Ecological Characterisation of the Environmentally Sensitive Areas of England

THE

ENVIRONMENTALLY

SENSITIVE

AREAS

SCHEME

ECOLOGICAL CHARACTERISATION OF THE ENVIRONMENTALLY SENSITIVE AREAS OF ENGLAND

FINAL REPORT TO DEFRA

P.D. Carey, S. Hulmes, Peter Nuttall, R. Large, L. Hulmes, P. Croxton, W. MacFarlane, R. Holder, R. O’Shea, D. Spence, C. Pinches, R.A. Garbutt, A. Burrows, R.A. Wadsworth

Monks Wood

July 2005

EXECUTIVE SUMMARY

This report presents the results of a project to assess the ecological quality of land being managed within the Environmentally Sensitive Areas Scheme (ESAs), an agri-environment scheme designed to enhance the environmental quality of farmland in England.

The objective of the project was to assess the ecological quality of a sample of agreement land in terms of its vegetation characteristics and the presence of habitats listed within the UK Biodiversity Action Plan (BAP). The detailed objectives were to:

provide national estimates of the extent of BAP Broad and Priority Habitats under ESA agreement;

provide national estimates of vegetation character, and hence the ecological ‘quality’, of ESA agreement land, with particular emphasis on Priority Habitats;

analyse the spatial distribution and vegetation characteristics of the habitats surveyed (with special reference to Priority Habitats) with regard to geographic location, agreement age and management tier; and

establish a baseline for future evaluation of changes in ecological ‘quality’.

A secondary objective of the project was to develop and test the use of hand held electronic mapping devices in the field. The use of such devices was found to have potential and the experience from this project will prove valuable for future development.

A main objective of this study was to determine the contribution of the ESAs to the protection of Priority Habitats. As it was thought that a random sample of ESA agreements would not detect sufficient examples of rare and important habitats, a stratification system was designed for this survey to target management tiers in the ESA schemes where it was thought more likely that Priority Habitats would be present.

A stratified random sample of 450 agreements was selected for survey covering an area of approximately 70,000 ha, representing approximately 12.5% of the land under ESA agreement in England. The stratification was initially based on identifying eleven habitat divisions that were considered likely to be found in ESA agreements. ESA management tiers were then assigned to each habitat division. The survey sample was subsequently drawn from a list of ESA agreements containing the target management tiers.

This project was also designed so that a comparison with a previous project, the ecological characterisation of the Countryside Stewardship Scheme, and also the Countryside Survey could be made.

A survey of all land cover in sample agreements was undertaken to map UK Biodiversity Action Plan Broad and Priority Habitats using a vegetation key. Where appropriate, habitat mosaics were mapped. As well as conventional mapping and subsequent digitisation from paper maps using Arc-View GIS, electronic data gathering in the field was trialled, involving digitising in the field using hand held computers with POCKET GIS.

A random 200 m2 vegetation quadrat was recorded within each agreement surveyed using Countryside Survey methods. In addition, a quadrat was recorded in each Priority Habitat identified at the site, excluding any that had been recorded by the random quadrat. The quadrat positions were mapped and permanently marked in the field to allow precise relocation. Each quadrat was classified in terms of the National Vegetation Classification (NVC) and the Countryside Vegetation System (CVS); the number of species and the presence of rare and scarce species were also quantified. The quadrats were co-located with the spatial data in the field and database by using GPS and the software POCKET GIS.

In addition, a variety of observations were taken (e.g. photographs and target notes on rare species and/or weed infestations) to aid interpretation of future surveys. A species list was also produced for each habitat encountered. These data probably provide the largest resource linking Broad Habitats and the species found in them. However, whilst the position of target notes has been entered into the database, entry of the species lists is outwith the remit of this project.

The most widespread Broad Habitats found were Acid Grassland (20%), Improved Grassland (19%) and Neutral Grassland (14%). Based on a projection from the stratification of management tiers used, the survey suggests that approximately 143,000 ha of Improved Grassland, 114,000 ha of Neutral Grassland and 77,000 ha of Acid Grassland are under ESA agreement.

