The impacts of deer on woodland butterflies: the good, the bad and the complex

Transcription

1 The impacts of deer on woodland butterflies: the good, the bad and the complex R.E. FEBER 1 *, T.M. BRERETON 1, M.S. WARREN 1 AND M. OATES 2 1 Butterfly Conservation, Manor Yard, East Lulworth, Wareham, Dorset BH20 5QP, England 2 The National Trust, 33 Sheep Street, Cirencester, Gloucestershire GL7 1RQ, England * Present address: Wildlife Conservation Research Unit, Department of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, UK Summary Deer grazing is an important feature of many key butterfly habitats in Britain, yet few data are available on its impacts. Butterfly populations can be affected in a number of ways, through effects on the local availability of larval food-plants or nectar sources, to larger-scale changes in habitat structure and management. Many woodland butterflies have historically relied on clearings in coppiced woodland, but current high numbers of roe deer (Capreolus capreolus), and to a lesser extent fallow deer (Dama dama), can severely reduce tree regrowth and are now a major disincentive to the maintenance of this traditional form of management. In contrast, deer grazing may be very beneficial in some woodland habitats. In Scotland, most colonies of the threatened pearl-bordered fritillary (Boloria euphrosyne) rely to some extent on grazing by deer, to slow down natural regeneration and maintain open bracken (Pteridium aquilinum) habitats in canopy gaps and along woodland edges. Such areas have recently been targeted by schemes to encourage native woodland, but the fencing out of deer and other grazing animals can lead to rapid loss of these crucial woodland habitats. Far more research is needed to determine the full extent of these impacts and implications for conservation programmes. Introduction Whilst many invertebrates depend on mature woodland and high forest, a number of others rely on the early successional stages within woodlands or woodland edges for their survival. Nearly three-quarters of Britain s resident species of butterfly regularly breed in woodland, and about one-third are found exclusively or primarily in woods through much of their British range (Warren and Key, 1991). Of these true woodland butterflies, around half are strongly associated with early successional habitats within woodlands. They include several species which are either rare or rapidly declining, such as the high brown fritillary (Argynnis adippe), the pearlbordered fritillary (Boloria euphrosyne), and the heath fritillary (Mellicta athalia), all of which are listed as priority species in the UK Biodiversity Action Plan Steering Group Report (Anon., Institute of Chartered Foresters, 2001 Forestry, Vol. 74, No. 3, 2001

2 272 FORESTRY 1995). In addition to woodland butterflies, many typically grassland species will also use rides, glades or clearings within a wood. Most work to date on the conservation of woodland butterflies has concentrated on the role of management of rides and coppiced areas in creating and maintaining such early-successional habitats within woodlands (e.g. Robertson et al., 1995). However, while it is well known that the type, level and timing of grazing pressure plays a key role in the management of grassland habitats for butterflies (Butterflies Under Threat Team, 1986; Oates, 1993), similar data are not available for woodlands. The current increases in deer densities in many parts of Britain have had ecological and economic impacts on both commercial and conservation woodlands (Putman and Moore, 1998), but, although the implications for butterflies are likely to be considerable, few studies have quantified them. In this paper we review information on the impacts of deer on butterflies, and consider the implications of grazing and browsing activities of deer for the conservation of woodland butterflies and their habitats. Importance of grazing and browsing Taken as a whole, butterflies in the UK are highly dependent on grazing and browsing: over threequarters of resident species breed in habitats that are maintained to a greater or lesser extent by vertebrate herbivores. Many are grassland species where grazing is essential to maintain open conditions needed for breeding (e.g. Butterflies Under Threat Team, 1986). Grazing affects the structure and composition of plant communities, with consequences for all stages of the butterfly s life cycle. For example, the adonis blue (Lysandra bellargus) and silver-spotted skipper (Hesperia comma) require warm conditions and are restricted to short-grazed vegetation (Thomas, 1991). Other species such as the marsh fritillary (Eurodryas aurinia) are associated with extensive cattle