干旱区生态环境知识资源中心

Biogeochemistry of terrestrial ecosystems is controlled by interactions among factors operating at several spatial and temporal scales. The purpose of this study was to evaluate the relative importance and interaction of relatively static landscape factors and more dynamic factors in a shortgrass steppe landscape. The landscape factors examined were topographic position, and soil texture. The dynamic factors studied were seasonal climate and the localized effects of individual plants on soils. Patterns were evaluated by sampling soil between and under individual Bouteloua gracilis plants in paired upland (erosional) and lowland (depositional) plots at eight locations at the Central Plains Experimental Range (CPER), Colorado. We quantified five organic C and N pools (total, fine and coarse particulate organic matter [POM], mineral-associated organic matter [MAOM], and potentially mineralizable C and N), and we estimated seasonal patterns of in situ N dynamics with three methods (extractable inorganic N, net N mineralization in uncovered cores, and N adsorbed on ion exchange resin [IER] bags).

Topographic position and soil texture each explained much of the landscape-scale variation of C and N pools and vegetation structure. Most lowland plots were enriched in silt, clay, C, and N relative to adjacent upland plots, and topographic position affected most pools significantly. Most vegetation and biogeochemical variables were strongly correlated with soil sand content. Across the range of sand content encountered (40-83%), the fraction of area in bare soil openings >5 cm across increased sevenfold, and most C and N fractions increased by 2-4.5 times. Plant-induced, microscale heterogeneity of soil C and N was comparable in magnitude to landscape-scale heterogeneity for pools with more rapid turnover (POM and mineralizable C and N), but presence or absence of plants did not affect more stable, mineral-associated organic matter. Plant-induced heterogeneity was significant in all locations, but its importance likely decreases with decreasing sand content as cover becomes more continuous, particularly in lowlands.

Total extractable inorganic N and nitrate, N adsorbed on resin bags, and the proportion of mineralized N that was nitrified during incubations increased with increasing soil water content or precipitation, but net N mineralization did not vary systematically with precipitation. Inorganic N availability was greatest during relatively moist spring periods, and these were the only times when indices of in situ N availability followed the spatial patterns expected from laboratory assays of C and N pool distribution.

The relatively weak spatial patterns for N dynamics contrast with the substantial landscape and individual-plant scale variation in C and N pools. Landscape patterns of N mineralization and availability may be tied to N pools, such as POM N, that are not strongly related to topography or texture. Particulate organic matter appears to be especially important to N retention and availability on sandy soils, which typify most sites at the CPER; the proportion of total N residing in POM was high and POM C:N ratios were low in sandy upland soils.

We suggest that soil texture is a key proximal control over biogeochemical processes and is largely responsible for observed landscape-scale patterns, including topographic differences. Models that integrate effects of texture and topography have the potential to link landscape biogeochemical patterns to short-term processes that directly influence SOM dynamics and to long-term geomorphic processes that influence soil distribution, including eolian redistribution of soil materials, which is important in many dry regions.