不同尺度苔藓结皮土壤性状的空间分布特征

苔藓结皮是生物结皮的高级阶段,它在防风固沙、改善土壤水分、养分等方面具有十分重要的生态作用,但前期研究多是对结皮区和无结皮覆盖区土壤性状的比较,对结皮斑块内土壤理化因子的空间分布情况尚不明确。运用经典统计学,地统计学以及克里格插值法探讨了土壤水分、有机质、全氮、全磷、总盐在样方和斑块两个尺度的空间变异特征,旨在阐明不同尺度藓类结皮土壤空间异质性的强度,明确不同尺度土壤各性状合适的取样距离。研究结果表明藓类结皮土壤性状在两种尺度均表现出中等程度的变异,随着取样尺度减小,自相关性在空间异质性中占的比重增加,结构因素的影响增强,随机因素的影响减弱。样方尺度的空间异质性大于斑块尺度异质性,且土壤性状的变程较大,采样时可适当增大取样间距;两种尺度下5个土壤性状中,全磷的变程最小,实际取样时应适当缩小取样距离,总盐变程大,自相关程度低,因而取样间距可适当放大。斑块尺度,藓类结皮土壤性状由边缘向中心递增(磷递减)的现象与斑块的边缘效应有关。

Spatial distribution of soil properties covered by moss crusts on different scales

Biological soil crusts are the organic complexes composed of bacteria, cyanobacteria, algae, lichen and moss with soil particles. They play critical roles in desert ecosystems, including enhancing soil stability and fertility, preventing soil erosion by water or wind, facilitating seed germination. Moss crust is the latest successional stage of biological soil crusts in arid and semiarid land, which perform important functions. Most of the studies focused on the difference of soil properties between moss crust and lack of moss crust. Generally, moss crust distributes in patches in interdune areas of sand dunes, it is a small scale. Little is known about its nutrients distribution at patch scale. The aims of this research are to (1) make sure the strength of soil heterogeneity for moss crust at both quadrate scale and patch scale; (2) support basic data for moss crust according to patch size. In this study, patch's area is calculated by Photoshop CS 5.0. Five soil physiochemical properties were analyzed with standard soil analytical methods. Classical statistics, geostatistic techniques, semivariogram model fitting and mapping methods were used to describe the spatial distribution of soil properties at both scales. The results indicated that there were medium variations for five soil properties of moss crusts at both scales. Coefficient variations of five properties at quadrate scale varied from 12% to 30%. With the exception of total phosphorus, most coefficient valuations of the other four properties at patch scale were less than 20%, which indicated the soil spatial heterogeneity of moss crust at quadrate scale is much stronger than that at patch scale. The spatial autocorrelation of five soil properties at quadrate scale ranged from 65% to 92%. For patch scale, except for total phosphorous, spatial autocorrelations of the other four parameters were over 90%, which means, that the proportion of spatial autocorrelations of moss crust increase with the decreasing of sampling scale. Additionally, the effect of structure factors increased on patch scale with the decreasing proportion of the random factors, which was consistent with the result of spatial autocorrelation. In this study, values of soil water, organic matter and total nitrogen decreased from center of moss patch to edge, however, total phosphorus increased from center to edge at patch scale, which indicated that there was heterogeneity for moss crust patch at micro scale. There was irregular change for total soluble salt at patch scale, but at quadrate scale, total soluble salt content with mosses was much higher than that without mosses, which implied that with the decreasing of sampling scale, the heterogeneity of soil salt disappeared. The autocorrelation ranges of water, organic, total nitrogen and total phosphorus at patch scale were smaller than the diameter of the moss crusts patch, autocorrelation range of total soluble salt was out of patch size, on the whole, autocorrelation ranges in creased with the sampling size. Thus, sampling distance can be increased at quadrate scale, but for patch scale, sampling distance should be reduced to avoid missing the tiny heterogeneity. At both scales, the lowest autocorrelation range of total phosphorus was observed. Therefore, the sampling distance must be short for variation of phosphorus, while for total salt, sampling distance can be increased properly due to its big variation range.