Soil Quality Dynamics and Spatial Heterogeneity in Grasslands and Cropping Systems in Western Canada

Abstract

Managing the land properly should lead us to conserve the soil which is a critical substance for sustaining our life and the global society. Quantifying the effects of land management and land-use conversion on soil physical and biological properties can aid to identify best management practices that can reverse the trend of declining soil quality. Additionally, underlying heterogeneity in soil properties can impose and amplify existing management challenges in common land use systems. Therefore, this study was conducted to identify sensitive soil quality indicators among contrasting land managements and also to characterize the spatial heterogeneities of key soil attributes using ordinary kriging (OK), regression-kriging (RK), cokriging (coK), and regression-cokriging (RcoK) geostatistical approaches. Our results demonstrated an improved hierarchical fractal aggregation (Dm) in soils covered by perennial legumes and grasses (Dm = 0.97) compared to nonfractal aggregation under fallow phases (Dm = 0.99). Our results proved that complex crop rotations including perennials enhanced soil quality which was concurrently associated with higher crop production. Long-term cattle manure additions had strong positive effect on nutrient cycling, while balanced fertilization beneficially influenced soil-water relationships and physical condition. Our results suggest that the Dm value, S-index, plant available water (PAW), soil organic C (SOC) concentration, and microbial biomass C (MBC) are highly responsive soil quality indicators, and hence, useful for evaluating management options. Collectively, our results indicate that conversions from either natural forest or native grassland to cultivated lands detrimentally alter the soil structural characteristics through substantially lowering macroporosity, saturated water content, and S-index. Such land-use conversions into cultivated lands also appear to increasing MBC in soils. Furthermore, fungi and Gram-negative bacteria were found to be distinctive biomarkers in the natural forest and native grassland soils, respectively, perhaps suggesting a paradigmatic shift in microbial community composition when long-term cultivation is established. Comparison of OK, RK, coK, and RcoK approaches at our field scale revealed that the combination of kriging with certain terrain covariates [e.g., elevation and depth-to-water (DTW)] as implemented in the coK method delivers enhanced soil mapping while reducing prediction uncertainty.