Characteristics Of Undergrowth Herbaceous Species Diversity And Its Inversion Model Of Picea Schrenkiana Var.Tianschanica Morest In Tianger Forest Park,Xinjiang

Species diversity is the basis of maintaining the stability of ecosystem structure and function,and is a comprehensive quantitative index to measure the richness,dominance and evenness of species in the habitat,which plays an important role in the service function of forest ecosystem.In order to reveal the characteristics of understory herbaceous species diversity of Picea schrenkiana var.tianschanica forest and its relationship with stand structure and environmental factors,this paper studied the understory herbaceous species diversity of Picea schrenkiana var.tianschanica forest in Tianger Forest Park,Xinjiang,analyzed the relationship between the herbaceous species diversity and the three factors of stand structure,topography and soil,established an inversion model for herb species diversity of Picea schrenkiana var.Tianschanica,it is expected to provide theoretical reference for biodiversity protection of natural forest and ecological benefit evaluation of forest resources in Xinjiang.The main conclusions are as follows:There are 115 species of plants belonging to 39 families and 94 genera in the communities of Picea schrenkiana var.tianschanica.Herbaceous plants have obvious quantitative superiority,single genus and single species are more in the communities,and the composition of understory herbaceous vegetation is more complex.The species diversity of the community was high,the evenness was large,and the habitat conditions were stable.The species diversity of the community increased firstly and then decreased with the increase of the elevation gradient and stand age.The factors that significantly affected the understory herbaceous species diversity of Picea schrenkiana var.tianschanica forest were stand age,mean height,mean diameter at breast height,stand density,altitude,slope,soil bulk density,soil total nitrogen content,alkali hydrolytic nitrogen content and total phosphorus content.The explanation rate of all variables for species diversity was 53.51%,among which stand factor accounted for 19.54%,topographic factor accounted for 9.49%,soil factor accounted for 17.83%,and interaction among the factors accounted for 6.65%.The unexplained part may be affected by human disturbance and random factors such as logging and grazing.The herbaceous species diversity of Picea schrenkiana var.tianschanica forest was determined by stand and environmental factors,the BP neural network model of herbaceous species diversity was constructed based on the above significant variables.The results showed that the mean square error of the model was 0.0086,and the mean relative error between the theoretical value and the simulated value was0.71%～4.72%,and there was no significant difference between the two values(P>0.05).It is feasible to estimate the understory herbaceous species diversity of Picea schrenkiana var.tianschanica forest by BP neural network model.