Arthropod data
Arthropods were collected with suction sampling, using a modified leaf blower (Stihl™ BG85), for 45-60 seconds. In each site, three suction samples, less than 1 m apart, were taken in the central pixel. Before each sampling, a plastic ring (Ø = 45 cm) was haphazardly placed in the central pixel to delimit the suction area. Arthropods were collected in a cotton bag mounted at the front of the suction device and this bag was emptied into a plastic container. All specimens were transferred to Eppendorf tubes filled with 75% ethanol. In the laboratory, we sorted and counted spiders and beetles. All specimens were sorted to family, and partly identified to genus or species (spiders – all families, beetles – only Staphylinidae and Carabidae) by ourselves or experts. In some cases, mainly juvenile spiders, specimens could not be identified to lower taxonomic levels. Those individuals were included in total density estimates but only in species counts if no individuals of that genus were identified to a lower taxonomic level. We thereafter rarefied species number for each site using the alpha-command in BAT (Cardoso, et al. 2015).
For further community analyses, we scored traits; main habitat preference (beetles and spiders), feeding guild (beetles) and hunting guild (spiders). Habitat preference was obtained from artfakta.se, based on expert assessment, using the approach of Åhlén et al. (2024). In short, species are scored as wetland specialists if indicated habitat preferences mainly include wetland habitat types, as terrestrial specialist is preferences mainly include terrestrial habitat types, and as habitat generalist when the preference include both wetland and terrestrial habitat types. Beetle feeding guilds (carnivore, herbivore, other feeders) were based on artfakta.se and spider hunting guilds (ambush hunter, ground hunter, orb web weaver, sheet web weaver and space web weaver) were based on Cardoso et al. (2011). For lists of traits and species classifications, see Appendix 1.