SID 5 (Rev. 05/09) Page 1 of 34

General enquiries on this form should be made to: Defra, Procurements and Contracts Division (Science R&D Team) Telephone No. 0207 238 5734 E-mail: research.competitions@defra.gsi.gov.uk

SID 5 Research Project Final Report

Note In line with the Freedom of Information

Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects.

• This form is in Word format and the boxes may be expanded or reduced, as appropriate.

ACCESS TO INFORMATION The information collected on this form will

be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code WC0786/CR0479

2. Project title

Understanding the status, taxonomy and ecology of UK Biodiversity Action Plan Priority Species.

3. Contractor

organisation(s) Hymettus Ltd

4. Total Defra project costs £ 97 125

(agreed fixed price)

5. Project: start date ................ 15 April 2010 end date ................. 30 September 2012

SID 5 (Rev. 05/09) Page 2 of 34

6. It is Defra’s intention to publish this form. Please confirm your agreement to do so. ................................................................................... YES NO

(a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.

Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.

In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain

Executive Summary

7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.Objectives Hymettus Ltd, in discussion with members of several recording schemes and societies, identified 45 invertebrate species where there was potential for short studies of distribution and natural history to address gaps in knowledge that were hindering the implementation of conservation measures. The main objectives of the research included clarifying the distributions of the species, identifying their habitat preferences, investigating aspects of nesting and egg-laying and determining the type of pollen utilized by different types of bee and the type of prey collected by different solitary wasps to feed their young. Two studies aimed to clear up confusion regarding the precise identity of the species occurring in Britain. Four of the studies aimed to assess specific habitat management or improvement techniques that could be used in the conservation of individual species. Methods Species distributions were clarified by a combination of desk study, in which existing records were obtained from a variety of published and other sources, and field survey. The investigations of habitat preferences were more varied in their methods. For a few species, that were relatively easy to find at a number of different locations, attempts were made to systematically record presence / absence or abundance alongside one or more environmental variable. More often, simple field observation, combined with a literature search on occasion, were all that were realistic. Pollen utilization was investigated by removing pollen loads carried by individual bees and identifying the grains using standard microscopical techniques. Analysis of the prey collected by solitary wasps was attempted either by removing each prey item from the female wasps as they returned to the nest or by investigating the contents of nests for prey. Scanning electron microscopy images and comparison of DNA sequences were used to confirm the identity of one species of pill millipede. Adult insects were reared from a large sample of cigar galls on reed to determine the identity of the fly species capable of causing these galls. The success of different management techniques in providing suitable habitat for ground beetles on heaths was assessed by noting the number of the target species collected in plots where winter burning, cutting, grazing or turf-stripping followed by soil disturbance had been applied. The feasibility of implementing and assessing habitat improvement measures on specific sites for the two species of solitary bee was discussed and plans drawn up with the relevant conservation bodies. Results and conclusions Polyzonium germanicum (a millipede): The distribution was confirmed as being restricted to Kent but new records extended the range further east and west than previously recognised. A preference for habitat with relatively high soil moisture levels was demonstrated. Trachysphaera lobata (a pill millipede): The identity of the species was confirmed. In England, the millipede is known as a single population at one site on the Isle of Wight.

SID 5 (Rev. 05/09) Page 3 of 34

Metaiulus pratensis (a millipede): The distribution was confirmed as being restricted to Kent. A population was rediscovered at a single location in flood meadow and carr woodland. Nothogeophilus turki (a centipede): The species was not located by field survey. This suggested that the centipede is locally extinct on the Isle of Wight. It may survive still on the Isles of Scilly. Cicindela sylvatica (Heath Tiger Beetle): The confirmed distribution was restricted to patches of lowland heath from Dorset to Surrey. A poor ability to disperse to patches of newly created habitat may be limiting its conservation. Poecilus kugelanni (a ground beetle): Controlled winter burning appeared to be the most generally applicable technique for maintaining suitable habitat for the species. Cutting vegetation to less than 10cm every 3-4 years (and removing the cut vegetation) or turf-stripping and ground disturbance every four years had been used successfully on specific sites. Amara fusca (a ground beetle): The main ecological factor controlling the distribution of the beetle appeared to be the presence of its main food plant, Field Wormwood Artemisia campestris. Pogonus luridipennis (a ground beetle): The current distribution appeared to be centred on the coasts of Lincolnshire and North Norfolk. It may be rediscovered yet in Dorset but appeared to have been lost from most of its previous range. Harpalus froelichii (Brush-thighed Seed-eater): Bare ground plots created by turf-stripping appeared to be an effective method for creating beetle habitat for as long as they contained stands of Fat Hen Chenopdium album. The beetle appears to be well adapted to using small patches of ephemeral habitat. Ophonus laticollis (a ground beetle): The confirmed distribution was centred on the Norfolk Brecks although existing records showed that the beetle occurred very locally across southern England. Ophonus melletii (a ground beetle): The beetle was not located during field survey. Existing records suggested that it is unpredictable in occurrence and has a very localised distribution with a relatively stable population at just one site in southern England. Anisodactylus nemorivagus (a ground beetle): The current distribution appeared to be in thinly scattered populations on suitable pockets of habitat over large areas of heath from Dorset to Sussex. The beetle appeared to prefer low growing heather with bare, sandy patches and often a thin layer of peaty soil on south-facing slopes. Philorhizus quadrisignatus (a ground beetle): The species was not located during field survey. No new information on habitat preferences was obtained. Philorhizus vectensis (a ground beetle): The beetle occurred at a low density along the coast from Kent to Cornwall and around into the Severn Estuary. It appeared to be a thermophilous species typically found on sparsely vegetated, south facing maritime slopes. Doratura impudica (a leafhopper): The preferred habitat was shown to be the extreme seaward edge of coastal dunes where Sand Couch, Elymus farctus, formed single species stands or dominated the plant community. This grass was assumed to be the bug’s host plant. Chrysis fulgida (a ruby-tailed wasp): The wasp was not located by field survey. Existing records showed a distribution centred on Surrey, Hampshire and Dorset. Tapinoma ambiguum/erraticum (ants): T.erraticum was not located during field survey. The study provided evidence that T.ambiguum may prefer nesting on sandy soils on heath where there is a high proportion of bare ground, soil temperature is high and the soil is free draining. Formica exsecta (Narrow-headed Ant): The ant was found to incorporate older, dead wood stumps into existing nests but confirmation of its ability to colonise artificially created stumps requires longer term studies. Anergates atratulus (an ant): The species was not located during field survey. The preferred habitat of its host, the ant Tetramorium caespitum, was shown to be Arrhenatherum grassland and lichen heath with exposed flints and pebbles. It was not certain that the same habitat associations would be exhibited by the parasite as by its host. Formicoxenus nitidulus (Shining Guest Ant): A genuine decline was noted in the more northerly part of the British range of Formica rufa, one host of Fx. nitidulus. By contrast a second host, Formica lugubris, appeared to be thriving throughout its range. As Fx. nitidulus shows no preference for either host this was not expected to affect its status and distribution. Temnothorax interruptus (an ant): The preferred habitat on coastal shingle sites was shown to be Arrhenatherum grassland and lichen heath with exposed flints and pebbles. It appeared to benefit from light winter grazing. Homonotus sanguinolentus (Bloody Spider-hunting Wasp): The current distribution was found to be restricted to lowland heath in Dorset, the New Forest and Surrey. Pseudepipona herrichii (Purbeck Mason Wasp): The current distribution was found to be restricted to a small number of heaths in Dorset where the wasp appeared to be well established. Odynerus melanocephalus (Black-headed Mason Wasp): The study demonstrated that larvae of the wasp were able to develop on larvae of at least two different species of Hypera weevil. Odynerus simillimus (Fen Mason Wasp): The study provided some evidence that hours of sunshine, rainfall levels and ground temperature may be important in controlling nest establishment and pupal development. Rhopalum gracile (a solitary wasp): Its presence in the Norfolk Broads and the effectiveness of trap nests

SID 5 (Rev. 05/09) Page 4 of 34

for monitoring the wasp were confirmed. No new information was obtained on the larval prey items utilized. Passaloecus clypealis (a solitary wasp): The species was not located by field survey. No new information was obtained on the larval prey items utilized by the wasp. Cerceris quadricincta and Cerceris quinquefasciata (solitary wasps): Larvae of both wasps appeared to be fed on adults of common, small to medium-sized weevils. No evidence was found of geographical variation in prey items taken. Colletes floralis (a solitary bee): Factors characterising the preferred habitat of the bee included bare ground, short and floristically-rich turf and varied topography. The bee utilized a variety of pollen types but appeared to have a clear preference for Apiaceae. Pollen from Asteraceae, Fabaceae, Rosaceae and Lamiaceae was utilized also. Colletes halophilus (a solitary bee): Nesting aggregations appeared to favour sandy substrates in Marram dunes or at a dune/saltmarsh transition but the bee was able to exploit a wider range of substrates and artificial habitats potentially increasing the opportunities for its conservation through coastal realignment schemes. Andrena ferox (a mining bee): No clear evidence was obtained that lack of genetic diversity in oak trees affected the availability of pollen resources through the flight period of the bee. It was considered probable that other factors such as site aspect and woodland structure were important too. Andrena tarsata (a mining bee): Tormentil Potentilla erecta appeared to be the main pollen resource utilized by the bee in North Yorkshire. Lasioglossum angusticeps (a solitary bee): The sole pollen resource utilized by the bee in Britain appeared to be Bird’s-foot Trefoil Lotus corniculatus. Osmia parietina (a solitary bee): The bare rock surfaces and calcareous conditions favoured by English populations of the bee were considered atypical of European populations. The sole pollen resource utilized by the bee in Britain appeared to be Bird’s-foot Trefoil Lotus corniculatus. Eucera longicornis (Long-horned Bee): The species was not located during field survey. No insight was gained into the influence of different grassland management regimes on the conservation and foraging activity of the bee. Anthophora retusa (a solitary bee): No evidence was obtained to support the theory that additional appropriate forage resource provision has a positive effect on the area occupied by a population of the bee. An awareness of the issues around the species has been raised within relevant conservation bodies. Bombus ruderarius (Red-shanked Carder Bee): Queens occurred in a variety of habitats but workers tended to forage in more floristically-rich sites. Where it grew, Kidney Vetch Anthyllis vulneraria appeared to be the favoured pollen source of workers. Pollen from other Fabaceae, Rosaceae and Lamiaceae was utilized also. Asindulum nigrum (a fungus gnat): The species appeared to prefer a fen-meadow habitat type with tall, dense, grassy sward and tussocks untrampled by livestock rather than a wet fen habitat. Neoempheria lineola (a fungus gnat): The species could not be located by field survey but on evidence from existing records its English distribution is restricted to the New Forest. The few existing records provided no information on preferred habitat or foraging resources. Lipara similis and other Lipara species (Cigar-gall flies): The confirmed distribution of L. similis was restricted to East Anglia. L. pullitarsis was identified as a species new to Britain. A dichotomous key to adults of British Lipara species was produced but differentiation of galls was not possible. No clear differences in habitat preferences between species were identified. Dolichopus laticola and D. nigripes (long-legged flies): Both species appeared to favour a fen-meadow habitat type; intermittently cut, high in plant diversity and with low dominance by tall reed.

Project Report to Defra

8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the scientific objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and

SID 5 (Rev. 05/09) Page 5 of 34

any action resulting from the research (e.g. IP, Knowledge Transfer).

INTRODUCTION Following the publication in August 2007 of the updated UK Biodiversity Action Plan List of Priority Species and Habitats, a workshop was organised in December 2008, involving members from UK BRAG and BRIG and species group experts, to categorise and prioritise research actions for UK Priority Species. Attendees categorised the unfulfilled actions listed in Species Action Plans as survey / monitoring or ecological research. Some new actions were identified also that included a requirement for some taxonomic research. The work reported here attempted to complete some of these actions and thus address gaps in the knowledge needed to promote conservation of some of the 413 invertebrate Priority Species listed. The project resulted from a research proposal submitted by Hymettus Ltd that addressed some of these issues in relation to invertebrates. In discussions with representatives from a number of recording schemes and societies (Auchenorrhyncha Recording Scheme, British Myriapod & Isopod Group, Bees, Wasps & Ants Recording Society, Carabidae Recording Scheme and Dipterists Forum) 45 species were identified where a combination of desk studies, field surveys, taxonomic studies and autecological studies had a realistic chance of delivering positive outcomes for species conservation over the relatively short period of three field seasons. Advances in knowledge of taxonomy and species distribution were considered necessary prior to studies of the ecology of certain species. However, it was the autecological studies that were expected to provide an insight into appropriate conservation management of a small number of ecological assemblages of species. Although, even as the research was taking place, the UK Biodiversity Action Plan was replaced by the UK Post-2010 Biodiversity Framework for the period from 2011 to 2020, the lists of priority species agreed under UK BAP are still the basis of much biodiversity work in the four countries of the UK and thus the findings reported are still relevant to the practical conservation of invertebrates. OBJECTIVES 1. To determine the current English distribution of the millipedes Trachysphaera c.f. lobata, Polyzonium germanicum and Metaiulus pratensis; the centipede Nothogeophilus turki; the beetles Cicindela sylvatica, Pogonus luridipennis, Ophonus laticollis, Ophonus melletii, Anisodactylus nemorivagus and Philorhizus vectensis; the flies Lipara similis and Neoempheria lineola and the wasps Chrysis fulgida, Homonotus sanguinolentus and Pseudepipona herichii. 2. To assess localised changes in status and distribution of the ant Formicoxenus nitidulus and the bee Andrena tarsata. 3. To investigate the taxonomy of the Lipara species occurring in Britain and prepare new keys to the adults and galls of these species. 4. To determine the identity of the Trachysphaera species occurring in Britain by comparison with continental material. 5. To determine the environmental conditions necessary for nest establishment / egg laying by the flies Asindulum nigrum and Neoempheria lineola; the ant Formica exsecta; the wasp Odynerus simillimus and the bee Colletes halophilus. 6. To determine the environmental and ecological factors characteristic of sites inhabited by the millipedes Polyzonium germanicum and Metaiulus pratensis; the bug Doratura impudica; the beetles Cicindela sylvatica, Amara fusca, Anisodactylus nemorivagus, Philorhizus quadrisignatus and Philorhizus vectensis; the flies Lipara similis, Dolichopus laticola, Dolichopus nigripes, Asindulum nigrum and Neoempheria lineola; the ants Tapinoma ambiguum/erraticum, Temnothorax interruptus and Anergates atratulus and the bees Colletes floralis, Andrena ferox, Osmia parietina, and Bombus ruderarius. 7. To determine the nature (if any) of the foraging resources utilized by adults of the fly Neoempheria lineola; the wasps Odynerus melanocephalus, Rhopalum gracile, Passaloecus clypealis, Cerceris quadricincta and Cerceris quinquefasciata and the bees Colletes floralis, Lassioglossum angusticeps, Andrena tarsata, Osmia parietina and Bombus ruderarius. 8. To assess the feasibility and / or success of trialling specific habitat management / improvement techniques in the conservation of the beetles Hapalus froelichii and Poecilus kugelanni and the bees Eucera longicornis and Anthophora retusa. 9. To develop and test a monitoring protocol for detecting the presence of Andrena tarsata and a method of recording habitat quality. The research is reported below as autecological studies of the target species grouped within seven ecological / geographical assemblages. INVERTEBRATES OF SAND DUNES AND SHINGLE 1. Doratura impudica Introduction Doratura impudica is a medium-sized leafhopper. All verified records of the species in England to date come from coastal sand dune habitats. Putative records from other habitats have turned out to be mis-identifications or the result of record coding errors. In continental Europe, the species appears to occur in a wider range of habitats than in England including clearings of pine forests, inland dunes, sandy ruderal sites and sparsely-vegetated brownfield sites. The host plant of D. impudica has been reported to be Calamagrostis epigejos and perhaps also Elymus species in Germany, to be Elymus athericus in northern Italy and Peter Kirby has suggested that the host

