populations and gap makers were analyzed by using point pattern analysis. The results showed that there were a total of 13 species with diameters at breast height greater than 2 cm in tree layer, and great differences were observed in the densities of main populations. The importance values of A. nephrolepis, P. koraiensis, Betula platyphylla and Acer ukurunduense were ranked in the first 4 in the plot. The diameter class structure of their populations presented an inverse 'J' curve. The spatial distribution patterns for A. nephrolepis and P. koraiensis were similar, which changed from aggregated, random to uniform distribution with the spatial scale. For B. platyphylla, the distribution was aggregated at ≤40 m scale, and random at >40 m scale, whereas A. ukurunduense presented an aggregated distribution pattern at the whole research scale. Except that the negative correlation between B. platyphylla and A. ukurunduense existed at the whole research scale, positive correlation between the other populations at small scale and negative correlation at large scale were observed. Only A. nephrolepis and B. platyphylla had significant positive correlation, and generally no significant correlation existed between other populations. Spatial distribution pattern of gap makers was characterized as aggregated distribution at small and middle scales, and random distribution with increasing scale. Spatial point pattern of gap makers formed by uprooting exhibited unimodal type distribution, and random, aggregated, and uniform distribution also occurred. Spatial point pattern of gap makers formed by breaking overall presented a little fluctuation, random and aggregated distributions alternatively appeared at small scale, and random distribution happened at large scale. Their spatial correlation analysis indicated that significant positive correlation existed at ≤32 m scale, and non-significant negative correlation at >32 m scale.