Numerical analysis of inter-specific relationships in Alpine steppe community in Bayanbulak

Highland ecosystems are believed to be of vital importance in ecological study,partly due to the rapid climate change worldwide and the highest sensitive feature of this special system,not only its productivity and biodiversity but also its inter-specific relationship,to environmental alterations.The Bayabulak Steppe distributes in the middle part of Tianshan Mountains,and is situated at 42°18′—43°34′ N,82°27′—86°17′ E with altitude between 2460—2760 m.Climate here belongs to typical alpine cold climate with yearly mean temperature of-4.8 ℃,yearly mean precipitation here is 276.2 mm,yearly evaporation is(1 135.2±112.3) mm,the snow period in Bayabulak is about 150—180d,and there is no frost-free time in this region.Stipa purpurea and Festuca ovina grassland are the dominant vegetation types in our study site. In order to clarify the species composition and inter-specific association of the alpine grassland community of Bayanbulak in Xinjiang,China,variance ratio(VR) of the overall association,association coefficient(AC),Ochiai index(OI) and Spearman rank correlation coefficient,based on the 2 × 2 contingency table,were used in our quantitative analysis.Field data were collected in July 2009.Three parallel transect were firstly set up and the distance between each of them was 50 m,then 20 plots(0.50 m × 0.50 m for each unit) were randomly set within each transect,in so doing total of 60 plots were involved in the present work.Our results showed that: 1) There are 36 species with 14 common species of plants in Bayinbulak alpine steppe,and they belong to 26 Genus of 16 Family.W = 61.52,a statistic index to test whether the VR is significant or not,indicated that the overall correlation of all plant species was positive but not significant.2) Inter-specific associations of keystone species with others were neutral or not significant.Keystone species showed greater independence and occupied prior ecological niche in the community,and their distributions were not susceptible to other species.Values of OI were basically identical to those of association coefficient;Based on the measuring of OI,19.79% of total 91species-pairs composed by 14 common species was significant or very significant,and 72.21% of them showed positive correlation while 27.79% showed negative correlation for those significantly related species-pairs;3) Inter-specific associations of accompanying species or weeds with others were significant(negative or positive),suggesting higher dependency of accompanying species or weeds on other species,namely they were susceptible to other species and only occupied inferior ecological niche in the community.Furthermore,populations of these plant species are instability in the process of ecosystem development;4) This study provides new dimensions for the theory of species association.According to their adaptability to the environment and the relationships between plants and environment,the 14 common species were divided into three ecological species groups.Positive correlations occurred among the species in the same ecological species groups,while the negative correlations were found in the different ecological species groups.It mainly due to their different biological features induced by different adaptability to habitats or separated niches;5) Several suggestions were proposed to promote the ecological restoration of degraded steppe systems in the study region.