Data and code from Local landscape elements enhance flying insect biomass in a complex agricultural landscape

We collected insect samples from 24 agricultural field edges of which 9 had a hedgerow, 8 a flower strip, while 7 field edges did not have a natural element. Flying insects were caught using Malaise traps (British Museum type, SKU: 10.501). The highest point of these traps was 1.75m at the front, sloping down to 0.95m at the back. The width was 1.12m and depth was 1.70m. Collection jars were filled with 96% ethanol as a preservative fluid. The research was carried out over three consecutive years (2021 – 2023). Each of the 24 locations was sampled 8 times, with one exception being sampled only 6 times. In 2021 locations were sampled in June, July, August and September. From 2022 onwards the sampling scheme was adjusted to sampling efforts in May, June, July and August until a total of 8 samples per location was achieved. We investigated the effects of local factors in the form of vegetation characteristics and farming practices on flying insect biomass. Vegetation height was measured directly around the Malaise traps and in the adjacent agricultural field. At the time of placement, all flowering plant species in a 10m radius of the Malaise trap were recorded. The vegetation cover and composition of the agricultural fields and flower strips was determined by vegetation surveys. The species diversity of hedgerows was captured in linear surveys (100m). Linear surveys consisted of scoring all plant species in the hedge and its undergrowth at 5-meter intervals for the 100m survey length. Ultimately, we used vegetation heights, both next to the trap and in the adjacent agricultural field, total plant diversity and diversity of non woody species as local vegetation covariates. For local management practice covariates, we consider whether the field is used for livestock or crop cultivation, fertilized with manure or chemicals, and whether pest control is applied. We retrieved this information from questionnaires filled in by all cooperating farmers and landowners. As landscape context plays an important role in determining the possibilities for insect populations to inhabit an area, we quantified landscape scale data at different scales (radii 50m, 250m and 500m) using landcover maps (Top10NL and Natuurnetwerk Nederland). While there is some overlap at larger radii, the circles around locations are largely independent at 250m with a mean of 90.9% independent area on average (sd = 14%). At 500 meters there is more overlap, with a mean independent area of 57.7% (sd = 26%). We determined percentages of arable cropland, arable grassland, forest and protected natural area. These four land cover types provide close to 100% cover of the areas surrounding the traps (only roads and settlements are excluded). To prevent overlap between land-cover types, we excluded forest patches from the protected natural area cover. Protected natural areas in this study region consist of flood plains, semi-natural grasslands and several water bodies with their associated vegetation. As a measure for landscape heterogeneity and configuration, we used edge density: landscapes with small fields and patches of natural area have a higher density of edges than landscapes with large fields. We quantified edge density as the total length of edges per area (m/m2) at all radii for each sampling location. Furthermore, an extensive map of all hedgerows surrounding the sampling locations (radius of 1000m) was hand drawn using a combination of satellite images and height maps in QGIS version 3.22.5. We calculated the total hedge length (m) for all radii. .