Managing soil to support biodiversity is important to sustain the ecosystem services provided by soils upon which society depends. There is increasing evidence that functional diversity of soil biota is important for ecosystem services, and has been degraded by intensive agriculture. Importantly, the spatial distribution of reservoirs of soil biota in and surrounding arable fields is poorly understood. In a field experiment, grass-clover ley strips were introduced into four arable fields which had been under continuous intensive/conventional arable rotation for more than 10 years. Earthworm communities in arable fields and newly established grass-clover leys, as well as field boundary land uses (hedgerows and grassy field margins), were monitored over 2 years after arable-to-ley conversions. Within 2 years, earthworm abundance in new leys was 732 ± 244 earthworms m-2, similar to that in field margin soils (619 ± 355 earthworms m-2 yr-1) and four times higher than in adjacent arable soil (185 ± 132 earthworms m-2). Relative to the arable soils, earthworm abundance under the new leys showed changes in community composition, structure and functional group, which were particularly associated with an increase in anecic earthworms; thus new leys became more similar to grassy field margins. Earthworm abundance was similar in new leys that were either connected to biodiversity reservoirs i.e. field margins and hedgerows, or not (installed earthworm barriers). This suggests that, for earthworm communities in typical arable fields, biodiversity reservoirs in adjacent field margins and hedgerows may not be critical for earthworm populations to increase. We conclude that the increase in earthworm abundance in the new leys observed over 2 years was driven by recruitment from the existing residual population in arable soils. Therefore, arable soils are also potential reservoirs of biodiversity.