Materials and Methods
The experiment was run for 2 weeks from 13-26 May 2019. 120 artificial birds nests made from conifer leaves and metal wire were used in this experiment. 30 trees were selected at the forest edge and interior. Trees at the forest edge were selected using a linear transect across the forest edge, with trees selected at 10m intervals. Due to the size of the fragment, a linear transect was unable to be used for the forest interior. After arbitrarily choosing an initial tree in the centre of the interior, trees were chosen using a random number generator, which generated a bearing and a tree in that bearing was chosen if i) it was not within 5m of any other tree and ii) if it was still within the forest interior. Each tree contained 2 nests, 1 ground and 1 arboreal. In total, 60 nests were placed at the forest edge (<10m away from the edge) and remaining 60 were placed at the forest interior (>100m away from the edge). Where necessary, arboreal nests were attached to trees using steel wire. Each tree was a minimum of 5m away from any other selected tree, and no attempts were made to disguise or hide the nests. Since nest markers may cue predators to the presence of the nest, no nest markers were used.
3 eggs were placed in each nest, 1 quail egg and 2 plasticine eggs. The pilot study indicated a high risk of artificial eggs being removed or destroyed, for this reason 2 plasticine eggs were used instead of 1 (S1). Plasticine eggs were made by hand to have the dimensions of a reference quail egg (3.5cm long, 2.8cm diameter), and were aerated for 3 days after manufacture to mitigate the strong scent. To reduce the scent further, plasticine eggs were coated with a thin layer of rubber (PlastiDip®) before aeration, which has been shown to reduce overestimation of mammalian predation rates \cite{Purger_2012}. Eggs were marked on their base to identify the nest they were placed in, to assists in cases where an egg was removed from the nest but able to be recovered. A nest was considered to be predated if any of the eggs were damaged or missing. Nests were examined daily and predated eggs were replaced. Predator tooth marks in the plasticine eggs were identified using reference photos of woodland predators skulls. For spatial analysis, nest locations were recorded using a Garmin eTrex 10, and co-ordinates were averaged from 5 measurements over the course of 1 day for increased accuracy.