作 者 :欧阳资文1,2,3;彭晚霞1,2,3;宋同清1,2;曾馥平1,2;王克林1,2;关欣3;吴海勇1,2,4
期 刊 :应用生态学报 2009年 20卷 06期 页码:1329-1336
关键词:土壤;
摘 要 :通过网格(10 m×10 m)取样,运用地统计学方法研究了喀斯特峰丛洼地4类典型干扰区表层土壤(0~20 cm)有机质的空间变异、分布,及其生态学过程和机制.结果表明:随着干扰强度降低,植被由农作物(Ⅰ)—人工林(Ⅱ)—次生林(Ⅲ)—原生林(Ⅳ)顺向演替,土壤有机质逐步提高,且达到了显著水平(P<0.05).4类干扰区均具有良好的空间自相关性,不同干扰区空间变异特征不同,除Ⅲ类干扰区土壤有机质半变异函数优化符合指数模型外,其他3类干扰区均符合高斯模型;Ⅰ类区土壤有机质的空间自相关呈中等程度,C0/(C0+C)值达26.5%,其他3类干扰区C0/(C0+C)值在9.0%~22.6%,呈强烈的空间相关性;由于人类干扰强烈,Ⅰ和Ⅱ类区呈低能量匀质状态,变程及空间自相关范围较大,Ⅳ类区植被覆盖率较高,变程也较大;Ⅲ类区干扰强度中等,植被类型多且分布不均,变程最小;Ⅱ和Ⅳ类区的分维数(D)值较小,土壤有机质的空间依赖性较强;而Ⅰ和Ⅲ类区D值较大, 土壤有机质空间分布的随机变异较大;Ⅰ和Ⅱ类区土壤有机质呈单峰分布,Ⅲ类区土壤有机质呈凹型分布,Ⅳ类区呈凸型分布.减少干扰是喀斯特峰丛洼地脆弱生态系统土壤质量改善、植被迅速恢复及生态重建的重要保障.
Abstract:By using geostatistic methods, this paper studied the spatial variation and dist
ribution of soil organic matter as well as its ecological processes and related
mechanisms in four typical disturbed areas (cropland, man-made forest,
secondary forest, and primary forest) of karst peak cluster depressions in nort
hwest Guangxi of China. Eighty soil samples (0〖KG-*2〗-〖KG-*7〗20 cm) were coll
ected from an aligned grid of 10 m×10 m for the analysis of soil organic matter
. The soil organic matter content increased significantly (P<005) with the
decrease of disturbance and the vegetation succession from crop to man-made fo
rest to secondary forest to primary forest. Soil organic matter content had good
spatial autocorrelation in all of the four typical disturbed areas, but its spa
tial heterogeneity differed. Gaussian model fitted best to the semivariance fun
ctions of soil organic matter content in the study areas except secondary forest
area where exponential model fitted well. In cropland area, the spatial autocor
relation of soil organic matter was at medium level, with the C0/(C0
+C) being 265%; while in the other three areas, the spatial autocorrelatio
n was at high level, with the C0/(C0+C) being 90%〖KG-*2〗-〖K
G-*7〗226%. The range and scale of the spatial autocorrelation of soil orga
nic matter in cropland and man-made forest areas were larger than those in the
other two areas, possibly due to the strong human disturbance and the homogeneit
y of low energy. The range of the spatial autocorrelation of soil organic matter
in primary forest area was large due to the high vegetation coverage, while tha
t in secondary forest area was the lowest due to the diverse vegetation communit
ies and their uneven distribution. The low fractal value (D) of semivariance fu
nctions of soil organic matter in man-made forest and primary forest areas sugg
ested that a strong spatial dependence existed, while the high D in cropland and
secondary forest areas suggested a great random variance of spatial distributio
n of soil organic matter occurred. The spatial pattern of soil organic matter pr
esented a unimodal distribution in cropland and man-made forest areas, a concav
e distribution in secondary forest area, and a gibbous distribution in primary fo
rest area. To reduce human disturbance would be helpful to the soil quality impr
ovement, rapid vegetation restoration, and ecological reconstruction of karst de
generative ecosystems.