SID 5 (Rev. 05/09) Page 6 of 34

plants in England were ‘probably grasses’, in coastal sand dunes, without specifying any particular species. The objective of this study was to determine the environmental and ecological factors characteristic of sites inhabited by D. impudica. Methods Twelve sand dune sites along the coasts of SE England and East Anglia between West Sussex and north Norfolk were visited in 2010-11. Detailed sampling was carried out at 22 sub-sites by standardized sweep-netting (vacuum sampling was used initially but found not to be any more productive for this species). The number and sex of all adult D. impudica were recorded (no nymphs were found). The relative cover of all plant species and bare ground at each sub-site was recorded on the DAFOR scale with the data being used to allocate each sub-site to an NVC community. The height of the vegetation and the density of sand couch, Elymus farctus stems in 50x50cm sample quadrats were also recorded. The total area of potentially suitable habitat was estimated. A number of environmental variables were recorded at each site in order to characterize the abiotic micro-habitat, including aspect, slope and soil pH. Sand particle size and the presence of obvious disturbance (in the form of trampling) were recorded on ordinal scales. Site locations were recorded using a GPS to an accuracy of +/- 10m. Photos of the habitat were taken at each site. Results The species composition of the plant community at sub-sites where D. impudica was either found or not found was compared using ordination (Detrended Correspondence Analysis). The eigenvalue (0.875) and the gradient length (4.79) for the first axis indicated a high turnover and therefore good separation of species along this axis, although this was strongly affected by one anomalous site with a very high loading on axis 1. The second axis also had a moderately high eigenvalue (0.403) and a reasonable gradient length (2.90). The ordination plot is complicated by the superimposition in the lower left-hand corner of six sub-sites that were pure stands of E. farctus; three at which D. impudica occurred and three where it did not. The ordination shows that the species composition of the vegetation at sites with D. impudica was essentially a small sub-set of those at which it was absent. This is partly because all the sub-sites with D. impudica contained E. farctus, as did the majority of those without the bug (9 out of 12). This might suggest that sub-sites without D. impudica had a more species rich plant community than those that supported the bug. However, this difference was not statistically significant (mean species richness for sub-sites with and without D. impudica was 4.10 and 5.17 respectively; t=0.77; P=0.454. Logistic regression was employed to explain the occurrence of D. impudica in terms of environmental variables using its presence or absence as the binary response variable. Of the environmental variables measured at each sub-site, only one (% cover of E. farctus) showed a significant (positive) relationship with the presence of D. impudica (Regression coefficient=0.179 , Z-score=1.96, P=0.050). Neither the total area of E. farctus habitat nor the mean height of the vegetation showed any significant relationship with the presence of the bug. Discussion Initial surveying at Sandwich Bay in 2010 indicated that D. impudica was exclusively associated with E. farctus, where this grew in sparse stands at the seaward margin of sand dunes. Typically, this habitat occupies a narrow strip (2-5m wide) of land running parallel to the shore between the strandline community (typified by plant species such as Atriplex laciniata, Salsola kali and Suaeda maritima) and the start of fixed dune communities on the landward side dominated by marram grass (Ammophila arenaria). These thin strips of E. farctus frequently include other ‘backshore’ species such as Euphorbia paralias, Honkenya peploides and Beta vulgaris. However, the stands are frequently monospecific, albeit with above-ground tillers that are sparsely distributed. This community conforms exactly with the NVC community SD4: Elymus farctus ssp. Boreali-atlanticus foredune community. At its landward margin, it often includes sea couch, Elytrigia athericus (with which E. farctus can hybridize) and occasionally lyme-grass, Leymus arenarius. This project has provided strong evidence that the principal (and perhaps sole) host plant for D. impudica in Britain is sand couch, E. farctus. It was found only in areas where E. farctus formed mono-specific stands or dominated the plant community. Closely related plant species, such as Elytrigia atherica, did not support the bug. However, extensive searching of some apparently suitable sites (e.g. Winterton Dunes) failed to detect the bug, suggesting that factors other than the presence of the host plant may be critical. It was most frequently encountered, and in greatest numbers, where its host plant grew in large, sparsely-stemmed, monocultural stands. Such stands usually occur in a narrow zone (maximum 10m) between uncolonized sand at the top of the beach and stabilized dune dominated by marram grass, Ammophila arenaria. As such, the preferred habitat, the extreme seaward edge of coastal sand dunes, is very vulnerable to erosion and fragmentation by wave action and to disturbance by human trampling. Conclusion It is clear that the preferred habitat for this species is the extreme seaward edge of coastal sand dunes where sand couch, Elymus farctus, forms mono-specific stands or dominates the plant community. This species is therefore assumed to be the bug’s host plant. The objective of this study was achieved in full. 2. Anergates atratulus and Temnothorax interruptus Introduction Investigations of the habitat preferences of the ants Anergates atratulus and Temnothorax interruptus on shingle beaches at Dungeness and Rye Harbour had been initiated in 2007 with the initial aim of providing information on the impact of grazing on the populations of the two species. Attempts to identify the habitat preferences of the parasitic A. atratulus are hampered by the difficulty of locating the species. Therefore, the work focussed on its host, Tetramorium caespitum, a common species of ant around the coasts of southern England. The methods

SID 5 (Rev. 05/09) Page 7 of 34

adopted involved excavating shingle within quadrats and turning over flints in different vegetation types to determine if ants were nesting there. The results were intended to inform management planning for the protected sites. Preliminary findings indicated that light winter grazing of Arrhenatherum grassland and of lichen dominated sward may be beneficial for Tx. interruptus but insufficient data was available for a robust test of any relationship. The highest densities of Tm. caespitum were found on ungrazed lichen heath with lower densities on winter grazed areas. As a parasite, A. atratulus can only be supported by larger colonies of its host and it has been suggested that the host queen must have been lost from the colony before the parasite can become established in a nest. The objectives of this study were to provide a robust analysis of the habitat preferences of Tx. interruptus and to determine if relationships existed between vegetation type and firstly the presence / absence of Tm. caespitum and secondly the size of a Tm. caespitum colony. Methods In July 2010 a network of individually numbered hanging tiles was placed in different vegetation communities (Cytisus scoparius scrub, lichen-rich acid grassland, shingle margin acid grassland or stable Arrhenatherum elatius grassland) on the shingle at Dungeness and Rye Harbour. This was considered a more efficient and less destructive method than excavating quadrats. As the tiles become warm under sunlight they attract a variety of ants rearing eggs, larvae and cocoons. The tiles were checked after three weeks and the species present was recorded. Colony size was categorised as absent, foragers only, small (tens of workers), medium (hundreds of workers) or large (thousands of workers). In August 2011 an alternative approach was tested using cheese bait set under over-turned plant-pot saucers. A total of 40 sets of bait were laid out in a grid 5m apart in homogenous stands of vegetation and left for two hours to attract ants before being checked. The number and species of ant visiting each set of bait was recorded. Results Only one nest of Tx. interruptus was found under a tile in each of 2010 and 2011. The tiles were in different vegetation communities. The cheese bait technique was no more effective with just three individuals of Tx. interruptus being attracted to one set of bait in Arrhenatherum grassland. The very low numbers of ants and nests observed precluded detailed analysis. Due to the categorical nature of the variables, analysis of the relationships between Tm. caespitum and vegetation type were carried out using a Chi-squared test of association. Data from the hanging tiles did not provide clear evidence of differences in species distribution between vegetation types. However, data from the use of cheese baits suggested that there was a significant association between the presence / absence of Tm. caespitum and vegetation type (X2=30.855, p<0.001, df=3). Further analysis suggested this relationship to be due to the low level of occurrence of the ant in the scrub and stable grassland vegetation types compared with the earlier successional grassland types (X2= 45.502, p<0.001, df=2). There appeared to be no association between nest size and vegetation type but large nests (>1000 workers) were more frequent than expected in areas with light winter grazing. The opposite was true in ungrazed areas where individual foragers, small (<100 workers) and medium nests (>100, <1000 workers) were more frequent than expected (X2= 17.533, p<0.001, df=3). A. atratulus was not found in any of the nests monitored. Discussion Although insufficient data were collected for robust statements to be made regarding the status of Tx. interruptus in the different communities, some general observations can be made. Since 2007 Tx. interruptus has been found in coastal Arrhenatherum grassland, which has abundant areas of exposed pebbles, and on areas of lichen heath with exposed flints. It nests under the pebbles unlike its close relative Tx. albipennis which has usually been found nesting in the hollow stems of dead vegetation. However, one mixed species nest was observed. On areas of lichen heath Tx. interruptus is likely to benefit from grazing if this exposes flint pebbles in amongst the vegetation. Tm. caespitum was very scarce in areas of unconsolidated shingle. As the coastal grassland becomes more established and the shingle more fixed, with finer material filling the voids between the pebbles, so Tm. caespitum starts to establish nests. Although the ant is frequently found on stable, long-established Arrhenatherum grassland it is more strongly associated with lichen-rich areas in shingle margin acid grassland and especially in mature lichen-rich acid heath. Thus the distribution of Tm. caespitum across the shingle is relatively predictable, with this ant occurring in stable coastal grassland and lichen heath, although each patch of shingle vegetation does not have a uniform density of the ant, presumably due to chance effects in colonisation. Without further knowledge of the relationship between Tm. caespitum and A. atratulus it is impossible to know if the distribution of the latter is similarly predictable. If the relationship is simply one based on colony density then the distribution of A. atratulus can be predicted but if colony size or even colony health are involved then no predictions are possible based on the results of this study. Conclusion Based on very limited data, the preferred habitat of Tx. interruptus on coastal shingle is Arrhenatherum grassland and lichen heath with exposed substrate. Tm. caespitum shows similar habitat preferences and both species appear to benefit from light winter grazing. It is not certain that the same habitat associations are exhibited by A. atratulus as by its host. The objectives of this study were achieved as far as was possible with the methods used and data gathered. More efficient methods of monitoring both Tx. interruptus and A. atratulus and / or detailed investigations of the relationship between A. atratulus and Tm. caespitum would be required to make further progress.

SID 5 (Rev. 05/09) Page 8 of 34

INVERTEBRATES OF THE NORTH SEA AND ENGLISH CHANNEL COASTS 1. Trachysphaera lobata Introduction The first British specimens of a pill millipede of the genus Trachysphaera were collected from East Cliff, Bembridge, Isle of Wight in 1984 by sieving sandy soil in a narrow, coastal belt of sycamore dominated woodland. Species of Trachysphaera are generally difficult to identify because numerous morphological characters are known to vary greatly both within and between populations and many species descriptions are based on a very small number of specimens. The East Cliff specimens were provisionally identified as T. lobata but it was noted that there were slight morphological differences from the type description. Over the next 20 years the continued existence of the East Cliff population was monitored by members of the British Myriapod and Isopod Group (BMIG). In 2010 two further populations of Trachysphaera were reported from South Wales. The objectives of this study were firstly to determine the identity of the Trachysphaera species occurring in Britain by comparison with continental material and secondly to confirm the current English distribution of the millipede. Methods Three members of BMIG visited East Cliff on 22nd February 2011 and again on 7th May 2012. Non-standardised hand searching was employed to search for Trachysphaera specimens. As the field workers were familiar with the millipede and the sandy, humus rich pockets of soil that were most likely to support it, this method was considered to be the most time-efficient. The extent of the occupied area of habitat was assessed by searching for the millipede whilst moving east and then west along the coast. Eight figure OS grid references were noted for the most easterly and most westerly points at which the millipede was found. Apparently suitable sites at Priory Woods and Centurions Copse on the RSPB Brading Marshes reserve were surveyed on 22nd February 2011 and 7th May 2012 respectively. Surveys were undertaken at several other locations along the south and west coast of the Isle of Wight on 23rd and 24th February 2011 and on 8th and 9th May 2012. A sample of 25 Trachysphaera specimens was collected into 98% ethanol from the East Cliff population in February 2011. A similar sample was collected by a BMIG member from near Swansea in March 2011. These samples were sent to the Research Museum Alexander Koenig in Bonn, where a revision of the genus is underway. SEM images of various morphological structures were taken from five specimens from the East Cliff population and eight specimens from the Welsh population. DNA was extracted and COI gene sequences obtained for comparison from 13 specimens from the East Cliff population, 12 from the Welsh population and two from Italy. Results The 2011 search at East Cliff extended the length of habitat known to be occupied by Trachysphaera by 120m to the east and by 10m to the west. This is a significant increase of over 50% on the 200m length of occupied habitat reported in 2005. Trachysphaera was found again at East Cliff in 2012 but at much lower densities than in 2011. The millipede was not found at any of the other locations surveyed in 2011 and 2012. SEM images of morphological characters on the last tergite before the anal shield confirmed the identity of both British Trachysphaera populations as T. lobata. However, multi-layer images taken of twelve specimens from East Cliff and of nine specimens from the Welsh population revealed large differences in colour and encrustration within each population that could not be correlated to differences in other morphological structures or to genetic differences. Comparison of DNA sequences show that the two British populations of T. lobata, while related to each other, have a different evolutionary history. Discussion The distribution of T. lobata within the East Cliff site is known to be uneven as the millipede is only found in pockets of humus rich, sandy soil with or without coarser pebbles and is absent from the clay soil that underlies most of the woodland and from slumps at the base of the unconsolidated cliff where the soil is often sandy but lacks any organic material. Therefore, the apparent extension of occupied habitat may not correspond to an increase in population. An apparent decrease in population between 2011 and 2012 may be due to differences in the timing of the surveys and aspects of the life cycle of the millipede that are not as yet understood. The question of whether the two disjunct populations are the last remaining relict populations of a Trachysphaera distribution spanning southern England and South Wales, with all intermediate populations either undiscovered or extinct, or if the Welsh and Isle of Wight populations are the result of two separate dispersal or introduction events from distant European mainland populations, cannot be answered at present because of the lack of samples from the European continent. Further surveys at coastal sites in south Devon and Dorset may be worth considering but the discovery of sites in South Wales suggests the coastal location may not be a key factor. Conclusion The taxonomic objective was fully achieved and the identity of the species in both England and Wales was confirmed as Trachysphaera lobata. On current knowledge it would seem that the distribution of T. lobata is restricted to a single site in England. 2. Nothogeophilus turki Introduction The centipede Nothogeophilus turki was first described in 1988 on the basis of specimens collected in the Isles of Scilly and the Isle of Wight (Freshwater Bay and Newport Docks). Reportedly it was collected from Cornwall in the mid1990s but no further data exists to support this record and there are no other more recent records. The objective of this study was to determine if the centipede could be found still on the Isle of Wight. Methods

SID 5 (Rev. 05/09) Page 9 of 34

Details of the Isle of Wight sites were requested from the original recorder who reported that the Newport Docks site had been developed. Three members of BMIG visited the Isle of Wight on 23rd and 24th February 2011 and 8th and 9th May 2012 when non-standardised hand searching was employed to search for N. turki. In 2011 the locations searched were Freshwater Bay, Colwell Chine, Compton Chine, Brook Chine, Knowles Farm and Bonchurch Landslips. Freshwater Bay and Colwell Chine were revisited in 2012 and further sites at Alum Bay, Sheppard’s Chine, Fort Victoria Country Park and the west bank of the River Medina between Newport and Dodnor were searched. Results N. turki was not found during searches at any of the sites on the Isle of Wight during February 2011or May 2012. Discussion If N. turki is still be present on the Isle of Wight then it is only likely to be detected by expensive, intensive extraction sampling or by random chance. Other small, coastal centipedes such as Pachymerium ferrugineum have been rediscovered by casual recording at sites after many years of apparent absence. This supports the view that not enough is known of its autecology to develop reliable detection and monitoring methods. Previous records suggest there is a greater chance that N. turki could be found again on the Isles of Scilly. A visit for the sole purpose of attempting to record this one species may be expensive but against this the centipede is apparently an endemic. Further searches on St Marys and Tresco could form a useful component of a broader based expedition such as has been suggested by Buglife in recent years. Conclusion Despite 12 man-days of field work by experienced workers, N. turki was not located and on current knowledge it would appear that the centipede is locally extinct on the Isle of Wight. 3. Pogonus luridipennis Introduction Pogonus luridipennis is a small ground beetle that inhabits saltmarsh. It was thought to have declined significantly in recent years and the objective of the study was to determine the current English distribution of P. luridipennis. Methods Existing records of the beetle were collated. Data were requested from the national Carabidae recording scheme and individual coleopterists with knowledge of the beetle were contacted for further details of their records and others they may know of. The most recent records were used to target surveys by local coleopterists. Field work focussed on known grid reference points and other areas of apparently suitable habitat. At Salthouse, Norfolk (28th August 2010 and September 2011) the shingle was searched by looking under stones and seaweed in the shallow lagoons. At Eype’s Mouth, Bridport, Dorset (24th February, 5th April, 27th April and 27th September 2011) searches were carried out around a salt inundated pool on the beach and on surrounding clay cliffs. Stones were embedded in the clay banks of the pool and lower edge of the clay cliffs to provide refugia that were checked on subsequent visits for any beetles inhabiting the area. There was no access to the saltpan habitat at Titchwell RSPB reserve, Norfolk (9th June 2011) but similar habitat was searched at Thornham Point, Norfolk (9th June 2011) and Morston, Norfolk (25th September 2011). An identification guide to Pogonus spp. was produced and a copy sent to Gibralter Point Nature Reserve in Lincolnshire in March 2011. It was hoped that this would inspire local naturalists to go out and look for the beetle. Two further locations with historic records of P. luridipennis were surveyed on 12th October 2010 (Weybridge, Dorset) and during May 2011 (Newnham, Gloucestershire). Results The desk study identified the historic range of P. luridipennis as the east and south coast of England from Lincolnshire to Dorset and the southern shore of the Severn Estuary / Bristol Channel from Gloucestershire to North Devon. However, there was just one site in Dorset, three sites in Norfolk and two sites in Lincolnshire with post-1990 records of the species. The beetle was not found by field work at any of these locations in 2010 or 2011. The identification guide was a success as two individuals were reported by a local naturalist from a saltpan at Seacroft Marsh, Lincs. in May 2011. The site was checked by experienced coleopterists on 10th September 2011 and 17th September 2011. Six individuals were found on each occasion. Discussion The results confirm a significant decline in the range of P. luridipennis over recent years. There now appears to be just one location in Lincolnshire where it can be found reliably. However, the beetle appears to be partly subterranean in habits so night searching with a torch between April and October may prove to be a more efficient way of locating it. Also the effectiveness of light traps could be explored and the identification guide may be a useful method of encouraging moth recorders to assist in further survey work including studies of its phenology. Studies of exact habitat preferences, prey and larval ecology are required but would ideally include more populations than just the one discovered at Seacroft Marshes Conclusion The current English distribution of P. luridipennis appears to be centred on the coasts of Lincolnshire and North Norfolk. The beetle may still be rediscovered in Dorset but appears to have been lost from most of its previous range. 4. Philorhizus vectensis Introduction

