Experimental Design

Ten 50m x50 m sampling areas (plots) were established in Terceira Island within pristine native forest. Plots locations were chosen based on accessible forest habitat that was deemed to be relatively undisturbed in terms of human impact and invasive plant species.  Sampling for all methods, with the exception of leaf litter, followed the semi-quantitative sampling using an adaptation of COBRA protocol.

In total the sampling protocol requires 16 hours of physical searching and collecting, approximately 7 hours for the placement of pitfall traps, approximately 5 hours for the removal of pitfall traps, and approximately 1 hour for the collection of leaf litter filtrate, representing a total of 29 person hours per plot.

Active aerial searching during the night (AAS):  Hand collection assisted with pooter, vial, small brush and forceps, from knee level to as high as the collector can reach. Using a high intensity head lamp (at least 100 lumen). 4 hours of searching, each hour is counted as 1 sample.

Searching under bark, lichens and mosses during the day (BLM): A small hoe is used to dislodge bark, mosses and lichens.  Material is dislodged onto a white beating tray to separate arthropods for collection.  2 hours of searching, each hour is counted as 1 sample.

Searching in decaying trunks, dead wood on the ground, and under stones during the day (TWS): A small hoe is used to break open decaying wood, using 1 m2 white sheet to collect debri from decaying wood to separate arthropods for collection.  The sampling region is from ground level to knee height.  2 hours of searching, each hour is counted as 1 sample.

Foliage beating during the day (FBD): A 110 x 80 cm while sheet, supported by a frame, should be used as a drop-cloth (beating tray).  A wooden pole of about 1 m is used to beat tree branches.  2 hours of searching, each hour is counted as 1 sample.

Foliage beating during the night (FBN): As for FBD, but with the assistance of a head lamp. 2 hours of searching, each hour is counted as 1 sample.

Foliage sweeping during the day (FSD): A round sweep net with an opening diameter of 30 to 50 cm is used to sweep bushes and herbs from waist height to ground level. 2 hours of searching, each hour is counted as 1 sample.

Foliage sweeping during the night (FSN): As for FSD, but with the assistance of a head lamp. 2 hours of searching, each hour is counted as 1 sample.

Pitfall traps (PIT): Fortyeight pitfall traps are placed immediately outside the perimeter of a sampling plot, 12 along each edge of the plot, spaced equidistantly.  Pitfall traps should be approximately 8 cm wide at the top, and approximately 12 cm deep (European standard 33cl cps).  Traps are filled with 3 - 4 cm of 100% propylene glycol and left in the field for 14 days.  Traps should be protected from predation, inundation with rain water, and unwanted vertebrate capture (i.e. reptiles). See Supplementary Figure 1 for a suggested trap cover.  A group of 4 contiguous traps is counted as 1 sample.

Leaf litter sampling (LLS): A winkler apparatus with an opening of about 30 cm is used to sieve litter in the field. The minimum area to be sampled is two square metres of forest floor, to be sampled as eight 0.5 x 0.5 m quadrats.  All leaf litter within a 0.5m x 0.5m quadrat is collected and shaken vigorously in the winkler apparatus to collect the filtrate in the bottom of the winkler device, which is then transferred to sealable plastic bags for transport to the lab.  Sampling a total of two square metres of very dense leaf litter generates sieved material that will fill four light extraction boxes (see Supplementary Figure 2), to a height of approximately 2cm. In areas with a lesser depth of leaf litter, a larger total area should be sampled, such that four light extraction boxes can be filled.  This sampling facilitates comparison across plots by using either the unit area of leaf litter sampled, or the total volume of sieved material.  Sieved material is distributed across four extraction boxes with a mesh size of 5mm x 5mm (see Supplementary Figure 2 for details).  Specimens are collected into propylene glycol using a lamp to generate a temperature gradient for extraction.  The filtrate from leaf litter is 1 sample.

ADDITIONAL SAMPLING

SLAM Traps - The capturing system used was a passive flight interception trap called SLAM. These are about 1,1x1,1x1,1m, and once the animal is intercepted by hitting the black mesh, if it crawls up the net will fall into the sampling bottle, filled with propylene-glycol.