Interspecific Segregation in the Main Species of Evergreen Broad-leaved Forest in Nanling National Nature Reserve, Guangdong Province

Analyzing interspecific segregation helps reveal species interactions, community structure and dynamics. We established 36 quadrats and drew a distribution map for 27 species numbering 10 or more. Using ArcView GIS software and the nearest-neighbor method, interspecific segregation was studied with an N ×N nearest-neighbor contingency table, a 2 ×2 nearest-neighbor contingency sub-table, and Pielou's coefficient of segregation for the individual tree species. Distribution patterns were measured by a revised Clark-Evans index (CE) and the widely-used χ² method. Results showed that:(1) three tree species were clumped, and the others were random;(2) Most species pairs were positively segregated (59.83%), some species pairs were randomly segregated (40.17%), and no species pairs were negatively segregated;(3) Some differences existed among different species in the interspecific segregation of all species. Larger individuals representing dominant species in the communities tended to be positively segregated with other large trees, though sometimes tended to random segregation;(4) Interspecific segregation was closely related to distribution patterns. There was a larger proportion of positive segregation between clumped species and other species than there was negative segregation. Most random segregation occurred between random species, which also caused slight positive or negative segregation;(5) The 27 species overlapped in distribution and were characterized by overall non-segregation.