SID 5 (Rev. 05/09) Page 10 of 34

Philorhizus vectensis is a small ground beetle which inhabits sparsely vegetated sand and shingle and maritime cliffs. It has a limited global distribution with non English populations restricted to the Channel coast of France, along the Atlantic coasts of France, Portugal and Spain and around to the Mediterranean coast of France and an outlier site in Wales. The objectives of this study were to determine the current English distribution of P. vectensis and to characterise more precisely the micro-habitats the beetle inhabits. Methods Existing records of P. vectensis were collated and examined for information that could be used to characterise its habitat. Data were requested from the national Carabidae recording scheme and individual coleopterists with knowledge of the beetle were contacted for further details of their records and others they may know of. The most recent records were used to target surveys by local coleopterists. Field work focussed on known grid reference points and other areas of apparently suitable habitat. Fieldwork involved visits to Queener Point and Gribbin Head, Cornwall (20th October 2010), Lamorna, Cornwall (26th October 2011), Torcross, Devon (18th May 2010, 16th September 2010, 2nd November 2010, 25th March 2011, 20th October 2011), Dun Point, Devon (18th May 2010, 16th September 2010), Strete Gate Beach, Devon (16th September 2010), Bolberry Down, Devon (22nd October 2010, 5th March 2011), Great Mattiscombe, Devon (25th March 2011), Chesil Beach and Church Ope Cove, Dorset (12th October 2010), Studland, Dorset (27th September 2011), West Bay and Seatown, Dorset (11th October 2011) and Cavenham Heath, Suffolk (22nd September 2011). Tussocks of vegetation on rocky cliffs were searched for two to four hours by hand and by suction sampling at the sites along the southern coasts of Cornwall and Devon. Similar methods were used on the shingle and heathland sites in Dorset and Suffolk. Target notes were produced at all locations where P. vectensis was found or had previously been found. These notes included details of vegetation, substrate, aspect and management. Results The desk study identified the historic range of P. vectensis as the south coast from the Thames Estuary around to the Severn Estuary. Two inland, outlier records from North Essex and the Suffolk Brecks were identified also. Modern (post-1990) records were located from four sites in Cornwall, seven sites in Devon, one site in Somerset, two sites in Dorset, one site on the Isle of Wight and two sites in Kent. The only beetles recorded during the field work were from grass tussocks on south facing maritime cliffs at three sites in South Devon. Six specimens of P. vectensis were shaken from grass tussocks on cliffs at Dun Point, a new location for the species, on 18th May 2010. One beetle was collected at Bolberry Down on 5th March 2011 and another was collected at Torcross on 25th March 2011, the only specimen found during five visits to this normally reliable site. Discussion The results suggest the current range of P. vectensis is probably similar to its historic distribution. It probably has always occurred at a low density in suitable locations throughout its range. The apparent concentration of the most recent records in the South West can be explained by increased recorder effort in that area. To enable a wider number of people to identify and search for the beetle it would be valuable to produce an identification guide to Philorhizus species similar to that produced for Pogonus species and use this to target searches for the beetle more effectively. Suction sampling of martime habitats ought to be considered as the most effective method for finding this species in any future survey. Conclusion The current English distribution of P. vectensis is along the southern coasts from Kent to Cornwall and around into the Severn Estuary to Somerset. It probably occurs at a low density on sea cliffs and other coastal habitats throughout this range. The limited information available suggests that P. vectensis is a thermophilous species typically found on sparsely vegetated, south facing maritime slopes in the mild climate of south west England. However, coastal shingle and heathland may also provide suitably warm and sparsely vegetated conditions. 5. Colletes halophilus Introduction The coasts of England, especially those of East Anglia, hold a significant proportion of the known world population of the mining bee Colletes halophilus. Although, the bee is perceived to be threatened by sea level rise, coastal realignment schemes could usefully contribute to the conservation of this species if appropriate nesting habitat was included within the design of such schemes. However, relatively little is known about the nest site preferences of the bee. The objective of this study was to determine the environmental conditions necessary for nest establishment by C. halophilus. Methods Previously known nesting aggregations of C. halophilus at Colne Point nature reserve and Alresford Creek railway embankment, Essex were visited in September 2010 and Alresford Creek was revisited in October 2011. Attempts to locate other nesting aggregations were made at three other sites in Essex (Stone Point salt marsh north of Walton Hall Marshes, Middle Beach at Dovercourt and Mersea Stone) and along the Orwell Estuary, Suffolk in autumn 2010. This involved walking potential habitat, looking for females returning to their nests or for males scouting low over the ground for mating opportunities. In autumn 2011 the sites at Stone Point and Middle Beach were revisited and further attempts were made to locate nesting locations in south Suffolk. At confirmed nesting locations, the characteristics of the soil at the surface were assessed using the Bryophyte Ecology Group soil texture key, photographs of the location were taken and the vegetation present in the immediate vicinity of the nests was recorded. In addition to field surveys, details of other nest aggregations in Essex, Lincolnshire and Norfolk were obtained from other hymenopterists.

SID 5 (Rev. 05/09) Page 11 of 34

Results Field data was collected from a total of 15 nesting aggregations in Essex, Lincolnshire and Norfolk. The substrate utilised was assessed as sand at nine, sandy loam at two, loamy sand at two and as silty clay at two nesting aggregations. Six of the aggregations were in level substrate, six were in near vertical faces and the remaining three were in moderate slopes. Two distinct niches were exploited in natural habitat, low-lying transition zones between sand dune and saltmarsh habitat (four nest aggregations) and also dry Marram Grass dune vegetation (five nest aggregations). Nesting aggregations of C. halophilus also exploited a variety of artificial locations including a railway embankment, silt lagoons and seawall embankments. Discussion There is clearly a good deal of variation in the nesting sites selected by C. halophilus. The majority of the nesting locations are, however, in a more or less sand-dominated substrate and it appears that most of the particularly large nesting aggregations are in this medium. All of the substrates would be capable of holding a stable tunnel, so that bare, loose soil is unlikely to be favoured. Even in sand dune habitats, it is postulated that the fibrous root stock of Marram Grass and other vegetation present is playing a role in helping to support the tunnel walls. At some sites the substrate remains dry but at most sites it is exposed to significant levels of moisture at some times of the year. The natural sites all have a similar juxtaposition of sand dune and saltmarsh habitats i.e. a seaward ridge of sand with saltmarsh behind or are in the eroding margin of dry Marram Grass dunes. An interesting aspect of this study is the wide range of slopes, including near vertical artificial slopes similar to the eroding dune margins, within which the bee nests. Conclusion Nesting aggregations of C. halophilus appear to favour sandy substrates in Marram dunes or at a dune/saltmarsh transition. However, the bee is able to exploit a wider range of substrates and artificial habitats potentially increasing the opportunities for its conservation through coastal realignment schemes. INVERTEBRATES OF HEATHLAND 1. Cicindela sylvatica Introduction The distribution of the Heath Tiger Beetle Cicindela sylvatica in England has been well documented, however, some of the published records are considered contentious, potentially giving a false impression of the species’ historic range. There is ongoing survey effort in Surrey but there has been little recent survey work from other counties within the species’ recent historic range, such as Sussex, Hampshire and Dorset. The autecology of tiger beetles per se is also well documented but there appears to be a paucity of literature pertaining specifically to C. sylvatica. The objectives of this study were to ascertain the current English distribution of C. sylvatica and to determine the environmental and ecological factors characteristic of sites inhabited by the beetle. Methods A combination of desk top study and field survey was used to investigate the current distribution of C. sylvatica. Data were requested from the Carabidae recording scheme; the county Wildlife Trusts and county record centres in Dorset, Hampshire, Surrey and Sussex; National Trust (Purbeck); RSPB (Dorset Heathlands Project); Buglife and Natural England. County Coleoptera recorders from Surrey, Hampshire, Sussex, Dorset, Gloucestershire, Kent and Suffolk, along with the current and former national recorders for the Carabidae were consulted. Label data were garnered from pinned and carded specimens at the British Museum of Natural History, London; National Museum Wales, Cardiff; World Museum, National Museums Liverpool: Hope Collections, Oxford University Museum; Colchester & Ipswich Museums; Tullie House Museum & Art Gallery, Carlisle; Herbert Art Gallery & Museum, Coventry; Leicestershire County Council Museum; Hampshire Museums Service, Winchester and the British Entomological and Natural History Society collections, Reading. A literature search, including standard identification works, county atlases and unpublished research reports, was undertaken but was not comprehensive due to time constraints. Indexes to The Coleopterist, The British Journal of Entomology and Natural History, Entomologist’s Monthly Magazine, Entomological Record and Journal of Variation and The Entomologist were checked. Individual naturalists with a knowledge of the species also provided data and discussion via personal communications and internet forums. Finally, internet searches targeting natural history websites (e.g. iSpot, Wild about Britain) and photographic websites (e.g. Flikr) were used to locate casual observations. Field survey was restricted to Hampshire and Dorset. Sites with historic records and potential new sites in the Aldershot / Farnborough and New Forest and Christchurch areas were visited on 13th June 2010, 11th June 2011 & 10th September 2011. Numerous open access heathland sites in the Purbeck area of Dorset were visited on 16th - 17th May 2010, 9th - 10th August 2010, 12th -15th June 2011, 16th - 17th July 2011, 3rd - 4th September 2011 & 10th - 11th September 2011. Difficulties in gaining access to military training sites where the species is known to occur at Woolmer, Longmoor, Bovington and Lulworth prevented survey but recent records exist from access authorised ecological personnel. Walkover surveys of historic sites, focusing on known grid reference points and areas of otherwise apparently suitable habitat, were conducted and beetles captured with a 4’ handle fine mesh sweep net. Individual beetles were sexed in the field and released immediately afterwards. The autecology of C. sylvatica was investigated through a combination of literature search, field observations during distributional surveys for the beetle in southern England and mark-release-recapture studies of both C. sylvatica and C. campestris, at Thursley Common during 2011. Results

SID 5 (Rev. 05/09) Page 12 of 34

C. sylvatica has been recorded historically from as many as 36 hectads (although some of the records are questionable). It is believed to be still extant in14 of these hectads, a 61% decline in range. The desk study located contemporary (post-2009) records from sites in west Surrey, localised areas of north and south Hampshire and the Purbeck and Lulworth areas of Dorset. The narrowly defined ecological requirements listed by many authors suggest that C. sylvatica is a stenotopic species of patches of bare ground on drier, compact sandy soils within lowland heathland habitat. This was supported by distributional studies that found populations of the beetle were restricted to suitable habitat patches on lowland heathland, with a single extant population on coastal sand dune heath (at Studland, Dorset). In the mark-release-recapture study no adults of C. sylvatica were recaptured at a distance exceeding 200 metres from the original capture point and no exchange was detected between populations. Populations of C. sylvatica were smaller and more locally distributed on parched early successional heath than those of C. campestris. Discussion The historic data collected during the desk study show that the species was formerly more widespread with a population recorded as far north as Lincolnshire. The current distribution is thought to be restricted to a narrow area of occupancy across the lowland heathland of southern England from Surrey to Dorset. Modern (i.e. post1990) records exist from West Sussex but not after 1993 and the beetle is believed to be extinct in the county. The suitability of an individual patch of habitat would be expected to decline quite rapidly on many heaths necessitating frequent dispersal to new areas. Thus the poor dispersal ability of C. sylvatica suggested by the mark-release-recapture study may be a limiting factor for the species. Once published, extensive studies of habitat suitability for C. sylvatica in Germany and Spain should provide more insight into the environmental and ecological factors likely to be characteristic of the beetle’s habitat in England. Conclusion The current English distribution of C. sylvatica is restricted to patches of lowland heath from Dorset to Surrey. Within this habitat the beetle requires patches of bare ground on drier, compact sandy soils. A poor ability to disperse to patches of newly created early successional heath may be limiting its conservation. 2. Poecilus kugelanni Introduction Poecilus kugelanni is a medium-sized ground beetle that inhabits sandy heath lands in southern England. The beetle favours warm, south-facing slopes. The objective of this study was to assess the success of different techniques (grazing, burning, cutting and the creation of bare ground) in maintaining conditions suited to the conservation of P. kugelanni. Methods Seventeen localities in Dorset, Devon and Cornwall, where populations of the ground beetle P. kugelanni are known, were visited in 2010 and 2011. The beetle was searched for using two methods: grubbing amongst roots of heather; and setting pitfall traps baited with fermenting apple. Details of habitat condition and management techniques were noted for each site. At two of the larger locations, Aish Tor, Dartmoor and Aylesbeare Common in Devon, pitfall traps were set in a variety of suitable sites where winter burning, cutting and grazing had been applied as different habitat management techniques and the numbers of P. kugelanni caught recorded. At Aish Tor, a baited pitfall trap was set between 7th and 21st June 2010 and 3rd and 22nd May 2011 in each of six areas. At Aylesbeare Common, 11 baited pitfall traps were set between 8th and 22nd June 2010 in areas which had been subject to different management techniques. Winter burning was not used here. Five of these traps were set again in May/June 2011. Results At Aish Tor 57 of the 70 beetles captured came from just two traps in areas managed by controlled winter burning and cutting of fire breaks every 3 to 4 years. At Aylesbeare Common no beetles were trapped in areas grazed by cattle and ponies. 40 of the 94 beetles caught came from one trap in an area managed by cutting but there were no beetles found in another cut area. The vegetation here was cut back more severely than at Aish Tor and the cuttings removed. A third technique used at this site was scraping a thin layer of topsoil to disturb the ground and allow regeneration of heather every few years. The remaining seven traps were set in areas managed in this way. One trap caught 35 beetles and another caught 13 beetles but most caught none. Visits to the remaining localities confirmed that good populations of P. kugelanni were present where controlled winter burning was in use. Discussion Creation of bare ground and early successional heath land habitats are essential for P. kugelanni and also benefit many other specialist heath land invertebrates including bees, wasps and ants. Small-scale winter burning of heather and gorse between October and February is an excellent method of creating early successional habitats. It was noted that in an area at Aish Tor now managed purely by cutting, the number of beetles had fallen considerably since a survey on 2000/2001 when an accidental fire in the mid 1990s was presumed to have improved the habitat. Cutting firebreaks amongst heather and gorse on heath land was effective as long the cut vegetation was removed. The vegetation must be cut shorter than 10cm. On Dartmoor burning areas of habitat appeared to be a more effective management technique than cutting. Turf-stripping small areas and disturbing the ground every 4 years to maintain early successional habitats had proved to be an effective method of creating good habitat on the east Devon pebblebed heaths. Cutting rank gorse and heather and removing or burning cut material was an effective method for opening up habitats to allow access for grazing animals such as sheep, ponies, cattle and rabbits. It was essential that the cut vegetation was removed. Grazing was maintaining

SID 5 (Rev. 05/09) Page 13 of 34

suitable habitat on some sites but the numbers of P. kugelanni found were not as high as on sites managed in other ways. The small number of samples was a limitation on the validity of the findings. Conclusion Controlled winter burning appears to be the most generally applicable technique for maintaining the early successional heath required by P. kugelanni. However, cutting vegetation to less than 10cm every 3 to 4 years (and removing the cut vegetation) or turf-stripping and ground disturbance every four years has been used successfully on specific sites. 3. Amara fusca Introduction Although there are a very few historic records of the ground beetle Amara fusca from Kent and Crymlyn Burrows near Swansea and one recent record from Newport, this beetle has always been considered a specialist of the East Anglian Brecks. It can be found in litter at the base of plants during the day but the beetles climb the food plant after dark to feed on seeds. The objective of this study was to determine the environmental and ecological factors characteristic of sites inhabited by A. fusca. Methods All existing British records of A. fusca were examined for information on the ecological requirements of the beetle. Sites in the Suffolk Brecks that were known to have supported A. fusca [Brandon Artemisia Reserve (Suffolk Wildlife Trust), Maidscross Hill and Wangford Road verge] were visited on 24th August 2010 and again in June and September 2011. Diurnal hand-searching was used to locate populations and the presence and abundance of potential food plants were recorded. Also a 10 hour diurnal and nocturnal search at Crymlyn Burrows was undertaken on 21st September 2010. Results Single specimens of A. fusca were collected on each of the three visits to Brandon Artemisia Reserve by hand-searching in moss at the base of Field Wormwood Artemisia campestris plants. An independent worker collected another specimen from this site by sweeping vegetation about an hour after dusk on 15th October 2011. The beetle was not seen at Maidscross Hill or Wangford Road verge or at Crymlyn Burrows during the study. A PhD student recorded a single A. fusca from Weeting Heath, Norfolk in 2011. Discussion The existing records clearly indicated that Artemisia campestris was the main food plant of the beetle. Only one of the sites visited, Brandon Artemisia Reserve, supported stands of this plant and this was the only site where A. fusca was found in both years. A few existing records indicated that the beetle will use Mugwort (Artemisia vulgaris) but although this plant was present at the other two Suffolk locations the beetle was not found here. A few plants of Artemisia vulgaris were present at Weeting Heath but were unlikely to be sufficient to support a population of A. fusca and the beetle captured here was probably a vagrant from the nearby Brandon colony. It is possible that Artemisia crithmifolia (formerly Artemisia campestris subsp. maritima) was the food plant for the Crymlyn Burrows population. Artemisia vulgaris also grows here but neither plant appeared to occur in large stands any more. Only four plants of Artemisia crithmifolia could be found in 2011 so it is not surprising that the beetle could not be found. Conclusion The key factor determining the distribution of A. fusca appears to be the presence of its main food plant Field Wormwood, Artemisia campestris. Artemisa crithmifolia may be a suitable food plant also but is even more restricted in occurrence than Artemisa campestris. Although the beetle has been reported from Mugwort, Artemisia vulgaris, it seems likely that if such a common plant were suitable then the beetle would be more widely distributed in Britain. 4. Anisodactylus nemorivagus Introduction Anisodactylus nemorivagus is a medium-sized ground beetle that inhabits sandy heath lands. The objectives of this study were to determine the current English distribution of A. nemorivagus and to characterise more precisely the micro-habitats the beetle inhabits. Methods Existing records of A. nemorivagus were collated and examined for information that could be used to characterise its habitat. Data were requested from the national Carabidae recording scheme and individual coleopterists with knowledge of the beetle were contacted for further details of their records and others they may know of. The most recent records were used to target surveys by local coleopterists. Field work focussed on known grid reference points and other areas of apparently suitable habitat. Five sites in the New Forest (Hampshire) were each visited on at least two occasions between 26th May 2010 and 16th May 2011. Searches were made by grubbing at the base of heather and two baited pitfall traps were set for a week between 26th May and 2nd June 2010 on three of the sites (Parc Pale, Matley sandpit and Peel Hill) and again at Peel Hill between 4th and 26th April 2011. Three baited pitfall traps were placed at Cunniger Bottom, Deadman’s Hill, New Forest between 26th April and 19th May 2011 and four traps were placed at Picket Post, New Forest between 27th April and 19th May 2011. Two Surrey sites (Chobham Common and Brentmoor) were searched by grubbing at the base of heather on 13th October 2010 and Brentmoor was revisited on 17th May 2011. A site with older records at Cavenham Heath, Suffolk was visited on 22nd September 2011. Although there were no previous records, Iping Common, Sussex was surveyed

