A prototype tracing-technique to assess the mobility of dispersed earthworm casts on a cultivated hillslope soil under unconsolidated and crusted surface conditions

Abstract

Whilst evidence has been found that dispersed earthworm casts can act as a sediment-source on fully vegetated hillslopes, uncertainty still remains over the mobility, and hence, the ability of such material to similarly act as a sediment-source on cultivated hillslopes. This study reports preliminary findings from a prototype sediment tracing technique to assess the mobility of eroded earthworm casts on the same unconsolidated (i.e. uncrusted) and crusted hillslope soil using two different artificial radionuclides. A 0.75 * 0.75 * 0.1 m erosion plot, pitched at a 10% slope (5.7°) was filled with bulk soil and textured to simulate an agricultural-type seedbed finish. A total of 56 similarly-sized casts representing 197 g of sediment (dry weight) were labelled with 430 Bq of caesium-134 activity, evenly distributed across the unconsolidated soil and subjected to ca. 15 mm hr-1 of simulated rain until casts disintegrated, during which time surface runoff was collected continuously. Soil was air-dried to allow a surface crust to form and then a further 56 similarly-sized casts representing 198.5 g of sediment (dry weight) were labelled with 433 Bq of cobalt-60 activity, evenly distributed across the crusted surface and again, subjected to simulated rainfall until their complete disintegration. Surface runoff was continuously collected. An unmixing model was used to partition labelled sediment from unlabelled bulk soil. For the 1st rainfall event, 20.6 g of 134Cs-labelled material, representing 10.5% of the total mass deployed, was removed by surface runoff from the unconsolidated surface. During the 2nd rainfall event, a further 16.2 g of 134Cs-labelled sediment (8.2%) was removed from the crusted surface, along with 73.1 g (36.8%) of 60Co-labelled sediment. In comparison with 134Cs-labelled material, approximately double the amount of 60Co-labelled sediment was eroded from the crusted soil in approximately half the rainfall exposure time. These findings, although very preliminary, firstly provide proof-of-concept confirming the efficacy of the tracing technique and secondly, suggest that cast-derived material is not only mobile on cultivated hillslopes, but the degree of mobility and hence the source-potential of the eroded material increase markedly when soil surface conditions change from unconsolidated to crusted.