以新疆博斯腾湖周围分布的唐古特白刺纯林、唐古特白刺+红柳混交林、唐古特白刺+胡杨混交林等3种林型为对象,测定不同林型的唐古特白刺生物量,唐古特白刺灌丛沙包的长度、宽度、高度等指标,应用线性回归方法,分别对不同生境的唐古特白刺灌丛的地上生物量、沙包内生物量、粘土内生物量和总生物量建立了回归预测模型。结果表明,和硕县唐古特白刺灌丛沙包生物量总特征主要表现为:地上生物量、沙包内生物量、地下生物量和防风固沙功能由高到低排序是唐古特白刺纯林〉唐古特白刺+红柳混交林〉唐古特白刺+胡杨混交林,粘土内生物量则相反。本研究所建立的回归预测模型,在唐古特白刺纯林,除了粘土内生物量以外,各回归方程的p值均小于0.05的显著性水平;在唐古特白刺+红柳混交林,除了灌丛沙包的长度、宽度、高度等指标对沙包内生物量和粘土内生物量的作用分别是显著和不显著以外,其他3个回归方程的p值均小于0.05的显著性水平;在唐古特白刺+胡杨混交林,各回归方程的p值均大于0.05的显著性水平;从整体上看,本项研究所建立的回归模型是可用的。
Three forest types including Nitraria tangutorum forest, N. tangutorum + Tamarix ramosissima mixed forest, N. tangutorum + Populus diversifolia mixed forest around the Bosten Lake in Xinjiang province were taken as research objectives. The biomass of N. tangutorum, the length, width and height of Nitraria shrub sand dune were measured. Regression prediction models of aboveground biomass, biomass of internal dune, biomass in clay and total biomass were established by applying linear regression method. The biomass characteristics of N. tangutorum shrub sand dune in Heshuo county presented as follows : the order of aboveground and underground biomass, biomass of internal dune and the ability of wind protection and dune fixation is N. tangutorum pure forest 〉 N. tangutorum + T. ramosissima mixed forest 〉 N. tangutorum + P. diversifolia mixed forest, while the biomass in clay is opposite. For regression prediction model of N. tangutorum pure forest, exception for biomass in clay, the p-value of all regression equations was less than significant level of 0. 02. For N. tangutorum + T. ramosissima mixed forest, in addition to the effects of sand dune length, width and height on biomass of internal dune, biomass in clay were significant and not significant, the p-value of other three regression equations was less than significant level of O. 02. For N. tangutorum + P. diversifolia mixed forest, the p-value of each regression equation was more than significant level of 0. 05. From an overall point of view, the regression model built in this paper is feasible.