Study On Soil Arthropods Diversity In The Fragmented Landscape Of The Thousand Island Lake Region

Soil arthropods, as a bioindicator of different types of ecosystems and animportant object of study, play an important role in ecosystem health assessment,biodiversity conservation, ecosystem management, degradation of ecosystemrestoration and reconstruction. Habitat fragmentation which considered as the mostimportant factor in the loss of species diversity, is become the hotspot of studingecology in recent years. The land-brigde lands which created by a new dam is anideal platform for the study of habitat fragmentaion. land-brigde lands offer severalpotential advantages over terrestrial habitat patches for studying the effects offragmentation because large numbers of patches are created over a relatively shorttime period by a single known disturbance event. In addition, land-bridge islandshave well-delineated boundaries and are surrounded by water.So, the experimentalcontrollability with land-brigde islands is easyer than the terrestrial habitat patches. Itcan easyer to solve the theoretical and practical problems. The Thousand Island Lakeis a large man-made hydroelectric reservoir in western Zhejiang province, it is a―model‖or―experimental‖system for studying habitat fragmentation. In this study,we investigated species richness and genetic diversity of soil arthropods on16islands in the Thousand Island Lake. The aim is to study the contribution of theisland attributes to the effects of habitat fragmentation. The contents and the resultsare mainly as followings:1species diversity:(1) Two years, we got a total of166267individuals in16islands whichbelonged to25oders respectively. The dominant groups were Acariformes,Collembola and Hymenoptera, the common groups contained Psocoptera, Diptera,Coleoptera, Araneae, Isoptera, Isopoda, Scutigeromorpha, Scuti-gerellidae andPauropodidae. Rare groups accounted only2.32%of the total individuals but thenumber is the most.(2) The taxn of the soil arthropods is increasing with the island area increasing(F=4.17, P=0.03), but the density of soil arthropods do not show this rule (F=3.12, P=0.07). The result of the SAR supported the classic island biogeography theory(LnS=0.0219*LnA+2.8527, R2=0.4692, P<0.05).(3) The cluster analysis results showed a high relationship with island area. Thelargest island can separate from other islands. Cluster analysis revealed the highcorrelation between geographic and cluster distance model. ANOVA analysisshowed that all three kinds of diversity index were not significant (P>0.05)correlation with island area.(4) The result of RDA analysis showed that island area, island isolation, islandelevation, island shape index and the PAR influenced the soil arthropodscommunities greatly. Soil moisture, soil depth and thickness of the litter are the mainfactor affecting the distribution of soil arthropods.(5) The seasonal dynamics analysis showd that the density of the the soilarthropods is the most in July each year, and least in November.all islands arefollowed this rule, but the small islands have a most dramatic fluctuations. The taxnof soil arthropods showed no obvious pattern with the season changes.2Genetic diverstiy(1) We used SRAP markers to analysis genetic diversity of Pachycondylaluteipes and Lycosa coelestri in14islands.The size of scored bands ranged from50to800bp, a lot of bands were observed from5sieved primers, respectively. Thisindicated that SRAP markers could detect a lot of polymorphic bands in thepopulations of Pachycondyla luteipes and Lycosa coelestris.(2) Molecular variance analysis (AMOVA) revealed that the genetic variationamong populations is greater than the genetic variation within populations in bothspieces, which indicated that the island isolation have a great influence on the geneticvariation of both spieces. The coefficient of population differentiation (Gst) showed ahigh level degree of genetic differentiation among each populations, which indicatedthat the land-brigde islands caused by a dam can influence the genetic variation ofthe Pachycondyla luteipes and Lycosa coelestri. The gene flow (Nm) revealed thatthe gene flow among populations have a weak level of degree (Nm<1), But the Nmvalues obtained from some adjacent islands suggested that high level of migrationmay have occurred between some adjacent populations (Nm>4). But most Nmvalues are less than1, meaning that there might have a genetic drift in most populations.Cluster analysis (UPGMA) of the Jaccard similarity coefficients revealed the highcorrelation between geographic and genetic distance model.(3) Spearman correlation and regression analysis showed a significant corelationbetween island isolation and genetic diversity. The result of Mantel test gives a moreinformation about the correlation between geographic distance and genetic distance.