SID 5 (Rev. 05/09) Page 14 of 34

on 14th October 2010 and 26th September 2011. The general habitat present in the location of previous records was noted as were the details of the precise location of any specimens of A. nemorivagus found during the field visits. Results The desk study showed records of A. nemorivagus from across southern and eastern England from Cornwall to Norfolk with the great majority of records coming from lowland heath in Dorset, Hampshire and Surrey. Modern (post-1990) records were located from one site in Dorset, four sites in Hampshire, two sites in Surrey and one site in Sussex. The only A. nemorivagus caught in the traps was a single male at Matley sandpit. Three beetles were collected by grubbing at Peel Hill in 2011. These field records and limited information available from the desk study suggested that the beetle occurs in low growing heather on south-facing slopes. Areas of bare sandy ground and often a thin layer of peaty soil were characteristic of the sites. Discussion A. nemorivagus is a difficult species to study as it seems to occur in low density throughout its range though this may be the result of a lack of knowledge about its ecology and precise habitat requirements. Certainly it appears to be a difficult species to locate even at sites where it is known to occur. It has a rather early season of adult activity from late March to May when coleopterists may be less active in field recording. Maintaining suitable ecological and environmental conditions for the survival of A. nemorivagus will require site management. Observations at the sites visited indicate that heather may be maintained in a suitable condition by grazing by cattle, ponies or rabbits, by cutting to form firebreaks or by turf-stripping to produce areas of regenerating heather. Conclusion The current English distribution of A. nemorivagus is as thinly scattered populations in suitable pockets of habitat over large areas of heath land in Dorset, Hampshire, Surrey and Sussex. The limited information available suggests that the beetle occurs in low growing heather with areas of bare sandy ground and often a thin layer of peaty soil on south-facing slopes. 5. Harpalus froelichii Introduction Previous studies of the ground beetle Harpalus froelichii have shown the food plant of the beetle to be Fat Hen Chenopodium album. It has been suggested that soil disturbance in areas of the Brecks close to existing colonies of H. froelichii may encourage growth of C. album and colonisation by the beetle. The objective of this study was to assess the success of turf-stripping and soil rotavation in the creation of habitat suitable for H. froelichii. Methods In recent conservation work targeted at the Grey Carpet Moth Lithostege griseata Butterfly Conservation had established approximately 60 bare ground plots by turf-stripping and rotavation. These plots varied in size from a few square metres to narrow strips over 200m long. Butterfly Conservation staff advised on which of the plots had C. album established on them and so were potential new habitat for H. froelichii. A total of eleven plots at Aspall Close, Ramparts Field, Kings Forest and Maidscross Hill (all in Suffolk) and Cranwich Heath, Norfolk were searched by hand on 23rd or 24th August 2010. Aspall Close, Ramparts Field and Maidscross Hill were visited again in 2011 on 7th June and 22nd September. Results The plots at Aspall Close, Kings Forest, Maidscross Hill and Cranwich Heath supported only sparse growth of C. album and no H. froelichii were found here. C. album was growing in dense stands at Ramparts Field in 2010 when a single beetle was found. The plot had become overgrown in 2011 and there were few C. album plants. No H. froelichii were found. At Aspall Close there was a patch of bare ground on the site of a large bonfire sited near to the Butterfly Conservation plot. C. album was growing densely here and searching amongst the roots produced two H. froelichii in 2010 and a single beetle in June 2011. Discussion Creation of bare ground plots appears to be an effective method of creating habitat for H. froelichii as long as C. album is present in sufficient quantity to provide enough seed to support the beetle. This beetle is obviously very active and able to find small areas of suitable habitat quickly. It seems likely that suitable areas of habitat for the beetle are regularly created by agricultural activities in sandy, arable field margins in East Anglia. Unfortunately the food plant is considered a weed species by farmers. It may be worth investigating if the beetle can survive on food plants likely to be more acceptable to farmers. A single specimen of H. froelichii was found in Lincolnshire in 2005. This may have been a stray from Breckland or an indication that other populations remain to be discovered in that area. Conclusion Bare ground plots created by turf-stripping are an effective method for creating habitat for H. froelichii as long as they contain stands of C. album. The beetle appears to be well adapted to using small patches of ephemeral habitat. To maintain populations at these sites re-stripping or disturbing part or all of each plot during the winter every 2-3 years would be the most appropriate method for sustaining populations of the beetle and its food plant. 6. Ophonus laticollis Introduction

SID 5 (Rev. 05/09) Page 15 of 34

Ophonus laticollis is a medium sized, seed-eating ground beetle which inhabits partially vegetated ground, including field margins on chalky soil. The objective of this study was to determine the current English distribution of the beetle. Methods Existing records of O. laticollis were collated. Data were requested from the national Carabidae recording scheme and individual coleopterists with knowledge of the beetle were contacted for further details of their records and others they may know of. The most recent records were used to target surveys by local coleopterists. Field survey initially focussed on Gallows Hill, Thetford, the most reliable British site for finding the beetle, which was visited on 24th August 2010 and 7th June 2011. The survey method used was hand searching for one hour at the junction between the grass field margin below the hedge and the cultivated ground. Cranwich Camp, Norfolk, another breckland site, was surveyed on 25th September 2011. Results The desk study showed that historic records of O. laticollis were widespread across England south of a line from the Humber to the Severn Estuary. Modern records (post-1990) were located from six sites in West Norfolk, one site in East Norfolk, one site in East Suffolk, one site in Buckinghamshire, one site in Oxfordshire, one site in Somerset and one site in East Sussex. The only specimens of O. laticollis found during the field visits were two individuals at Gallows Hill in 2011. However, an independent survey of two farms adjacent to Gallows Hill provided several records of O. laticollis in 2010 and on 24th August 2011 a new site for the beetle was located at Lodge Hill, near Princes Risborough, Buckinghamshire. Discussion The results suggest that although O. laticollis appears to have modern records from across much of its historic range, East Anglia and in particular the Norfolk Brecks constitutes a stronghold for the species. Even where the beetle occurs it can have a very localised distribution and it would be useful to discover the reasons for this. No individuals were found during the first visit to Gallows Hill in 2010 as the specific field margin searched did not support the beetle whereas it does occur in nearby margins. To enable a wider number of people to identify and search for the beetle it would be valuable to produce an identification guide to Ophonus species similar to that produced for Pogonus species and use this to target searches for the beetle within its historic range. Conclusion The current English distribution of O. laticollis is centred on the Norfolk Brecks but it appears to occur very locally across southern England. This patchy occurrence is repeated at smaller scales, at least down to tens of metres. 7. Ophonus melletii Introduction Ophonus melletii is medium sized, seed-eating ground beetle which inhabits open and partially vegetated ground on chalky and sandy soils. The objective of this study was to determine the current English distribution of O. melletii. Methods Existing records of O. melletii were collated. Data were requested from the national Carabidae recording scheme and individual coleopterists with knowledge of the beetle were contacted for further details of their records and others they may know of. The most recent records were used to target surveys by local coleopterists. Field visits were made to Howell Hill, Cheam, the most reliable British site for finding the beetle, on 17th May 2011 and 22nd August 2011. Two other sites, both undercliffs supporting wild carrot, were surveyed. These sites were Branscombe, Devon (25th August 2011) and Tregantle, Whitsand Bay, Cornwall (10th May 2011 and 9th September 2011). All sites were searched by grubbing in vegetation and searching in the seed heads of wild carrot. Results The desk study showed that O. melletii has occurred in many English counties south of a line from the Humber to the Severn Estuary. Modern records (post-1990) were located from two sites in Kent, two sites in Cambridgeshire and one site in each of East Sussex, Surrey, Buckinghamshire and Norfolk. The only location where it has been found reliably in recent years is Howell Hill, Cheam in Surrey and two male specimens were found here again in 2011. The beetle was not found at either Branscombe, Devon or at Tregantle, Whitsand Bay, Cornwall. Discussion O. melletii appears to be a very localised species with most recent records occurring at new locations including at least one brownfield site. Difficulties in separating O. melletii from related species may be part of the reason for the paucity of records. To enable a wider number of people to identify and search for the beetle it would be valuable to produce an identification guide to Ophonus species similar to that produced for Pogonus species and use this to target searches for the beetle within its historic range. Conclusion Currently O. melletii appears to have a relatively stable population at just one location in southern England. Away from this location it appears to have a very localised distribution and is unpredictable in occurrence. 8. Formica exsecta Introduction Chudleigh Knighton Heath, Devon is the last remaining site in England for the Narrow-headed Ant Formica exsecta. The number of nests on the site has been monitored for over a decade. In recent years this has been

SID 5 (Rev. 05/09) Page 16 of 34

extended to monitoring nests on areas of the site subject to different management techniques. Although, the impact of swaling and mowing on the nests within one management compartment continued to be monitored over the two years of this study, the main objective was to investigate the conditions required for nest establishment by F. exsecta. Methods Twenty one artificial ‘stumperies’ were set up during 2009, and another 22 existing cut tree stumps were identified for future monitoring. Five additional existing cut stumps were identified in 2010 and during winter 2009/10 a line of over 20 stumps was set up on Highways Agency-managed land, on the A38 footpath verge. In August 2011 ten additional existing stumps were identified for future further monitoring. The presence of ant nests and the species involved were noted in all identified ‘stumperies’ and stumps in 2010 and again in May and August 2011. Results The majority of the newly created ‘stumperies’ had not been colonized by ants by 2011 but three showed signs of ant activity, in the form of chewed sawdust material present under bark, though no ants were seen. Also, two of the ‘stumperies’ placed near to existing F. exsecta nests had been incorporated into those nests. One of the older existing stumps supported a F. fusca nest in August 2011 and single workers were seen on two separate occasions amongst sawdust within the chainsaw slots of a cut stump in May 2011. Ten other existing stumps showed signs of ant colonization in 2011 and at least four of these were active Lasius nests. Discussion The stump found with the active F. fusca nest was noticeably older (possibly >10 years) and more decayed than any of the existing stumps or ‘stumperies’ on the site. This suggests the newly created ‘stumperies’, although these were created from pre-cut and partly decayed wood, 4-5 years old, are not yet sufficiently rotted for colonization and succession to occur. However, it is clear that existing nests can utilize these resources in suitable conditions and longer term monitoring may yet provide valuable information on encouraging recolonisation of sites by F. exsecta. The presence of Lasius nests in at least four existing stumps probably excluded F. exsecta as it is known to be outcompeted by Lasius species. Although nests were lost and new ones established each year, both swaling and mowing techniques produced net gains in both years. In 2010 there were two extra nests in the swaled area compared with four extra nests in the mown area. In 2011 the gains were six and nine nests respectively. Some scrub is required for collection of prey items by the ants but as shading increases the warm, sunny conditions required for nest development are lost. Swaling and mowing can both control scrub on the site but the very limited, initial results show a greater net gain of nests following mowing. From previous years it has been seen that swaling has the potential to damage or destroy nests where transient localised conditions, for example a change in wind direction or concealed accumulation of combustible litter, causes fires to burn downwards rather than sweep quickly across surface vegetation. Conclusion A longer term experimental trial and monitoring would seem to be needed to record succession of ant species colonizing standing dead wood stumps but this study has provided useful baseline data for further work. 9. Tapinoma ambiguum and Tapinoma erraticum Introduction The ants Tapinoma ambiguum and T. erraticum have been separated only recently and the taxonomy is still uncertain. Differences in the habitat preferences of the species pair are unclear although some British workers have suggested that T. ambiguum favours vegetation communities on wetter peat whereas T. erraticum favours drier sand. The objective of this study was to determine the ecological and environmental characteristics of sites inhabited by each of the ants. Methods The study was based at Godlingston Heath, Dorset as both species had apparently been recorded here in the past. However, the species can only be separated by relatively few experts so specimens from each nest studied were collected at the same time as the ecological and environmental variables of vegetation type and soil type were noted. Data were collected from 25 nests in 2010 and a further 10 nests in 2011. In 2011 further variables measured were the percentage of bare ground, average sward height, soil temperature and soil moisture. These variables were also measured at ten quasi-random sites where nests were not present. Results When the ants were eventually identified all were found to be T. ambiguum. A Pearson Chi-Square test on data collected in 2011 demonstrated a significant association between soil type and T. ambiguum nests (X=95.139, DF=3, P=0.001), with sandy soils supporting most T. ambiguum nests. Relationships with other variables were tested using a Kruskal-Wallis test. This suggested that nests were more likely to occur in areas with large amounts of bare ground (H=73.74, DF=1, P=0.001), where soil temperature was high (H=74.63, DF=1, P=0.001) but soil moisture was low (H=99.79, DF=1, P=0.001). Discussion The nests found on Godlingston Heath were all within heathland but no link was found with specific vegetation classes. This is supported by various European studies where the ant has been found to be strongly associated with a range of heathland types, both wet and dry. The results suggest that T. ambiguum is a typical lowland heath specialist depending on patches of bare, free draining soil to provide the high soil temperatures it requires. However, the small amount of data collected was a major limitation of the study and a larger data set from a wider range of sites would need to be collected before general conclusions can be drawn. The preference of T.

SID 5 (Rev. 05/09) Page 17 of 34

erraticum for drier sand could not be tested as it was not found at Godlingston and although it was found on two other dry heaths in Devon, these did not provide sufficient data for further analysis. The European studies have found T. erraticum occurs in a wider range of habitats than T. ambiguum and one respected researcher has stated that it does not occur on sand. Conclusion Limited results suggest that T. ambiguum prefers patches of bare soil on sandy heath to provide the high soil temperatures it requires but more data are needed to support these findings. Insufficient data could be collected to determine the habitat preferences of T. erraticum. 10. Homonotus sanguinolentus Introduction Previous research has shown that the larvae of the Bloody Spider-hunting Wasp Homonotus sanguinolentus are ectoparasitoids of the spider Cheiracanthium erraticum. The wasp has been recorded from lowland heaths in the Isle of Purbeck and Wareham Forest in Dorset; the New Forest in Hampshire and the Thames Basin heaths in Surrey but was assumed to have disappeared from the latter. The objective of this study was to determine the current English distribution of H. sanguinolentus. Methods A desk study involved collation of all previous records of H. sanguinolentus. The results of the desk study enabled the selection of sites for field survey. Five sites in Hampshire between 5th and 28th August 2010, six sites in Dorset between 12th July and 12th August 2010 and fourteen sites in Dorset between 30th June and 3rd August 2011 were surveyed. Suitable habitat in the vicinity of previous records was searched. As sightings of adult H. sanguinolentus are rare, efforts were concentrated on locating the egg-laying retreats of the host spider in the tops of grasses and other vegetation, primarily in bogs and wet or humid heathland. A small hole was made in the retreat using fine forceps and if a spider was present it was checked for the presence of a larva of H. sanguinolentus attached to its abdomen. The possibility of confusion with C. pennyi, which is almost indistinguishable from C. erraticum in the field, did not alter the survey methods although it does mean that the apparent absence of parasitized spiders on a site cannot be taken as a reliable indicator of the absence of H. sanguinolentus. Results In 2010 no larva of H. sanguinolentus were discovered in the 68 C. erraticum retreats checked but a pinkish pupa that may have been of this species was found in one retreat from each of Godlingston Heath, Dorset; Vales Moor, Hampshire and Kingston Common, Hampshire. In 2011 larva were found attached to spiders at three sites, namely Godlingston Heath (2), Creech Heath (2) and Wareham Forest (1), and a total of 55 Cheiracanthium retreats were found across twelve of the sites with almost half of these (21) at Holt Heath. No field survey was undertaken on the Surrey Heaths but a single female wasp was reliably reported from Witley Common, Surrey on 14th July 2011. Discussion H. sanguinolentus is a very difficult animal to detect using current techniques. The ectoparisitoid larva collected during this study probably belonged to this species but more work would be required to confirm this. The rediscovery of an adult wasp on the Surrey Heaths in 2011 after an apparent absence of many years, despite recent targeted searches by local entomologists, illustrates the problems of confirming species distribution with absolute certainty. It seems likely that H. sanguinolentus is still present in low density on lowland heath throughout its historical range in southern England. Conclusion The current distribution of H. sanguinolentus in England probably is restricted to lowland heath in Dorset, the New Forest and Surrey. 11. Pseudepipona herrichii Introduction Previous research has identified four resources that must be present on a site occupied by the Purbeck Mason Wasp Pseudepipona herrichii, namely exposed clay for nesting; water to aid in nest construction; Bell Heather Erica cinerea to feed adult wasps and caterpillars of the moth Acleris hyemana as the sole known larval prey item. These resources can be found widely in Britain but historical records suggest P. herrichii is restricted to lowland heath in the Poole Basin, Dorset. The objective of this study was to determine the current English distribution of P. herrichii. Methods A desk study involved collation of all previous records of P. herrichii. The results enabled the selection of sites for field survey. In 2010 six heaths were visited between 24th July and 12th August. In 2011 seventeen heaths, including all from which P. herrichii was known, were visited between 15th June and 14th July. On each visit the numbers of adult wasps and nesting burrows on the site were noted. Results The desk study located 50 historical records of P. herrichii, mostly from 11 extant heaths within the Isle of Purbeck, Dorset. A total of 4 adult wasps and 153 burrows were counted from five of the six sites surveyed in 2010. In 2011 a total of 69 adult wasps were found across seven sites and 365 burrows were counted on six sites. Both wasps and burrows were observed at five of the known sites, Stoborough Heath RSPB reserve,

