Synthesis of Earthworm Influence on Soil Hydrology across Scales

Abstract:

Introduction

Earthworms influence soil hydrology through their burrowing behavior. They create soil structure (macropores and macroaggregates) leading to preferential infiltration to different soil depths. Different earthworm species show different burrowing behavior which results in a different influence on soil hydrological processes, varying from more diffuse infiltration in the top soil to rapid preferential transport to deeper layers. There is a strong variability in presence and abundance of different earthworm species at various spatial scales. Therefore in this paper we synthesise the influence of earthworms on soil hydrology from plot scale to meso-catchment scale, based on various measurement campaigns at different spatial scales.

Study areas and methods

The first study area is the Weiherbach Catchment (Plate and Zehe, 2008) in the south west of Germany (Baden-Württemberg in South-West Germany, 49°8’N, 8°44’E), which is a well studied small catchment (3.6 km²) with a rather homogeneous silty loam soil texture. The land-use in this catchment is mainly agricultural land-use, in a small scale mosaic, and tillage practices have been changing from conventional to conservational over the past 20 years.

The second study area is the Attert Catchment (Juilleret et al., 2012) on the border between Luxemburg and Belgium, which is a meso-scale catchment (310 km²) and has been a research catchment of the Gabriel Lippmann Institute in Luxembourg since 2003. Within the catchment there are diverse geological settings (schist stone, lime stone, marl stone) and different land-uses: forest, agricultural areas and grassland, resulting in a large gradient in environmental conditions.

In both study areas and at different sampling scales and resolutions we performed a combination of earthworm samplings with macropore counts at different depths in the soil and dye-tracer experiments to reveal the infiltration patterns resulting from the different earthworm species compositions and abundance. In the following the results of the various campaigns will be summarized going from the meso-catchment scale earthworm distributions and influences on hydrological processes to within the field scale patterns of abundance and infiltration.

Meso-catchment scale distributions of earthworms and macropores

At meso-catchment scale, where the environmental gradients are large, including different geologies and land-uses, there is also a large difference in habitat suitability for earthworms. At this scale the macropore flow caused by earthworms is not always the dominant flow process, but funnelled flow and fingered flow or macropore flow along roots and through cracks in the soil can be the dominant flow process. The distribution of these flow processes depends strongly on geology and land-use, with fingered flow mainly in the forests in the sandstone area, funnelled flow on the grassland or agricultural fields on the hill tops of the schist area and macropore flow along roots in the forest in the schist area.

The first results of earthworm distribution models show that, using data derived from GIS maps, the land use, slope, heat load index and topographical wetness index are the main factors explaining the spatial distribution of anecic earthworms in the Attert catchment.

Catchment scale distributions of earthworms and macropores

Distributed throughout the Weiherbach catchment we performed measurements of earthworm abundance (grouped by ecological class and life-stage) and simulated rainfall experiments (with approx. 43 mm/h dye stained water) at 16 locations. Under the rainfall simulations we excavated vertical and horizontal soil profiles to study the infiltration pattern. Furthermore we counted macropore numbers (grouped by size class, <2mm, 2- 6mm and >6 mm, and hydrological activity, i.e. blue stained or not) on three horizontal profiles at different depths (10 cm, 30 cm and 50 cm).

Spearman rank correlation coefficients for different earthworm types and macropore classes in various soil depths showed that there is a significant relationship between earthworms and macropores only after classification, with the highest correlations between macropores >6mm and anecic earthworms. Macropores numbers decrease with depth and with size (Figure 1a and b) while the effectiveness of the macropores increases with size and generally showed a strong linear trend between effective macropore numbers and total macropore numbers. The > 6 mm macopores were nearly all stained at 10 and 30 cm depth.

A model of the earthworm distribution based on 65 samplings on agricultural fields throughout the Weiherbach Catchment (Palm et al., 2013), where the range of environmental gradients is short, showed that the main predictors for earthworm abundance are soil tillage, topography, soil conditions, and biotic interactions with other the other functional earthworm types, though they differ for the different functional types. However there was a relatively large measurement-based model uncertainty in these earthworm species distribution models (Palm et al., 2013), due to the high small scale variability in earthworm abundance, which posed a problem for representative sampling of earthworm abundance at the field scale. This is difficult to avoid when sampling for species distribution models as there is a trade-off between representative sampling within the field scale and the achievable number of sampled fields.