grazing, which maintains a varied sward structure and high densities of the larval foodplant, devil s-bit scabious (Succisa pratensis) (Warren, 1994). The impacts of grazing and browsing by deer on woodland invertebrates and their habitats must be considerable, but are the subject of few studies (see review by Stewart, 2001). One of the few studies of the effects of deer on butterflies by Pollard and Cooke (1994) reported an indirect effect of muntjac deer (Muntiacus reevesi) on egglaying by the white admiral (Ladoga camilla). They showed that a substantial proportion of potential egg-laying sites, the lower leaves of honeysuckle (Lonicera periclymenum), were removed by deer in Monks Wood in Cambridgeshire. Large-scale habitat changes and impacts on butterflies brought about by changes in grazing pressure are illustrated through historical records of the New Forest. The changing status of butterflies in the New Forest inclosures is well documented in the entomological literature and is summarized by Oates (1996) and Oates et al. (2000). From this wealth of anecdotal information, it is clear that both the presence and absence of deer over time has had profound effects on butterfly populations. The golden era of butterfly and moth collecting in the New Forest was ushered in by the New Forest Deer Removal Act of 1851, which virtually eliminated a fallow deer population some strong (Tubbs, 1986) from the entire forest region. In 1900 there were only an estimated 200 deer in the forest. The entomological literature gives the impression that the abundance of butterflies in the inclosures during the 1890s was orders of magnitude greater than has been recorded in any woodland system in Britain in recent decades (e.g. Castle Russell, 1952). Dramatic changes in the composition of the inclosures occurred during the twentieth century, as a result of widespread felling of broadleaved woodland and coniferization and the penetration of the inclosures by commoners stock in the late 1960s. The impact of the latter was compounded by a massive rise in the deer population. High levels of grazing and browsing have had adverse impacts on the shrub layer, both under the canopy and along ride edges: the impact of this on butterfly populations has been enormous. The white admiral, which formerly abounded in many broadleaved and conifer inclosures, has been reduced to a few inclosures where deer have been actively controlled (Oates et al., 2000). It breeds exclusively on honeysuckle, which is highly palatable to deer, especially fallow deer. Other butterfly species such as meadow brown (Maniola jurtina), and ringlet (Aphantopus

3 DEER AND WOODLAND BUTTERFLIES 273 hyperantus), had declined in the forest due to the loss of nectar sources along ride edges, particularly bramble (Rubus fruticosus) which was formerly abundant along many rides. The loss of bramble is largely attributable to forestry practices during the mid-twentieth century and browsing by deer. During the late 1990s, efforts to control deer activity in some of the inclosures have met with an encouraging response by butterfly populations (Oates et al., 2000). Coppice management The impacts of deer on butterflies of early successional habitats may be greatest in coppiced woodlands, for both ecological and economic reasons. Historically, coppicing provided ideal conditions for many woodland butterflies that depend on early successional vegetation in new clearings, and open areas such as rides and glades. The pearl-bordered fritillary, for example, requires violets (Viola spp.) in open sunny situations within woodlands, growing amongst a shallow layer of leaf or bracken litter, for egglaying, and an abundant supply of woodland herbs such as bugle (Ajuga reptans) to act as nectar sources (Brereton et al., 1999). Where coppicing has been abandoned, the fresh clearings that the butterfly needs to survive are no longer created, and woodlands quickly become overgrown and colonies shaded out. The demise of traditional coppice management and increased shading in woodlands have been implicated in the declines of this and many other woodland butterflies (Warren and Key, 1991; Warren, 1993). In recent years, coppicing is often practised for conservation rather than commercial reasons, as it has been shown to promote diverse communities of, for example, ground flora (Barkham, 1992) and birds (Fuller, 1992). The Coppice for Butterflies Challenge grant scheme was initiated by the Forestry Authority in 1996 with the aim of stimulating coppice restoration in eight regions of England where threatened woodland butterflies still survive (Warren, 1996). For the two years following coppice restoration, the sites were assessed for changes in vegetation, butterfly