SID 5 (Rev. 05/09) Page 18 of 34

Grange Heath, Hartland Moor Tramway, Slepe Heath and Godlingston Heath in both 2010 and 2011. Wasps and burrows were observed at Creech Heath in 2011 only. Neither wasps nor burrows were found at the remaining known sites in either year. A single female wasp was observed at Arne RSPB reserve in 2011. Discussion Despite sub-optimal weather conditions in 2011, the number of wasps and nests observed showed a significant increase on the data for 2010. There appears to be a core of five to six sites where can be found reliably and presumably these sites support the strongest populations of the wasp. Targeted field survey by Hymettus in 2009 had found the wasp also on four of these heaths. The species is seen less regularly at Upton Heath, Stoborough Heath NNR, Morden Bog, Wareham Forest and Knighton Heath. Suitable habitat for nesting appeared to be present still at Upton Heath, Morden Bog and Wareham Forest but site management could improve the suitability of these and the remaining sites. Arne is a new location for the species but no evidence of a breeding population in the form of nest burrows was located and the wasp observed may have been a vagrant. Holt Heath and Sopley Common & Ramsdown are considered likely to be unsuitable for P. herrichii because of the absence of clay seams, necessary for nesting, within the sandy to gravelly substrates. Conclusion Although very restricted in distribution, P. herrichii appears to be well established within its strongholds in Dorset. 12. Andrena tarsata Introduction An attempt to undertake an initial desk study showed that the necessary data to assess localised changes in the status and distribution of Andrena tarsata was was not available. In particular historical records of both the bee and of its main forage plant Tormentil Potentilla erecta had not been collected with sufficient precision to enable investigation of any correlation in distribution of the two species. In January 2011 it was agreed that the study of A. tarsata would be progressed through a new objective to develop and test a monitoring protocol for detecting the presence of the bee and a method of recording its habitat quality. The objective of determining the nature of the foraging resources utilized by adults of A. tarsata remained unchanged. Methods Pollen loads were collected from three females of the bee A.tarsata from sites on Fylingdales Moor, North Yorkshire in July 2010. The pollen was subjected to microscopic examination. In 2011 a monitoring protocol for A. tarsata and the quality of its habitat was drafted and tested at sites on the North York Moors. The protocol allowed for two different methods of assessing forage quantity and quality because of the wide variation in growth habit of the likely forage plant, Tormentil Potentilla erecta. Firstly, surveys were conducted using a small quadrat (0.25m²). Where foraging A. tarsata were identified, the quadrat was placed over the area of P. erecta they were foraging on, a photograph taken, and a count made of the number of fully open flowers. Measurements were taken of the height of each plant and of the maximum width of the plant. The aim of this sampling was to quantify the amount of bloom available on the plants the bees chose to visit. This was not designed as a random sampling of the whole area. A small size of quadrat was chosen because the individual plants were not large. Secondly, systematic belt transect surveys were conducted using a larger quadrat (0.50m²) at 10m intervals along the centre line of each site. The aim of this sampling was to look at the general distribution of P. erecta within each chosen site, as well as to quantify the available flowers. The height of each plant was measured also. Results Microscopic examination of the pollen indicated the bee was collecting only from plants of the genus Potentilla. A. tarsata was found in low numbers from just three sites in 2011. Discussion Although many more samples would need to be examined for a firm conclusion, from botanical survey it appeared likely that P. erecta was the main source of pollen, at least in the area surveyed. Silverweed P. anserina and another creeping species, possibly P. reptans, may be utilised at some sites. The bee does use other species of Potentilla in Europe as well as other Rosaceae (Fillipendula and Rubus) and has been observed visiting Calluna for nectar. The small number of bees observed during field work in 2011 prevented the collection of sufficient data for a statistical valid test of the monitoring protocol drafted to predict habitat quality for A. tarsata. Conclusion Tormentil P. erecta appears to be the main pollen resource utilised by A. tarsata in North Yorkshire. The monitoring protocol requires further validation to confirm that it can predict habitat quality for A. tarsata. INVERTEBRATES OF FLOWER-RICH GRASSLAND 1. Odynerus melanocephalus Introduction Previous observations of the mason wasp Odynerus melanocephalus on several post-industrial sites in Warwickshire resulted in the conclusion that the sole prey items collected by the wasp to feed its own larvae were larvae of the weevil Hypera postica taken from Black Meddick Medicago lupulina. The objective of this study was to assess the validity of this conclusion for populations of O. melanocephalus elsewhere in England. Methods Previous locations of colonies of O. melanocephalus at Alphamstone, Essex; Salisbury Plain, Wiltshire and Roughlands Cliff, Isle of Wight were visited. The only extant colony located was that at Roughlands Cliff. This site was visited on 4th, 17th and 18th June 2010. Larvae were collected from female wasps returning to their nest after

SID 5 (Rev. 05/09) Page 19 of 34

foraging and from the Birds-foot Trefoil Lotus corniculatus on which the wasps were foraging. A search for M. lupulina revealed the nearest patch to be 220m away. No wasps were observed visiting the patch and sweeping the plants produced no Hypera larvae. Results Provisional identification of the two paralysed weevil larvae taken from female wasps suggested that the species involved was H. plantaginis and rearing larvae collected from L. corniculatus confirmed the identification. Discussion H. plantaginis is the species of Hypera normally associated with Birds-foot Trefoil and was probably the prey item utilised by a colony of O. melanocephalus in North Wales that has been reported as foraging at Birds-foot Trefoil. The results suggest that the wasp larvae are able to develop on a diet of at least two different species of Hypera weevil. It is not clear whether different populations prey on only one species of weevil or if the prey taken is more dependent on availability of prey items. Conclusion O. melanocephalus larvae are able to develop on at least two species of Hypera weevil. 2. Cerceris quadricincta and Cerceris quinquefasciata Introduction The solitary wasps Cerceris quadricincta and C. quinquefasciata stock their nest burrows with adult weevils as food for their larvae. Previous studies report a wide variety of quite common larval prey weevils can be utilised by both wasp species and the main prey species may differ from year to year in any one location. The objective of this study was to determine if the distributions of the wasps could be related to any geographical variation in prey items taken. Methods Active nests of C. quadricincta and C. quinquefasciata were located by searching likely nest locations for signs of female wasps and by searching for the cleptoparasite Hedychrum niemelai in the hope that it would lead to nests. In 2010 C. quinquefasciata was sought at five sites in Suffolk and one in Essex from which it had been recorded previously. One site in Colchester, Essex was visited for C. quadricincta. Two sites in Colchester and one at Ramsgate, Kent were visited in 2011. At each site any female wasp returning to the nest was collected and any prey items carried was removed for identification before the wasp was released. Results In 2010 no nests were located at any of these sites and only a few individual adult C. quinquefasciata were observed. In 2011 weevils were obtained from females of C. quadricincta at one of the Colchester sites and at the Ramsgate site. All of these weevils were identified as the very common Sitona lineatus. Discussion The results of this study add little to previous research merely confirming that common weevils form an important resource. On this basis it appears unlikely that prey distribution is a limiting factor in the distribution of either wasps. What may be a limiting factor is the ability to find prey in sufficient quantity to support large colonies. Whilst data on nest characteristics are not available for C. quadricincta, it may be possible to draw some parallels from the closely related C. quinquefasciata. Nests comprise up to ten cells, each of which usually has about 50 prey items, although this can be as high as 82 in a single cell. Thus, a single finished nest may hold approximately 500 weevils. Multiplying that up by the number of females in a viable colony, it is clear that these wasps are likely to need a weevil-rich resource relatively close to the nesting site. This food resource can only be provided by a botanically rich and/or very extensive site. Conclusion There is no evidence that the distributions of C. quadricincta and C. quinquefasciata are related to any geographical variation in prey items taken. 3. Lasioglossum angusticeps Introduction The small mining bee Lasioglossum angusticeps has a very restricted distribution along the south coast of England from Dorset to the Isle of Wight. Females of L. angusticeps cannot be distinguished in the field from those of the closely related and much more common L. punctatissimum creating difficulties in field studies. The objective of this study was to determine the nature of the pollen resources used by L. angusticeps. Methods Burning Cliff, Ringstead Bay, Dorset was visited on 4th June 2010 to collect females of the bee Lasioglossum angusticeps carrying pollen loads. Eleven live females were collected. A second visit in August 2010 allowed males to be associated with the females thus confirming their identity. In addition pollen from seven museum specimens collected from Dorset (Chiddeock, Lyme Regis and Ringstead) and one specimen from the Isle of Wight was obtained for analysis. Results All of the pollen loads consisted almost entirely of Birds-foot Trefoil Lotus corniculatus. One of the museum specimens carried two pollen grains of a smallish Trifolium sp. in amongst thousands of Lotus grains but this was probably accidental. Discussion

SID 5 (Rev. 05/09) Page 20 of 34

No pollen records have been traced for the bee in mainland Europe or elsewhere within its world range. The species is tentatively regarded on available evidence as being at least narrowly oligolectic on Fabaceae and possibly even monolectic on Lotus corniculatus (though this would seem most unlikely given the bee’s vast geographical range). Female L. angusticeps may visit flowers of other species of Fabaceae but this has not been confirmed. It would appear that availability of foraging resource is not a factor limiting the distribution of L. angusticeps in England. Conclusion On available evidence the sole pollen resource utilised by L. angusticeps in Britain is Lotus corniculatus. 4. Colletes floralis Introduction The UK populations of the Northern Mining Bee Colletes floralis may represent as much as 50% of its global range. However, the species is most frequent in Scotland and Northern Ireland with an isolated English population in Cumbria. The objectives of this study were to determine the environmental and ecological features characteristic of the habitat of C. floralis and to identify the pollen resources utilised by the bee. Methods A large amount of research on C. floralis, mainly in Ireland and Scotland, has been undertaken over the last decade. A literature search was undertaken to identify the unpublished reports, theses and published papers resulting from this research. Information detailing aspects of the autecology of C. floralis was extracted from this literature. Field survey at Sandscale Haws, Cumbria in 2010 enabled pollen samples to be collected from 10 female bees for microscopic analysis. Results C. floralis is strongly associated with coastal habitats. It has been recorded from fixed dunes, fore dunes and machair grassland. Southerly aspects are preferred for nesting but are not essential as the Sandscale Haws population is found along a trackside bank with an easterly aspect. Nests are not found along the high tide mark but further inland within vegetated zones although coarse grass communities have been shown to be negative attributes of sites. Water and urban areas are significant barriers to dispersal of the bees over the medium to long distances. Colletes floralis has been regarded as polylectic though the proportions of pollen types within loads differ between populations. Apiaceae appear to play an important role as the most frequently foraged plant group (despite its proportion of availability at each site). Pollen from Fabaceae, Rosaceae and Lamiaceae has been reported to be utilized in relation to its proportion of availability at a site. Asteraceae pollen has been reported as being avoided but the pollen loads from Sandscale Haws comprised as much as 65% from this group, probably from Smooth Hawkbit, Crepis capillaris. Discussion All of the habitats which support C. floralis are flower-rich, open mosaic habitats with short turf, bare ground and varied topography. Varied topography appears to be essential in providing suitable nesting situations for the female bees. The varied topography may also serve to protect nesting and/or key foraging areas from prevailing winds. The location of nests away from the high tide mark suggests a need for some stability in the substrate and protection from the elements. Vulnerable nesting and forage sites may require protection from storm surges and from development as golf links. Coarser grasses are likely to be problematic as these communities encroach and inhibit more floristically-rich, short turf communities favoured by C. floralis. Genetic differentiation between bee populations has been suggested as playing a role in pollen selection. Conclusion The habitat preferences of C. floralis include bare ground, short and floristically-rich turf and varied topography. Coarse grass communities are negative attributes to sites. Although C. floralis is polylectic there appears to be a clear preference for pollen of Apiaceae. Asteraceae, Fabaceae, Rosaceae and Lamiaceae are also utilised and the proportions in pollen loads differ between populations. 5. Osmia parietina Introduction The Wall Mason Bee Osmia parietina is a scarce bee found sporadically in western and northern Britain but it has been recorded from only eleven English sites, all in Cumbria and Lancashire, in very low numbers. The objectives of this study were to determine the environmental and ecological features characteristic of the habitat of O. parietina and to identify the pollen resources utilised by the bee. Methods Four locations in North Lancashire and Cumbria where O. parietina had previously been recorded were visited during May to July 2011. Carnforth Ironworks Slag Banks, Yealand and Thrang End Allotments, Gait Barrows NNR and Hutton Roof Crags were each visited twice. Areas from which the bee had been recorded in the past were searched and patches of potential forage were monitored by a system of random walks. In addition Warton Crag was searched for O. parietina as part of a World Museum Liverpool survey for Lancashire Wildlife Trust. Results No specimens were recorded at the first four locations but a single female was seen at Warton Crag. No pollen was obtained from O. parietina during this study but samples collected previously from twelve females at three sites in North Wales and from four bees in NW England have all consisted of Birds-foot Trefoil Lotus corniculatus. Discussion

SID 5 (Rev. 05/09) Page 21 of 34

Previous records from NW England indicate that O. parietina is found mainly inland in sheltered warm situations with bare rock surfaces of damaged limestone pavement and sparse patches of L. corniculatus. Carnforth slag bank mimics this habitat with bare reflective slag surfaces rich in phosphates and carbonates favouring L. corniculatus. Records of the bee in Wales are predominately coastal and include a wider variety of natural and disturbed habitats. Limestone sites are limited to coastal quarries and abandoned copper mines. Sites on the Llyn Peninsula are predominately on hard metamorphic rock where nests are confined to rock fissures and dry-stone walls. One population was recorded nesting in old beetle burrows in the trunk of a fallen veteran oak, a situation similar to that reported from continental Europe. Although the bee is described as polylectic in Europe, it is not unknown for insects to have restricted larval diets at their northern range edge so it is not so surprising that this may also be the case for O. parietina. Conclusion The bare rock surfaces and calcareous conditions favoured by the English populations of O. parietina are clearly not typical of the bee’s habitat preferences elsewhere in Britain and Europe. A supply of suitable nesting cavities and a supply of pollen would seem to be important requirements. On current evidence O. parietina appears to be monolectic on L. corniculatus pollen in Britain. 6. Eucera longicornis Introduction In May 2007 the Long-horned Bee Eucera longicornis was discovered on an orchard site at Heightington, Worcestershire. This represented one of the most northerly records of the bee and raised the question of how the colony could be encouraged. The objective of this study was to investigate the influence of different grassland management regimes on the conservation and specifically the foraging activity of E. longicornis. Methods A set transect was walked on 7 occasions during June and July 2010 and again from May to July 2011, following the route used by the group responsible for the original records. Land adjacent to the path, the orchard and fields belonging to 3 farms in the surrounding area was included in the survey for nest sites and foraging activity. When actual or potential nest sites and forage resources were located a note was made of the management regime. Results E. longicornis was not seen at Heightington in 2010 or 2011. Several fields at Old House Farm supported a relatively rich flora including potential forage resources and one field contained a small vertical bank resulting from a cart track which supported nests of many mining bees and potter wasps. The fields were being lightly grazed by sheep. The fields at Chapel Farm were heavily grazed by sheep and provided few potential forage resources but a south facing landslip did provide potential nest sites. Most of the fields at Norchard Farm were improved grassland but one floristically rich hay meadow provided potential forage for the bee. Discussion In the absence of the bee no trial of grassland management regimes could be undertaken at the planned experimental site. The failure to find E. longicornis over two seasons suggests that this species may be extinct in the immediate vicinity of Heightington although the bee was recorded from a site less than 15km away in 2011. However, suitable foraging plants were available, though not as profusely in 2011 as in 2010, possibly as a result of drought, and the likely nesting sites remained available. Therefore, if, as has been suggested, E. longicornis can sustain a population over a wide area at low density it may still be around and very difficult to find. Conclusion The study was unable to provide any insight into the influence of different grassland management regimes on the conservation and specifically the foraging activity of E. longicornis. 7. Anthophora retusa Introduction Previous work on the Potter Flower Bee Anthophora retusa has demonstrated that a relatively high density of nectar-rich plants are required for suitable foraging habitat and that the loss of such habitat has been a major driver in the decline of this species. The objective of this study was to investigate the potential to improve habitat quality for A. retusa, specifically to explore whether additional appropriate forage resource provision had an effect on the area occupied by a population of A. retusa at Cuckmere Haven, East Sussex. Methods A meeting took place at Cuckmere Haven on 21st July 2010 with a representative of the South Downs Joint Committee (now replaced by the South Downs National Park Authority) to check for potential nesting habitat; to explore the possibility of establishing a more bee-sympathetic grazing regime in the area and to explore the possibility of establishing new forage areas under the Agricultural Stewardship programme using sown flower mixes. It was agreed to pursue the habitat improvement objectives as part of the development of a management plan for the South Downs. A field visit to the area on 15th May 2011 involved a walkover of potential forage and nesting habitat to check for the presence of A. retusa. A second meeting on 14th September 2011 included representatives from Natural England, Sussex Wildlife Trust, East Sussex County Council, and National Trust in addition to South Downs Joint Committee and Hymettus. Agreement was reached to targeted habitat improvement in 2012 at both forage and nesting components through the ESL/HSL agreements operating, or soon to operate, in the target area. A map of the suggested areas for improvements within these agreements was provided to Natural England. A costed plan for monitoring of the responses of the target bee species, forage and

