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: [email protected]

SID 5 Research Project Final Report

z Note In line with the Freedom of Information Act 2000, Defra aims to place the results Project identification of its completed research projects in the public domain wherever possible. The CR 0470 SID 5 (Research Project Final Report) is 1. Defra Project code designed to capture the information on the results and outputs of Defra-funded 2. Project title research in a format that is easily Understanding the role of woodland management in the publishable through the Defra website. A conservation of UK BAP moths SID 5 must be completed for all projects.

• This form is in Word format and the boxes may be expanded or reduced, as 3. Contractor appropriate. Wildlife Conservation Research Unit organisation(s) (WildCRU, Department of Zoology, z ACCESS TO INFORMATION University of Oxford) The information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual £ 97,460 researchers or organisations outside 4. Total Defra project costs Defra for the purposes of reviewing the (agreed fixed price) project. Defra may also disclose the information to any outside organisation 5. Project: start date ...... 01/06/2010 acting as an agent authorised by Defra to process final research reports on its 31/03/2011 behalf. Defra intends to publish this form end date ...... 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.

SID 5 (Rev. 07/10) Page 1 of 30 6. It is Defra’s intention to publish this form. Please confirm your agreement to do so...... YES (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

SID 5 (Rev. 07/10) Page 2 of 30

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. It has become clear that it is not only rare, localised, and specialist species that require conservation action, but common, widespread, and generalist species too. Over recent decades, severe declines have been recorded in a range of taxa for what were once very common species. Evidence suggests that these widespread but declining species need different conservation approaches to those required for rare species. In general, rare species have highly specific resource requirements, which demand targeted conservation management strategies. Because of their generalist character and often broader resource requirements, widespread species need a different approach to address their needs and reverse their population declines. What this approach should consist of is not clear for many taxa.

This dichotomy clearly shows up in the species-rich group of moths. In the UK, 81 specialist species are listed as Biodiversity Action Plan (BAP) Priority species and, in 2007, 71 widespread species showing evidence of rapid and severe declines (i.e. Rothamsted Research national abundance trends over 35 years: >69% decline) were added. This group of widespread but rapidly declining species were identified by species experts as requiring further research. Most of the species within the latter group are to a greater or lesser extent associated with woodland complexes, which include both closed and open areas. As forestry management has changed drastically during the last century, our research investigated the effects of some of these changes on this suite of species.

The main objectives of our study focused on increasing the understanding of the effects of currently typical woodland conservation management practices which aim to open up dense, closed woodland structures, in particular coppicing and woodland ride widening, on this study group of widespread but rapidly declining species. The exact cause(s) of their decline is still unknown, and research is urgently needed to inform conservation action at both local management and wider policy levels.

We placed light-traps at 36 fixed sites within the Tytherley woodland landscape (Hampshire/Wiltshire border), which represents a unique study area into management requirements for these species since the level of conservation management within this landscape has recently been increased and diversified as part of Butterfly Conservation’s South East Woodlands Project. Six ‘management’ treatments (young/medium/old coppice; wide/standard ride; non-coppiced deciduous ‘standard woodland’) were each represented by six sites, and these were all sampled nine times between July and mid-October 2010. We avoided bias by sampling all management groups every trap night, in equal numbers, and only under suitable weather conditions.

The total list amounted to 11,670 individuals from 265 species of larger moth. Thirty-eight of these were part of the study group (891 individuals). Seven woodland species that might have been expected to occur did not and their dramatic national declines (87-97% over 35 years) may well explain their absence. We trapped 249 individuals of fifteen nationally scarce / Red Data Book (RDB) species: Festoon Apoda limacodes, Triangle Heterogenea asella, Mocha Cyclophora annularia, Devon Carpet Lampropteryx otregiata, Great Oak Beauty Hypomecis roboraria, Small Black Arches Meganola strigula, Kent Black Arches Meganola albula, Double Line Mythimna turca, Mere Wainscot Chortodes fluxa, Waved Black Parascotia fuliginaria, White-line Snout Schrankia taenialis, Lunar Yellow Underwing Noctua orbona, Light Crimson Underwing Catocala promissa, and Dark Crimson Underwing Catocala sponsa. The last three of these are also BAP Priority species. A remarkable find was Clifden Nonpareil Catocala fraxini. Seven individuals, both male and female, were trapped on four nights, always near suitable habitat (i.e. mature Aspen trees), pointing strongly towards a local population of this species and confirming its recent re- establishment after it became extinct in Britain in the 1960s.

A key finding was that overall abundance was lowest in coppice (characterised by bare ground / exposed conditions) and wide rides (exposed conditions), whereas most moths were found in the sheltered standard rides and standard woodland. Overall species richness was lowest in coppice and highest at standard/wide rides and standard woodland. It has been shown that factors such as bare ground and exposedness impact on ambient temperature and levels of convective cooling for day-flying Lepidoptera. It is hence likely they are also having an impact on activity/occurrence of night-flying Lepidoptera. This overall picture hence contrasts with the one for heliothermic day-flying Lepidoptera (i.e. butterflies and day-flying moths), where open woodland conditions generally mean higher abundance and species richness, due to higher day temperatures, higher solar irradiance, and other key adult and larval resources.

A closer look, however, reveals that while this overall picture for larger moths applies to most species,

SID 5 (Rev. 07/10) Page 3 of 30 irrespective of their scarcity or national trend, the study group showed the smallest difference in abundance between sheltered and more open sites. Wide woodland rides in particular had relatively high abundance of these rapidly declining species. Also, when dividing the species into groups depending on their level of woodland affinity, we found that the species group without any direct affinity to woodland had much higher abundance and species richness in wide woodland rides and young coppice than in older coppice and more sheltered sites, compared to the other species groups, with some or strong woodland affinity.

Another key finding was that the total area of woodland surrounding coppiced plots had a strong positive effect on the abundance of the study group moths (i.e. those with national declines >69%). It also increased overall species richness (at least in medium/old coppice), particularly through its effect on the group of species with national declines up to 69% (Rothamsted Research trends over 35 years).

In addition to general patterns of abundance and species richness, we found that 124 species (47% of the total) occurred in lower numbers at sheltered than at more exposed sites (i.e. standard rides/standard woodland vs. wide rides/coppice). Of these 124 species, 49 (18% of the total) were in fact restricted to these open areas. This pattern was strongest for the study group moths, with 22 of the 38 species (58%) occurring in equal or higher numbers at managed (open) compared to unmanaged sites. Focusing on this group, we showed that more species occurred in young and medium coppice than in old coppice, although overall abundance increased from young, over medium, to old coppice, leading us to the conclusion that the whole coppice cycle is of value to this rapidly declining group of species. We also believe that widening woodland rides is valuable to them, as their overall abundance was higher in wide compared to standard rides.

The species-specific analyses also show that coppicing and ride widening are valuable conservation tools for a significant number of larger moth species, especially benefiting the study group species, which are not necessarily the most abundant species. We suggest the mechanism behind the pattern of increased total woodland species richness involves an increased structural and micro-climatic diversity of woods, which provides resources for species with an affinity for more open biotopes. We also show that the greatest benefits to increased coppicing and ride widening are likely to be had within larger woods and mainly so for nationally declining and rapidly declining species.

However, the complementary analyses on overall abundance and species richness show that the sheltered, dark, humid, late-successional high forest biotope is characterised by high numbers of both individuals and species, and is at the same time especially important for some scarce and RDB species. These complementary findings demonstrate the value of this woodland biotope and that this successional stage clearly needs to be accounted for within woodland management.

Based on the results of these complementary analyses, our main recommendations are (i) to enlarge existing high forest and create new ones, in order to have a sufficient extent of cores of typical dark woodland biotope, and additionally (ii) to ‘buffer’ these dark cores from the open ‘matrix’ by creating, at their edges, lighter woodland zones which can indeed be achieved by implementing coppicing or wide woodland rides. Such zoning will safeguard the dark environment needed by shade- or moisture-loving woodland specialists, whilst at the same time allowing easy access for species of mixed and more open biotopes to a light and open woodland biotope. We show that a conservation focus on either coppicing/ride widening or sheltered woodland will not deliver as much biodiversity value as the combined implementation of both approaches at once: this two-tier approach may well be vital both for endangered woodland specialists and declining, once-widespread species that are finding it ever more difficult to locate enough resources within current degraded landscapes.

Regarding dissemination of information, all records were passed to the National Moth Recording Scheme of Butterfly Conservation’s Moths Count Project and via this to the NBN-gateway. As such, they are available for future research and conservation planning, and we have helped to improve knowledge on the status (temporal trends in abundance and distribution) of larger moths, which is especially welcome for those species of conservation concern, as it helps to better understand the causes of their national declines.

The findings have been and will be disseminated as clear guidance on woodland management to conservation practitioners, woodland owners, relevant (governmental) bodies, and the wider scientific community. Specifically, by informing the design of conservation management within English woodlands (e.g. design of Woodland Improvement Grants), we have facilitated the translation of our findings into practical conservation, especially benefiting the study group of species. Ultimately, it is only such evidence-based conservation actions that are able to reverse negative trends observed for woodland biodiversity.

SID 5 (Rev. 07/10) Page 4 of 30 In conclusion, this study has significantly increased the autecological understanding of the study group by demonstrating clear differences in abundance/species richness between different woodland management regimes. In line with the observed results, we believe that populations of these widespread, but nationally rapidly declining species could be significantly increased by an increased and landscape-scale implementation of coppicing and ride widening within the outer zone of, preferentially, large woodlands, without compromising habitat availability and quality for endangered woodland specialists.

Now that this project has developed a standard set of sampling points which allow assessment of the importance of a range of management features, it would be interesting to see whether the results hold true by sampling an additional year (using the same methodology), and especially so by sampling during spring (i.e. generally colder conditions). Sampling in the current project only started in July which meant that species with an earlier flight season were excluded. The colder spring conditions may also result in different effects of the management treatments on moths compared to summer conditions. Additionally, a follow-up study would also allow more data to be obtained on habitat preferences of the re-established Clifden Nonpareil and the other 14 recorded nationally scarce/Red Data Book species.

