Abstract: Stepwise regression analysis and multiple numeric theory were used to analysis the relationship between height growth of Picea mongolica and the site conditions, and the multiple models of spruce growth are: y=8.0031+0.2087p 5+0.1286p 6, y=7.7836-7.9944s+0.6686a. The results showed that seedling tree height growth of the spruce on sandy land was closely related to the precipitation in May and June, however the relation to the temperature in June was not so significant. Precipitation, temperature and their combined conditions in main growth seasons were closely related to growth of the spruce. The results of multiple numeric theory analysis indicated that the position of slope and age of plants were also main factors to effect the growth of the spruce. Site conditions, age of plants and precipitation in May and June, temperature in June should be paid special attention to in management and introduction of the spruce to other area on sandy land.

Key words: climate change, environmental heterogeneity, tree species distribution prediction, landscape fragmentation, sampling plot data.