Within field scale distributions of earthworms and macropores

At small scales there is often a very high variability in earthworm abundances (Rossi, 2003b), with a patchy pattern at field scales (Rossi, 2003a; Whalen, 2004). A detailed sampling of earthworms on four fields in the Weiherbach catchment (with 150, 75, 95 and 115 samples per field) showed that the abundance and small scale pattern of earthworms depends on the soil tillage and the surrounding land use: on two of the fields conservational tillage was practiced over the past 15 years, however, the earthworm abundance was only relatively high on one of those fields. This field had a neighboring apple orchard, resulting in a decreasing trend of epigeic earthworms from that field border into the field and a small forest patch, with a decreasing abundance of anecic earthworms into the field. On the other field with conservational tillage but without surrounding sources of earthworms the mean abundance and pattern was similar to those on fields with regular ploughing.

At a limited number of plots ( 16 on each field) we performed sprinkling experiments with dye tracer and excavated horizontal profiles at 10 cm, 30 cm and 50 cm depth on which we counted the macropores of the different size groups ( 6 mm). Within the field scale there is also a significant correlation between the endogeic and epigeic earthworms with the smaller sized macropores in 10 cm and 30 cm soil depth, while the anecic earthworms are mainly correlated with the larger sized macropores (>6 mm) in all soil layers. These correlations between earthworms of different ecological types with the macropores these earthworms are expected to form become stronger when the average abundance of earthworms surrounding the sites of macropore counts are used instead of only the local counts. Thus it seems that the amount of macropores is relatively more stable than the local amount of earthworms. For predictions of spatial distributions of macropores the high small scale spatial variability in earthworms should be filtered out and a field scale smoothed earthworm distribution pattern may be more reliable to use.As for the influence of the differences in earthworm abundance on the hydrology in the different fields, generally the amount of macropores is much higher on the field with conservational tillage, than on the field with conventional tillage, except for the large macropores (>6, Table 1). The consequence for the infiltration is that on the field with high amount of macropores a higher diffuse infiltration in the topsoil is seen in the larger blue stained area at 10 cm depth and subsequently less infiltration to deeper layers through the macropores. Therefore not only the amount of deep macropores are important but mainly also the amount of shallow macropores which influence the diffuse infiltration in the topsoil and limit the supply of water to the deeper macropores. As the macropore abundance, size and depth distribution depends on the abundance of different earthworm species this shows that not only the single species abundance but rather the composition and abundance of the different species is important to understand the influence of earthworms on soil hydrological processes.

Figure 1: Effectiveness of macropores of different size classes a) 2-6 mm and b) >6 mm in different soil layers and typical infiltration patterns on horizontal soil profiles of c) field with regular ploughing d) a field with conservational tillage

Table 1: Mean macropore numbers of the different pore sizes in different depths in the soil and fraction of blue stained area on the horizontal soil profiles, both based on 16 horizontal profiles per soil depth per field as well as mean and standard deviation of earthworms per field.

	Conventional tillage (n= 16)			Conservational tillage (n= 16)
	10 cm	30 cm	50 cm	10 cm	30 cm	50 cm
	6	6	1	20	66	37
2-6 mm	72	63	54	91	251	143
> 6 mm	11	18	36	8	15	22
All	89	87	91	119	331	202
Blue stained area (%)	0,50	0,23	0,15	0,64	0,09	0,08
Earthworms/m2	Endogeic	Anecic	Epigeic	Endogeic	Anecic	Epigeic
mean (stdev)	19,2 (14,6)	6,8 (6,6)	0,3 (1,4)	66,9 (31,6)	6,6 (8,8)	80,3 (54,2)

Juilleret, J., IIffly, J.F., Hoffmann, L. and Hissler, C., 2012. The potential of soil survey as a tool for surface geological mapping: a case study in a hydrological experimental catchment (Huewelerbach, Grand-Duchy of Luxembourg). Geologica Belgica, 15(1-2): 36-41.

Palm, J., van Schaik, N.L.M.B. and Schröder, B., 2013. Modelling distribution patterns of anecic, epigeic and endogeic earthworms at catchment-scale in agro-ecosystems. Pedobiologia, 56(1): 23-31.

Plate, E.J. and Zehe, E., 2008. Hydrologie und Stoffdynamik kleiner Einzugsgebiete, Prozesse und Modelle. Schweizerbart, Stuttgart, Germany.

Rossi, J.P., 2003a. Clusters in earthworm spatial distribution. Pedobiologia, 47: 490-496.

Rossi, J.P., 2003b. Short-range structures in earthworm spatial distribution. Pedobiologia, 47: 582-587.

Whalen, J.K., 2004. Spatial and temporal distribution of earthworm patches in corn field, hayfield and forest systems of southwestern Quebec, Canada. Applied Soil Ecology, 27(2): 143-151.