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: z the scientific objectives as set out in the contract; z the extent to which the objectives set out in the contract have been met; z details of methods used and the results obtained, including statistical analysis (if appropriate); z a discussion of the results and their reliability; z the main implications of the findings; z possible future work; and z any action resulting from the research (e.g. IP, Knowledge Transfer).

Scientific background:

It has recently been demonstrated that the majority of common, widespread moth species have declined during the last decades1. In Britain, 71 species (i.e. 21% of common species) declined severely (i.e. national decline >69%) and were identified by species experts as requiring further research and listed as ‘widespread but rapidly declining’ BAP Priority species in 2007 as a result. Recent work on moths has focused on farmed landscapes and the design of agri-environment schemes to reverse declining trends. The uptake of certain management options (for example, wide, nectar-rich field margins and hedgerow trees) and the landscape-scale implementation of these schemes have been shown to be important factors, beneficial both for widespread moth species2,3,4 and possibly more localized species as well5.

However, the cause of declines in populations of widespread moth species is not well understood, and factors other than management of farmland are likely to be important. Work on other Lepidoptera (butterflies) has demonstrated that woodlands have lost a significant proportion of their original butterfly fauna during the second half of the last century6. As well as typical, specialised woodland species being lost, so also have been species that make use of open areas (e.g. rides, heaths, meadows) within woodlands7. The main reason for this loss is afforestation of open areas and intensive, high forest management practices that have almost completely replaced traditional coppicing

SID 5 (Rev. 07/10) Page 5 of 30 practices7, particularly since WWII. It seems very likely that it is a decline in habitat quality, associated with the loss of this more open and lighter woodland biotope, that has caused the decline in butterflies associated with woodland, as the total area of woodland has often remained stable or has even increased in western Europe8,7,6,9. Woodland conservation management targeted at extended-width rides and coppiced areas creates and maintains early-successional biotopes, providing sunny, yet somewhat sheltered areas of host plants and nectar sources for many butterfly species within woodland10,11,12,13,14,8,15. As coppicing is carried out rotationally (e.g. cycles of 7-10yrs), the amount of light and warmth reaching the forest floor in coppiced areas changes throughout the different successional phases16, with some species strongly favouring the first successional years, while other more shade- tolerant species preferring the later successional stages of the coppice cycle13,17.

Semi-natural, broad-leaved, lowland woodland is likely to be an extremely important habitat for widespread moth species, with an estimated 60% being highly or partly dependent on this biotope. However, while the impacts of contrasting woodland management have been relatively well studied, and are well understood for butterflies10,11,12,18,19,20,21,22, there has been little work done on the effects of woodland management for moths (but see very recently23). Studies only relate to effects of large-scale timber harvesting in boreal or tropical forests, with moth communities found to respond predictably to management practices, driven by changes in plant communities24,25.

In England, broad-leaved, semi-natural woodlands are embedded within agricultural landscapes, and these sheltered biotopes contrast strongly with the exposed agricultural landscape. Semi-natural woodland provides viable habitat resources (e.g. food plants, nectar sources, shelter) for large suites of moth species26,27. Some species are restricted exclusively to rather dense woodland and, although the majority of species are found in more open biotopes, a large proportion of them are also found within open woodland complexes as they provide resources to a greater or lesser degree28. However, no work has been done to quantify these associations.

The overall aim of this project was to quantify the effect on the study group of widespread but rapidly declining moth species of elements of woodland management practices that are known to be important for the conservation of butterflies (i.e. coppicing and ride management). These findings will have implications for the conservation of the total group of 71 widespread but rapidly declining BAP Priority moth species, and many other moth and invertebrate species too.

The project’s four key objectives:

(1) Quantify the presence/absence and abundance of a suite of 65 widespread but rapidly declining BAP Priority moth species (not all 71 species due to the seasonal and geographical restriction) between July and September in a lowland, broadleaved woodland landscape in southern England. (2) Investigate the impact of contrasting woodland management practices on overall species richness and abundance of individual species within the study group. Specifically, study the effect of two woodland conservation practices by comparing (i) coppiced woodland with standard ‘high-forest management’ woodland and (ii) extended-width rides with standard rides. (3) Compare the extent of differences in presence/absence and abundance of the study group species between three age classes of coppiced woodland: young (<3yrs), medium (3-6 yrs), and old (>6-9yrs). (4) Disseminate the findings via clear management guidance for local conservation practitioners and private woodland owners, via articles targeted at local and national governmental and non-governmental organizations, and via peer-reviewed papers to the scientific community. The results of this project will help deliver conservation targets for this group of widespread but rapidly declining BAP Priority moth species, in particular, by informing the design of conservation management within English woodlands.

SID 5 (Rev. 07/10) Page 6 of 30 Methods:

Presence/absence, abundance and species richness (i.e. number of species) data of macro-moth species were collected from early July until mid-October 2010 at 36 sites within the Tytherley woodland landscape (east of Salisbury, UK, with Bentley Wood ‘Site of Special Scientific Interest’ at its center) (Fig. 1), during a total of 27 trap nights, with 12 sites sampled at any one night. Each of the sites was sampled on nine different occasions (i.e. 324 single trap events). The Tytherley woodland landscape was chosen as it has a relatively diverse Lepidoptera fauna and has been the focus of long-term and ongoing woodland conservation management within Butterfly Conservation´s South East England Woodlands Project.

Data were collected from six experimental ‘woodland management’ treatments, each represented by six sites: young (1-2 years)/medium (3-6 years)/old (7-9 years) hazel coppice; wide (>20 m)/standard (<10 m) ride; non-coppiced, high deciduous oak forest ‘standard woodland’). We avoided bias by sampling all treatments every trap night, in equal numbers, and only under suitable weather conditions.

Light trapping is a passive method proven to be highly suitable to sample moth communities. Macro-moths were sampled using Heath pattern actinic light traps (6W)29, which operate on the ‘lobster-pot principle’, whereby individuals are drawn to an actinic tube secured vertically between baffles, fall unharmed down a funnel, and rest inside the trap2. Although this battery-run trap type does not result in samples as large as from generator-run traps, the advantage is that a larger number of identical traps can be deployed simultaneously, which is preferable when comparing relative abundances among sites. Their small attraction radius (typically only up to 10-15 m in both open and sheltered conditions; Slade & Merckx, pers. obs) results in a good reflection of local abundance, and avoids interference with other light traps, even if these were to be placed at a rather nearby distance. Light traps were placed during late afternoon and checked early next morning when we noted down the abundance of each macro-moth species. Prior to analyses, all species were classified in terms of conservation status and woodland affinity. We distinguished between (i) nationally scarce and Red Data Book (‘Scarce/RDB’) species28, (ii) severely declining common and widespread species (i.e. Rothamsted Research national abundance trends over 35 years >69% decline: UK Biodiversity Action Plan ‘Research Only’ Priority Species; Conrad et al. 2006) (‘Common Severely Declining’), (iii) declining common and widespread species (i.e. Rothamsted national 0-69% decline) (‘Common Declining’), (iv) increasing common and widespread species (i.e. Rothamsted national >0% increase) (‘Common Increasing’), (v) ‘Unknown Status’. With regard to woodland affinity we distinguished between (i) ‘ubiquitous’ species, (ii) ‘non-woodland’: species which main biotope is not woodland, (iii) ‘broad-woodland’: woodland species which main biotope is woodland but which also occur in other biotopes, (iv) ‘strict- woodland’: woodland species which only possible biotope is woodland28. GLM regression analyses were done to look at whether and to what extent the two woodland management practices (coppicing and ride management) differentially affect species richness and abundance of the species groups, and to test differences between coppiced woodland of the three different age-classes. The regression analyses were performed with log-transformed average values for abundance and species richness, across groups, calculated over the nine periods for each of the 36 sites, as same-site events are basically not statistically independent. Woodland patch size (log-transformed) was included as a continuous variable within all models. Model selection started with testing all three- and two-way interactions, and by performing backward selection of non- significant factors (and only in case of reasonably high p-values: > 0.38), to select final models. Since conservation status and woodland affinity variables were significantly 2 correlated (χ 12 = 44.53, p < 0.0001, N = 265), we did not include woodland affinity as a parameter within the models. Complementary to these analyses, we contrasted species-specific abundances between sheltered and more exposed sites (i.e. standard rides/standard woodland versus wide rides/coppice), corrected for the difference in sample size between sheltered and more exposed groups).

SID 5 (Rev. 07/10) Page 7 of 30

Details of key objectives and results / output:

(1) Quantify the presence/absence and abundance of a suite of 65 widespread but rapidly declining BAP Priority moth species between July and September in a woodland landscape in southern England.

All the fieldwork has been finalized successfully. As a result, we have presence/absence and abundance data of the focal 65 species for 36 sites (Fig. 1) within the Tytherley woodland landscape, collected during a total of 27 trapping nights. Each of the sites was sampled on nine different occasions. Out of a total of 324 single trap events, only two cannot be included into the analyses (one trap was vandalised, and another one was misplaced). The Tytherley woodland landscape (east of Salisbury, with Bentley Wood SSSI at its centre, and including several other woodlands) is ideally suited as the location for this study for three reasons: (a) it has an exceptionally diverse moth fauna (see below), partly because of its southern location, (b) Butterfly Conservation, one of the project’s sub contractors, has had a long-term involvement in the management of this area (as part of its South East England Woodlands Project), and so has extensive knowledge of which (parts of) woods within this landscape have been managed, how they are managed, and for how long they have been managed and (c) Butterfly Conservation has recently increased conservation management in this woodland landscape.