Mosaics of different habitats covered 25% of the whole area surveyed. The stratified projection is that they will be found in 18% of all ESA land which approximates to 114,000 ha. Most mosaics recorded were combinations of Acid Grassland, Bracken, Rocky Ground, and other upland habitat types.

Twenty one Priority Habitats were found, accounting for 17 % of the agreement land surveyed. In addition to the 17%, land within mosaics (6500ha) also contained one or more Priority Habitat, but the extra area of Priority Habitat within these mosaics is not calculable. Upland Heathland was the most abundant Priority Habitat covering approximately 5511 ha (8%) .

The stratified projection for Priority Habitats predicts that approximately 63,000 ha of all ESA land will be Priority Habitat representing about 12% of the total.

The analysis of vegetation data from quadrats revealed that 53 % of all randomly-placed quadrats were categorised as the CVS class Infertile Grassland and 24 % as CVS class Fertile Grassland. Approximately 700 different species of vascular plants and bryophytes were found in quadrats. The mean number of species per quadrat was 20. The most diverse quadrat was found in chalk grassland and had 69 species, and the least diverse quadrat, on recently cleared ground, had 1 species. 114 of the 450 Random Quadrats (25%) recorded were found to have been within Priority Habitats. No Red Data Book species was found in quadrats outside Priority Habitats. Two scarce species were found in Random Quadrats that were not Priority Habitats: Sesleria albicans was found in five quadrats and Hordeum marinum in two quadrats.

Quadrats located within Priority Habitats had a mean number of 28 species per quadrat.The number of species found in Priority Habitats was not that much higher than in the Random Quadrats because (i) some Priority Habitats are not diverse e.g. moorland or some Cereal Field Margins, and (ii) because the Random Quadrats also included some containing diverse Priority Habitats. The highest number of species found was 88, in a quadrat on the South Downs. One Red Data Book species, Thymus serpyllum, and possibly a second, Trifolium striatum were found in Priority Habitat quadrats. Seven different scarce species were found. The vegetation of these quadrats had a lower proportion of CVS Aggregate Class (AC) Fertile Grassland and Infertile Grassland, and a higher proportion of ACs Moorland Grass/ Mosaic and Heath/Bog than the randomly placed quadrats (note that randomly placed quadrats falling within Priority Habitats were double counted). Priority Habitat quadrats also contained NVC communities of conservation importance that were scarce or absent in quadrats falling outside Priority Habitats. They included more calcareous grassland, heathland and mire communities.

Coastal Floodplain and Grazing Marsh (CFGM) was treated separately to other Priority Habitats because, uniquely, it is not defined by its constituent vegetation. Fifty six quadrats were sited in CFGM, containing a mean of 21 species. The grassland present was typically infertile in nature.

Results are presented for each of the habitat divisions used in the sample stratification, for each individual ESA and also for each Priority Habitat.

The data have been compared to similar data from both the Countryside Survey 2000 (CS2000) and from the Ecological Characterisation of the Countryside Stewardship Scheme (CSS). The ESAs reflected the pattern found in CSS of having a much lower proportion of land in arable or horticulture than the countryside as a whole. The ESAs also have a lower proportion of improved grassland and a higher proportion of semi-improved grasslands than the countryside as a whole. The ESAs (10%) have a higher proportion of dwarf shrub heath than the countryside as a whole (3%) or the CSS (5%). The number of species found per random quadrat in the ESAs and CSS were almost identical as were the proportions of vegetation classes to which Random Quadrats were assigned. More species were found in Priority Habitat quadrats in the ESAs than in the CSS and many more quadrats were assigned to upland vegetation types in the ESAs.

Differences between the ESAs and the countryside as a whole clearly reflect the priorities of the ESAs, especially the high proportion of grassland and upland habitats. There are encouraging signs within this comparison that the Scheme has successfully included land of a different character than in the countryside as a whole and of a character likely to be considered of greater conservation value. The differences in the areas of some habitats found in the ESAs compared to the CSS demonstrates the targeting of ESAs to contiguous areas in England; especially areas of Upland and Wet Grassland in river valleys. The similarity in the botanical quality of vegetation in quadrats between the ESAs and CSS is striking.

