2.1 Study areas
For this study, water bodies surrounded with different proportions of forest versus agricultural land were selected: different landscapes to cover the whole gradient from fully forested to mainly agricultural were chosen. Landscape might affect not only the lake trophic status, but also the nest predator community, as stated by Holopainen et al. (2020a). Studies were conducted at lakes and ponds at two areas in Finland, Evo and Maaninka (Figure 1). Both areas have permanent lakes, which carry water through the summer. Lake shorelines may be affected by spring floods, but otherwise the water level is rather constant. The trophic level among these lakes vary from oligotrophic to eutrophic (see Holopainen & Lehikoinen, 2021). In addition to permanent lakes, in both areas the study included other wetlands, which were shallow and had temporally varying shorelines: seasonal ponds, beaver ponds and man-made ponds (hereafter ponds). Hereafter all lakes and ponds are called water bodies. The same mammalian predators and all common corvid species occur in both study areas (Lindén et al., 1996; Valkama et al., 2011).
Evo in southern Finland (61°12’N, 25°07’E) represents a typical boreal forest landscape in Finland. Human settlements are scarce, with few agricultural fields (hereafter fields) inside the study area and larger agricultural lands south from the studied water bodies. The lands within 1 km of the Evo water bodies consists on average 1 % of fields (range 0–6%).
In Evo duck surveys were undertaken at 45 water bodies (27 permanent lakes, 10 beaver ponds and 8 seasonal ponds) within a c. 39-km2 area. Water body size varied between 0.04–37 ha. The median shoreline length for permanent lakes was 1.0 km and for ponds 0.5 km. From the 45 study water bodies, six seasonal, six beaver ponds and 12 permanent lakes were chosen for nest predation experiments and invertebrate surveys (for more information, see Supplementary Appendix S1).
Maaninka in eastern Finland (63˚15’N, 27˚30’E) is a mosaic of agricultural land and forests with some internationally important bird-lakes (Natura 2000 and IBA -lakes; Leivo et al., 2002). The area represents typical agricultural landscape of Finnish lake district. Duck surveys were made at 18 lakes and seasonal ponds across c. 47 km2: 17 water bodies were used for the experiments, among them the important bird-lakes. All seasonal ponds were included that occurred during the study years and to which permission was granted from the landowners. The lands within 1 km of the water bodies consists on average 59% of fields (range 24–75%), the remainder mainly private forest. Water body sizes were 1–149 ha. The median shoreline length for permanent lakes was 3.0 km and for ponds 1.1 km. In total there were seven permanent lakes, two man-made ponds and eight seasonal ponds for nest experiments and invertebrate surveys.
To estimate the landscape structure (i.e. the field percentage) near the study water bodies, QGIS 2.18.7 (QGIS Development Team, 2017) and topographic vector map (National Land Survey of Finland, 03/2019) were used. The 1 km radius zones around the water bodies were measured from the shoreline. We used the 1km radius, as in the comparisons of nest predation patterns between 500 m and 250 m radius zones, the 250 m zone has shown to have habitat specific effects, while the effects disappear with a larger zone (Uusihakala, 2021). Thus we believe that 1 km radius zone describes the landscape scale foraging range of mammals. Waters from the zones were excluded in order to count the field percentage of surrounding land areas.