Worcestershire Record No. 21 April 2007 pp. 50-53

Terry Knight

Introduction
In February 2006 a survey was started to find out more about the distribution and abundance of land snails in south-east Worcestershire. Various habitats were searched and sufficient data are now available to give details of the findings from most roadside sites. These were derived from a sample taken from each of 120 sites.

Survey Area
I endeavoured to make a fairly even coverage of the parts of "Greater Worcestershire" (see Knight 2005) in OS grid squares SP03, SP13, SP14, and SP15 plus those parts south and east of the rivers Avon and Isbourne in squares SP04 and SP05 (see Figure 1). This consists of 85 whole and 71 part monads (1km by 1km squares). The area lies in the Vale of Evesham except for the south-eastern half of the tongue of land extending into Gloucestershire near Broadway. This small portion consists of the escarpment face of part of the Cotswold Hills and the plateau beyond.

The soils are generally slightly to fairly calcareous apart from two main areas in the north-west. From Bengeworth to Offenham the surface strata are formed from sand and gravel swept down from the north and deposited in river terraces. These are non-calcareous. In contrast, the soils in parts of Cleeve Prior and the Littletons are shallow, stoney and highly calcareous. These occur between the summit of Cleeve Hill and Littleton Brook where Lower Lias limestone/mudstones lie just below the surface.

The Vale was cleared of woodland and was intensively cultivated from early times. At the end of the nineteenth century and during the first half of the twentieth century a large part of the central area within square SP 04 was converted into orchards and market gardens. Very little of this remains as the industry declined from the 1960's onwards and the land reverted back to mostly arable with some pasture.

Survey Method
The majority of whole monads and part monads have at least one public road in them and one sample was taken from practically all of these monads. In addition, in a very few cases, a second sample was taken by mistake! At each sample site a convenient place was chosen to park and, from close by, the roadside verges, banks and dry ditches were searched for five minutes. In built-up areas samples were not taken where roadsides were completely surfaced over. Any obvious live snails or shells were collected but, due to practical considerations, the larger species (Helix, Cepaea and Arianta) were not collected but their presence was noted. Occasion finds of Oxyloma or Succinea were ignored because of the difficulties in identification.

The collected species were later identified, where possible, counted, and then the unoccupied shells were left for a month or two. The dried contents of the shells were then tipped out and occasionally further, smaller snails came to light. These were usually Vallonia costata, sometimes Cochlodina or Pupilla, and once, Azeca. The data includes these additional individuals. All sampling was done in March as after then the vegetation started to grow and conceal the shells.

Identifications
As collecting was done during the day in dry weather, only a few live individuals were spotted. Most individuals were empty mature shells but there were a significant number of empty juvenile shells, some of which could not be identified. Nearly all shells were in a fairly fresh condition and could be identified. Identifications were based on Kerney and Cameron (1979).

Individuals of the four larger species plus 3358 individuals of 21 other species were identified from the 120 five-minute samples. Details are given in Table 1 and Table 2 above. The minimum number of snails found at a site was one and the maximum was 98 plus two types of larger species. (The figure of 98 is getting near the most that could physically be collected in five minutes.) A map showing the number of individuals collected, from the 21 species, for each monad sampled is given in Figure 1. To enable an analysis to be carried out to ascertain if the type of road or its adjacent habitat had any significant effect on the presence of snails, each site was allocated two fairly arbitrary categories as given in Table 3. Histograms for each category were then drawn as shown in Figure 2. Table 4 lists the average percentage of the main constituent species for a selection of these categories. (Table 4 and figure 2 are at the end of the article). Table 4 also includes a category for adjacent land that was formerly orchard or market garden and also percentages for the four 10km squares SP03, SP04, SP13 and SP14. The average number of snails collected per sample in these four squares was 28.1, 28.6, 25.3 and 26.7 respectively.

Some road verge sites in the area are thick with snail shells and the list of 25 species spotted during the five minute collections in March is notable. Together with snails of more specialised habitats such as Lias stone walls, this must make the Vale of Evesham one of the best areas in the county for snails.

The topographical features and soil types generally run in a south-west to north-east direction. Figure 1 does not show any apparent correlation between this direction and the number of snails collected per sample. The former use of some of the adjacent land as orchard or market garden also seems to have had little or no effect on the percentages of the main constituent species as these were practically identical with arable and grassland sites elsewhere. However, although the calculated average number of snails collected was very similar for the 14 arable sites (at about 31), for the seven grassland sites it was somewhat larger (36.9 compared with 24.9 elsewhere).

Apart from the two most abundant species (Trichia striolata and Monacha cantiana) the percentage species composition was fairly constant for all sites with adjacent arable, grassland or built-up areas. But at these sites the percentage increased for the former species (30%, 49%, and 63% respectively) and decreased for the latter (40%, 20%, and 9%). The average number of snails collected per sample also decreased (30.8, 27.1, and 17.6) largely due to the number of the latter species decreasing (12.2, 5.4, and 1.6) while the former remained fairly constant (at around 11).

The data suggest that more snails can be collected in five minutes from the wider and more heavily trafficked roads that are salted in the winter (trunk, main and secondary) than the other roads which are not salted (see Figure 2 on a separate page). The average number collected being 36.3 for the former compared to 23.3 for the latter. This variation seems to be mainly due to Cernuella virgata which, in addition to being found in southern inland sites in Britain, is more widespread as a sea-cliff species (see Kerney 1999). This liking for seaside locations, together with the normally shorter vegetation on the verges, may be a contributory factor in the increased numbers. (Is Cernuella virgata going to become the faunal equivalent of Danish Scurvygrass Cochlearia danica and progress along major roads?)

Although 120 sites were sampled and the results are probably valid for general observations involving a reasonable number of these samples, those using less than, say, 20 are likely to be unreliable. This is particularly the case with the individual monad detail given in Figure 1 as each is only from a single sample.

Distribution maps
It is hoped to search more sites in the area, particularly those of other habitats, and produce species distribution maps at a later date.

Acknowledgements
I should like to thank John Meiklejohn for giving me advice on the identification of one or two difficult (to me!) species.