LAUSR

Abstract A method for characterizing aggregate stability with repeated laser diffraction measurements was tested on soils spanning the prairie-forest ecotone in northern Minnesota, USA. These soils formed in similar parent material but display a wide range of upper horizon morphology, organic carbon content, and chemistry, allowing assessment of the method's performance over a wide range of aggregate stability and its utility in identifying factors influencing aggregate behavior. Equations representing fine material release through breakdown of two aggregate populations as first-order processes were fit to experimental data. The best-fit parameters for these equations, and an additional index of persistent water-stable aggregate content, indicated distinct differences in aggregate behavior among the major horizons of Mollisols and Alfisols. Linear models were developed to explore the relationships between these parameters and soil physicochemical characteristics for the dataset as a whole and for subsets corresponding to four zones with different vegetation history, soil orders, and major soil horizons. The relationships identified were relatively weak (R2 = 0.30 to 0.70). The best predictors for the parameters representing early disintegration of less stable aggregates were cation exchange capacity (CEC) and effective CEC (ECEC) for the whole dataset, although organic carbon and nitrogen contents also emerged as predictors for forest and Alfisol subsets. The best predictors for the index of persistent water-stable aggregate content were organic carbon content, base saturation, or exchangeable Ca/Mg ratio, depending on the particular subset and fine material size fraction used in the analysis. The relatively weak explanatory power of organic carbon content as a predictor of aggregate behavior in these experiments was somewhat surprising, given prior work on aggregate stability. Both CEC and ECEC may serve as proxies for the various combinations of organic matter and clay content that influence aggregate stability in these samples, explaining their importance as predictors. It is likely that other factors not examined in this research contributed to aggregate stability, including carbonate content, clay mineralogy, and differing frequency and types of pedoturbation under grassland and forest. The results of this study are relevant to reconstructing the development of texture-contrast profiles as forest invaded grassland over the past 4000 in the study area, as documented by paleoecological research. In particular, loss of organic matter below a thin A horizon may have facilitated initial development of an E horizon in which weak aggregation favored clay eluviation; loss of clay would then have weakened aggregate stability still further. We suggest this new method for assessing aggregate stability can also be applied to research on soil erosion and runoff potential as affected by land use and management.