abundance, and coppice structure (Clarke, 1999). In all areas where deer were present, deer fencing was a requirement of the Challenge grant. The results of the survey showed that deer fencing had a strong effect on stool growth, ensuring good coppice development necessary for the long-term viability of the coppice (Table 1). However, in the fenced areas, coppice regrowth was usually extremely rapid. While this is highly desirable from a commercial viewpoint, it has potentially adverse consequences for butterfly populations. Coppice only provides suitable breeding habitat for early successional species for a limited time, often as little as 2 3 years, and it is possible that the strong growth seen on the Challenge sites may shorten this window of suitability even further. However, it is clear from this and many other reports that, where deer are present, they can cause high levels of damage to stools in unfenced plots, and regrowth of stools can be checked to levels where the ground vegetation may become dominated by grasses and sedges (e.g. Cooke, 1994; Putman, 1996). This reduces suitability of the habitat for butterflies needing the earliest successional vegetation, as well as damaging the commercial value Table 1: Comparison of coppice growth inside and outside deer fencing at Coppice for Butterflies Challenge sites (taken from Clarke, 1999) Mean coppice Mean coppice height (m): height (m): Significance Site fenced unfenced level (t-test) Bentley Wood 1.3 ± ± 0.1 t = 10.1, P < Pitton Copse 2.0 ± ± 0.1 t = 5.6, P < Hunts Copse 1.3 ± ± 0.1 t = 0.4, P > 0.1 Redlynch Plantation A 2.1 ± ± 0.1 t = 11.8, P < Redlynch Planatation B 1.2 ± ± 0.1 t = 10.1, P < Stockton Wood 2.0 ± ± 0.05 t = 14.8, P < 0.001

4 274 FORESTRY and long-term viability of the coppice crop. Clarke (1999) suggests that a compromise could be reached on some sites, allowing limited access to deer after, perhaps, the second year of growth. This could be achieved by leaving gates open, and the resulting grazing, browsing and creation of tracks might encourage greater vegetation variation and retain some patches of bare ground. However, much would depend upon the intended commercial future of the coppice crop, the biodiversity interest at the site, and the local density of deer. Young conifer plantations Where young conifers are planted on ancient woodland sites, the conditions prior to canopy closure can be quite similar to those found in coppiced areas. At Cirencester Park Woods in Gloucestershire, pearl-bordered fritillary populations were intensively studied between 1994 and Eight out of 10 colonies were in young conifer plantations and, significantly, five of these were in areas of young conifer plantation favoured by the woods sizeable fallow deer population (Oates, 1995). Here, the deer had extensively damaged young trees and encouraged flushes of violets by grazing competitive grasses, browsing the bramble and poaching the ground. In many of the plantations, violets were most common along the edges of deer runs. Seven of the colonies were in deer-damaged conifer plantations >10 years old, and it is unusual for plantations to support the butterflies for so long. The largest colony was in an area showing three obvious replanting phases due to repeated deer damage, and records show that the butterfly was breeding in this plantation in 1982, suggesting colony longevity of at least 14 years. Other causes of localized crop fail, notably the difficulty of establishing young trees in free-draining soil, may also be significant factors in prolonging habitat suitability. In addition to their effects on egg-laying habitats, Oates (1995) suggests that browsing of the underwood layer by deer may facilitate the butterfly s powers of colonization throughout the woodland, by removing vegetation that might be an obstacle to dispersal. Open woodlands The impacts of deer, even on one species alone, may differ widely according to habitat, and over the short- to medium- or long-term. Impacts of deer on pearl-bordered fritillary can be complex in lowland English woods, but in Scotland recent research suggests that deer grazing may be vital in ensuring the long-term survival of the species. A survey of past, present and potential pearlbordered fritillary sites, conducted in 1997 and 1998, showed that woodland colonies were located either in clearings between tree cover or on the woodland edge (Brereton et al., 1999). An important feature of all occupied sites was that they were composed of fairly short vegetation, which, as a result of grazing pressure, lacked significant natural tree regeneration. Although the west of Scotland holds a large proportion of the UK s remaining Atlantic woodlands, they are