Sampling was carried out during May 2013 over four sampling periods, each a week apart, with a minimum of six replicate samples for each microhabitat. Each replicate consisted of a 1m2square which for the deadwood microhabitats included a stump or log. All the stumps sampled were from trees felled in the 1950s so were of uniform size and decay stage, logs were of a similar decay state and size (0.5-1.0 m length). Effort was to vary the time of day and month microhabitats to minimise variability due to weather differences and temporal effects. The methods followed previous work by the Soil Biodiversity Group with the exclusion of pitfall trapping, since that sampling method is strongly biased towards groups which actively move on the surface and is therefore unsuitable for a small spatial scale project, as it is likely to record arthropods originating from outside the microhabitat.
Soil moisture, temperature and pH were measured, probes being placed in the centre of the 1m2 square or in the case of logs and stumps as close to the wood as possible. Some soil temperature and pH measurements were not possible due to equipment failure, in these cases the mean value for that microhabitat was used in analysis.
This was carried out based on the 1m2square, a record was made of each plant species present at ground level, the understory (up to 2 m height) and canopy (above 2 m height). Species level identification, including grasses, sedges and conspicuous mosses, was possible for most samples.
For each microhabitat with the exception of seeps (were litter was absent) and logs and stumps, leaf litter and the loose uppermost topsoil from the 1m2 square was sieved using a 1cm2 mesh and hung in Winkler bags for three days. The method was modified for logs and stumps with litter being sieved from on, underneath and the area immediately adjacent to the wood to limit sampling to macrofauna living in close proximity to the wood rather than the surrounding leaf litter. Extracted invertebrates were preserved in 80% Industrial Methylated Spirits (IMS).
With the exception of stumps, where digging a pit was impossible, a 25 x 25 x 10 cm deep soil pit was dug in the centre of the 1m2 quadrat and invertebrates hand sorted into 80% IMS. For logs the pit was dug directly underneath after it had been rolled over and the leaf litter had been sieved.
Adult specimens were counted and identified to species in seven major groups: ants, centipedes, millipedes, woodlice, earthworms and beetles. For beetles it was possible to identify true and broad-nosed weevils, ground beetles, leaf beetles, throscids and rove beetles (except subfamilies Aleocharinae and Tachyporinae). Analyses were conducted on the combined datasets rather than sub-analyses of individual groups as this provides the most complete model for biodiversity and avoids difficulties with analysing groups with only a few species.