SPATIAL AUTOCORRELATION OF GENETIC STRUCTURE IN A POPULATION OF CYCLOBALANOPSIS GLAUCA IN HUANGSHAN, ANHUI

CHEN Xiao-yong

2001-01-01

Abstract:Genetic structure is defined as non-random distribution of genetic variation, and this study utilised spatial autocorrelation analysis to analyse variation among genotypes in a single population as, especially for within deme analysis, it is better than F-statistics or GST. The genotype of all individuals ≥2.5　cm DBH in a western Huangshan population of Cyclobalanopsis glauca was detected using allozyme analysis, and spatial autocorrelation of each allele was then analyzed using Moran's I. Most Moran's I values for alleles were larger than expected values, but two of the alleles had significant positive spatial autocorrelation. Detailed analysis showed that most alleles were significantly spatial-autocorrelated at short distances. A correlogram showed positive spatial autocorrelation in all alleles except Est-2-d at short distances (0-5 m). In general, Moran's I indices decreased as distance interval increased. However, the trends of different alleles varied, and therefore no dominant evolutionary factors could be identified.