Throughout the whole fieldwork period (start July – mid October 2010) moths were sampled using Heath pattern actinic light traps (6W)29. Heath light traps operate on the ‘lobster-pot principle’, whereby moths are drawn to an actinic tube secured vertically between baffles, fall unharmed down a funnel, and rest on the inside of the trap or on pieces of egg-tray provided2. This type of light trap is ideal as they are light and battery- run. Although they do not result in catches as large as from traps powered by generators, their advantage is that a larger number of traps can be run simultaneously. Hence, rather than increasing the catch of a single trap, which is not essential for this project’s aims, we preferred to sample at an increased number of sites and hence better compare relative abundances among sites. Another advantage of these less bright traps is that they have a smaller attraction radius (typically only up to 10-15 m in both open and sheltered conditions; Slade & Merckx, pers. obs.), which results in a better reflection of the local abundances, rather than luring individuals from further away. The light trapping technique is proven to be highly suitable to sample moth communities as it is a passive sampling method that is standardized by using identical equipment (traps), enabling simultaneous sampling at several sites. Light traps were placed during late afternoon and checked early next morning when we noted down the abundance of each larger moth species. Moth sampling was carried out by two experienced persons (Martin Townsend and Dr. Thomas Merckx), which allowed to increase sampling effort and guaranteed the quality of obtained data.

SID 5 (Rev. 07/10) Page 8 of 30 Fig. 1: Location of 36 fixed trap sites (six from six experimental groups each) and sampled woodlands within the Tytherley woodland landscape

The complete dataset contains a total of 11,670 individuals from a total of 265 different larger moth species (including 4 species groups: i.e., Acronicta sp., Hoplodrina sp., Mesapamea sp., and Oligia sp.). From the study group of 65 widespread but rapidly declining species, 38 species (891 individuals) are represented (as remarked in the original project proposal, not all 65 species were expected to turn up, and the absence of certain species is informative as well). The species list also contains 12 nationally scarce species (one of which is a BAP Priority species, and another one is a former/downgraded BAP Priority species) (238 individuals), three Red Data Book species (two of which are also BAP Priority species) (9 individuals), and one species listed as extinct in the UK (3 individuals, and 4 additional individuals recorded via extra trapping effort outside the experimental set-up).

Our project initially reduced the group of 71 widespread but rapidly declining BAP Priority species to a focal group of 65 species due to the project’s seasonal and geographical restriction. However, as we eventually extended the light trapping into October, we managed to trap a further two such species, namely Figure of Eight and Sprawler. As remarked in the original project proposal, not all 65 species were expected to be present, and the absence of certain species is informative. The total number of trapped study group species amounts to 38 species (891 individuals), with a total of 29 focal species not trapped (Table 1, which contains the species´ scientific names). However, we did most likely trap both Grey dagger and Rustic, but as we opted not to carry out genitalia examination for some species groups (see proposal), we grouped these species together with Dark dagger and Uncertain, respectively, and treated them as non-BAP species within the analyses. The absence of the remaining 27 species can easily be explained for 19 of them: 7 species occur (mainly) on heaths and moors (Anomalous, Broom Moth, Broom-tip,

SID 5 (Rev. 07/10) Page 9 of 30 Dark Brocade, Heath Rustic, Neglected Rustic, Streak); 4 species are mainly coastal (Galium Carpet, Brindled Ochre, Mullein Wave, Rosy Minor), 4 species need damp areas (e.g. moors and marshes) (Crescent, Haworth’s Minor, Oblique Carpet, Shoulder-striped Wainscot); 2 species are dependent on currants and/or gooseberry (Spinach, V-moth); 1 species needs woods with Elm (Dusky-lemon Sallow); and 1 species needs open ground with Ragwort (Cinnabar). The reason behind the absence of the remaining 8 species is less obvious: although 4 species are not directly linked to woodland (Dark Spinach, Garden Dart, Large Nutmeg, Latticed Heath), two species should indeed partly occur in woodland (Double Dart and White-line Dart), and two are labelled as ubiquitous species (Dot Moth and Mottled Rustic). We believe their recent severe declines may well explain their absence, and we believe that some of them may have locally disappeared altogether. According to IUCN-categories, 4 of these species are indeed labelled as ‘Endangered’ (i.e. % decline over 35 years > 90%: Double Dart (97%), Garden Dart (97%), Dark Spinach (95%), White-line Dart (92%)), while three species display very strong declines too and almost made it into the ‘Endangered’ category (Dot Moth (88%), Large Nutmeg (88%), Latticed Heath (87%)). We have no clear explanation for the absence of Mottled Rustic, which could be either due to its rather high rate of decline as well (73%) or due to chance, as it is a ubiquitous species that is generally common to abundant.

Table 1: List of the project’s focal species, listed as widespread but rapidly declining BAP Priority species, in alphabetical order, with observed abundance (N).

common name genus species name N Anomalous Stilbia anomala 0 August Thorn Ennomos quercinaria 1 Autumnal Rustic Eugnorisma glareosa 4 Beaded Chestnut Agrochola lychnidis 3 Blood-vein Timandra comae 16 Brindled Ochre Dasypolia templi 0 Broom Moth Melanchra pisi 0 Broom-tip Chesias rufata 0 Brown-spot Pinion Agrochola litura 6 Buff Ermine Spilosoma luteum 17 Centre-barred Sallow Atethmia centrago 5 Cinnabar Tyria jacobaeae 0 Crescent Celaena leucostigma 0 Dark Brocade Blepharita adusta 0 Dark Spinach Pelurga comitata 0 Dark-barred Twin-spot Carpet Xanthorhoe ferrugata 6 Deep-brown Dart Aporophyla lutulenta 1 Dot Moth Melanchra persicariae 0 Double Dart Graphiphora augur 0 Dusky Brocade Apamea remissa 2 Dusky Thorn Ennomos fuscantaria 4 Dusky-lemon Sallow Xanthia gilvago 0 Ear Moth Amphipoea oculea 2 Feathered Gothic Tholera decimalis 25 Figure of Eight Diloba caeruleocephala 7 Flounced Chestnut Agrochola helvola 1 Galium Carpet Epirrhoe galiata 0 Garden Dart Euxoa nigricans 0 Garden Tiger Arctia caja 3 Ghost Moth Hepialus humuli 2 Grass Rivulet Perizoma albulata 1 Green-brindled Crescent Allophyes oxyacanthae 130 Grey Dagger Acronicta psi - Haworth’s Minor Celaena haworthii 0 Heath Rustic Xestia agathina 0 Hedge Rustic Tholera cespitis 2

SID 5 (Rev. 07/10) Page 10 of 30 Knot Grass Acronicta rumicis 4 Lackey Malacosoma neustria 1 Large Nutmeg Apamea anceps 0 Large Wainscot Rhizedra lutosa 1 Latticed Heath Chiasmia clathrata 0 Minor Shoulder-knot Brachylomia viminalis 267 Mottled Rustic Caradrina morpheus 0 Mouse Moth Amphipyra tragopoginis 6 Mullein Wave Scopula marginepunctata 0 Neglected Rustic Xestia castanea 0 Oak Hook-tip Watsonalla binaria 9 Oak Lutestring Cymatophorima diluta 62 Oblique Carpet Orthonama vittata 0 Pale Eggar Trichiura crataegi 98 Pretty Chalk Carpet Melanthia procellata 38 Rosy Minor Mesoligia literosa 0 Rosy Rustic Hydraecia micacea 10 Rustic Hoplodrina blanda - Sallow Xanthia icteritia 47 September Thorn Ennomos erosaria 6 Shaded Broad-bar Scotopteryx chenopodiata 22 Shoulder-striped Wainscot Mythimna comma 0 Small Emerald Hemistola chrysoprasaria 2 Small Phoenix Ecliptopera silaceata 74 Small Square-spot Diarsia rubi 2 Spinach Eulithis mellinata 0 Sprawler Asteroscopus sphinx 1 Streak Chesias legatella 0 V-moth Macaria wauaria 0 White Ermine Spilosoma lubricipeda 3 White-line Dart Euxoa tritici 0

Our species list also contains 11 nationally scarce species (one of which is a BAP Priority species, and another one a former/downgraded BAP Priority species) (237 individuals), 3 Red Data Book species (two of which are also BAP Priority species) (9 individuals), and 1 species listed as extinct in the UK (3 individuals) (Table 2). These 15 species are grouped into one conservation status class ‘scarce and RDB species’ within the analyses (249 individuals).

Table 2: Overview of abundance (N) of trapped Nationally Scarce and RDB species common name genus species conservation status N Festoon Apoda limacodes Nationally Scarce 6 Great Oak Beauty Hypomecis roboraria Nationally Scarce 46 Kent Black Arches Meganola albula Nationally Scarce 3 Mere Wainscot Chortodes fluxa Nationally Scarce 48 Mocha Cyclophora annularia Nationally Scarce 94 Small Black Arches Meganola strigula Nationally Scarce 2 White-line Snout Schrankia taenialis Nationally Scarce 7 Lunar Yellow Underwing Noctua orbona Nationally Scarce / BAP 1 Waved Black Parascotia fuliginaria Nationally Scarce / Declining 8 Double Line Mythimna turca Nationally Scarce / former BAP 1 Devon Carpet Lampropteryx otregiata Nationally Scarce / Increasing 21 Dark Crimson Underwing Catocala sponsa RDB2 (Vulnerable) / BAP 1 Triangle Heterogenea asella RDB3 (Rare) 2 Light Crimson Underwing Catocala promissa RDB3 (Rare) / BAP 6 Clifden Nonpareil Catocala fraxini Former extinct UK resident / Nationally 3 Scarce (conservation statuses according to Waring & Townsend 2009)

SID 5 (Rev. 07/10) Page 11 of 30

Our species list contains a further total of 90 species that are nationally in decline (3564 individuals), 73 species that are nationally on the increase (5986 individuals), and 49 species for which the national trend is unknown (980 individuals) (trends based on Rothamsted national abundance data over 35 years). The total list amounts to 265 species.

(2) Investigate the impact of contrasting woodland management practices on overall species richness and abundance of individual species from the study group of widespread but rapidly declining species. Specifically, study the effect of two woodland conservation practices by comparing (i) coppiced woodland with standard high forest and (ii) extended- width rides with standard rides.

