---

作 者 ：宋于洋,李园园,张文辉*

期 刊 ：生态学报 2010年 30卷 16期 页码：4317~4327

关键词：梭梭(Haloxylon ammodendron)；不同发育阶段；RipleysK(r)函数；g(r)函数；分布格局；关联性；

Keywords:Haloxylon ammodendron (Haloxylon ammodendron), different developmental stages, Ripley′sK(r) function, g(r) function, distribution pattern, spatial association,

摘 要 ：运用点格局法中的Ripley’sK(r)函数的变形L(r)函数和g(r)函数，对古尔班通古特沙漠不同生境下不同发育阶段梭梭种群的空间格局及关联性进行了研究。结果表明：L(r)函数显示梭梭种群格局倾向于聚集分布，且集中分布在0-25m尺度范围内，而g(r)函数分析在小于3m的尺度上呈聚集分布，大于3m的尺度后波动较小，最大聚集尺度表现在0-10m范围内；梭梭在不同发育阶段过程中，L(r)和g(r)函数都显示由幼苗、幼树的聚集分布变为成年树的随机分布，甚至在某些尺度上变为均匀分布，同时幼苗幼树向成年树过渡过程中，梭梭的聚集强度呈逐渐减弱趋势。在关联性分析中，L(r)函数分析中正关联维持的尺度范围较g(r)函数大。L(r)函数分析幼苗与幼树、成年树在0-15m尺度内呈现正关联，g(r)函数在0-5m范围内表现为正关联，而幼树与成年树在0-10m尺度内多呈负关联 ，且两个大小级的形体大小差异越大，它们的正关联关系越弱，甚至表现为负关联或无关联。幼苗幼树的聚集分布和正关联是梭梭种子的传播和密度制约的一个平衡，对梭梭种群的生存和发展是有利的。另外，梭梭种群分布格局的强度在不同地形也存在差异，奎屯平地比五家渠的聚集强度和关联性大，缓坡差异较小，这说明地形对各种资源的再分配间接地影响了梭梭的格局。总体上看，同时应用L(r)和g(r)函数进行梭梭空间格局与关联性分析时结果不尽相同。L(r)函数的最大值可以反映典型的聚集尺度，而g(r)函数中出现的第1个最大值可以表示植株间典型距离。在小尺度下两种函数分析所得空间格局是一致的，而在大尺度上有较大差异。g(r)函数在小尺度范围内的分析结果更接近实际情况，有利于揭示出梭梭空间格局与生态过程有联系的“关键尺度”，说明梭梭为了适应恶劣环境往往表现为聚集分布，这种聚集生长现象有利于个体的生存与繁衍。因此，联合使用L(r)函数和g(r)函数更有利于揭示植物个体间的关系。

Abstract:L(r) function and g(r) function (the derivation of Ripley′s K(r) function) are used in analyzing the spatial pattern and spatial correlation of Haloxylon ammodendron population in different developmental stages in Gurbantonggut desert. The L(r) function revealed an aggregated distribution trend as a whole and the scale ranges from 0m to 25m, while g(r) function revealed an aggregated distribution mainly within 3 meter′s scale, beyond which the distribution becomes smooth and the aggregation trends extended to no more than 10m. With the development of the Haloxylon ammodendron from seedling and sapling into the adult, both the L(r) function and g(r) function show that the distribution trend changes from aggregated to stochastic or even uniform distribution in some scale. As to the spatial correlations, positive correlation is shown in both the L(r) and the g(r) functions, but the scale is larger in L(r) function than that in g(r) function. In L(r) function, the relationship between seedling and sapling or the adult tree is positive within 15m′s scale, while in g(r) function positive correlation is shown between the seedling and the sapling within 5m′s scale, whereas negative correlation is shown between the sapling and the adult within 10m′s scale. That is to say, the greater the difference in the age of the plants, the weaker the positive correlation is shown. As a balance confined by the amount and the spreading of the seed, aggregated distribution and positive spatial correlation in seedlings are beneficial to the survival and development of the Haloxylon ammodendron population. The distribution pattern of the Haloxylon ammodendron population is also affected by topography. The aggregative intensity and association in Kuitun interdune are greater than that in Wujiaqu but their ramp difference is small. This shows that topography has an indirect influence on the distribution pattern of the Haloxylon ammodendron population because of its redistribution function to all the resources. On the whole, in analyzing the spatial pattern and spatial correlation of Haloxylon ammodendron population different results are shown in L(r) function and g(r) function. The maximum value in L(r) function shows the typical aggregation scale while the first maximum value in g(r) function shows the typical distance between plants. The spatial patterns shown in L(r) function and g(r) function are consistent in small scale, but differ a lot in large scale. In small scale the result shown in g(r) function is closer to the reality and it truly reflects the "key scale" in the correlation between the spatial pattern and the developmental stage of the Haloxylon ammodendron population. The result shows that to adapt to the unfavorable environment, aggregated distribution which is beneficial to the survival and development of the individual, is the main trend for the Haloxylon ammodendron population. Consequently, the combination of the L(r) and g(r) is favored in reflecting the spatial correlation among plant individuals.

---