Study site The study took place in the commune of Mauguio near Montpellier (34), which has a Mediterranean climate with hot, dry summers. The average annual temperature in Mauguio is 14.8°C and the average annual rainfall is 607 mm (Climatik, INRAE). The DIAMs (Dispositif Instrumenté en Agroforesterie Méditerranéenne sous contrainte hydrique) experimental site was set up in 2017 on an INRAE experimental station Diascope. It consists of three land uses:

• crops: cereals or legumes ;

• agroforestry: linear lines of black locust (Robinia pseudoacacia, Linnaeus 1753) planted in 2017 and 2018, separating alleys cultivated with cereals or legumes similar to those in the crop plots;

• black locust (R. pseudoacacia) tree plantation.

Farming practices are considered to be conventional but low-input, and are identical in the cultivated areas of the agroforestry plots and the field crop plots, both in terms of tillage and inputs (fertilisation, crop protection). In the forest plots herbaceous plants grew spontaneously following the initial sowing of an alfalfa cover in 2015.

Samplings and identification At the end of winter (February-March), systematic sampling was carried out within each bloc in 5 habitat types corresponding to the 3 land uses:

• tree plantations,

• annual crops,

• agroforestry, subdivided into 3 habitats: the tree line, 1 metre from the tree line, and the middle of the cultivated alley.

For each modality, in each of the 3 blocs, 10 locations were studied, for a total of 150 sampling points. In addition, 18 points (2 transects of 3 points per block) were added in the treeless grass strips.

A monolith of soil measuring 25 x 25 x 20 cm deep was extracted, then sorted manually in the field. Once the individuals had been washed in water, they were placed in 99% alcohol to fix them. Finally, they were counted and identified in the laboratory under a binocular. Species identification was carried out using the interactive key developed by the UMR Eco&Sols for adult individuals, i.e. those in which the clitellum was visible. If the clitellum was not visible, the individuals were considered to be juveniles.