SID 5 (Rev. 05/09) Page 22 of 34

nesting habitat components was drawn up and forwarded to Natural England and the South Downs National Park Authority. Results Little progress was made towards planned habitat improvement objectives in 2010 nor were the planned improvements in forage resources agreed in September 2011 progressed in summer 2012. A small number of nests and individuals of A. retusa were noted during the field visit in 2011. Discussion The election of a new Government and a decision by Eastbourne Council to take all tenanted farmland back into its direct management introduced sufficient uncertainty to land ownership and conservation management payments to halt progress on planned habitat improvement objectives in 2010. The improvements in forage resources arising out of the new agreement were not implemented due to the very wet conditions in summer 2012. Nesting resource improvement through scrub removal is still planned for winter 2012/13. Conclusion The study was unable to provide any evidence that additional appropriate forage resource provision has an effect on the area occupied by a population of A. retusa. However, an awareness of the issue has been raised within relevant conservation bodies and the work is being progressed by a consortium on interested organisations. 8. Bombus ruderarius Introduction The Red-shanked Carder Bee Bombus ruderarius is the least well understood of the British bumblebees. Some previous work has suggested that B. ruderarius may be a woodland edge species rather than a grassland species. The objectives of this study were to determine the environmental and ecological features characteristic of the habitat of B. ruderarius and to identify the pollen resources utilised by the bee. Methods A desk study of existing records of B. ruderarius allowed information on habitat and sex / caste to be collated and locations of historic records from Warwickshire to be identified. Known sites in Warwickshire were searched by carefully checking different grassland swards and looking for any workers flying in such areas. Locations with active B. ruderarius were photographed and the vegetation described. Pollen samples were obtained for microscopic analysis from 11 live B. ruderarius workers in 2011 and from a pinned specimen collected in 2010. Results A single B. ruderarius nest was discovered, which was examined and photographed on several dates. The nest was constructed of moss and located within sparsely vegetated grassland with kidney vetch and small hawthorn saplings. No spring queens were observed and there were few observations of workers during the field work. The desk study provided records of queens from woodland rides (both broadleaved and coniferised), old quarries, arable margins, road verges, chalk downland, limestone grassland, chalk-heath, coastal dune, heathland/acid grassland, unimproved neutral grasslands and suburban gardens. There were fewer records of workers, mainly from chalk downland, grassland and scrub. The pollen samples indicated that Kidney Vetch Anthyllis vulneraria was a very important pollen source for the bee at both of the locations where the plant grew, forming at least 90% of the pollen collected by eight of the ten bees sampled and 15-20% of the pollen collected by the other two (Lotus sp. or rose / bramble formed 80% of these two samples). At two locations where A. vulneraria was not available, then the two bees had made heavy use of rose / bramble or Lamiaceae, probably Marsh Woundwort Stachys sylvatica, pollen. Other pollen types were found as much rarer components of the samples. Discussion Examination of existing data suggested that records of queens, which wander much further than workers, may have created a false impression of habitat preference. Workers appeared to be much more strongly attached to floristically-rich, unimproved meadows and other flowery grasslands with plentiful Fabaceae, A. vulneraria being an especially favoured pollen source. However, the dataset is relatively small and more information on the foraging of workers outside of Warwickshire is required, Conclusion On the evidence of this study B. ruderarius queens may be found in a wide variety of habitats but workers tend to forage in the more floristically-rich sites. A. vulneraria appears to be a favoured pollen source of workers where the plant grows but other species are used especially other Fabaceae, Rosaceae and Lamiaceae. INVERTEBRATES OF EAST ANGLIAN WETLANDS 1. Odynerus simillimus Introduction The Fen Mason Wasp Odynerus simillimus was considered extinct in the UK for much of the 20th century. It was rediscovered at Hickling Broad in 1986 and has since been collected from a handful of coastal wetland sites in Essex, Suffolk and Norfolk. Nesting aggregations have only been recorded in Tendring District, Essex and the Norfolk Broads. The objective of this study was to determine the environmental conditions necessary for nest establishment by O. simillimus. Methods Information on the phenology and nest building activities of O. simillimus was gathered through weekly monitoring visits to Hickling Broad and to Sutton Fen (Norfolk) from June to August 2010 and May to August 2011, to Martham Broad (Norfolk) from July to August 2010 and less regular visits to Cattawade (Suffolk/Essex) from May

SID 5 (Rev. 05/09) Page 23 of 34

to August 2010. The number of active nests was recorded on each visit. Nests were identified by their distinctive architecture, “active nests” being defined as burrows with at least some “chimney” around the hole. After rain, chimneys were often washed out and then some judgement had to be made over what evidence of a chimney was required. Holes that were very round and with any remnant of a chimney or with a female wasp visible were generally included. The number and sex of adult wasps observed on each visit were also recorded. Maximum daily air temperature, daily sunshine hours and daily rainfall were obtained for Buxton, Norfolk and for Wattisham, Suffolk. In May 2011 one data logger measuring soil temperature was installed 10cm below the surface at each of Cattawade, Hickling Broad and Sutton Fen. The data loggers were collected from Hickling Broad and Sutton Fen in October 2011 but the Cattawade instrument went missing. Results The earliest adult wasps in 2011 were 3 males observed at Sutton Fen on 10th May, over a week earlier than the previous earliest recorded date in 2010. This was preceded by two months of unusually dry, warm and sunny weather. The earliest females were seen on 26th May 2011 at Sutton Fen when the first active nests were observed both here and at Hickling Broad. Discussion During fieldwork it was apparent that wasp activity was very sensitive to the weather, and it might be expected that local weather data for the spring and summer would account for both the early emergence and the fluctuations in nesting activity. However, there was no obvious correlation between activity and air or soil temperature. Sunshine and rainfall appear to be more important factors. However, it must be questioned how relevant the weather station data were to conditions in the immediate vicinity of the O. simillimus colonies. Ground temperature is likely to be important pre-emergence for pupal development but the temperature needs to be recorded over a number of years to gain sufficient data to draw firm conclusions. It may also be important during the nesting period, as there may be a minimum temperature below which the female wasp, resting in the ground overnight, cannot start foraging regardless of the air temperature. Conclusion The study provided some evidence that hours of sunshine, rainfall levels and ground temperature may be important factors in controlling nest establishment and pupal development in O. simillimus. 2. Rhopalum gracile and Passaloecus clypealis Introduction The larvae of the solitary wasps Rhopalum gracile and Passaloecus clypealis are believed to develop on small Diptera with which the nest is stocked by the adult but the species involved have not been identified. These wasps are rarely observed and wasps carrying prey have never been collected. The objective of this study was to determine the larval prey resources utilized by R. gracile and P. clypealis. Methods Trap nests were constructed from cut reed stems packed into 15cm lengths of 5cm diameter plastic water pipe and mounted on bamboo canes. In May 2010, 40 traps were placed in reed fringing a ditch and 40 in a sedge bed at Woodbastwick Fen NNR, Norfolk. The reed stems were collected and the traps replaced in March 2011. The stems were placed into plastic ziplock bags and kept close to outside temperatures for a period of 4 months. They were monitored every two to three days for emergent aculeates. They were also opened to allow air to circulate and moisture to escape in order to dry contents as most bags were waterlogged. Random reeds were taken from the nests, cut and examined for pupae throughout the monitoring period. The second batch of traps was collected in March 2012. The stems were immediately placed into ziplock bags and processed within two days. All stems were dissected and any pupae placed in gelatine capsules and left outside to develop. Pupae were monitored daily from 1st May until 6th August 2012. Results At least three adult R. gracile emerged from lengths of reed in 2011, although body fragments found could suggest a higher number. No P. clypealis were recorded and no pupae of R. gracile were found in the 150 reed stems sampled. No specimens of R. gracile or P. clypealis were amongst the aculeates that emerged in 2012. Discussion The study confirmed the presence of R. gracile in the Norfolk Broads and the effectiveness of trap nests for monitoring the wasp. It is not possible to say whether P. clypealis can be monitored using trap nests as it may not have been present in the area sampled. It was not possible to match any diptera remains with wasp nests and thus add to current information on the prey items collected by either wasp. Conclusion The study provided no new information on the larval prey items utilized by R. gracile or P. clypealis. 3. Asindulum nigrum Introduction The fungus gnat Asindulum nigrum has a very scattered distribution in southern England but is very local with populations confined to small meadow or fen sites. Some of these sites are of the highest quality but others appear to be fairly typical water-meadows, but its distribution suggests that it must have very specialised requirements. The objectives of this study were to determine the ecological and environmental conditions characteristic of sites inhabited by A. nigrum and to determine the conditions it requires for egg-laying. Methods

SID 5 (Rev. 05/09) Page 24 of 34

A. nigrum was searched for by sweeping at Woodbastwick Fen and Bure Marshes NNR in the Norfolk Broads over two days in July 2010 and again in July 2011 when six days of field work allowed for brief visits to Geldeston Meadows and Upton Broad and a speculative visit to Horse Fen, Ludham which appeared to have suitable habitat for the fly. The precise location of specimens collected was plotted on aerial photographs and descriptions of the relevant habitat compartments were compiled in the field. Any association of the flies with fungi was noted as was the presence of larvae. Results A single female A. nigrum was captured in 2010. In 2011 thirteen or fourteen (one was probably double counted) A. nigrum were observed at Woodbastwick Fen in three discrete areas at two places in one meadow and along or close to a central main drove. Nearly all the observations were from areas dominated by, or at least with substantial amounts of Calamagrostis or Holcus; or in a few instances Juncus dominated. These were botanically diverse meadows so at nearly all places where A. nigrum was noted there was also Juncus, Carex, Phragmites, Gallium, Equisetum and many other species. Although there were dead Betula and Alnus scattered across the survey site, A. nigrum displayed no clustering around these features and the lignicolous fungi developing on them. At one of the A. nigrum colonies a few very delicate terrestrial cap-fungi, possibly a species of Psathyrella, were found with nematocera larvae. Discussion Both of the areas where A. nigrum was found were very slightly elevated so less prone to winter inundation, the meadow because it was at the “inland” margin of the fen (i.e. away from the river) where the land starts to rise slightly and the drove because it has been built up over centuries to provide access through the fen. Also all observations were in areas that were not being summer grazed so not suffering any visible trampling from herbivores. These associations indicate the general habitat these flies were seen in, and should not be treated as being very precise. Often an individual fly was watched flying though the habitat, rapidly moving from Juncus to Calamagrostis or Holcus dominated areas, although they were not observed flying into dense Phragmites or Eupatorium. The nematocera larvae found with a possible species of Psathyrella could indicate the larval development site for A. nigrum, but it will be necessary to rear the larvae to confirm this. The greatest likelihood is that A. nigrum larvae develop by feeding on fungal hyphae amongst rotting grass litter at the bases of dense grass tussocks, perhaps of a species of fungi that produce fruiting bodies only rarely or that are ephemeral. Barring a series of lucky, chance encounters with the species as eggs are laid, further, resource intensive observation, probably over a number of years would be necessary to unravel the life history of A. nigrum. Conclusion On the limited evidence of this study, A. nigrum appears to be associated with fen meadows with tall, dense, grassy sward and tussocks untrampled by livestock rather than with wet fen. The study provided no clear evidence for the egg laying site of A. nigrum. 4. Dolichopus laticola and Dolichopus nigripes Introduction The distribution of the flies Dolichopus laticola and D, nigripes is almost confined to the fens of the Norfolk Broads. The objective of this study was to characterise the habitats occupied by D. laticola and D, nigripes. Methods Six Norfolk fens at Barton, Sutton, Catfield (Great Fen), Woodbastwick, Horning Marsh Farm and Ebb & Flow were selected for studies of the relationship between environmental variables and the flies Dolichopus laticola and D, nigripes. At least one of the species had been found at each of these sites between 2007 and 2009. Samples were obtained using a standard 10 minute sweep-net sample which consisted of repeatedly sweeping vegetation for about 25 sweeps and inspecting the net. All obvious dolichopodids were removed using a pooter, The patch of sampled vegetation swept was usually about 30-40m in diameter when in the centre of a stand, or a 50-80m length of ditch margin. Environmental variables were measured at each sampling point. Vegetation structure was scored using the DAFOR scale for each ‘layer’ in the sense used for Natural England’s Common Standards Monitoring of invertebrates. Soil wetness was scored on a subjective scale from 1 (dry) to 4 (standing water). Management of fen vegetation was measured as recent cut, old cut, neglected or grazed. 183 samples were collected over ten consecutive days between 20th and 29th June 2010. Results Initial exploration of the results showed that a lower than expected frequency of samples containing either of the species were from carr. Beyond that there were no obvious patterns in the data. As neither abundance data nor explanatory environmental variables were normally distributed, parametric multiple regression or generalised linear regression could not be applied. Instead, the relationship of each Dolichopus species with the explanatory variables was examined using tree models as these are unaffected by lack of normality or linearity in the data and are easy to interpret. Abundance data were investigated using a regression tree and presence-absence data using a classification tree. The 19 samples from carr were excluded to prevent this appearing as the first split in the tree. The management variable also had to be excluded as it was a construct rather than a true measurement. Discussion A classification tree, obtained using presence-absence data, with just two branches was the best fit for D. laticola. Wetness was responsible for the first split of the tree, the fly being more frequent in drier sites, and this split can be regarded as statistically significant using the ‘one standard deviation rule’. Further useful information was

SID 5 (Rev. 05/09) Page 25 of 34

found in further splits and in a regression tree using the abundance data even if these could not be fully supported by cross-validation. Although the evidence was weak, it suggested that D. laticola preferred ‘old cut’ fen, characterised by a larger proportion of tall herb vegetation and moderate amounts of leaf litter. It appeared to avoid places dominated by tall, dense reed to the exclusion of tall herb vegetation. This may be more characteristic of frequently cut and wetter commercial reedbed. Higher densities of flies were found in wetter areas, but this result conflicted with an apparent increased frequency of occurrence in slightly drier places (although still damp). Neither tree fitted the model well for D. nigripes. Again the evidence was weak but suggested that D. nigripes preferred more open vegetation, in which highest densities were reached, and strongly avoided vegetation dominated by tall reed and sedge. The lack of a relationship with soil wetness may have been due to the similarly damp conditions in all the sites. Conclusion On current evidence it seems possible that both Dolichopus species are more associated with intermittently cut fen, high in plant diversity and with low dominance by tall reed. Such areas are often considered to belong to a fen-meadow habitat type. 5. Lipara similis and other Lipara species Introduction Cigar galls on reed in England are known to be caused by three different species of flies of the genus Lipara and a fourth species is believed to be involved but has not yet been recognised as British. This uncertainty in the taxonomy combined with a need to rear adults from the galls for certain identification has resulted in a lack of information on the distribution and ecology of the Lipara species. The objectives of this study were to investigate the taxonomy of the Lipara species occurring in Britain and prepare new keys to the adults and galls of these species; to determine the current English distribution of Lipara similis and to determine the environmental and ecological factors characteristic of sites inhabited by the fly. Methods Between February and April 2011 samples were collected from 84 subsites across ten sites within East Anglia and the New Forest; an additional sample was taken from Thame, Oxfordshire. At each site, concentrations of galls were located and provisional identifications to species made on the basis of the appearance of the galls. Ecological measurements were taken from within a 0.5m square quadrat and the number of each gall type was noted. Percentage of shade cover, water level and density of reed stems were estimated within the quadrat. Direction and distance to the nearest scrub was estimated. Information on habitat management was noted also if it was available. It was necessary to rear the adults and identify these to confirm the characters of the galls. The basal and broadest diameter of each gall collected was measured and then galls were stored individually in sealable bags and moistened twice per week, a system which has worked well in the past when rearing adults. After emergence the adults were then frozen or stored in denatured alcohol. Results In total 556 specimens were reared from 247 galls and of these 347 specimens, including all Lipara specimens were identified. 54 galls produced no insects and the remainder produced 88 Lipara, 378 inquilines, 44 parasitic Hymenoptera, 4 predators (Cleigastra apicalis) and 43 Cecidomyiidae. Within Lipara, 34 specimens of L. lucens were reared, 15 specimens of L. similis, 13 specimens of L. rufitarsis and 4 specimens of L. pullitarsis. However, the last two can only be identified by examination of the male genitalia and 22 females could not be assigned to one of these two species. L. similis was reared from galls found close to running water, either streams or a river except at one site where this was not noted. It was found close to trees or shrubs, although mostly not in the shade. Discussion A key to adults of the species, including to L. pullitarsis, has been produced. A description of the galls is included but these cannot be used for identifications; L. lucens only emerged from ‘lucens-type’ galls and L. similis only emerged from ‘similis-type’ galls, but L. rufitarsis and L. pullitarsis emerged from all four gall types. Adult rearing seems to be the only reliable method of recording the species except for L. lucens, a rather common species. L. pullitarsis will be officially added to the British list in a separate paper. The small number of L. similis reared provided insufficient data for any statistically robust analysis of the environmental variables. Further rearing is necessary to determine the habitat preferences of the species. However, the findings combined with existing records indicate that L. similis is confined to a maximum of six localities in England, mostly in East Anglia. A recent study in a rather localised area in Poland, has suggested that more than one species of Lipara can oviposit on one reed stem but that only one adult usually emerges. The study also reported differences in the morphology of eggs, larvae and pupae of all four species, but this needs further testing in other countries. The same applies to reported differences in habitat preference of all four Lipara species. Conclusion The current distribution of L. similis appears to be confined to Cambridgeshire, Huntingdonshire and Suffolk. The presence of L. pullitarsis in Britain has been confirmed and a key has been produced to the adults of all four species of Lipara but differentiation of galls is not possible on the basis of current knowledge. Due to the small number of L. similis that emerged from galls the habitat preferences of the fly remain unclear and more data collection is required. INVERTEBRATES OF OPEN WOODLAND

SID 5 (Rev. 05/09) Page 26 of 34

1. Polyzonium germanicum Introduction The millipede Polyzonium germanicum is a south-eastern species in Britain with all verified records from 13 hectads in East Kent. Data submitted to the national recording scheme suggested there had been a significant population decline over the last 25 years with post-1980 records from just four hectads. The objectives of this study were to ascertain the current English distribution of P. germanicum and to determine the environmental and ecological factors characteristic of sites inhabited by the millipede. Methods Between 14th and 17th April 2011 the British Myriapod & Isopod Group organised a standardised survey in East Kent. Kent Wildlife Trust assisted in gaining permission to survey 51 locations in 27 hectads across the known range including all locations where the species had been recorded previously.Each location was visited by a team of two or three surveyors who identified patches of suitable habitat for collecting time-standardized samples. A sample comprised five minute sub-samples collected at each of six locations within a 20x30m quadrat. Sampling involved searching for millipedes by grubbing in moss and in vegetation tussocks, turning logs and stones and sieving through leaf litter. The number of P. germanicum collected was recorded along with details of litter type, vegetation structure and estimates of soil moisture, litter depth, and bare ground. Results P. germanicum was recorded from 22 of the 92 samples collected. P. germanicum was recorded from 16 different hectads. Logistic regression was employed to explain the occurrence of P. germanicum in terms of the environmental variables measured using its presence or absence as the binary response variable. Soil moisture was the only environmental variable that showed a significant (positive) relationship with the presence of the millipede Discussion The millipede is now known to survive throughout its previous range and the records have extended its known range both eastwards and westwards. It is unlikely that the distribution of P. germanicum extends further east as there are few suitable sites but it is possible that further survey could locate populations to the west in areas of West Kent and East Sussex. In England P. germanicum had been assumed to be a species associated with woodland, especially sweet chestnut coppice but the results of this work suggest that it is more likely a species requiring high soil moisture (or possibly humidity) levels and will occur in a wider range of habitats including heaths and scrubby fen which provide suitable conditions. Conclusion The current English distribution of P. germanicum is restricted to Kent but the millipede appears to be more widespread there than previous records suggest. There is no evidence of a decline in populations of the species. P. germanicum appears to be a species requiring high soil moisture levels. 2. Metaiulus pratensis Introduction The millipede Metaiulus pratensis was described as new to science from specimens collected between 1939 and 1956, mainly from Wye, Kent. All subsequent British records have been from the same county, most recently from the Medway Valley, but it was last collected in 1988 and feared extinct. The original records of M. pratensis originated from newly ploughed pasture and arable land but it has also been recorded from caves in France and the most recent records from Kent were from woodlands and wetland. The objectives of this study were to ascertain the current English distribution of M. pratensis and to determine the environmental and ecological factors characteristic of sites inhabited by the millipede. Methods On 17th and 18th April 2011 a team of three surveyors from the BMIG visited four sites along the upper Medway floodplain to search for M. pratensis. A second team visited a fifth site near Maidstone on 17th April 2011. The surveyors spent between one and three hours hand searching each site by grubbing in vegetation tussocks, turning logs and sieving leaf litter. Results M. pratensis was located at only one of the sites visited. Yalding Fen supported a very strong population of the millipede amongst leaf litter and at the soil-litter interface in seasonally wet grassland and carr. Further survey work along the upper Medway planned for 2012 was prevented by flooding. Discussion On current knowledge it would appear the distribution of M. pratensis is restricted to a single site. However, it is more likely that a lack of knowledge of the autecology of the species means that it is under recorded. The recording of the species from this site confirmed its association with habitats similar to those previously noted in the Medway Valley but did not provide sufficient data for a new analysis of the environmental and ecological factors characteristic of sites inhabited by the millipede. An extension of the work beyond the end of the project is required to understand the details of the autecological requirements of M. pratensis and thereby determine its true distribution. Conclusion The current English distribution of M. pratensis appears to be restricted to one site in the Medway Valley, Kent. The micro-habitat occupied at this site was amongst leaf litter and at the soil-litter interface in seasonally wet grassland and carr but there is insufficient evidence to confirm the autecological requirements of the millipede.

