2.2 Beetle sampling and taxonomic assignment
We explored beetle communities across the country in the two distinct habitat types, applying different trapping methods (pitfall traps, window traps, light trapping, and hand collection, in order to capture different aspects of community composition) throughout four consecutive years (2018-19-20-21, depending on sampling site). Pitfall traps consisted of plastic cups (300 cm3) with a diameter of 8 cm at the top, buried in the ground so that the top rim was flush with the soil surface. The window traps were made of 1.5-liter transparent plastic drinking water bottles with one window and were suspended upside-down. The traps were hung on wooden poles at approximately 150 cm above ground level. Both trap types were filled with about 50 ml of 70% alcohol and had a cover for rain protection during the collections carried out in the rainy season. Traps were left five days in the field, then all captured insects were brought back to the laboratory, where beetles were separated from other species and debris and sorted into families under a microscope. Light trapping consisted of a 125v bulb and white clothing of 4.5 m2 hung up between trees. All beetle specimens were directly stored in 95% alcohol in the tube after sorting. Furthermore, hand collection was done by the first author during visits to the field. All beetle specimens were assigned to the family level, like several other studies which have reported the use of higher insect taxa (family level) to investigate habitats and environmental scenarios (Parikh et al., 2021; González et al., 2015; Báldi, 2003). We then described diversity and community composition across landscape contexts based on the identity and abundance of beetle families.
Some sites were sampled each month for several years using different trapping methods and many traps, whereas in some sampling sites collection occurred only occasionally during a few days in a single year using one or a couple of complementary trapping methods, resulting in a small number of traps. For example, four sampling sites were equipped with pitfall (2 × 30) and window (2 × 30) traps in 2020. Then, in these sites, sampling occurred each month of the year for 5 days, resulting in a particularly large sampling effort (60 traps × 5 days × 12 months). In contrast, a single sampling site was equipped with 15 window traps for nine days in 2018. Six sites were sampled by hand collection during a few days in 2019 and 2021. Finally, light trapping was used for three to five days in four sampling sites in 2019 and 2021 (Table 1).