弄岗喀斯特季节性雨林枯立木多度的空间分布及影响因子  

Spatial distribution of standing dead trees abundance and its impact factors in the karst seasonal rain forest, Nonggang, southern China

作　　者：王斌[1,2] 向悟生[1,2] 丁涛[1,2] 黄甫昭[1,2] 文淑均[1,2] 李冬兴[1,2] 郭屹立[1,2] 李先琨[1,2] 

机构地区：[1]中国科学院广西植物研究所,桂林541006 [2]广西友谊大森林生态系统定位观测研究站,凭祥532600

出　　处：《科学通报》2014年第35期3479-3490,共12页Chinese Science Bulletin

基　　金：国家自然科学基金(31270469,31300359);国家科技支撑计划(2011BAC09B02)资助

摘　　要：枯立木(standing dead tree)是森林的重要结构组成部分.对枯立木空间格局与其影响因子等进行探讨,有助于揭示森林树木死亡机理及森林群落动态规律.本研究将弄岗喀斯特季节性雨林15 hm2样地内胸径大于1 cm的枯立木划分为不同胸径等级,并在不同空间尺度上统计了枯立木多度、空间相邻因子、地形因子及群落类型因子等数据,用零膨胀负二项模型(zero-inflated negative binomial models)解决了多度数据的非正态性且零值过多等统计问题,分析了该森林枯立木多度的空间分布规律及影响因子.结果表明:(ⅰ)弄岗15 hm2样地内胸径大于1 cm的枯立木有2254株,平均胸径6.21 cm,最大胸径61.86 cm;胸径分布呈反"J"型,无明显断层;小径级枯立木在海拔较高地区分布较多,大径级枯立木在海拔较低地区分布较多;(ⅱ)小径级枯立木呈聚集型空间格局,大径级枯立木趋向于随机型空间格局;枯立木多度存在多尺度的空间自相关结构;(ⅲ)随着径级增大,枯立木多度与凹凸度、空间相邻因子的相关性逐渐减弱,与坡度、海拔的相关性逐渐增强;随着单位取样尺度增大,枯立木多度与凹凸度、海拔、坡度、群落类型因子的相关性逐渐减弱,与空间相邻因子相关性逐渐增强.研究表明,喀斯特森林中枯立木分布格局由多尺度的空间结构所构成,不同空间尺度、地形条件、群落类型等对枯立木的径级结构及数量组成有显著影响,而枯立木结构的空间异质性将会影响森林光资源及木质残体储量等的空间动态,进而影响群落物种组成.Standing dead trees are an important part of the composition of forest structure. Therefore, studying the spatial distribution of standing dead trees abundance and its impact factors has provided much needed insights to understand the mechanism of tree death and the succession of community. According to a 15-hm2 study plot investigation in northern tropical karst seasonal rain forest, Nonggang, southern China, the standing dead trees with DBH（diameter at breast height） greater than 1 cm were classified. And the spatial distribution of standing dead trees abundance and its impact factors like spatial, topographic and communities’ were analyzed using zero-inflated negative binomial models. The results showed that：（1） there were 2254 standing dead trees in the plot with an average DBH of 6.21 cm. And the maximum DBH was 61.86 cm. The DBH size-class structure of standing dead trees exhibited an inverse J-shaped pattern. Dead trees with smaller DBH distributed more densely in the high elevation areas, whereas those with larger DBH had a higher density in the valley bottoms.（2） Standing dead trees with smaller DBH displayed an clumped distribution, while those with larger DBHs had a random distribution. And the abundance of standing dead trees displayed different spatial autocorrelation at different spatial scales.（3） As DBH size-class increased, the correlation between abundance and elevation, abundance and slope gradually increased, while that of abundance and convex, abundance and spatial adjacent factors gradually decreased. The correlation between abundance and spatial adjacent factors increased as unit spatial scale increased, and the correlation between abundance and convex, elevation, slope and community type factors gradually decreased as unit spatial scale increased. There were different spatial distribution of standing dead trees abundance existed as the scales changed. And the spatial, topographic factors, different communities have significant impact on these spatial patterns of standin

关 键 词：零膨胀模型 负二项分布 空间自相关 地形因子 群落类型 

分 类 号：S718.5[农业科学—林学]