贺兰山油松生长的气候响应机制初步探讨

应用VaganovShashkin模型对贺兰山油松的生长变化作了拟合,发现4～8月的降水量大小决定了该年树木生长的宽度变化,另外,除5月外的1～8月份温度和树木生长的负相关是因为较高的温度不利于土壤中水分的保持,间接地导致树木生长的减慢。VS模型在贺兰山地区取得了良好的拟合效果,更为符合树木生长的生理学意义。通过拟合序列和3个采样点树轮宽度序列的对比发现,拟合序列代表着一个较大范围树轮宽窄的变化。坡向对树木生长的影响较大,甚至超过了海拔高度和距离的影响,故建议树木年轮气候学采样应沿着同一坡向采样。

A PRIMARY DISCUSSION ON THE CLIMATIC RESPONSE OF Pinus tabulaeformis IN THE HELAN MOUNTAIN

The traditional statistical analysis that is widely used in tree-ring research could establish the preliminary relationship between tree-ring indices and climatic factors, but it is very difficult to describe tree ring climatic response on cellular scale. It is, therefore, necessary to apply process-based models, such as the Vaganov-Shashkin model, to link them directly and to give tree-ring growth response to climate an explicit physiological meaning. In this paper, we use the Vaganov-Shashkin model to simulate climatic response of Pinus tabulaeformis in the Helan Mountain. Through the simulation of Pinus tabulaeformis growth, it has been found that the variation of tree-ring width is mainly controlled by both the total precipitation from April to August and the mean temperature from January to August of the current year. We observed a positive relationship between ring width and rainfall, but a negative correlation between ring width and the mean temperatures from January to August except for May due to the loss of the soil moisture, which is caused by increasing of the mean temperature. In another word, the high temperature indirectly results in slow growth by increasing the loss of soil moisture. The reason of the positive correlation between May temperature and ring width is that the beginning time of cell division is decided by the temperature in May. Of course, ring widths benefit from early beginning of cell division. The Vaganov-Shashkin model has a good simulation result in the Helan Mountain area and has better biological meaning than statistical method. The comparison between the simulation series and the three tree-ring standard series indicates that the simulation series could present a regional pattern of tree-ring width variation. Meanwhile, we found that the direction of slope has stronger influence on tree growth than that of both elevation and distance. Thus, sampling along the same slope within a site is suggested in dendroclimatology. It is very important to apply this model to reconstruct precipitation and temperature in the future study.