Widespread declines in farmland biodiversity have led to state-funded schemes which take land out of production to create (semi-)natural habitats for biodiversity (e.g. EU agri-environment schemes; US Conservation Reserve Program). Common features of such schemes are grassland strips at the edges of agricultural fields, and we examine potential co-benefits of these biodiversity set-asides for contributing to grassland connectivity. Although set-aside strips had negligible impact on landscape-scale species persistence (using metapopulation models parameterized for flying insects run on 267 landscapes of ~30 000 ha across England), they nonetheless improved connectivity in 74% (198/267) of landscapes (comparing landscapes with and without set-asides), as shown by range expansion rates increasing by up to 100%. Benefits of set-aside strips varied according to species type (high/low dispersal, high/low population density), but had little benefit for species with low dispersal and small population sizes, which  enerally failed to expand. High dispersal/high density species were already successful expanders regardless of set-asides (> 75% of simulations were successful without set-sides) although expansion rates were still improved when set-asides were added. Whilst alternative strategies for placement of set-aside strips (more/less aggregated), revealed no consensus ‘better’ strategy across species types, set-aside benefits were generally greatest in landscapes with intermediate availability of semi-natural grassland (0.5-4% over). We conclude that small-scale set-asides have the potential to improve connectivity, which we expect to help some species track climate change, and connect habitat patches within existing climate space for others. However, set-asides are unlikely to benefit low dispersal species which are probably at greatest risk from agricultural intensification.

Habitat layers used in this analyses were derived from semi-natural grassland cells from the 25 m raster of the 2015 UK Land Cover Map (available from UKCEH https://www.ceh.ac.uk/services/land-cover-map-2015) and option points from the Environmental Stewardship Scheme published by Natural England (available from data.gov https://data.gov.uk/dataset/6c0f19e7-9a2d-4c50-b548-3b7d4b9c18bb/environmental-stewardship-scheme-options-england). These layers were combined and used to produce raster layers of habitat area at 500 m resolution.

Models were run within 267 20 km diameter, non-overlapping circular landscapes arranged on a regular grid (included).

The construction of alternative habitat arrangement scenarios is outlined in the original paper.

IFM_results.zip

Zip file containing (1) IFM_results.csv, the outcomes of IFM simulations for each landscape/scenario/species type combination and (2) README.txt explaining column headings

MPC_results.zip

Zip file containing (1) MPC_results.csv, the outcomes of metapopulation capacity calculations for each landscape/scenario/species type combination and (2) README.txt explaining column headings

landscapes.zip

Zip file containing the ESRI shapefile of landscapes as used in this analysis

R code for both metapopulation models is included within the Supplementary Materials of the original paper.