AND

(3) Compare the extent of differences in presence/absence and abundance of widespread BAP moth species between three age classes of coppiced woodland: young (<3yrs), medium (3-6 yrs), and old (>6-9yrs).

By sampling intensively at 36 fixed sites (see below for the experimental design), characterized by different management practices and histories, we were able to test the impact of coppicing management (including length of coppice rotation: three coppice age classes covering the whole coppice rotation length) compared to current standard high forestry practices, and the impact of extended-width ride management compared to standard rides.

Experimental design:

Six Experimental Groups:

1. Standard woodland: standard high forestry practices 2. Coppiced woodland: age class young (<3 yrs) 3. Coppiced woodland: age class medium (3-6 yrs) 4. Coppiced woodland: age class old (>6-9 yrs) 5. Standard woodland rides: narrow (i.e. <10 m, mostly canopy-covered) rides within ‘standard woodland’ 6. Wide woodland rides: wide (i.e. >20 m) within ‘standard woodland’

Each one of these six experimental groups is sufficiently replicated as it is represented by a total of six light trap sites each. The occurrence of a limited number of small inter-trap distances is a result of restrictions imposed by the availability of treatments within the study landscape and by accessibility. Nevertheless, inter-trap distances were invariably larger than 50 m, which still avoids possible light trap interference as a result of the characteristically small attraction radius of the light traps used (typically only up to 10-15 m; see above). Moreover, inter-trap distances between traps of the same experimental group were invariably larger than 100 m in order to avoid spatial auto-correlation. Furthermore, we located sites from the same experimental group as much as possible in different woods. These measures favour the statistical independence of each of the six replicates for each treatment. Note also that nearby trap sites were not necessarily being sampled during the same nights, and that the large majority of inter-trap distances were considerably larger than these minimum values (Fig. 1). The sample size of 36 trap sites is based on a power analysis (using the GPOWER software) exploring plausible effect sizes, and taking into account the logistics of sampling. These 36 sites were sampled during each of nine consecutive 10-day periods between the start of July and mid October. During each 10-day period three trap night efforts were deployed, which resulted in a total of 27 trap nights over the whole fieldwork period. In each night of trapping 12 sites were sampled.

SID 5 (Rev. 07/10) Page 12 of 30 Given the successful data collection at all experimental groups (i.e., several woodland management classes) we were able to statistically analyse the recorded dataset in depth. These analyses provide a powerful insight into the efficacy (extent of effect, species affected) of popular woodland management practices for widespread but rapidly declining BAP Priority moth species and other moths.

GLM regression analyses were done to look at whether and to what extent the two woodland management practices (coppicing and ride management) increase species richness and abundance of the focal BAP and other species, and to test differences between coppiced woodland of three different age-classes (objective 3). We did not control for weather-related variables as we made sure to sample a similar number of all experimental groups during each and every trap night, and as we only sampled during suitable weather conditions too.

Explanatory Variables: • Treatment / Experimental group (6 classes): young/medium/old coppice / narrow/wide ride / standard woodland • Conservation status (5 classes): based on Rothamsted Research national trend data: (i) Study group of widespread but rapidly declining BAP Priority species (ii) Declining species (iii) Increasing species (iv) RDB / scarce species (according to Butterfly Conservation and Waring & Townsend 2009) (v) Unknown status • Woodland affinity (4 classes): based on Waring & Townsend (2003) (i) ‘Narrow’ woodland species: woodland is their only biotope (ii) ‘Wide’ woodland species: species use mainly woodland, but also other biotopes (iii) ‘Other’: species’ main biotope is not woodland (iv) ‘Ubiquitous’: species with a ubiquitous distribution

Since the Conservation status and Woodland affinity variables were correlated, they could not be tested within the same model, and separate tests were required to look into the effects of these variables (see below).

Analyses: GLM regression analyses were done to look at whether and to what extent the two woodland management practices (coppicing and ride management) differentially affect species richness and abundance of the species groups, and to test differences between coppiced woodland of the three different age-classes. The GLM regression analysis was selected and preferred over other types of analytical techniques (such as Principal Component Analysis) because, from the onset, the experimental design was set up to be analysed via GLM: we restricted the amount of parameters to be tested to those we were a priori interested in (conservation/management-related) and controlled for within the experimental design. Here, we believe this to be a more sound approach than testing a multitude of parameters and grouping these into new, but often less specific variables based on the power with which they explain the data variance. The regression analyses were performed with log-transformed average values for abundance and species richness, across groups, calculated over the nine periods for each of the 36 sites, as same-site events are basically not statistically independent. Woodland patch size (log-transformed) was included as a continuous variable within all models. Model selection started with testing all three- and two-way interactions, and by performing backward selection of non- significant factors (and only in case of reasonably high p-values: > 0.38), to select final models. Since conservation status and woodland affinity variables were significantly 2 correlated (χ 12 = 44.53, p < 0.0001, N = 265), these parameters could not be included within the same models.

SID 5 (Rev. 07/10) Page 13 of 30 Complementary, we contrasted species-specific abundances between sheltered and more exposed sites (i.e. standard rides/standard woodland versus wide rides/coppice), corrected for the difference in sample size between sheltered and more exposed groups.

Results overall analysis: The landscape-scale light-trap experiment recorded a total of 11,670 individuals from 265 macro-moth species (including 4 species couples: i.e., Acronicta, Hoplodrina, Mesapamea and Oligia sp.): 15 ‘Scarce/RDB’ species (249 individuals) (Table 2), 38 ‘Common Severely Declining’ species (891 individuals) (Table 1), 90 ‘Common Declining’ species (3564 individuals), 73 ‘Common Increasing’ species (5986 individuals), and a further 49 ‘Unknown Status’ species (980 individuals). A key finding was that overall macro-moth abundance was lower in coppice and wide rides than in standard rides and standard woodland (treatment: F5, 24 = 8.70; p < 0.0001) (Fig. 2). Although differences were not significant, average levels of overall abundance increased from young, over medium, to old coppice and wide rides. Overall species richness was lowest in coppice and highest at standard and wide rides and standard woodland (treatment: F5, 24 = 7.39; p = 0.0003) (Fig. 2).

Central areas of coppiced plots are particularly exposed, whereas some shelter is provided by the adjacent woodland edges for wide rides, and both standard ride and within- woodland environments characterised by high levels of shelter. Furthermore, especially young coppice, but also to some extent medium and old coppice are characterised by plenty of bare ground which, when not irradiated by the sun, lowers the temperature of the air layer directly above it. It has been shown that such factors impact on levels of convective cooling for day-flying endothermic butterflies31, and so it is likely they impact on activity levels and occurrence of these night-flying endothermic Lepidoptera too. This overall picture contrasts strongly with the one for day-flying Lepidoptera (i.e. butterflies and day-flying moths), where open woodland conditions generally mean higher abundance and species richness, due to higher day temperatures, higher solar irradiance, and other key adult and larval resources12,13.

Fig. 2: Overall macro-moth abundance (number of individuals) and species richness (number of species) (log10-transformed mean(SE)) for six experimental woodland management treatments, ordered according increasing shelter conditions.

SID 5 (Rev. 07/10) Page 14 of 30 While these overall pictures apply to all macro-moth species groups, irrespective of their scarcity or national trend status, the ‘Common Severely Declining’ species showed the smallest difference in abundance between the woodland management treatments. Wide woodland rides in particular were characterised by relatively high abundance of these rapidly declining species (treatment x status: F20, 96 = 2.05; p = 0.011) (Fig. 3).

Fig. 3: Abundance (log10-transformed mean(SE)) of two high profile conservation status species groups for six experimental woodland management treatments, ordered according increasing shelter conditions.

Another key finding was that the total area of the woodland patch surrounding young/medium/old coppiced plots had a strong positive effect on the abundance of ‘Common Severely Declining’ species (treatment x status x area: F20, 96 = 2.21; p = 0.0056). Surrounding woodland area also significantly increased overall macro-moth species richness at medium/old coppice (treatment x area: F5, 24 = 5.70; p = 0.0013), particularly through its effect on the group of ‘Common Declining’ species (status x area: F4, 136 = 3.63; p = 0.0077).

Species-specific analyses

The multi-species analyses described above generally show a lower overall abundance and species richness in coppiced areas and wide rides (although this was less pronounced for the study group). However, such exposed sites may have a different species composition altogether. If this were the case, species from coppiced plots and wide rides would add to the typical woodland species composition found in more sheltered sites within the same woodland patch. Adding species that preferentially occur in wide rides and coppiced sites may increase the total species richness of woodlands (at the woodland scale that is). Also, some of these ‘extra’ species may be species of conservation concern, which find it increasingly difficult to survive in human-dominated landscapes due to agricultural intensification and other pressures. An ‘opening up’ - via coppicing and ride widening - of woodland borders may allow such species to enter woodland complexes and encounter their specific set of habitat resources within the woodland biotope. Hence, woodland border conservation management (coppicing, ride widening) may become increasingly important for species of conservation concern that are not typically associated with woodland (high forest).

SID 5 (Rev. 07/10) Page 15 of 30 In order to get a better insight in such a process we performed additional species-specific analyses, complimentary to the multi-species analyses described above. These basically consist of an overview of the species numbers and percentages with regard to uniqueness to (i.e. strictly restricted to) and preference for (i.e. numbers higher at) ‘open’ sites (i.e. coppiced sites and wide rides) vs. sheltered sites (i.e. standard woodland and standard rides) (corrected for the difference in sample size between non-sheltered and sheltered groups) (Table 4).

Table 4: Overview of species unique to or preferring sheltered (i.e. standard woodland and standard rides) versus more open sites (i.e. coppiced sites and wide rides).

