Vincent de Leijster

92 Chapter 5 FNC - Cenficafé, 1994), extractable P (Bray method; Bray and Kurtz, 1945) and K (ammonium acetate method; Schollenberger and Simon, 1945) and soil organic matter (SOM; Walkley- Black method; Walkley and Black, 1933). In each 20x20 m plot we also visually assessed soil and soil surface characteristics (understory vegetation cover, litter cover, soil stability and soil surface roughness) of eight 1x1 m sub-plots. In these sub-plots we visually assessed the percentage of soil covered with herbaceous vegetation, the cover of litter and dead woody material, and we collected eight additional soil samples to examine soil stability using the slake test (Tongway and Hindley, 2004). Following the Tongway and Hindley slake test, we used the categories: (0) no coherent fragments available, (1) fragment collapses in less than 5 seconds, (2) fragment substantially collapses in 5-10 seconds, (3) surface crust remains intact with less than 50% slumping of sub-crust, and (4) whole fragment remains intact with no swelling. In each sub-plot we measured the soil surface roughness in five classes varying from less than 3 mm to more than 100 mm relief, following Tongway and Hindley (2004). 5.2.3.4 Arthropod community properties Data on butterfly (Lepidoptera) communities were collected in 2019, using the transect count method (Caldas and Robbins, 2003). Two 50 m transects in close proximity to the vegetation plots were walked twice spending 5 minutes per transect. The total length covered per plot was 200 m in 20 minutes. We chose for relatively short transects and more repetitions for practical reasons as slopes were steep (average 30% steepness). Due to the rainy season —in which data were collected— the butterfly community observations were done only without precipitation and after 10:00 AM, as early mornings were too cold and humid resulting in low butterfly activity. While walking the transect, butterflies that crossed the transect within a proximity of 3 m (left, right and above), were identified based on wing characteristics and morphology. When identification was not possible, the butterflies were netted, photographed and then released to be identified by local butterfly specialists. From this data we calculated butterfly species richness, abundance and diversity, the latter being calculated using the Shannon-Wiener diversity index. 5.2.3.5 Abiotic conditions We recorded air temperature (˚C) and relative air humidity (%) every 30 minutes while being on a farm using a thermo-hygrometer (TFA, Maxim II), usually between 10:00 AM and 3:00 PM. We recorded the elevation of each plot using a GPS device (Garmin GPS 62s) and the