Exploring visual soil evaluation and examination methods on highly-weathered tropical soil

Abstract

National and international initiatives have been undertaken in Uganda to improve soil quality and increase crop production. However, means to evaluate and examine soil quality, particularly soil physical quality, is lacking in the country. In this study, visual soil evaluation and examination (VSEE) spade and core tests, which comprise rapid and simple methods to semi-quantitatively assess soil structure, have been tested. The derived soil quality scores Sq were compared with soil quality indicators SQi derived from traditional lab-based methods of soil structure analysis. Tests were conducted and samples taken in Uganda on highly-weathered soils with sandy clay loam texture. Both the 0–15 cm topsoil and the 15–30 cm moderately compacted subsoil were considered. Test and sampling sites comprised 18 farmers’ fields (maize, Zea mays L.) that were under conventional tillage, permanent planting basins and rip lines for three years, as well as four locations in a natural forest. All VSEE approaches tested showed a significantly better Sq score in the natural forest (good quality) as compared to maize fields (fair/moderate quality), with the subsoil always showing lower quality than the topsoil. Methods based on Visual Evaluation of Soil Structure (VESS) were more responsive to differences in soil quality than the Visual Soil Assessment (VSA) approach. Statistical analysis showed that there was a good to moderate correlation between the VSEE-based Sq scores and lab-derived SQi values, with Pearson r correlation coefficients of 0.52–0.69 for bulk density, 0.66-0.78 for air capacity, 0.53–0.73 for air permeability, 0.52–0.72 for hydraulic conductivity, 0.18-0.48 for mean weight diameter under fast wetting. The correlation with an overall integrated index of soil quality SQI ranged between 0.56 and 0.77. Minimizing the potential effect of local variability by averaging Sq scores and SQi or SQI values per treatment and depth, improved the correlation, with e.g., Pearson r ranging from 0.84 to 0.95 when relating Sq to SQI. We also found a significant correlation between VESS Sq scores and the shape of the water retention curve, particularly in the wet range (r > 0.50). Our results show that in general, VSEE methods are promising alternatives to evaluate differences in physical soil quality of highly-weathered soils in a rapid, intuitive, practical and cheap way.