SID 5 (Rev. 05/09) Page 27 of 34

3. Philorhizus quadrisignatus Introduction Philorhizus quadrisignatus is a small, arboreal ground beetle that inhabits broadleaved trees and shrubs. It is found in woodlands and parkland in southern England. The objective of the study was to determine the environmental and ecological factors characteristic of sites inhabited by P. quadrisignatus. Methods All available records of P. quadrisignatus were collated and examined for information that could be used to characterise its habitat. Field visits to sites with recent records were undertaken at Bushy Park, Middlesex on 13th October 2010 and 26th September 2011; Queen Mary Reservoir, Middlesex on 26th September 2011 and Bookham Common, Surrey on 18th May 2011. Up to four hours was spent searching under bark of hawthorns and other trees. Visits to sites with historical records were undertaken at Deadman’s Hill, New Forest on 15th October 2010; Ockham Common, Surrey on 18th May 2011; Bedfont Country Park, Surrey on 26th September 2011; Hayling Park, Hampshire on 27th September 2011 and St Leven and Porthgwarra, Cornwall on 26th October 2011. Results The beetle was not found during fieldwork at any of the locations visited. A new location for the beetle was discovered by an independent worker at Bishop’s Stortford, Hertfordshire on 10th May 2011. The existing records noted that P. quadrisignatus has been found under the bark of deer grazed hawthorns, under sycamore bark and in the woodland canopy but no further information on micro-habitats was available. Discussion Work on the autecology of P. quadrisignatus will be difficult until a more reliable method of locating the beetle is developed allowing a larger volume of data on habitat preference to be collected. Philorhizus and related genera (formerly lumped under Dromius) can be difficult to find and the technique of searching under bark is destructive. Night searching may prove to be a more efficient technique for finding it. Conclusion Existing records indicate that P. quadrisignatus is a subcortical species. It is a difficult species to locate and field work was unable to provide further information on its ecological and environmental requirements. 4. Formicoxenus nitidulus Introduction The Shining Guest Ant Formicoxenus nitidulus is a social parasite reliant on wood ant hosts, in Britain primarily Formica rufa and F. lugubris. Analysis of historical records shows an overlap in distribution of the two wood ants covering the whole of northern England from Staffordshire north to Cumbria and Northumberland. As the conservation of this species relies entirely on the health of wood ant populations, it was decided that a survey to assess the state of F. rufa and F. lugubris was an important first step towards assessing localised changes in the status of Fx. nitidulus, the objective of this study. Methods Historical records of foragers or nests provided sufficient detail to enable 30 F. rufa populations and 49 F. lugubris populations to be identified for field survey. Site visits were carried out between May and October 2010 and between April and October 2011 on warm sunny days, as far as possible. Nests were located from existing grid reference data combined with observing foraging worker ants and following them to the nest. Warm weather (leading to high ant foraging activity) was important, because although visual search for nests was also made, on large sites with dense undergrowth this was much less effective than following foragers. Where wood ants were present, at least two nests were sampled from each population and three workers were taken from each nest for confirmation of species identification under the microscope. The position of the nest was recorded using GPS and notes were made on the size and appearance of the nest. The surface of the nest was observed for Fx. nitidulus and a 0.5l sample of nest material was removed from the nest and spread on a white surface to search for Fx. nitidulus. In autumn 2011, the sites visited earlier in 2011 with confirmed wood ant populations (8 sites) were re-visited to check for Fx. nitidulus males on the surface of the nests. Results For both wood ant species, 60-70% of historical records were confirmed. Fx. nitidulus was found very rarely during the survey. Discussion The two species have fairly distinct distributions. In northern England and the Midlands, Derbyshire is the only county with historical records of both F. rufa and F. lugubris and even here there is doubt over the single F. rufa record. The separation of the two species suggests that intraspecific competition is not likely to account for any changes in distribution. For F. rufa, there seems to be a genuine decline within the region surveyed, that is, the more northerly part of its British range. By contrast F. lugubris appears to be thriving throughout its range. As Fx. nitidulus shows no preference for either species this should not affect its status and distribution. The low number of records of Fx. nitidulus should not be taken as an indication of its actual occurrence. To thoroughly survey Fx. nitidulus, wood ant nests should be visited on warm autumn days, when workers and sexuals are commonly seen on the nest surface. Due to the scarcity of warm autumn days, it was not possible to re-visit all the wood ant sites to check for Fx. nitidulus, although the list of extant wood ant populations should help to target future studies. The high detection rate of Fx. nitidulus when surveying conditions were suitable (80%) may be taken to be

SID 5 (Rev. 05/09) Page 28 of 34

encouraging, but given the very low sample size (5 sites), clearly a more wide-ranging survey is required before general statements about the prevalence of Fx. nitidulus can be drawn. Conclusions On the evidence of the distribution of F. rufa and F. lugubris there would appear to be no significant localised changes in the status of Fx. nitidulus. However, given the very low sample size, a more wide-ranging survey is recommended. 5. Chrysis fulgida Introduction The jewel wasp Chrysis fulgida has been shown to be a parasitoid of the wasp Symmorphus crassicornis which in turn hunts larvae of the leaf beetle Chrysomela populi to stock its nest. S. crassicornis is typically an aerial stem nester and can be induced to nest in holes of suitable diameter drilled in wooden blocks. The objective of this study was to determine the current English distribution of C. fulgida. Methods All available historical records of C. fulgida were extracted from the data held by the Bees, Wasps and Ants Recording Society. Wooden trap nests were constructed and three traps were located at each of nine suitable sites in Surrey, North Hampshire and Dorset in April 2010. The traps were set in an elevated position in all but one locality (i.e. Sheet’s Heath, Surrey). The sites chosen supported C. populi populations and most were known localities for S. crassicornis. The traps were collected in October 2010 and kept outdoors over-winter with shelter from excessive rain and snow. In spring 2011 any pupal cases present in the over-wintered traps were extracted for rearing individually in gel capsules. Results The desk study identified records of C. fulgida scattered across south eastern England. The wasp was not found during the field survey. A single male S. crassicornis was reared from Eelmoor. Human disturbance to traps had occurred on several sites in Surrey and the majority of the traps sited at Corfe Common, Dorset and Eelmoor, N.Hampshire suffered woodpecker damage. Discussion The existing records of C. fulgida show a distribution centred on Surrey, Hampshire and Dorset. The apparent absence of the wasp from these areas during the field survey was thought to be due to problems with the vulnerability of the traps to interference and the requirement for a much greater number of traps to be used across a greater range of sites. Conclusion The field survey could not confirm the distribution of C. fulgida but there is no reason to believe it is still present across central southern England. 6. Andrena ferox Introduction Females of the bee Andrena ferox forage almost exclusively for oak pollen. Thus the species potentially could be excluded from an area if the flowering period of all the oaks in that area was highly synchronised and pollen is not available over the several weeks required to raise the brood. The objective of this study was to investigate this theory. Methods Three known sites for A. ferox in East Sussex and Kent were compared with three nearby control sites where the bee was presumed absent. At each site a transect was established that included 30 mature trees. Each transect was surveyed three times at one week intervals between 6th and 21st May 2010. On each occasion the condition of the canopy (ignoring the lowest branches) was assessed using binoculars and assigned to one of five categories • in bud – no significant opening of leaves • just in leaf – many buds bursting, but no catkins • green catkins – catkins developed, but not bearing pollen • yellow catkins – catkins producing pollen • brown catkins – catkins over, brown and shrivelled Results The results were inconclusive. Catkins producing pollen were only observed on a single visit and so it was not possible to say that there was long period of availability of pollen at any of the sites. The progression through the various stages of development was not linear and the speed of change between the second and third visits suggests that the flowering period at all sites was fairly short. Discussion At sites at which A. ferox has been recorded there might have been a longer flowering period of oak than on sites where the bee was absent, although a further visit after another week would have been needed to confirm this. Frost damage to some trees was evident from the blackened and shrivelled leaves, and this probably affected the catkins too. This is a hazard inevitable in an early flowering species, but might be mitigated by genetic diversity which could confer differing degrees of frost resistance to individual trees, and by the woodland structure where solitary, isolated trees might be more susceptible than those in a closed canopy. It was observed that there were considerable differences between catkin production of individual trees. Some trees produced copious catkins and

SID 5 (Rev. 05/09) Page 29 of 34

relatively few leaves while others produced many leaves but very few catkins. This may be further evidence of genetic diversity, although it is not known whether trees that are “heavy” catkin producers do so every year. Conclusion No clear evidence of suitability or unsuitability for A. ferox was shown between the sites. While lack of genetic diversity, for example in a plantation derived from a common stock, may be reflected in synchronisation of flowering period, it is probable that other factors such as site aspect and woodland structure are important too. 7. Neoempheria lineola Introduction There are modern (post-1990) records of the fungus gnat Neoempheria lineola from three sites, all in the New Forest namely The Knowles, Denny Wood and Brinken Wood. There are older records from one other site in the New Forest (1905) and one at Oakley Wood, Gloucestershire (1986). The objectives of this study were to determine the current English distribution of N. lineola, to determine the environmental and ecological factors characteristic of sites inhabited by N. lineola, to determine the conditions necessary for egg laying and to determine nature of the foraging resources utilised by adults of N. lineola. Methods Permission to survey at Oakley Wood was refused so field work was undertaken on 31st May 2011 and 16th May 2012 at Denny Wood, 2nd June 2011 and 18th and 28th May 2012 at The Knowles and 3rd June 2011 and 17th and 23rd May 2012 at Brinken Wood. In advance of visiting each site, aerial photographs were consulted to locate areas with large numbers of large broad-leaved trees. These areas were then walked searching for fallen or standing dead wood or mature trees with holes or signs of fungal infection, especially beech. All of these that could be accessed were swept with a white-bag net, above and if possible below fallen branches, along the boles of standing stumps and dying trees and over any fungal fruiting bodies associated with dead wood. If any Neoempheria specimens were found then a 10-figure grid reference was taken and it was potted separately for later examination. Results No specimens of N. lineola were found in 2011 or 2012. Examination of existing records provided no information on preferred habitat, beyond the fact that all sites were woodland in the broadest sense, or on the foraging resources utilised by adults. Discussion Fungus gnat abundance and diversity seemed to be particularly poor in 2011 and it seemed possible that the very warm and dry spring had had an effect on their abundance. In 2012, after a very wet April and early May, it was hoped that the fungus flora would be healthy and the chance of finding N. lineola potentially enhanced. Despite the very wet period preceding field work, neither fungi nor fungus gnats appeared to be any more abundant than in 2011. It seems unlikely that no specimens of N. lineola at all would be found if it occurred in sufficient numbers to enable study of its autecology except very infrequently and unpredictably. Only if a population can be found that reliably occurs in significant numbers will it be possible to pursue ecological work. Conclusion N. lineola could not be located by field survey but on evidence from existing records its English distribution is restricted to the New Forest. The few existing records of N. lineola provided no information on preferred habitat, beyond the fact that all sites were woodland in the broadest sense, and provided no information on the foraging resources utilised by adults DISCUSSION Distributional survey and autecological study have demonstrated that, of the species included in this project, only one can be considered a truly stenotopic sand dune species. The leafhopper Doratura impudica is obligate on sand couch Elymus farctus growing in a narrow strip between the strandline and the start of the fixed dunes. D. impudica is part of a larger assemblage of sand dune invertebrates but these other species are found in the more stable areas, especially where dune heath develops, and they also occur as part of other assemblages, often on coastal shingle and heaths. Conservation management for D. impudica involves reducing the threats to its habitat where it survives. The natural processes of erosion or deposition of sand by winter storms cannot be prevented but reduction of impact from developments including coastal defences that have indirect effects on natural processes can be considered. Also, most of the remaining occupied habitat is under considerable visitor pressure and injudicious conservation management (e.g. inappropriate positioning of structures such as boardwalks) as well as trampling. Due to the limited powers of dispersal of D. impudica, such factors could jeopardise small populations already vulnerable to any fragmentation and isolation of habitat. With the knowledge gained of its micro-habitat, other sand dune systems further west along the south coast, in South Wales and further north along the east coast can now be surveyed with more efficient use of resources. The existence of an assemblage of invertebrates containing species common to sand dunes, coastal vegetated shingle and heathland was noted above. Although representatives of many invertebrate groups make up this assemblage, those considered in this project were all ants, namely, Temnothorax interruptus and Tetramorium caespitum along with its parasite Anergates atratulus. These invertebrates are associated with the early successional stages of the habitats where patches of bare ground with high insolation generate warm conditions.

SID 5 (Rev. 05/09) Page 30 of 34

Thus slopes with a south facing aspect are especially favoured. The members of the assemblage disappear from the habitat as the sward becomes closed and taller vegetation leads to more shading and therefore cooler conditions. The importance of temperature as an environmental factor is further demonstrated by the distribution of the species included in this project; most are restricted to the southern coastal counties of England. Management techniques to control succession and maintain exposures of bare ground on heaths are discussed below. General advice on management of sand dune and shingle habitats to maintain invertebrate diversity is readily available (e.g. www.buglife.org.uk/conservation). However, in addition to this published information, the reported work on the coastal vegetated shingle at Dungeness suggested that on areas of lichen heath Temnothorax interruptus was likely to benefit from light grazing. The aim of such management would differ subtly from most grazing where the objective is to control succession. Here the benefit would be an improvement in the nesting resources available to the ant brought about by exposing flint pebbles in amongst the vegetation. The carabid beetle Philorhizus vectensis occasionally occurs on shingle on the south coast where as a thermophile it benefits from the same warm conditions as the ants discussed previously. The beetle more usually occurs on coastal cliffs and is further considered in this context below. Eleven of the invertebrates studied in the project are associated with lowland heathland habitats. The bee Andrena tarsata also shares similarities in the requirement for nesting resources on bare sandy soils but it is found on upland sites. The majority of these species share with the shingle and sand dune assemblages the requirement for patches of unshaded areas of warm bare ground. The bare earth provides for nesting and / or forage resources for the majority of the beetles, wasps and ants studied but the suitability of the substrate varies between species. Areas of free draining sandy soil are favoured for ease of excavating burrows by species such as Andrena tarsata. More clay content in the soil provides for a more stable burrow structure such as is required by the wasp Pseudepipona herrichii and the tiger beetle Cicindela sylvatica. Disturbance of any areas supporting nesting aggregations will obviously damage the current generation and protection from damaging activities, including trampling by humans and livestock, may be necessary. Disturbance of the nesting aggregation through management aimed at maintaining the bare ground should not be necessary on most sites. A planned timetable of rotational management can produce alternative nesting resources nearby that the colony will utilise once the original area has become unsuitable through the growth of vegetation. Creation of bare ground and early successional heathland habitats may be achieved by a variety of means. The project has shown that small-scale winter burning (swaling) of heather and gorse between October and February can be an excellent method of creating early successional habitats in Cornwall (where conservation was targeted at Large Blue Maculinea arion), Devon, Dorset and Hampshire, specifically to benefit ground beetles such as Anisodactylus nemorivagus and Poecilus kugelanni. However, care is required to ensure the conditions are right for the fire to burn swiftly across the surface rather than downwards when it will be more damaging to invertebrates e.g. nests of Formica exsecta. Previous work by the Aculeate Conservation Group (now Hymettus) has shown that controlled burning can be used to benefit the wasp Homonotus sanguinolentus. This species is associated with a later stage of heathland succession as the spider that it parasitizes requires tall grasses for spinning its egg laying retreats. The habitat will take a longer time post burn before it becomes suitable for the wasp than for those species associated with the early successional heath. Mowing or cutting heather and gorse is effective at creating suitable habitat for early successional species also but the vegetation must be cut shorter than 10cm and the cut vegetation must be removed. Turf-stripping small areas and disturbing the ground every 4 years to maintain early successional habitats has proved to be an effective method of creating good habitat on the east Devon pebblebed heaths. The same approach in Breckland used in the conservation of Grey Carpet moth Lithostege griseata has benefitted Harpalus froelichii also. This more regular disturbance may be required to maintain foraging resources for herbivores such as Harpalus froelichii and Amara fusca, both seed-eating beetles with food plants that are annuals or short-lived perennials. Current records for some of these species e.g. Ophonus laticollis suggest that the necessary soil disturbance may be being achieved best in field margins as part of the agroecosystem rather than in true heathland settings. As the provision of the food plant is clearly an important factor, at least on a local scale if not always in a wider context for common arable weeds, it may be possible to look at strengthening the populations of the food plants in regions where the beetles occur. Grazing by cattle, ponies and rabbits maintains a suitable habitat of closely cropped heather and bare earth at some sites in the New Forest but trampling by livestock, nutrient enrichment from dung and over grazing especially by deer were identified as problems for invertebrates and their habitat at other sites. Flower-rich grassland has been identified as the habitat of nine of the invertebrates studied in the project. Six of these species are bees (Anthophora retusa, Bombus ruderarius, Colletes floralis, Eucera longicornis, Lasioglossum angusticeps and Osmia parietina), the other three are solitary wasps (Odynerus melanocephalus, Cerceris quadricincta and Cerceris quinquefasciata). Although this grouping includes species occurring over very different geographical ranges, the project has shown that these differences are unlikely to be due to resource availability in the form of pollen type for the bees or prey species for the wasps. Analysis of pollen taken from the bees either demonstrated that the species was polylectic (Bombus ruderarius, Colletes floralis and Eucera longicornis) or, if oligolectic, that it was dependant on a common and widespread plant species such as Lotus corniculatus (Anthophora retusa, Lasioglossum angusticeps and Osmia parietina). Similarly, the solitary wasps appear to be able to raise larva successfully on a diet of common weevil species, switching between species according to availability at any one place and time. There is a strong coastal component to the distribution of most

