Genetic Structure Of Ardisia Crenata Var. Bicolor (Myrsinaceae) In Fragmented, Ever-green Broad-leaved Forests

Theoretical predictions indicate that forest fragmentation will result in the loss of plant genetic diversity and the increase of population differentiation. However, some researchers have found that plant population doesn't response to fragmentation accordingly. Other factors, such as life-span, population density and fragmented time plant population has experienced, may also influence the genetic consequences of fragmentation.Genetic consequences of forest fragmentation might be obvious in Ardisia crenata var. bicolor due to its short generation length (about 3 years) and relatively low density. However, there are different opinions about its classification. It's morphology is quite similar to that of A. crenata, and the only difference between them is that the lower side of leaves of A. crenata var. bicolor is deep purple. We also found some individuals that are intermediate to them morphologically in the field.By cross-pollination between the two taxa, we got some seeds, indicating that the two taxa shouldn't be treated as two distinct species biologically. Five allozymic loci (ADH, EST, SOD, AAT and POD) were analyzed, of which 4 loci (EST, SOD, AAT and POD) are the same and can't be used to distinguish A. crenata var bicolor from A. crenata. ADH is quite different, it can be used to diagnose them. DNA sequences of A. crenata var. bicolor and A. crenata are the same in a fragment of cpDNA trnL-trnF. Above results indicate that A. crenata var. bicolor and A. crenata should be treated as two different taxa of one species.Using RAPD markers, we studied the genetic structures of 16 populations in continuous and fragmented forests of Tiantong National Park and adjacent areas. The 10 primers chosen for analysis generated a total of 76 bands. The genetic diversity of A. crenata var. bicolor is low compared to other species also based on RAPD markers (species level: P100: 67.11%, H: 0.1916, SI: 0.2463; population level: P100: 5.26%-47.37%, H: 0.0260-0.1747, SI: 0.0318-0.2531). Significantly positive relationship was found between measures of diversity (P, H and SI) and the log of estimated population size. The relatively low genetic diversity might be explained by populations' geographical positions. The studied populations are located on theeastern margin of A. crenata var. bicolor' distribution in China. Due to effects of founder events, genetic drift and inbreeding, marginal populations usually possess relatively low genetic variation. Restricted geographical range might be another explanation of the low genetic diversity. Though 16 populations were sampled, their spatial distances were small. The largest distance of pairwise populations is 4.59 km. If populations had been sampled in a large range, more genetic variation should be found. Forest fragmentation has influenced populations of A. crenata var. bicolor. Because of sampling effect, rare alleles might be lost in small populations. Small population effects (inbreeding, genetic drift) can also lead to the decrease of genetic diversity. Maybe due to effects of bottleneck and inbreeding after fragmentation, High genetic differentiation was observed among populations (G_st: 44.5%). No significant relationship was found between genetic distance and spatial distance. This is usually interpreted that selection or drift plays a more significant role than gene flow. In this study, no obvious difference was found in their habitats, we concluded that genetic drift led by fragmentation contributed to the differentiation.Above results showed that short life-span shrubs and herbs with low or moderate density may be more sensitive to fragmentation, could be used as indicators of forest fragmentation.