大兴安岭森林火烧恢复年限对土壤磷及其有效性的影响

林火是大兴安岭地区森林生态系统的重要影响因子，对地下土壤养分，特别是磷及其有效性变化长时间序列研究，将有助于林火后植被恢复管理、模型模拟及科学评价。选择呼中、南瓮河、双河、图强、塔河、加格达奇、满归等地区火后恢复2-50年样地48块，并以火烧区域周边相似地形地貌、未经火烧区域为配对对照样地，野外调查记录地形地貌和地理位置数据，室内分析土壤全磷和速效磷。配对t检验、冗余排序分析、趋势分析、离散性分析等处理相结合，试图找出火烧后土壤磷及其有效性变化特征。研究结果表明：（1）整体数据平均值及离散程度表明，火后全磷及速效磷均值均稍低于对照，火后二者的离散程度为37%-49%，低于对照样地6-22个百分点；区分不同恢复时间，配对t检验未发现火烧样地土壤磷与对照间的显著差异。（2）以火烧对照比值及差值进行趋势分析发现，相对于对照，火烧后全磷呈现先增加后降低的变化趋势（R²=0.89-0.95，P=0.07-0.15），火后20-30年间开始低于对照样地。速效磷也呈类似变化趋势，但是相关性较低（R²=0.44-0.67，P=0.44-0.66）；速效磷占比年代变化趋势不明显（R²=0.08-0.12，P > 0.95）。（3）RDA排序分析表明，火烧导致恢复年限成为影响磷变化的最主要因素（15.3%），其次是经度（10.7%）；而在对照样地中，纬度是土壤磷变化最关键影响因素，能够解释其变化总量的27.6%。研究结果为森林火后土壤养分管理以及森林演替生态评价提供重要参数。

Effects of after-burning rehabilitation times on soil phosphorus and its availability in the Daxing'anling forests

Forest fire is an important factor affecting the forest ecosystem in the Daxing''anling Mountains. A long-term comprehensive analysis of the changes of subsurface soil nutrients, especially phosphorus and its availability, will contribute to forest fire vegetation restoration management, model simulation and scientific evaluation. Based on historical fire data in Daxing''anling, this study selected 48 plots that were 2-50 years post-fire, at Huzhong, Nanwenghe, Shuanghe, Tuqiang, Tahe, Jiagedaqi, and Mangui, using the neighbouring unburned site, of similar terrain, as the control. Total phosphorus and available phosphorus were determined. A paired t-test, redundant ordination analysis, temporal trend analysis, data average and variation analysis were used to analyse the data. We found that (1) the pooled data mean value showed that total phosphorus and available phosphorus in fire sites were slightly lower than in the control site and the coefficient of variation of the data in burned sites was 37%-49% for two phosphorus forms. This was 6%-22% lower than the non-burned control site. When classifying the data into different rehabilitation times, the paired t-test did not find significant differences, (2) the temporal trend analysis found that total phosphorus first increased and then decreased post-fire (R²=0.89-0.95, P=0.07-0.15), and the threshold values were in 20-30 years. The available phosphorus showed a similar trend, with much lower correlation (R²=0.44-0.67, P=0.44-0.66). However, it was not obvious in the available phosphorus fraction to total phosphorus at the tested period of time (R²=0.08-0.12, P > 0.95), (3) the RDA ordination showed that fire changed the controlling factor for soil phosphorus changes. At fire sites, soil phosphorus variations could be explained by rehabilitation times after fire (15.3%) and longitude (10.7%), while in the control plots, latitude was the most important contributing factor to soil phosphorus change, explaining 27.6% of the variation. Our findings provide important parameters for the management of soil nutrients after forest fires, as well as the ecological assessment of forest fired-induced succession, especially the analysis of soil nutrient changes.