【作者】 喻泓；

【导师】 慈龙骏；

【作者基本信息】 中国林业科学研究院 ， 水土保持与荒漠化防治， 2009， 博士

【摘要】 火是许多陆地生态系统重要的生态因子,为林火干扰生态系统主要的驱动力。樟子松(Pinus sylvestris L. var. mongolica Litv.)是中国北方针叶林的建群树种,以耐寒、耐旱、耐贫瘠、防风固沙和速生丰产而著称,在中国“三北”地区防护林营造和荒漠化防治中得到广泛应用。内蒙古呼伦贝尔沙地天然樟子松林位于中国森林高火险发生区,因此林火也在天然樟子松林中频繁发生,并成为其结构、功能和演替的重要驱动力。自20世纪50年代以来,樟子松引种繁育和人工林营造得到了广泛关注,然而关于天然樟子松林结构、功能和演替的研究却并不多见,也缺乏樟子松对林火干扰响应方面的研究。本研究从呼伦贝尔沙地樟子松林最近的两次林火事件(1994年和2006年)入手,探讨樟子松林分结构、空间格局、林木竞争及林下植物多样性对地表火干扰的响应。利用全林木定位方法,在1994年和2006年两次林火事件中地表火干扰及1970年以来无林火干扰记录的樟子松林中,共调查了6块边长100 m、面积1 hm2的正方形典型样地;按不同树木种类分别调查记录样地中残桩、倒木、枯立木和2年生以上(包括2年生)活立木的位置坐标、地径或胸径及林火干扰后树皮熏黑高度等因子。另外,利用改进的Whittaker样方在不同林分中共调查9个樟子松林下植物多样性样方。通过Kruskal-Wallis和Duncan法,进行林火强度、林分胸径、林木竞争和林下植物丰富度的非参数方差分析及多重比较,分析研究樟子松林火和林分结构特征、林木竞争及林下植物多样性;利用Ripley’s K函数和成对相关函数g(r),通过单变量或双变量空间格局分析,揭示林火干扰前后、林分不同组分及其相互间的空间格局;Ripley’s K的线性化函数L(r)有时也用于单变量分析中,并与成对相关函数g(r)相互比较以揭示格局的变化;另外,利用物种-面积曲线和间接梯度分析(DCA)探讨不同时期地表火干扰下林下植物种的多样性及其变化。结果表明,(1)不同时期林火干扰林分的地表火强度间没有显著差异,地表火强度与林木胸径间表现出显著的正相关且具有显著的回归关系;(2)地表火烧死了大多数下层林木、小径阶林木和更新幼树,减少了树木种类,使林分密度降低了一个数量级;随着林火干扰后时间的推移,由于大量更新苗的出现,林分密度又明显增大;而无林火干扰林分密度的变化较小;(3)地表火干扰后,林分平均胸径成倍地增加,胸高断面积仅略微降低,樟子松胸高断面积在林分中仍占绝对优势且其比重略有上升;(4)地表火干扰下,樟子松林的空间格局呈现出从显著性聚集分布向随机分布变化的趋势,且随着时间推移这种趋势尤为明显;无林火干扰下,林分稀疏主导下的空间格局基本上保持相对稳定状态;(5)火后存活的幼树和更新幼苗均在小尺度上表现为显著的聚集分布,而无林火干扰下幼树在小尺度上为聚集分布或为双尺度聚集分布;火后存活大树则趋于随机分布,无林火干扰林分的大树则为小尺度上聚集分布或双尺度聚集分布;从全部径阶到较大径阶上,林木空间格局呈现出聚集—随机—均匀的变化趋势;(6)地表火干扰下,大树和幼树在小尺度上相互排斥,随着大量更新幼苗的出现,二者又相互吸引;无林火干扰林分,大树和小树间既有相互排斥也有相互吸引;地表火干扰下,大树和死树、幼树和死树分别在小尺度上表现为显著的正相关,随着时间的推移却又呈现相互独立状态;无林火干扰下,大树和死树、幼树和死树分别表现出不同尺度上的正相关;(7)樟子松林的种内竞争较种间竞争更为激烈;地表火干扰显著的降低了樟子松林的竞争强度,火干扰后持续的时间序列上,樟子松林竞争强度持续地降低;对象木竞争强度随其胸径的增加而减小,且二者具有显著的幂函数关系;(8)林火干扰后,短期内、小尺度上显著的增加了林下植物的α多样性而降低了植被盖度,较大尺度上林下植物的α多样性又趋于一致;林下植物的β多样性沿着火后1年—火后12年—无火林分的方向不断增大;间接梯度分析也表明,有更多的植物种趋于出现在火后1年的林分中;火后1年林分更趋向于湿生生境,有较多的偏湿生种。总之,地表火干扰显著地降低了樟子松林的密度,减小了林木个体间的竞争强度,促进了林下植物多样性的变化,推动了林分空间格局向着成熟林方向发展,表现出了地表火所具有的强烈的林分稀疏作用。所以,在樟子松天然林的保护和森林资源的经营管理中,需要充分考虑林火特别是地表火干扰对樟子松生长发育及林分演替的作用,在林火生态学的指导下,趋利避害,以实现天然樟子松林的可持续发展。更多还原

