The most commonly used method for sampling epigeal ground fauna is pitfall trapping (e.g. Chiverton and Sotherton, 1991; de Snoo, 1999; Schmidt et al., 2005; Jimenez-Valverde and Lobo, 2007) (see Figure 3.1). Although there is literature highlighting the limitations of this method, e.g. digging in effects; body size of species attracted; taxa attracted to the trap, ( Luff, 1975; Nilsson et al., 1988; Jarosik, 1992; Duelli and Obrist, 1998; Raworth and Choi, 2001) it remains the most cost effective way and popular way of sampling. Pitfall trapping does not sample all epigeal taxa equally; rather, it is more a measure of relative activity and relative abundance, hereafter referred to as abundance in this study. Many of the limitations highlighted in the literature can be minimized by applying a few simple steps in the methodology, as discussed below.
Figure 3.1 Example of a pitfall trap in situ. (Source: Cary Institute of
3.7.2 Timing of trapping
As it would be unfeasible to sample continuously throughout the summer due to gaining access to the plots issues and the vast number of individual organisms that would be collected, a compromise had to be made. The Isopoda and Myriapoda are most active in spring and autumn (Richards, 1995) whilst many spiders become most conspicuous by May (Roberts, 1996). Den Boer and Den Boer-Daanje (1990) had long term pitfall data sets (>10 years) from which it could be deduced that sampling in May and September in the current study would be expected to result in the capture of all the beetle families likely to be encountered. It was therefore decided that sampling in May and September would be the optimum times to set the pitfall traps on the allotment plots on the site discussed in Section 3.4, and that these data would be pooled as a
representative sample of the ground fauna of the sites in subsequent analyses.
3.7.3 Trap position
Each of the forty-two allotment plots was mapped using an appropriate scale on graph paper, split into squares representing 2 x 2m, usually equating to around 1-30 squares, depending on plot size. A random number was generated from this range for each plot. This number then became the starting point for the location of the pitfall traps. The traps were set 2m apart in a 2 x 3 array. If the proposed location for a pitfall trap fell on a hard surface, e.g. concrete path or a shed, the nearest suitable location was chosen. It must be borne in mind that these traps were set on working allotments, belonging to different individuals on each plot, therefore disruption to their working of the plot had to be treated sensitively. In most cases, the plot-holders gave complete free reign as to where the pitfall traps could be set and as the chosen random square was of a
reasonable size, there was some flexibility in how close they were to existing crops. Only in one case did the plot-holder ask for the traps to be set in a section other than the one allocated by the random number technique.
3.7.4 Trapping technique
In May 2006, each plot had six plastic vending cups (max. diameter 7.3 cm x 9.6 cm depth) sunk to ground level over a number of days prior to making them active. The cups had been dug in over the previous few days to try and avoid the effects of ‘digging in’ (Greenslade, 1973) i.e. some species may initially be attracted to the trap because the site has recently been disturbed. Lids were placed over the cups to prevent rain getting in and to deter larger animals such as toads from falling in and to lessen the chances of the contents being drunk by larger mammals such as dogs. The lids were made from Petri dishes with short lengths of pea canes super-glued on to form a small supporting tripod (see Appendix 3.3). As traps were set, these lids were pushed into the ground so that they remained closed until needed.
On the 1st May, 2006 the traps were set as ‘live’ with c.100 ml of propylene glycol at 20% solution, with 1% detergent and salt added to break the water tension and break down slug/snail slime, respectively. Propylene glycol was deemed to be the least environmentally harmful chemical that could realistically be used, which was important as half of the plots to be sampled were managed organically.
The live traps were left in situ for a week, before removing them on the 8th May 2006 to collect their contents. The contents of each trap were then rinsed with
water and put into 4% formalin for four days to ‘fix’ the samples. Fixed samples were then stored in 70% IMS prior to identification. The process was repeated in September 2006, giving a total of 528 samples. Due to the large volume of invertebrates collected, only three samples per plot were analyzed, whilst the other half formed an “insurance policy” in the event of vandalism or other disturbance to the pots. Due to some vandalism and suspected tampering with some of the pitfall traps, data from ten plots could not be used. In the BUGS studies (e.g. Smith et al., 2006a,b.), only three pitfall traps were used per garden and the data were pooled (ave. garden size 79.5 + 81.5m2; Loram et al., 2007), therefore three pots were deemed to be an acceptable sample size in this study, pooled per allotment plot.
All collected organisms were removed from the pots and recorded to morpho- species level. The target taxa that were examined were: Arachnida (Araneae); Isopoda (Oniscidea); Chilopoda; Diplopoda; Coleoptera (Carabidae) and
Gastropoda (the same taxa that were studies in the BUGS project (Smith et al., 2006 a,b.)). Some taxa found were relatively rare, e.g. bugs, bees, worms and pseudoscorpions. These species are not traditionally caught using pitfall traps (e.g. see Paoletti, 1999; Standen, 2000; Eremeeva and Sushchev, 2005; Smith et al., 2006a), so they may be seen as ‘incidental by-catch’, and were therefore not considered any further in this study. Similarly, although there were quite high numbers of Diptera, again they are not usually counted as epigeal and would therefore not usually be sampled by pitfall traps (e.g. Wheater and Cook, 2003). A few pots contained high numbers of ants, but these were not included in the analysis, as ants tend to follow each other’s trails, so if a pot initially contained a few ants, it is likely that many more would follow them in, inflating
the numbers, creating bias (Davis and Utrop, 2010). Again, they would not normally be sampled using only pitfall traps (e.g. York, 2000).
Collembola made up c.23% of the catch; however, this number only represents the number of individuals remaining after the fixing process. Some individuals were small enough to pass through the sieving process, therefore final numbers recorded do not constitute a true reflection on the total numbers present,
therefore the data on Collembola will not be considered in any further detail.
The specimens were identified using stereo-microscopes, a camera microscope and the following keys:
• Arachnida: Roberts, M.J. (1996) Spiders of Britain and Europe, Collins, London.
• Isopoda: Hopkin, S. (1991) A Key to the Woodlice of Britain and Ireland, FSC, Preston Montford.
• Myriapoda – Lee, P. (2005) Provisional Keys to British Millipedes v. 2 (unpublished draft); Barber, A.D. (2003) A guide to the identification of British centipedes: Aidgap Test Version 2003, FSC, Preston Montford (unpublished draft).
• Coleoptera: Luff, M.L. (2007) The Carabidae (ground beetles) of Britain and Ireland. RES Handbooks, Vol. 4, Part 2 (2nd Ed.) Field Studies Council, Shrewsbury; Forsythe, T.G. (2000) Ground Beetles, Naturalists’ Handbooks 8, Richmond Publishing Co. Ltd., Slough; Unwin, D.M. (1988) A key to the families of British beetles. Field Studies Council, Shrewsbury.
• Mollusca – Cameron, R. (2003) Land Snails in the British Isles, Aidgap and FSC, Preston Montford; Kerney, M. and Cameron R.A.D. (1979) A Field Guide to the Land Snails of Britain and North-west Europe, Collins, London; Sankey, J. (1987) How to begin the study of Slugs and Snails, Richmond Publishing Co., Richmond.
For the first stage of the data analysis, taxa were identified mainly to Order, except in the case of the centipedes, millipedes or spiders, (identified to Class). Hereafter, these groups of species will collectively be generally referred to as ‘taxonomic richness’ or simply ‘taxa’ or ‘morphospecies’ where appropriate. (Subsequent analysis in Chapter 5 will involve identification of the taxa down to Genus or species where possible.)