Data from: Agroecological transition increases arthropod diversity and decreases herbivore abundance on field margins
Field margin arthropods and vegetation data from peri-urban horticultural farms near Córdoba city (31.30- 31.50ºS and 64- 64.20ºW), in central Argentina. At each site, arthropods were sampled on the field margins using a 1m long vertical beat-sheet. Field margin was considered as the non-crop area adjacent to the vegetable crops, with at least some presence of herbaceous vegetation and a width of approximately 1m. Random samples were collected on each of the four sides of the field. Each sample consisted of two adjacent sub-samples that were obtained by beating the vegetation against the sheet a fixed number of times (3), and then collected from its base with a modified hand vacuum. Afterwards, samples were taken to the laboratory and preserved in 70% ethanol for further processing. Collected arthropods were classified into different orders and families using dichotomous keys. Adults were classified into morphospecies (further referred to simply as ‘species’). Specimens were identified at the family level, assigned to feeding guilds according to the dominant family habits, and for families with multiple feeding habits, they were identified at the subfamily level. On this basis, herbivores and natural enemies (predators and parasitoids) were considered for further analysis. For each sample, richness and abundance were determined by guild and taxonomic order, and biodiversity was estimated using the Shannon-Wiener index. Immature insects were considered only in terms of their abundance. For vegetation sampling, five 1x1 m quadrats were placed along each field margin with a minimum distance of 20m. Within each quadrat, maximum vegetation height was measured, and vertical photographs were taken and later processed in the software ImageJ to determine plant cover percentage. Alongside the quadrats, 6x1m transects were placed and used to determine floral resource abundance. All plant species in both squares and transects were accounted to determine total vegetation species richness. Due to logistic impediments and to avoid the noise of seasonal variation, these measurements were done once, between arthropod sampling dates, and in 7 out of the 10 fields.