【Abstract】 Fire, a ubiquitous important disturbance, is a crucial ecological factor in most terrestrial ecosystems. Mongolian pine (Pinus sylvestris L. var. mongolica Litv.), a dominant tree species in boreal forests of China, have been widely applied in combating desertification in the northern China and is renowned for being available in windbreak building, sand dune fixation, timber productivity, and tolerance of cold, drought, and innutrition. Natural Mongolian pine forests mainly sit on the southeast margin of the boreal forest in the Hulun Buir sandland, Inner Monglia, P. R. China, where belongs to one part of high risk area of wildland fire in China. Therefore, wildland fire is a frequent disturbance and a dynamic driving force in the forests.The seedling breeding and replantation of Mongolian pine have been intensively considered since 1950s. However, natural Mongolian pine forests have been least concerned, especially, its responses to wildland fire disturbance. In this study, responses of stand structures, spatial patterns, competitive intensity, and understory vegetation species of Mongolian pine forests to surface fire are explored in the last two wildland fire events in 1994 and 2006, respectively. Stumps, fallen logs, standing dead trees and living trees old than two years in six 1 hm2 plots, which had been attacked by the last two wildland fire events or with fire-free for 38 years, were fully mapped. Diameter at basal, diameter at breast height (DBH) and bark char height of trees were also measured and each individual was identified to species. Understory vegetation species were also surveyed by nine 0.1-hm2 multi-scale modified-Whittaker plots in the fire-mediated forests with different fire intervals. One-way nonparametric analysis of variance (ANOVA) (Kruskal-Wallis) and Duncan’s multiple range tests were conducted to detect the differences of surface fire intensity, stand structure, competitive intensity, and understory vegetation species richness of stands surveyed. The univariate and bivariate spatial pattern analysis of pre- or post-fire stand and the components of the fire-mediated forests were explored using the pair correlation function g(r). The linear form of Ripley’s K function, L(r), was also conducted to explore univariate spatial patterns of agents concerned once in a while and provide a reference to the pair correlation function, g(r). Species-area curve and Detrended Correspondence Analysis (DCA) were also conducted to explore the composition and variations of understory plant species in the fire-mediated forests.The results show that 1) surface fire intensities had some common characteristics significantly, and positively correlated with DBHs in three burned plots. Thus, the positive regressions could also be constructed between them. 2) Surface fire mainly excluded the small size classes and recruits, and got the kinds of tree species decreased. Thus, the tree densities decreased sharply to a lower order of magnitude postfire immediately and regained again for a great number of recruits occurred as time had lapsed since the stand burned. However, the tree densities in the fire-excluded stands had little changes post-mortality. 3) Mean DBHs of the burned plots increased multiplely, while its basal areas only decreased slightly postfire. Nevertheless, the basal areas of Mongolian pine were still predominant and its proportions in the corresponding plots increased slightly. 4) Spatial patterns of Mongolian pine forests had a trend from aggregation to randomness and were strengthened as time lapsed postfire. However, variations of spatial patterns in the fire-excluded plots were relatively stable post-mortality. 5) Saplings and recruits were all significantly aggregated at small scales postfire. Survival adults were randomly and aggregatedly distributed in the burned and fire-excluded plots, respectively. Spatial patterns of survival size classes had a trend of aggregation—randomness—repulsion significantly from all to larger size classes in both burned and unburned plots. 6) Adults and saplings were repulsively distributed postfire, and positively correlated with a good deal of recruits occurred as time had lapsed postfire. However, adults and saplings exhibited repulsively or aggregatedly in the fire-excluded plots. Adults and dead trees, saplings and dead trees were all positively correlated at different scales in the stands surveyed, respectively. 7) Intraspecific competition was more intense than interspecific competion in the stands surveyed. Surface fire had the competitive intensities decreased significantly in the fire-mediated forests, and the power functions could be constructed between them significantly. 8) The understory plant speciesα-diversity increased postfire instantaneously and converged with fire-free for several dozen years, while, theβ-diversity increased along the trend from one year postfire to 12 years, then to fire-free for several dozen years. More species, especially hygrophytes species, occurred instantaneously postfire in the 2006-burned stand analysed by the Detrended Correspondence Analysis.In a word, surface fire, as a dynamic driving force of Mongolian pine forests, has the tree density of stand reduced greatly, structures the stand, changes the composition and richness of understory plant species, directs spatial patterns of the forests evolving to mature, and exhibites an intensive thinning force to the forests. Thus, we should seriously take into consideration surface fire in the conservation and resources management of the natural Mongolian pine forests and realize sustainable development in practice under the changed fire regime and the background of increasing anthropogenetic activities and climate changing.更多还原