Species numbers

Status Stotal OnlyOpen PrefOpen No Pref PrefShelt OnlyShelt Study group 38 10 7 5 12 4 RDBscarce 15 5 0 1 8 1 Declining 90 14 23 5 44 4 Increasing 73 6 20 2 36 9 Unknown 49 14 10 2 19 4

Percentages Status OnlyOpen PrefOpen No Pref PrefShelt OnlyShelt Study group 26 18 13 32 11 RDBscarce 33 0 7 53 7 Declining 16 26 6 49 4 Increasing 8 27 3 49 12 Unknown 29 20 4 39 8

Hence, overall, a total of 124 species occurs in equal or higher numbers in open sites (i.e. coppiced sites and wide rides) than in sheltered sites. Out of these, 109 species have a preference for these open sites. A total of 49 species did even occur at these open sites only, and never at sheltered sites.

This pattern is strongest for the study group (i.e. 58%) (and is also strong for the ‘Unknown’ group: 52%). The study group differs significantly with the ‘Increasing’ species group (i.e. only 38% of species which occur in equal or higher numbers at open sites) in terms of number of species equal or more abundant in coppiced and wide rides versus sheltered 2 sites (χ 1 = 3.85; P = 0.05).

Note that the apparent contradiction between these species-specific results (lots of species with a preference for open sites) and the results from the overall analysis (overall preference for sheltered sites) is not a real contradiction. It is explained by the much larger weight of the abundant/common species within the overall analysis, where the species level is not taken into account, but only overall abundance of all moth species lumped together. The additional species-specific analyses are hence valuable as they give a complementary picture to the results from the overall analyses and are essential to understand the full picture of the results.

When focusing on the study group alone, we observe overall that there is quite a spread over the three coppice age classes, though old coppice has the lowest number of species: The 10 species restricted to open sites do so mostly at young coppice. The 7 species with a preference for open sites occur basically evenly at all coppice age classes. The 5 species that occurred equally in open and sheltered sites occur in all coppice age classes. The 12 species with a preference for sheltered sites occur mostly at medium coppice (only 2 species occurred mostly at old coppice). Overall, when summing abundances, we get a different picture, skewed by abundant species: 82 individuals in young coppice, 131 individuals in medium coppice and 145 individuals in old coppice. We hence conclude that the whole coppice cycle is of value.

SID 5 (Rev. 07/10) Page 16 of 30

Summary of results of species-specific analyses: Species-specifically, 124 species (i.e. 47% of the total) occurred in lower numbers at sheltered than at more open sites (standard rides/woodland versus coppice/wide rides). Of these 124 species, 49 (i.e. 18% of the total) were in fact restricted to these open areas. This pattern was strongest for the ‘Common Severely Declining’ species, with 22 of the 38 recorded species (i.e. 58%) occurring in equal or higher numbers at managed, open sites compared to unmanaged, sheltered sites. In this respect, they differ with the ‘Common Increasing’ and ‘Scarce/RDB’ species groups where only 38% (N = 28) and 40% (N = 6) of 2 species, respectively, occur in equal or higher numbers at open sites (χ 1 = 3.85, p = 0.05; trend only; respectively). Moreover, and although overall abundance within the ‘Common Severely Declining’ species group increased from young (N = 82), over medium (N = 131), to old coppice (N = 145), species richness decreased (young: S = 22; medium: S = 18; old: S = 15).

(4) Disseminate the findings via clear management guidance for local conservation practitioners and private woodland owners, via articles targeted at local and national governmental and non-governmental organizations, and via peer-reviewed papers to the scientific community. The results of this project will help deliver conservation targets for this group of widespread but rapidly declining BAP moth species, in particular, by informing the design of conservation management within English woodlands.

Our research findings directly contribute to the conservation of this group of widespread but rapidly declining BAP Priority moth species, because they will inform the design of optimum conservation management within English woodlands for these BAP species. This is achieved through the production and dissemination of management guidance for local conservation practitioners and private woodland owners, as well as through the wide dissemination of our findings within the scientific conservation community, and both local and national governmental bodies. We hope that, for example, our findings may enhance the Forestry Commission's Woodland Improvement Grants, and Natural England's HLS.

We ensure that the financial support of DEFRA is acknowledged in any of the following publications and information for wide dissemination.

Three key audiences for the project information were identified: (a) conservation planners, managers (statutory agencies, national database managers, BAP groups) and practitioners (e.g. woodland site managers); (b) scientific community; and (c) public. Butterfly Conservation is the Lead Partner for 81 specialist and 71 widespread but rapidly declining BAP Priority moth species. It already has in place an effective network for delivering results and information to statutory agencies, NGOs and academics. Results are incorporated into the National Biodiversity Network.

Both WildCRU and Butterfly Conservation have strong track records of disseminating research results through the scientific literature, publishing ca. 60 peer-reviewed papers per year. The results of this project will be reported in a high quality peer-reviewed journal too (see below). Butterfly Conservation is a UK leader in communication, education and public awareness of conservation messages. It runs over 700 training events a year aimed at both the general public and more specialist audiences, and has over 10,000 volunteers undertaking surveys, monitoring and habitat management on a wide range of sites. Butterfly Conservation has more than 30 reserves, and 14,000 members receive the magazine ‘Butterfly’ which is published three times a year and distributed to members, institutions, conservation bodies and others interested or involved in the conservation of butterflies, moths and related wildlife (estimated readership: 35,000) and the website receives over 317,000 hits per year. Raising awareness of the conservation of moths and their habitats is a key priority for Butterfly Conservation, and this is done through press releases, website, articles in its member’s magazine, and National Moth Nights. These

SID 5 (Rev. 07/10) Page 17 of 30 avenues will be used for some of the project’s communication of results, using BAP moths as a flagship group for issues surrounding the management of woodland.

Information for wide dissemination - realised publications:

o Merckx, T. (2011) Coppicing and woodland ride management effects on UK BAP moths. In: Noake, B., Rosenthal, A., Parsons, M. & Bourn, N. (eds.) Lepidoptera Conservation Bulletin Number 11: April 2010 – March 2011, Butterfly Conservation, Wareham, pp. 47-48.

This publication gives a short overview of the project and describes the main findings and implications of the study. For the full version, see (Annex 1).

Since publication (March 2011), it is online available via the following web link:

http://www.butterfly- conservation.org/uploads/Lepidoptera%20Conservation%20Bulletin%20No.%2011%20%2 8April%202010-March%202011%29.pdf

The Lepidoptera Conservation Bulletin is an annual publication and summary produced by Butterfly Conservation that reflects some of the diverse work carried out towards the conservation of moths and butterflies in the UK each year. It also summarises the resulting advances in our understanding of the BAP Priority species, both butterflies and moths, including information on distribution, habitat requirements and management.

o Merckx, T. (2011) Woodland management and rapidly-declining moths. In: Moths Count Newsletter 2011, Butterfly Conservation, Wareham, pp. 12.

This publication gives a short overview of the project and describes the main findings and implications of the study. For the full version, see (Annex 2).

Since publication (July 2011), it is online available via the following web link:

http://www.mothscount.org/uploads/Moths%20Count%20newsletter%205%202011%20%2 8book-style%29.pdf

Butterfly Conservation´s Moths Count project aims to encourage interest in moths throughout the UK and to run the National Moth Recording Scheme to improve knowledge and conservation of the 900+ species of larger moths.

o Hoare, D., Kelly, C. & McLellan, L. (2011) South East Woodland Project Report 2007- 2010. Butterfly Conservation Report, Butterfly Conservation, Wareham, UK (in prep).

This publication is the final report of the larger South East Woodland Project from Butterfly Conservation (deadline 31/03/2011). It puts our project within this larger framework and gives an overview of our project together with its main findings and implications for general consumption. The South East Woodlands Project draws to a close in April 2011 with the publication of a series of national resources promoting best practice woodland management for butterflies and moths. The project had a particular focus on three extensive demonstration areas, one of which was the area where we carried out this project, i.e. the Tytherley woodland landscape.

o Kelly, C. & Hoare, D. (2011) Tytherley Woods Project, Exit Report 2011. Butterfly Conservation Report, Butterfly Conservation, Wareham, UK (in prep).

This other Butterfly Conservation publication will shortly be sent to all local woodland owners in the Tytherley woodland landscape (i.e. this project’s study landscape). It also

SID 5 (Rev. 07/10) Page 18 of 30 refers to our project by means of a basic summary of results and a list of BAP Priority species found. The subset of local landowners who allowed us to carry out our research on their land have already received a list of species found on their land, with some explanation on the project, and they too will be informed about the project’s results (via the publication above) and be given conservation advice and management guidance in person in line with these results during early 2011.

Information for wide dissemination – production in preparation:

o In order to communicate the project's results to the wider scientific community, we have prepared an article to be submitted early November to the high-impact peer- reviewed conservation journal Conservation Biology.

o We are preparing an article on the project and its main findings for the ‘Conservation News’ section of Butterfly, which is Butterfly Conservation’s member magazine with readership of approximately 35,000. This will allow the wider public and practitioners to understand why specific woodland management is needed to not only increase populations of woodland butterflies and other wildlife, but also of moth BAP Priority species.

o Dr. Nigel Bourn represents Butterfly Conservation, one of the project’s sub- contractors, on the Woodland Biodiversity Integration Group (BIG), which is supported by DEFRA and Natural England, and is part of the UK BAP process. This facilitates rapid and effective dissemination of results directly to both commercial and conservation sectors, e.g. Confederation of Forest Industry (CONFOR), CLA, RSPB. Dr. Nigel Bourn is planning to give a talk on the project and its main findings and implications at the next meeting of the Woodland BIG.

o Dr. Thomas Merckx is planning to give talks on the research findings during spring 2012 at (i) the 3rd international symposium Future of Butterflies in Europe (Wageningen, The Netherlands), (ii) the Centre of Environmental Biology, University of Lisbon, Portugal, and (iii) Ecology and Biodiversity Research Group, Univeristy of Louvain-la-Neuve, Belgium.

o We are making sure that the key findings are communicated to the relevant government bodies and NGOs via the contacts both WildCRU and BC have developed over many years. Butterfly Conservation and the WildCRU are producing a 2-4 page illustrated summary of the main implications of the findings. This will be specifically addressed to the user community to support conservation action (non- technical audience), and will take shape as a separate Policy Advice Note. Also, the results of the project are summarised within this final report to DEFRA.