still considered to be under threat. The rarity of large areas of woodland and the low levels of tree regeneration has led to a government-funded Woodland Grant Scheme (WGS) and Farm Woodland Premium Scheme to encourage the management, regeneration and extension of existing woodland areas. The standard practice is to exclude all grazing animals by fencing and to eradicate bracken by spraying with Asulam to encourage new growth. Schemes last for either 10 or 15 years, after which stock and deer can be reintroduced. In a sample of 40 woodland sites occupied by the pearl-bordered fritillary in Highland and Argyll in 1997, it was found that a significant proportion (~50 per cent) of colonies were located in woodland compartments that had been recently fenced to exclude stock and deer and encourage woodland recovery. In May and June 1998, a mark release recapture programme of adults was carried out across a pearl-bordered fritillary site in Argyll, which extended over 6 km and included two WGS established in 1995 and The largest number of adults was found in woodland edge and open woodland habitats that had been recently fenced against deer (Table 2). Vegetation survey work at the site showed that, in the short term, exclusion of stock, and a resulting reduction in grazing pressure, had created a slightly ranker sward which, in turn, had facilitated a greater flowering abundance of preferred

5 DEER AND WOODLAND BUTTERFLIES 275 Table 2: Relative density of pearl-bordered fritillary butterflies in different habitat types at Lochawe, Argyll, Scotland Total area No. of adults Relative adult Habitat N* (ha) caught density (no./ha) Wood edge: deer fenced Wood edge: unfenced, sheep grazed Open woodland: deer fenced Open woodland: unfenced Data from mark release recapture (MRR) studies conducted in 1997 (see text for details) * N is number of MRR compartments. nectar sources. In particular, the mean density of bugle flowering spikes was significantly higher within stock-fenced woodland edges than elsewhere at the site (Brereton and Warren, 1999). However, grazing exclusion is likely to create problems for pearl-bordered fritillary in the future. Although, in the short-term, release from grazing increases the supply of adult nectar sources and larval food-plants, bracken litter soon builds up in the absence of trampling and becomes too dense to either support violets or provide a warm micro-climate for development of the immature stages. Where chemical bracken control is successful, canopy gaps may soon be infilled by regenerating trees. Survey work at one west Highland site in 1997 indicated that relative adult densities in WGS plots fenced for 5 years, were ~75 per cent lower than those in adjacent unfenced plots (12 adults hour 1 ha 1 compared with 40 adults hour 1 ha 1 ; see Warren et al. (1984) for details of timed count methodology). The decline was primarily caused by habitat loss, due to natural regeneration in ungrazed canopy gaps. The development of tree cover may cause further problems, with remaining colonies becoming increasingly isolated due to the species poor powers of dispersal. This could impact on the butterfly s potential to maintain viable metapopulations at a landscape scale. The activities of deer have beneficial effects on other butterfly species as well. For example, roe deer are present at chequered skipper (Carterocephalus palaemon) sites in Scotland all year round, and are considered to be very important to the species by preventing succession in larval areas, and keeping adult areas open where scrub invasion is detrimental. The culling of roe deer on one reserve has had the desired beneficial effect on tree and shrub regeneration, but has necessitated artificial scrub clearance to maintain areas for chequered skipper (Ravenscroft, 1995). Conclusions Although information is currently patchy, it is clear that deer are having a profound effect on several woodland butterfly species. In some circumstances the impact is detrimental, but in others it can be highly beneficial, for example where it helps to maintain woodland edge and bracken habitats. However, the ecological impacts of deer on butterflies and other insects seem to be quite complex and require considerable further research to elucidate. Much depends on the size of deer populations and factors such as seasonal diet availability and diet preferences, activity patterns, and the impact of other large herbivores and human intervention. Many studies have looked at the effects of total relief from the effects of grazing and browsing following exclosure, but future work should concentrate on the effects of different levels of grazing, and grazing by different species of deer, on floral and faunal communities. Because most of the butterfly species affected by deer are high conservation priorities, such research is needed urgently both to inform deer control programmes and to ensure sound habitat management in the future. Acknowledgements Much of Butterfly Conservation s work on fritillaries has been funded by English Nature (Species Action Programme), Scottish Natural Heritage, Countryside

