Land use and soil property effects on aggregate stability assessed by three different slaking methods

Abstract

Aggregate stability is an important structural feature of soils, since it controls surface erosion, water infiltration, plant growth and carbon stabilisation. As such, it might be considered as a potential descriptor of soil health in repeated national to continental-scale soil monitoring programmes, which is, as of now, rarely the case. This might be related to (i) the conception that it can be predicted reasonably well by standard soil parameters, and (ii) the lack of a high-throughput method. Here, we used a paired plot approach with 50 cropland and adjacent grassland field margin plots to specifically test (i) if measuring aggregate stability is added value over its mere estimation based on soil properties, and (ii) if a high throughput image recognition method can bear comparison with more classical methods. We evaluated the mean weight diameter (MWD), water stable aggregates (WSA) using classical setups, as well as the slaking index (SI) via imagine recognition. Methods were compared regarding their sensitivity to considered parameters as well as their reproducibility. Soil organic carbon (SOC) as well as aggregate stability were significantly higher under grassland than under cropland soils. Remarkably, the specific design of the study could reveal that the difference in aggregate stability between land use types was not solely affected by SOC content and quality, derived from mid-infrared spectroscopy. Quality and spatial distribution of organic matter inputs, absence of disturbance, as well as biotic parameters might all be relevant factors. Nevertheless, an important finding was that SOC quality had a higher explanatory power than SOC content alone for two out of three methods. Overall, the MWD was the most sensitive to the assessed drivers and together with the WSA the most reproducible method, with coefficients of variation below 6%. By contrast, those of the SI were as high as 192%, which hampered the detection of clear patterns along the soil property gradient and between land use types. For high-quality scientific applications, 2D image recognition cannot be recommended. Instead, we recommend the use of the MWD or WSA method for scientific purposes with a low number of technical replicates in larger-scale assessments to further unravel the importance of aggregate stability for healthy soils, and to better determine the underlying factors.