In line with the missions and current practice of both WildCRU and Butterfly Conservation, effective and wide dissemination of the project's findings are a top priority. It is our main aim to see our findings translated into woodland habitat improvements on the ground. Thus, in addition to the above listed ways of communicating our results to relevant audiences, we will integrate our findings in Butterfly Conservation's existing and future landscape-scale woodland restoration projects, and give direct advice to governmental bodies closely involved with woodland management. We will also highlight the potential role of both Woodland Improvement Grants and Higher Level Stewardship in the future conservation of UK BAP moths. Butterfly Conservation has eight people working on the South East England Woodlands Project and therefore able to immediately transfer knowledge gained from this project into conservation practice.

Main implications:

SID 5 (Rev. 07/10) Page 19 of 30 The species-specific analyses show that coppicing and ride widening are valuable conservation tools for a significant number of larger moth species, especially benefiting the study group species of widespread but rapidly declining species, which are not necessarily the most abundant species though. We suggest the mechanism behind the resulting pattern of increased total woodland species richness involves an increased structural and hence micro-climatic diversity of woods and a provision of additional resources for species with an affinity for more open biotopes. We also show that the greatest benefits to increased coppicing and ride widening are likely to be had within larger woods and mainly so for nationally declining and rapidly declining species. The complementary analyses on overall abundance and species richness show that the sheltered, dark, humid, late-successional high forest biotope is characterised by high numbers of both individuals and species, and is at the same time especially important for some scarce and RDB species. These complementary findings demonstrate the value of this woodland biotope and that this successional stage clearly needs to be accounted for within woodland management.

Based on the results of these complementary analyses, our main recommendations are (i) to enlarge existing high forest and create new ones, in order to have a sufficient extent of cores of typical dark woodland biotope, and additionally (ii) to ‘buffer’ these dark cores from the open ‘matrix’ by creating, at their edges, lighter woodland zones which can indeed be achieved by implementing coppicing or wide woodland rides. Such zoning will safeguard the dark environment needed by shade- or moisture-loving woodland specialists, whilst at the same time allowing easy access for species of mixed and more open biotopes to a light and open woodland biotope. We show that a conservation focus on either coppicing/ride widening or sheltered woodland will not deliver as much biodiversity value as the combined implementation of both approaches at once: this two- tier approach may well be vital both for endangered woodland specialists and declining, once-widespread species that are finding it ever more difficult to locate enough resources within current degraded landscapes.

By passing all records to the National Moth Recording Scheme of Butterfly Conservation’s Moths Count Project, and hence to the NBN-gateway, all records are available for future research and conservation planning, and we have helped to improve the knowledge on the status (temporal trends in abundance and distribution) of larger moths, which is especially welcome for species of conservation concern, as it helps to better understand the causes of their national declines.

The project has disseminated and will disseminate the findings as clear guidance on woodland management to conservation practitioners, woodland owners, relevant (governmental) bodies, and the wider scientific community. Specifically, by having informed the design of conservation management within English woodlands, we believe to have facilitated the translation of our findings into practical conservation, especially benefiting the study group.

In conclusion, one of the study’s main benefits is that it has contributed substantially to filling the clear knowledge gap which has been identified by the UK BAP working groups as a barrier to the delivery of conservation targets for the group of widespread but rapidly declining BAP Priority moth species. The significantly increased autecological understanding of this under-studied group and the elucidated efficiency of specific woodland conservation management practices are major outputs of this project and will directly contribute to the conservation and population increase of these widespread, but nationally rapidly declining species within English woodlands, by informing the design of optimum conservation management within English woodlands for these and other BAP Priority species. As such, our project may help the Woodland Biodiversity Integration Groups (established to bring together woodland biotope and associated species interests at the national level) on how best to integrate BAP Priority species requirements into woodland biotope management. Such a biotope-based delivery of UK biodiversity targets is in line with the recent approach to BAP delivery across the UK (see Natural England Research Report NERR02430), which resulted from the large increase in the number of

SID 5 (Rev. 07/10) Page 20 of 30 species and habitats requiring conservation action within the 2007 review of the UK BAP Priority species list.

Possible future work:

Now that this project has developed a standard set of sampling points which allow assessment of the importance of a range of management features, it would be interesting to see whether the results hold true by sampling an additional year (using the same methodology), and especially so by sampling during spring (i.e. generally colder conditions), as sampling in the current project only started in July which meant that species with an earlier flight season were excluded. The colder spring conditions may also result in different effects of the management treatments on moths compared to summer conditions. Additionally, a follow-up study would also allow more data to be obtained on habitat preferences of the re-established Clifden Nonpareil and the other 14 recorded nationally scarce/Red Data Book species.

SID 5 (Rev. 07/10) Page 21 of 30 Bibliography: (With names of people closely involved with the project in bold)

1. Conrad KF, Warren MS, Fox R, Parsons MS, Woiwod IP (2006) Rapid declines of common, widespread British moths provide evidence of an insect biodiversity crisis. Biol. Cons. 132: 279-291. 2. Merckx T, Feber RE, Riordan P, Townsend MC, Bourn NAD, Parsons MS, Macdonald DW (2009a) Optimizing the biodiversity gain from agri-environment schemes. Agric. Ecosyst. Envir. 130: 177-182. 3. Merckx T, Feber RE, Dulieu RL, Townsend MC, Parsons MS, Bourn NAD, Riordan P, Macdonald DW (2009b) Effect of field margins on moths depends on species mobility: field-based evidence for landscape-scale conservation. Agric. Ecosyst. Envir. 129: 302- 309. 4. Merckx T, Feber RE, Mclaughlan C, Bourn NAD, Parsons MS, Townsend MC, Riordan P, Macdonald DW (2010a) Shelter benefits less mobile moth species: the field-scale effect of hedgerow trees. Agric. Ecosyst. Envir. 138: 147-151. 5. Merckx T, Feber RE, Parsons MS, Bourn NAD, Townsend MC, Riordan P, Macdonald DW (2010b) Habitat preference and mobility of Polia bombycina: are non-tailored agri- environment schemes any good for a rare and localised species? J. Ins. Cons. 14: 499- 510. 6. van Swaay C, Warren MS, Lois G (2006) Biotope use and trends of European butterflies. J. Ins. Cons. 10: 189-209. 7. Gorissen D, Merckx T, Vercoutere B, Maes D (2004) Changed woodland use and butterflies. Why did butterflies disappear from woodlands in Flanders? J. Land. Ecol. Environ. Sci. Fland. Neth. 21: 85-95. 8. Sparks TH, Greatorex-Davies JN, Mountford JO, Hall ML, Marrs RH (1996) The effects of shade on the plant communities of rides in plantation woodland and implications for butterfly conservation. Forest Ecol. Manag. 80: 197-207. 9. Van Dyck H, Van Strien AJ, Maes D, van Swaay CAM (2009) Declines in common, widespread butterflies in a Landscape under intense human use. Cons. Biol. 23: 957-965. 10. Warren MS (1987) The ecology and conservation of the heath fritillary butterfly Mellicta athalia. 3. Population dynamics and the effect of habitat management. J. Appl. Ecol. 24: 499-513. 11. Warren MS (1991) The successful conservation of an endangered species, the heath fritillary butterfly Mellicta athalia, in Britian. Biol. Cons. 55: 37-56. 12. Warren MS (1993) A review of butterfly conservation in central southern Britain. 2. Site management and habitat selection of key species. Biol. Cons. 64: 37-49. 13. Warren MS, Thomas JA (1992) Butterfly responses to coppicing. In: GP Buckley (ed.) The Ecological Effects of Coppice Management. London, Chapman & Hall, pp. 249-270. 14. Young MR (1992) Conserving insect communities in mixed woodlands. In: MGR Cannel, DC Malcolm, PA Robertson (eds.) The ecology of mixed-species stands of trees. London, Blackwell Scientific Publications, Special Publication Series of the British Ecological Society 11, pp. 277-296. 15. Smallidge PJ, Leopold DJ (1997) Vegetation management for the maintenance and conservation of butterfly habitats in temperate human-dominated landscapes. Land. Urb. Plan. 38: 259-280. 16. Barkham JP (1992) The effects of coppicing and neglect on the performance of the perennial ground flora. In: GP Buckley (ed.) Ecology and Management of Coppice Woodlands. London, Chapman & Hall, pp. 115-146. 17. Greatorex-Davies JN, Sparks TH, Hall ML, Marrs RH (1993). The influence of shade on butterflies in rides of coniferised lowland woods in southern England and implications for conservation management. Biol. Cons. 63: 31-41. 18. Feber RE, Brereton TM, Warren MS, Oates M (2001) The impacts of deer on woodland butterflies: the good, the bad and the complex. Forestry 74: 271-276. 19. Benes J, Cizek O, Dovala J, Konvicka M (2006) Intensive game keeping, coppicing and butterflies: The story of Milovicky Wood, Czech Republic. Forest Ecol. Manag. 237: 353- 365.