6 276 FORESTRY Council for Wales, and ICI. The monitoring programme for the Coppice for Butterflies Challenge is supported by the Forestry Commission. Our thanks to Butterfly Conservation volunteers and other conservation partners for helping to collect survey data. Richard Fox, Nigel Bourn, Neil Ravenscroft, Rachel King and Charlie Rugeroni helped with research work in Scotland in References Anon Biodiversity: the UK Steering Group Report. Volume Two, Action Plans. HMSO, London. Barkham, J.P The effects of coppicing and neglect on the performance of the perennial ground flora. In The Ecological Effects of Coppicing. G.P. Buckley (ed.). Chapman & Hall, London, pp Brereton, T.M. and Warren, M.S Ecology of the pearl-bordered fritillary butterfly in Scotland and possible threats from bracken eradication measures in Woodland Grant Schemes. In Bracken Perceptions and Bracken Control in the British Uplands. J. Taylor (ed.). International Bracken Group Special Publication No. 3, Lancaster. Brereton, T.M., Roberts, E. and Warren, M.S Action for the Pearl-bordered Fritillary. Unpublished report to Butterfly Conservation, Wareham, Dorset. Butterflies Under Threat Team 1986 The Management of Chalk Grassland for Butterflies. Focus on Nature Conservation No. 17. Nature Conservancy Council, Peterborough. Castle Russell, S.J The New Forest in the nineties and after. Entomol. Rec. 64, Clarke, S.A Coppice for Butterflies Challenge Grant Scheme: Butterfly Monitoring Programme. Report to Butterfly Conservation and the Forestry Authority. Butterfly Conservation, Wareham, Dorset, UK. Cooke, A.S Colonisation by muntjac deer Muntiacus reevesi and their impact on vegetation. In Monks Wood National Nature Reserve, the Experience of 40 Years M.S. Massey and R.C. Welch (eds). English Nature, Peterborough, pp Fuller, R.J Effects of coppice management on woodland breeding birds. In The Ecological Effects of Coppicing. G.P. Buckley (ed.). Chapman and Hall, London, pp Oates, M. R The management of southern limestone grasslands. Br. Wildl. 5, Oates, M A study of the Pearl-bordered Fritillary in Cirencester Park Woods. West Midlands Butterfly Report. Oates, M.R The demise of butterflies in the New Forest. Br. Wild. 7, Oates, M.R., Taverner, J. and Green, D The Butterflies of Hampshire. Pisces Publications, Newbury. Pollard, E. and Cooke, A.S Impact of muntjac deer Muntiacus reevesi on egg-laying sites of the white admiral butterfly Ladoga camilla in a Cambridgeshire wood. Biol. Conserv. 70, Putman, R.J Deer on National Nature Reserves. Problems and Practices. English Nature Research Reports 173. English Nature, Peterborough. Putman, R.J. and Moore, N.P Impact of deer in lowland Britain on agriculture, forestry and conservation habitats. Mammal Rev. 28, Ravenscroft, N.O.M The conservation of Carterocephalus palaemon in Scotland. In Ecology and Conservation of Butterflies. A.S. Pullin (ed.). Chapman and Hall, London, pp Robertson, P.A., Clarke, S.A. and Warren, M.S Woodland management and butterfly diversity. In Ecology and Conservation of Butterflies. A.S. Pullin (ed.). Chapman and Hall, London, pp Stewart, A.J.A The impact of deer on lowland woodland invertebrates: a review of the evidence and priorities for future research. Forestry 74, Thomas, J.A Rare species conservation: case studies of European butterflies. In The Scientific Management of Temperate Communities for Conservation, I.F. Spellerberg, F.B. Goldsmith and M.G. Morris (eds). Blackwell Scientific Publications, Oxford, pp Tubbs, C.R The New Forest. Collins New Naturalist, London. Warren, M.S A review of butterfly extinction in central southern Britain: I. Protection, evaluation and extinction on prime sites. Biol. Conserv. 64, Warren, M.S The UK status and suspected metapopulation structure of a threatened European butterfly, the Marsh Fritillary Eurodryas aurinia. Biol. Conserv. 67, Warren, M.S Giant step to save woodland butterflies: the Coppice for Butterflies Challenge. Butterfly Conservation News 63, Warren, M.S. and Key, R.S Woodlands: past, present and potential for insects. In The Conservation of Insects and their Habitats. N.M. Collins and J.A. Thomas (eds). Academic Press, London, pp Warren, M.S., Thomas, C.D. and Thomas, J.A The status of the heath fritillary butterfly, Mellicta athalia Rott., in Britain. Biol. Conserv. 29,