Methods
We conducted the experiment in the Maliau Basin Protected Area in a region of lowland old growth dipterocarp forest in Bornean Malaysia (the area is described in (Griffiths et al. 2018) and in (Ashtonet al. 2019). In order to test the method, we buried 60 green tea bags and 60 rooibos tea bags (of a known weight) at random within the forest at a depth of 8 cm as recommended by Keuskamp et al. (2013). We collected 20 of each type sacrifically at 7, 14 and 60 days. We weighed each of the bags after collection and noted if they had been chewed opened by macrofauna. We left them out for 60 days rather than the recommended 90 days as our preliminary work showed that by 90 days most of the bags had holes in them and many were emptied. Keuskamp et al 2031 recommended reducing incubation time in “extreme sites… with extremely high k values (e.g. sites with high temperature and precipitation like site 13 in their study – which was rainforest in Panama).
In order to examine how effective the method is in tropical rain forest, we investigated the relationship between weight loss, the time that the bags were left out, and whether the bags had holes in them (i.e. opened by macrofauna). We expressed this as a % mass loss per bag, which was logit transformed for analysis (following (Warton & Hui 2011) so that we could use standard Gaussian linear models. Using R (R core team, 2013), we also tested the following full-interaction model, using thelm command: