Influence of climatic and geographic factors on the spatial distribution of Qinghai spruce forests in the dryland Qilian Mountains of Northwest China

The effect of climate variables (temperature and precipitation) on forest spatial distribution is more prominent in dryland high mountains, where forest distribution is inherently very sensitive to and strongly limited by the substantial spatial heterogeneity of site conditions. Thus, a more reliable prediction of forest distribution under changing environment depends upon an understanding of the joint influence of climatic and topographic factors and their thresholds. This study was conducted on Qinghai spruce forests as dominant tree species in the Qilian Mountains of northwest China. The spruce forest distribution was surveyed by remote sensing in Dayekou watershed and by field investigation in a nested smaller watershed. Analyses showed that mean annual air temperature and precipitation, which vary with elevation, are the key climatic factors determining forest distribution, but slope aspect also plays an essential role. The potential core distribution area of denser forests and potential distribution area including sparse forests are between the axes of elevation (2635.5-3302.5 and 2603.4-3325.8 m a.s.l.) and slope aspect (-74.4-61.2 degrees and - 162.6-147.1 degrees deviated from north). The corresponding threshold of mean annual air temperature at the upper elevation boundary is -2.59 and -2.73 degrees C, while the threshold of mean annual precipitation at the lower elevation boundary is 378.1 and 372.3 mm, respectively. Using these thresholds and the elevation gradients of climatic factors, the shifting of elevation boundaries under climate change scenarios can be predicted. However, the forest distribution is also limited by a soil thickness of >= 40 cm; and by slope position of lower-, lower- and middle-, and entire-slope within the elevation ranges of < 2800, 2800-2900, and > 2900 m a.s.l., respectively. This study showed that adding geographic factors will greatly improve the prediction of changes in forest distribution area in dryland mountains, in addition to the influence of climatic factors. (C) 2017 Elsevier B.V. All rights reserved.