SID 5 (Rev. 07/10) Page 22 of 30 20. Spitzer L, Konvicka M et al. (2008) Does closure of traditionally managed open woodlands threaten epigeic invertebrates? Effects of coppicing and high deer densities. Biol. Cons. 141: 827-837. 21. Konvicka M, Novak J et al. (2008) The last population of the Woodland Brown butterfly (Lopinga achine) in the Czech Republic: habitat use, demography and site management. J. Ins. Cons. 12: 549-560. 22. Hodgson JA, Moilanen A, Bourn NAD, Bulman CR, Thomas CD (2009) Managing successional species: Modelling the dependence of heath fritillary populations on the spatial distribution of woodland management. Biol. Cons. 142: 2743-2751. 23. Broome A, Clarke S, Peace A, Parsons M (2011) The effect of coppice management on moth assemblages in an English woodland. Biodiv. Cons. 20: 729-749 24. Fiedler K, Schulze CH (2004) Forest modification affects diversity (but not dynamics) of speciose tropical pyraloid moth communities. Biotropica 36: 615-627. 25. Summerville KS, Crist TO (2008) Structure and conservation of lepidopteran communities in managed forests of northeastern North America: a review. Canad. Entomol. 140: 475- 494. 26. Usher MB, Keiller SWJ (1998) The macrolepidoptera of farm woodlands: determinants of diversity and community structure. Biodiv. Cons. 7: 725-748. 27. Summerville KS, Crist TO (2004) Contrasting effects of habitat quantity and quality on moth communities in fragmented landscapes. Ecography 27: 3-12. 28. Waring P, Townsend M (2009) Field Guide to the Moths of Great Britain and Ireland. 2nd edn. British Wildlife Publishing. 29. Heath J (1965) A genuinely portable MV light trap. Entomol. Rec. J. Var. 77: 236-238. 30. Webb JR, Drewitt AL, Measures GH (2010) Managing for species: Integrating the needs of England’s priority species into habitat management. Part 1 Report. Natural England Research Reports, Number 024. 31. Merckx T, Van Dongen S, Matthysen E, Van Dyck H (2008) Thermal flight budget of a woodland butterfly in woodland versus agricultural landscapes: An experimental assessment. Bas. Appl. Ecol. 9: 433-442.

SID 5 (Rev. 07/10) Page 23 of 30

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.

o Merckx, T. (2011) Coppicing and woodland ride management effects on UK BAP moths. In: Noake, B., Rosenthal, A., Parsons, M. & Bourn, N. (eds.) Lepidoptera Conservation Bulletin Number 11: April 2010 – March 2011, Butterfly Conservation, Wareham, pp. 47-48.

This publication gives a short overview of the project and describes the main findings and implications of the study. For the full version, see Annex 1.

Since publication (March 2011), it is online, available via the following web link: http://www.butterfly- conservation.org/uploads/Lepidoptera%20Conservation%20Bulletin%20No.%2011%20%28 April%202010-March%202011%29.pdf

o Merckx, T. (2011) Woodland management and rapidly-declining moths. In: Moths Count Newsletter 2011, Butterfly Conservation, Wareham, pp. 12.

This publication gives a short overview of the project and describes the main findings and implications of the study. For the full version, see Annex 2.

Since publication (July 2011), it is online available via the following web link: http://www.mothscount.org/uploads/Moths%20Count%20newsletter%205%202011%20%28 book-style%29.pdf

o Hoare, D., Kelly, C. & McLellan, L. (2011) South East Woodland Project Report 2007- 2010. Butterfly Conservation Report, Butterfly Conservation, Wareham, UK (in prep).

o Kelly, C. & Hoare, D. (2011) Tytherley Woods Project, Exit Report 2011. Butterfly Conservation Report, Butterfly Conservation, Wareham, UK (in prep).

SID 5 (Rev. 07/10) Page 24 of 30 ANNEX 1: Merckx, T. (2011) Coppicing and woodland ride management effects on UK BAP moths. In: Noake, B., Rosenthal, A., Parsons, M. & Bourn, N. (eds.) Lepidoptera Conservation Bulletin Number 11: April 2010 – March 2011, Butterfly Conservation, Wareham, pp. 47-48.

• Coppicing and woodland ride management effects on UK BAP moths

The Wildlife Conservation Research Unit (University of Oxford) has teamed up with Butterfly Conservation to conduct a 10-month study which aims to increase our understanding of the effects of woodland conservation management practices on widespread but nationally rapidly declining Priority (‘Research Only’) UK BAP larger moth species (Rothamsted Research trends: >69% national decline in abundance over 35 years).

Thanks to Butterfly Conservation’s South East Woodlands Project, we placed light-traps at 36 fixed sites within the Tytherley woodland landscape, on the Hampshire/Wiltshire border. Six ‘management’ treatments (young/medium/old coppice; standard/wide ride; woodland) were each represented by six sites, and these were all sampled nine times between July and mid-October. We avoided bias by sampling all management groups every trap night, in equal numbers, and only under suitable weather conditions.

The total list amounted to 11,670 individuals from 265 species of larger moth. Thirty-eight of these were Priority (‘Research Only’) species (891 individuals). Seven woodland species that might have been expected to occur did not and their dramatic national declines (87-97% over 35 years) may well explain their absence. We trapped 249 individuals of fifteen nationally scarce/Red Data Book species: Festoon Apoda limacodes, Triangle Heterogenea asella, Mocha Cyclophora annularia, Devon Carpet Lampropteryx otregiata, Great Oak Beauty Hypomecis roboraria, Small Black Arches Meganola strigula, Kent Black Arches Meganola albula, Double Line Mythimna turca, Mere Wainscot Chortodes fluxa, Waved Black Parascotia fuliginaria, White-line Snout Schrankia taenialisI, Lunar Yellow Underwing Noctua orbona, Light Crimson Underwing Catocala promissa, and Dark Crimson Underwing Catocala sponsa. The last three of these are also UK BAP Priority species. A remarkable find was Clifden Nonpareil Catocala fraxini. Seven individuals, both male and female, were trapped on four nights, always near suitable habitat, pointing strongly towards a local population of this species.

A key finding was that overall moth abundance and species richness differed considerably between the management treatments, and appeared to be closely linked to the amount of shelter and bare ground associated with them. It is very likely that such factors impact on levels of convective cooling, and hence activity/occurrence for these night-flyers. Overall abundance was lowest in young coppice (characterised by plenty of bare ground), intermediate in medium/old coppice and wide rides, whereas most moths were found in the sheltered standard rides and woodland. Similarly, overall species richness was lowest in coppice and highest at sheltered sites. Wide rides, though characterised by lower overall abundance, nevertheless had richness levels similar to sheltered sites. This overall picture hence contrasts with the one for day-flying Lepidoptera, where open woodland conditions generally mean higher abundance and species richness, due to higher day temperatures and other key resources.

A closer look, however, reveals that while this overall picture for larger moths applies to most species, irrespective of their scarcity or national trend, the Priority (‘Research Only’) group of UK BAP species showed the smallest difference in abundance between sheltered and more open sites. Wide woodland rides in particular had relatively high abundance of these rapidly declining species. Also, when dividing the species into groups depending on their level of woodland affinity, we found that the species group without any direct affinity to woodland had much higher abundance and species richness in wide woodland rides and young coppice than in older coppice and more sheltered sites, compared to the other species groups, with some or a strong woodland affinity.

Another key finding was that the total area of woodland surrounding coppiced plots had a strong positive effect on the abundance of Priority (‘Research Only’) moths. It also increased overall species richness (at least in medium/old coppice), particularly through its effect on the species group which is

SID 5 (Rev. 07/10) Page 25 of 30 declining nationally (Rothamsted Research trends: 0-69% national decline in abundance over 35 years).

In addition to general patterns of abundance and species richness, we found that 124 species (i.e. 47% of the total) occurred in lower numbers at sheltered sites compared to coppice and wide rides. Of these 124 species, 49 (18% of the total) were in fact restricted to these open areas. This pattern was strongest for the Priority (‘Research Only’) moths, with 22 of the 38 species (i.e. 58%) occurring in equal or higher numbers at managed (open) compared to unmanaged sites. Focusing on this group, we showed that more species occurred in young and medium coppice than in old coppice, although overall abundance increased from young, over medium, to old coppice, leading us to the conclusion that the whole coppice cycle is of value to this rapidly declining group of species. We believe that widening woodland rides is also valuable to this species group, as their overall abundance was higher in wide compared to standard rides.

Our work shows that coppicing and ride widening are valuable conservation tools for moths as well as butterflies; they increase the total species richness of woodlands by increasing the micro-climatic and structural diversity of woods and providing additional resources for species with an affinity for more open biotopes, especially benefiting Priority (‘Research Only’) species. Although there are benefits to increased coppicing and ride widening within woodlands, we show that the greatest benefits are likely to be had within larger woods and mainly so for the nationally declining and rapidly declining species groups.

However, our results also show that the sheltered, dark, humid, late-successional high forest biotope is characterised by high numbers of moths and species, and is especially valuable for some scarce and RDB species. This study has demonstrated the importance of this valuable woodland biotope and that this stage of the succession clearly needs to be accounted for within woodland management.

Our main recommendations are (i) to enlarge existing high forest and create new ones, in order to have cores of typical dark woodland habitat, and (ii) to ‘buffer’ these dark woodland cores from the open matrix by creating lighter woodland zones at their edges. This can be done by coppicing or the creation of wide woodland rides. Such zoning will safeguard the dark and humid environment needed by shade- or moisture-loving woodland specialists, whilst at the same time allowing easy access for species of mixed and more open biotopes to a light and open woodland biotope. This two-tier approach may well be vital both for woodland specialists as well as the declining, once-widespread species that are finding it ever more difficult to locate enough resources within current degraded landscapes.

This work has been funded by DEFRA (Project CR 0470: Understanding the role of woodland management in the conservation of UK BAP moths). The involvement of the South East Woodlands Project in this work was supported by The Heritage Lottery Fund.

Contributed by Dr. Thomas Merckx, post-doctoral researcher, [email protected]

SID 5 (Rev. 07/10) Page 26 of 30 ANNEX 2: Merckx, T. (2011) Woodland management and rapidly-declining moths. In: Moths Count Newsletter 2011, Butterfly Conservation, Wareham, pp. 12.

• Woodland management and rapidly-declining moths

Our understanding of the conservation of woodland moths has been increased by a collaborative project undertaken by Butterfly Conservation and the Wildlife Conservation Research Unit at the University of Oxford. The 10 month study investigating the effect of woodland management on moths produced interesting results. The target species were the ‘Research Only’ UK BAP Priority macro- moths, designated as such because they have suffered >69% national declines in abundance in the Rothamsted Insect Survey over 35 years.

Light-traps were set within six different stages of woodland management in the Tytherley Woods complex, on the Hampshire/Wiltshire border; one of Butterfly Conservation’s South East Woodlands project demonstration areas.

Almost 12,000 individuals of 265 species were recorded between July and mid-October, 38 of which were ‘Research Only’ BAP species. Fifteen nationally scarce / Red Data Book species were trapped, including Triangle, Devon Carpet, Great Oak Beauty, Small Black Arches, Double Line, Mocha, Mere Wainscot and White-line Snout. Three threatened UK BAP Priority species were also trapped; Lunar Yellow Underwing, Light Crimson Underwing and Dark Crimson Underwing. Seven individuals of Clifden Nonpareil were also trapped.

Moth abundance and species richness differed between the woodland management stages and were highest in sheltered rides and mature woodland. Overall, fewer individual moths and fewer species were found in open areas, such as young coppice, where there is a high proportion of bare ground.

This pattern is opposite to that of day-flying Lepidoptera, where open areas have higher abundance and species richness, perhaps due to higher daytime temperatures. Wide rides, however, showed a mixed pattern for moths in this study, with relatively low abundance of individuals but high species richness (on a par with mature woodland).

The ‘Research Only’ group of UK BAP species showed some different trends to the overall results. Wide woodland rides in particular had relatively high abundance of these rapidly declining species. In addition, more than half of the ‘Research Only’ moth species recorded in the study occurred in equal or greater numbers in open, managed woodland habitats compared to more sheltered, less managed areas.

Sheltered, mature woodland areas were found to be important habitats for some scarce and Red Data Book species. Therefore maintenance of this successional stage should also be considered when undertaking woodland management. Coppicing and ride widening are valuable conservation tools for moths as well as butterflies. They increase the total species richness of woodlands, provide resources for species adapted to more open habitats, and benefit the ‘Research Only’ Priority species. However, in order to assist both the ‘Research Only’ Priority moths and the woodland specialists, management should also retain or enlarge core areas of existing mature forest.

Thomas Merckx, Wildlife Conservation Research Unit, University of Oxford

Acknowledgements This work was funded by DEFRA (Project CR 0470: Understanding the role of woodland management in the conservation of UK BAP moths). The South East Woodlands project was supported by the Heritage Lottery Fund.

SID 5 (Rev. 07/10) Page 27 of 30 ANNEX 3:

Model descriptions and interpretations:

Model 1: ABUNDANCE = Treatment + Woodland patch size + Conservation status + Treatment *Woodland patch size + Treatment*Conservation status + Woodland patch size*Conservation status + Treatment*Woodland patch size*Conservation status

Model output and interpretation (see also Fig. 2): Treatment: Overall observation: Exposed sites (i.e. wide rides and coppiced sites, which are characterized by less shelter and lower night temperatures) are highly significantly characterized by lower overall abundance than more sheltered sites (i.e. standard rides and standard woodland) (F5, 24 = 8.7; P<0.0001): • Coppiced sites have lower overall abundance than standard woodland and woodland rides (but not significantly so between medium or old coppice and wide rides) • No difference in overall abundance between the coppice age classes, though lowest abundance at young coppice (which has lots of bare ground and is characterized by a low vegetation height) • Lower abundance at wide rides than at standard woodland, and than at standard rides (trend only) • No difference in overall abundance between standard ride and standard woodland

Treatment*conservation status: The main effect of treatment (sheltered sites have higher abundance) differs in strength according to the conservation status group concerned, and is absent for the ‘Research Only BAP’ group (F20, 96 = 2.05; P=0.011) (Fig. 3; Table 3) (see also 3- way interaction below):

Table 3: Overview of model estimates for the treatment*conservation status interaction (sheltered sites in bold)

status Coppice1 Coppice2 Coppice3 RideS RideW Woodland BAP ‘Research Only’ 0.39 0.52 0.46 0.54 0.58 0.68 Declining 0.99 0.95 0.94 1.18 1.1 1.18 Increasing 1.04 1.13 1.25 1.38 1.23 1.4 RDBscarce 0.12 0.21 0.14 0.31 0.24 0.31 Unknown 0.49 0.55 0.54 0.71 0.5 0.66

The main effect of treatment (higher abundance at sheltered sites) is most visible for: • ‘Increasing’ species (increase in abundance with coppice age; standard ride and woodland similarly high abundance, and wide ride abundance situated in between sheltered and coppiced levels). • Then for ‘Unknown’ species and ‘RDB/scarce’ species: no difference in abundance between coppice classes but young coppice lowest abundance; sheltered sites similarly high abundance; wide ride abundance similar to coppiced abundance levels. • Then for ‘Declining’ species: no difference in abundance between coppiced classes; sheltered sites similarly high abundance levels; wide ride abundance situated in between. The only group that actually follows a significantly different pattern is the ‘Research Only BAP’ group which has the smallest difference in abundance between sheltered and non-sheltered sites. The overall effect of treatment, visible for all other species groups, is hence not clearly pronounced because the abundance at standard rides is relatively low: although standard rides are characterized by significantly higher numbers than young coppice (though no significant differences are present with medium and old coppice), abundance is lower (though not significantly so) than at wide rides, which are less sheltered. Sheltered woodland is characterized by high abundance though, but this is nevertheless not significantly different from wide ride abundance, and we only observe a non-significant trend for a difference with standard ride abundance.

SID 5 (Rev. 07/10) Page 28 of 30

Woodland patch size*treatment*conservation status: The main pattern of this significant three- way interaction (F20, 96 = 2.21; P=0.0056) is that woodland size - surrounding coppiced plots - only seems to significantly positively matter for the ‘Research Only BAP’ species group. Also, but to a much lesser extent, woodland size matters for the ‘Increasing’ species (trends for medium and old coppice) and the ‘Unknown’ group (significantly for old coppice), and not for the other species groups.

Other observations are: • No significant difference in the strength of the effect between the coppice age classes (i.e. overlapping error bars). • Trend for a higher abundance of ‘Research Only BAP’ species when standard rides are situated within larger woods.

Model 2: SPECIES RICHNESS = Treatment + Woodland patch size + Conservation status + Treatment*Woodland patch size + Woodland patch size*Conservation status

Model output and interpretation: Treatment: Overall interpretation: (see also Fig. 2) Species richness is highly significantly lower at coppiced sites than at standard woodland and woodland rides (F5, 24 = 7.39; P=0.0003).

Remarks: • There are no significant differences in species richness among coppice ages, nor are there among rides and sheltered sites. • Contrary to the observations with regard to abundance levels, wide rides have similar species richness levels than sheltered sites.

Woodland patch size*treatment: Main observation: Woodland size has a significant positive effect on species richness for medium and old coppice only (F5, 24 = 5.70; P=0.0013).

Woodland patch size*conservation status: Main observation: Woodland size has a significant and positive effect on the ‘Declining’ group only, and is not important for the other conservation status species groups (F4, 136 = 3.63; P=0.0077). .

Model 3: ABUNDANCE = Treatment + Woodland patch size + Woodland affinity + Treatment*Woodland patch size + Treatment*Woodland affinity + Woodland patch size*Woodland affinity + Treatment*Woodland patch size*Woodland affinity

Model output and interpretation: Treatment: Overall interpretation: Similar pattern to Model 1 (F5, 24 = 7.04; P=0.0004) (Fig. 2).

Treatment*Woodland affinity: The main effect of Treatment applies very well to the ‘Mixed’ and ‘Woodland’ group, while the ‘Ubiquitous’ species follow this trend, except with regard to the old coppice class which is characterized by a relatively high abundance (i.e. almost as high as standard ride and higher than wide ride abundance). The group that is not linked to the woodland biotope at all (i.e. ‘Other’) breaks this main pattern completely and significantly (F15, 72 = 2.81; P=0.0017), and occurs most abundantly in young coppice and wide rides, then in medium coppice, then in standard rides and woodland, and the group occurs least abundantly in old coppice.

SID 5 (Rev. 07/10) Page 29 of 30

Woodland patch size*Treatment: Wood size has a significantly positive effect on abundance within old coppice only (F5, 24 = 5.62; P=0.0015).

Woodland patch size*Treatment*Woodland affinity: (F15, 72 = 2.04; P=0.023) • The abundance of species with no affinity to woodland (i.e. ‘Other’ group) is significantly negatively affected by increasing woodland size, surrounding young and old coppice plots (and medium coppice too but not significantly so). • Increasing woodland size increases (though non-significantly so) the abundance of all other woodland affinity groups, especially so at medium coppice for the ‘Mixed’ and ‘Ubiquitous’ species groups.

Note: the group of ‘Other’ species only occurs in low abundance, and this observation should obviously not inform to make woods smaller. It is interesting from an ecological point of view to notice that their ecology (biotope preference) hinders them to infiltrate into large woods. In case one would want such species within woods, they will need open corridors connected to the outside of the woodland to draw them inside too.

Model 4: SPECIES RICHNESS = Treatment + Woodland patch size + Woodland affinity + Treatment*Woodland patch size + Treatment*Woodland affinity + Woodland patch size*Woodland affinity + Treatment*Woodland patch size*Woodland affinity

Model output and interpretation: Treatment: Overall interpretation: Similar pattern to Model 2 (F5, 24 = 5.66; P=0.0014) (Fig. 2).

Treatment*Woodland affinity (F15, 72 = 2.37; P=0.0078). • The main effect of Treatment applies well to ‘Mixed’ and ‘Ubiquitous’ species groups (but less pronounced for the latter group, with a gradual decrease in species richness with coppice age). • Woodland species numbers are lowest at young coppice and highest at standard woodland (obviously), but - interestingly - we see that species richness at medium coppice is better than at old coppice and is at the same level as both types of ride and almost as high as standard woodland. • The species with no affinity to woodland display a completely and significantly different pattern, with highest richness in wide rides and young coppice, then at medium coppice, then at standard rides and woodland, and lowest species richness at old coppice.

Woodland patch size*Treatment*Woodland affinity: (F15, 72 = 2.17; P=0.015) • The only significant positive effect of woodland size surrounding the plots is for the ‘Mixed’ group at old coppice plots. • Similarly as before, the ‘Other’ species group (no direct affinity to woodland) has mainly negative effects of woodland size on species richness, but not significantly so.

SID 5 (Rev. 07/10) Page 30 of 30