SID 5 (Rev. 05/09) Page 31 of 34

of these species which may be linked to temperature but also appears to reflect a general pattern of loss of suitable flower rich habitat on many inland sites. The availability of nesting resources close to foraging sites is important. This means that, when managing for the conservation of Colletes floralis within a site, it is suggested that improvement of the foraging resource within a maximum of 400m of a known nesting aggregation is necessary. The distances involved for the other bees, except Bombus ruderarius, will be the same or more likely even lower. As with the heathland group, nesting resources often means areas of suitable bare substrate in which to construct burrows. Bombus ruderarius is the exception being a surface nesting bee and the one nest detected during the project was within the grassland providing foraging for the colony. Control of vegetation at most of the study sites was achieved by grazing either by livestock or by rabbits and deer. Trampling by livestock may also be beneficial in producing patches of bare ground for nesting. However, over intensive grazing will destroy the value of both foraging and nesting resources by preventing flowering and damaging nest burrows as soon as they are created. The density and timing of livestock grazing on a site needs to be carefully planned and may need continual reassessment to ensure the necessary resources are available throughout the flight period of the insects. Grazing by deer and rabbits may require control by exclosures and / or culling. Management by mowing or cutting may also be effective in maintaining forage resources but again the timing is important. The studies in this project only encountered the use of cutting in a minor way when it was used to clear a small length of railway embankment that was becoming overgrown and unsuitable as habitat for Cerceris quadricincta. Scraping of surface vegetation may be used to create nesting resources on south-facing slopes but the relative benefits of scraping and cutting compared with grazing of flower rich grassland were not assessed in this project. In some circumstances it may be beneficial to use seed mixes in the improvement of foraging resources. The project’s findings suggest that enhancement of sites for Small Blue Cupido minimus through planting of kidney vetch could be beneficial to Bombus ruderarius on chalk and limestone areas in southern England. Legume rich pollen-nectar mixes developed for agri-environment schemes have been shown to benefit populations of bumblebees and may be useful for solitary bees including Anthophora retusa in East Sussex. Other solitary bees reliant on a wider range of forage species, e.g. Colletes floralis, may benefit from the use of wild flower mixes or pollen-nectar mixes with adjusted compositions to suit individual bee species and geographical locations. A number of the solitary wasps and true flies studied in this project are dependent on wetland habitats, mainly in East Anglia. This group comprises the solitary wasps Odynerus simillimus, Passaloecus clypealis and Rhopalum gracile and the true flies Asindulum nigrum, Dolichopus laticola, Dolichopus nigripes and Lipara similis (including three related cigar gall forming Lipara species). Within this group, conservation management for Odynerus simillimus is most readily achieved. It is highly dependent on agricultural, land drainage and conservation management activities to maintain the combination of food plants for its prey, nectar sources and suitable nest sites within close proximity. Thus management needs to ensure stands of lesser water-parsnip or fool’s watercress exist in ditches, ponds, fen or swamp, for collecting prey; tufted vetch is available as a proven nectar resource and suitable bare, dry ground is present for nesting. The nesting resource appears to be the limiting factor so some form of disturbance to create or maintain bare or lightly vegetated ground is key. Disposal of spoil from routine maintenance or restoration of watercourses, construction of banks and other engineering works all provide opportunities to create nesting habitat without additional expense. Introduction of extensive grazing schemes in fen habitats and clearance of trees or scrub from otherwise suitable banks can also improve nesting opportunities. Further work is required on the remaining species before conservation management prescriptions are feasible. The project found that Asindulum nigrum, Dolichopus laticola and Dolichopus nigripes appeared to avoid areas dominated by tall reed. Although the evidence was limited, open, species-rich fen meadow rather than permanently wet fen appeared to be the preferred habitat. Management of such sites was by intermittent cutting but the interval was not determined. Grazing was not considered a suitable management technique, at least during the flight period. The remaining species are associated with reed fen but the project was unable to determine whether any particular age of fen or other factors were important. Five of the species studied are found along the North Sea and English Channel coasts. Despite this there appears to be little similarity in the habitats occupied. The beetle Pogonus luridipennis and the bee Colletes halophilus are both associated with saltmarsh. However, the beetle is found on drying saltpans around lagoons and the bee utilises forage resources on the upper marsh. Both species may be threatened by coastal management schemes but with careful design realignment schemes could create valuable new habitat for these and other saltmarsh and brackish grassland transition invertebrates. Although the remaining three species have been found in habitats that could broadly be described as eroding soft cliff / undercliff this misrepresents the reality. The millipede Trachysphaera lobata occurs in an undercliff on the Isle of Wight but this is very different from the habitats in Wales and different again from those in France. Management prescriptions may be premature without a fuller understanding of requirements but the potential threat to the IOW site from coastal erosion and the construction of a board walk through the site require action. Awareness-raising amongst local and regional ecologists and the site owner / manager followed by site specific discussions are recommended. No conservation action is feasible for the centipede Nothogeophilus turki at this point in time. Much more survey and autecological study is required and it is likely to be many years, if ever, before the necessary understanding of this elusive species is gained. Philorhizus vectensis has been referred to above in the discussion of species associated with shingle. It is more typical of vegetated ledges and landslips on south-facing maritime cliffs in south-west England.

SID 5 (Rev. 05/09) Page 32 of 34

The remaining species included in the project are associated with open woodland although the millipede Metaiulus pratensis probably does not belong within this assemblage (or any of the others above) but further work on its ecology is required. During this study it was found in open scrub on floodplain of the River Medway. Previous records have come from similar sites but also from grassland and in France it is known from caves. As a result of the project little insight was gained into the key aspects of woodland structure for these invertebrates. Possibly the range of species considered was too diverse to enable general conclusions to be drawn, certainly most were difficult to find and therefore study. No records were made of the fungus gnat Neoempheria lineoola, the bee Andrena ferox or the beetle Philorhizus quadrisignatus. The ecology of the millipede Polyzonium germanicum is now better understood and there is no evidence of a threat to its survival. Although previously believed to be typical of sweet chestnut coppice in East Kent the project has established that soil moisture / humidity is a more important environmental factor than woodland type. Indeed the species is not restricted to woodland although in a county such as Kent with a relatively dry climate suitable conditions are only occasionally found outside woodlands. This finding also helps to explain the fact that observers have previously commented on the absence of the millipede from an area of coppice in the first few years after cutting. The guest ant Formicoxenus nitidulus was recorded only rarely but it was not directly targeted. One of its hosts, Formica rufa, appears to declining at the northern edge of its range in northern England but another host species, Formica lugubris, is thriving. As Formicoxenus nitidulus shows no preference for either species this should not affect its status and distribution. However, investigations into the reasons for the decline of Formica rufa in northern counties would help in assessing the future conservation status of Formicoxenus nitidulus. CONCLUSION The 2008 BRIG and UK BRAG workshop identified a number of knowledge gaps that formed a barrier to the effective delivery of conservation targets for Priority Species of invertebrates. This project has attempted to improve knowledge of the distribution and status, clarify the taxonomy and determine the ecology of the species studied. The ultimate aim has been to provide management prescriptions for the conservation of species (or assemblages of species) or at least to provide a foundation of knowledge on which further research can build to develop such prescriptions. Where a need for further survey and monitoring had been identified, the project was generally able to clarify species distributions by collation of existing records but field survey added little information in the form of significant new locations for most species. The main exceptions to this were the millipedes Metaiulus pratensis and Polyzonium germanicum. The discovery of a strong population of Metaiulus pratensis at a new location, 23 years after it was last seen, opens up the opportunity for ecological studies on the millipede. A better understanding of its ecological requirements is vital to future habitat conservation work and would improve the likely success of future survey to locate further populations. Field survey resulted in the discovery of several new locations for Polyzonium germanicum as well as confirmation that the species was present throughout its historical range. This updated knowledge of the distribution of the species will contribute to a reassessment of its conservation status. It had been feared that the millipede was in decline but clearly this is not the case and the improved knowledge of its habitat requirements gained from this work is likely to lead to the discovery of more populations in future surveys. In contrast, the records collated, including from field survey, show a 61% decline in the range of the tiger beetle Cicindela sylvatica when comparing post 1990 with pre 1990 records. Any trends indicated in the data for the other species surveyed are less clear but all of the records collected have been submitted to the relevant recording schemes and so will be available through BRC and NBN for use in future assessments of conservation status. The status of Formicoxenus nitidulus and the host wood ants in the north of England is now better understood and the survey results provide an excellent baseline from which to monitor changes in distribution and population resulting from climatic and land use changes. The project answered the questions regarding the taxonomy of populations of Trachysphaera pill millipedes in England and Wales. It is now clear that the same species Trachysphaera lobata is present at all sites. This species occurs also in France so is not endemic and conservation of the English and Welsh populations becomes less of an international issue. However, the millipede is still known from just three locations in Britain and is at potential risk from stochastic events that could lead to local extinction. During work at Eastcliff, Bembridge three potential threats were identified. The narrow strip of undercliff that is the habitat here is subject to erosion from the sea; slumps of the soft earth cliffs backing the undercliff smother existing habitat and do not become suitable for colonisation by the millipede for an unknown period of time; potentially damaging human activities and management of the area have been noted including tree felling, bonfires, fly tipping, and footpath construction. A series of boardwalks had been installed on the last site visit in 2012. The very low population density of the millipede observed on this occasion may have been due to the timing of the visit within the life cycle of the animal but habitat disturbance may have been important also. The findings of the project will enable information to be provided to Bembridge Parish Council and the IOW county ecologist on the importance of the site and the need for consultation on the species as part of any management plan for the area. Many of the gaps identified in the old UK BAP processes for invertebrate conservation related to inadequate knowledge of the ecology of individual species and species assemblages. This was reflected in the large proportion of the objectives of this project that addressed the resource requirements of species and how these

SID 5 (Rev. 05/09) Page 33 of 34

could be provided for species assemblages through different management techniques. Progress towards achieving the objectives varied between species and was often limited by the difficulty of locating populations of rare species that were large enough to study. Work on Doratura impudica means that the autecology of this species is much better known as the food plant and precise microsite it inhabits have been identified. It will be possible now to advise on dune management to protect the insect and to target future survey efforts more effectively. Knowledge of the autecology of Odynerus simillimus has already enabled advice on nesting habitat creation and general management for the wasp to be provided to ecologists, site managers and land owners within the Norfolk Broads, the Broads Authority, Environment Agency, Broads Internal Drainage Board, Norfolk Wildlife Trust and RSPB. Collating all that was known on the ecology of Colletes floralis has enabled a copy of the report and support on habitat management for the bee to be provided to the site manager at Sandscale Haws NNR in Cumbria. Evidence from work on the autecology of Andrena tarsata has been used in conjunction with the Planning Department of the North York Moors National Park to demonstrate potential damage to foraging areas by an extension of exploratory oil drilling. Aside from limiting the damage to the species from the specific application this was useful also in raising awareness of invertebrate conservation generally, and of Andrena tarsata specifically, with National Park staff. County Wildlife Trusts have been closely involved in several of the studies of habitat management and have gained knowledge of species’ requirements from this. For example, Devon Wildlife Trust hosted work on nest establishment in Formica exsecta and also heathland management for Poecilus kugelannii, work on the ecology of Polyzonium germanicum took place in collaboration with the Kent Wildlife Trust and Surrey Wildlife Trust was closely involved in work on Cicindela sylvatica. Information exchange with other conservation organisations including Butterfly Conservation, National Trust and RSPB has occurred due to research including their reserves as study sites and input into site management plans. Other species for which improved autecological information gained from this project will enable improved habitat management advice to be provided and more efficient future survey and monitoring are Amara fusca, Harpalus froelichi, Odynerus melanocephalus, Cerceris quadricincta, Cerceris quinquefasciata, Lasioglossum angusticeps, Colletes halophilus, Osmia parietina and Bombus ruderarius. In some studies, a greater insight into of the ecology of the species concerned was achieved but still insufficient knowledge is available to be able to recommend management prescriptions. Many of the species studied as part of the wetland group would fall into this category as would Temnothorax interruptus, the two Tapinoma species and Philonthus vectensis. The provisional information on habitat preferences gained from studying Temnothorax interruptus at Dungeness and Tapinoma at Godlingston Heath would become more valuable if incorporated into further studies of these ants at other locations in southern England. The apparent unpredictability of occurrence of Philorhizus vectensis has hampered studies of the beetle and an information guide targeted at general naturalists in southern coastal counties, similar to that produced for the Pogonus species and targeted at naturalists along the east coast during the project, may be an effective way of gaining further information on the ecology of the beetle beyond the end of the project. Much of the knowledge of habitat management for individual species will also be applicable to the conservation of species assemblages. The creation and maintenance of early successional heathland is important in the conservation of a wide diversity of invertebrates. The work with Poecilus kugelannii in Devon demonstrated the importance of different approaches to creating bare ground resources in different circumstances. The conservation of early successional assemblages of invertebrates on Breckland heath will benefit also from these techniques and the work on Amara fusca, Harpalus froelichi and Ophonus laticollis suggest that it may be possible to provide feeding resources in active farmland. Future work could investigate the impact of deliberate actions for increasing food plant availability on Breckland beetles, test the ability of the beetles to find new resources and investigate the extent to which Cross-compliance and Stewardship margins provide occupied habitat. The resource requirements identified for species associated with flower rich grassland suggest that although management advice may need to be site specific on occasion, it would generally involve controlling mowing and grazing regimes to ensure flowering of a diversity of plant species throughout the flight period of the target invertebrates. This approach would benefit a wider group of species than those studied. Investigations of the ecology of the wetland species were inconclusive and more research would be required in order to provide conservation advice for the assemblage. An extension to the project would involve further study of Dolichopus laticola and Dolichopus nigripes and their ecological associations with intermittently cut/grazed old fen-meadow. Asindulum nigrum may form part of this assemblage also. Highlighting the potential of this rather neglected component of fen vegetation systems would be an important outcome of the project and is being discussed with ecologists at the Broads Authority. References to published material

9. This section should be used to record links (hypertext links where possible) or references to other published material generated by, or relating to this project.

SID 5 (Rev. 05/09) Page 34 of 34

Ausden, M. and Gilbert, G. 2012. RSPB Reserves 2012: A review of our work. p.39. [www.rspb.org.uk/Images/rspbreserves2012_tcm9-326414.pdf] Bantock, T. and Stewart, S. 2012. Wildlife Reports: True bugs, leafhoppers and allies. British Wildlife. 23(4):277-279. Knowles, A. 2010. Some observations on the nesting behaviour of Colletes halophilus Verhoeff (Hymenoptera: Apidae). British Journal of Entomology and Natural History. 23(3):217-218. Lee, P. 2011. Ground beetle studies in the Brecks. Suffolk Biodiversity Partnership. Partners’ News September 2011. [www.suffolkbiodiversity.org/content/suffolkbiodiversity.org/PDFs/news-and-events/SBP%20newsletter%20September%202011.pdf] Lee, P. 2011. BMIG business. British Myriapod & Isopod Group Newsletter. 23. [www.bmig.org.uk/sites/default/files/bmignews23.pdf] Lee, P., Read, H.J. and Gregory, S.J. 2011. The re-discovery of Metaiulus pratensis Blower & Rolfe, 1956 (Diplopoda, Julidae) in Kent. Entomologist’s Monthly Magazine. 147:156. Robinson, E. 2011. The conservation status of Formica rufa and Formica lugubris in northern England and the Midlands. Abstract of presentation given to Wood Ant Symposium 18th – 19th October 2011. [http://www.woodants.org.uk/node/21] Stewart, A. 2012. Species focus: Doratura impudica. LEDRA Newsletter of the Auchenorrhyncha Recording Scheme for Britain & Ireland. 2. [www.ledra.co.uk/newsletters/LEDRA_Issue2_Spring2012.pdf] Walters, J. and Telfer, M.G. 2011. Guides to British Beetles. 9. Driftliners Pogonus. Ground Beetles (Coleopera: Carabidae). [http//dl.dropbox.com/u/48019879/websiteDownloads/gbb/9%20Pogonus.pdf] Several peer reviewed papers are in preparation. The Status and Ecology of Polyzonium germanicum in England Taxonomy of British populations of Trachysphaera lobata Habitat differences in Dolichopus nigripes and D. laticola The status and ecology of Doratura impudica in England Lipara pullitarsis new to Britain including a key to adults of the Lipara species and a description of the galls The status and ecology of Lipara similis and other Lipara species in Britain