Following concerns over an earlier draft of this report on the low figures for Blanket Bog and Dwarf Shrub Heath compared to national estimates, reanalysis was carried out and a comparison made between this survey and the Landcover Map 2000 and Landcover Map of GB. The reanalysis showed large areas of land that could potentially be degraded Blanket Bog in mosaics containing Upland Heathland and Acid Grassland. LCM2000, Countryside Survey and this survey all have much lower estimates for Blanket Bog than the national estimate produced by JNCC.

Projections for the probability of finding the four most common Priority Habitats: Floodplain & Coastal Grazing Marsh; Lowland Calcareous Grassland; Lowland Heath; and Upland Heath were created. Using maps of the ESAs and the distribution of management tiers it was possible to map these projections. The projections could be used to target future agreements or monitoring.

Overall, the results show that the Scheme has targeted grassland vegetation at higher proportions than found in the countryside as a whole. Moreover, this grassland tends to be less fertile than grassland in the countryside as a whole, suggesting an increased conservation quality. Furthermore, the presence of a high proportion of Priority Habitats, and the presenceof scarce NVC communities, suggests that the Scheme has successfully targeted land of high conservation value.

The real value of this study will become apparent if agreements and/or quadrats are resurveyed in the future, ideally at the same time as another Countryside Survey; only then will it be possible to judge the ecological value added to agreement land through time.

1INTRODUCTION 1

2THE SAMPLE 3

3METHODOLOGY13

4RESULTS17

Whole Sample17

Analysis of Different Management Within the ESAs29

The Results By ESA34

Comparison between the ESAs, the Countryside

as a whole and the Countryside Stewardship Scheme43

The results by priority habitat461

5DISCUSSION53

Acknowledgements60

References60

LIST OF APPENDICES

1.ESA SURVEY (AE02) - OVERALL PROTOCOL61

2.COMPLETE LIST OF ALL BROAD HABITATS AND MOSAICS

SURVEYED77

THE CVS CLASS, ELLENBERG AND CSR SCORES, AND NVC

CLASS OF EACH QUADRAT SURVEYED91

4.NATIONAL VEGETATION CLASSIFICATION (NVC) OF QUADRATS LOCATED WITHIN ESA LAND 111

5.BROAD AND PRIORITY HABITAT TABLES FOR DIFFERENT MANAGEMENT TIER DIVISIONS 114

6.VEGETATION CHARACTER OF QUADRATS IN INDIVIDUAL

ESAS129

THE BROAD AND PRIORITY HABITATS IN EACH

INDIVIDUAL ESA140

8.MANAGEMENT TIERS WITHIN ENGLISH ESAS196 9. PROJECTIONS OF PRIORITY HABITATS 173

10.COMPARISON OF SURVEY DATA WITH LCM2000 AND

LCMGB

1INTRODUCTION

Background

1.1The England Rural Development Programme (ERDP) has provided an opportunity for increased support for agri-environment measures, including the Environmentally Sensitive Areas (ESA) and Countryside Stewardship (CS) schemes. To inform policy decisions on development of these schemes they need to be evaluated.

1.2The recent CS monitoring programme (1997-2000) adopted a new approach to evaluating the ecological ‘quality’ of agreement land. This study enabled an estimate of the area of each habitat under agreement to be made and was compared with the wider countryside using data from Countryside Survey 2000 (CS2000). Broad and Priority Habitats were mapped at agreement scale and further data gathered from quadrats, using methods compatible with CS2000, to provide an indication of the ecological characteristics of land under agreement. The data also provide a baseline record for monitoring change. Various classifications of the data were made using systems such as the National Vegetation Classification (NVC) and Countryside Vegetation System (CVS).

1.3The primary aim of the project is to carry out a similar study of a sample of ESA agreements.

Environmentally Sensitive Area Schemes

1.4Twenty two areas in England have been designated as Environmentally Sensitive Areas (ESAs). This type of scheme was introduced in 1987 to encourage farmers to help safeguard areas of the countryside where the landscape, wildlife and historical interest is of national importance and is dependent on the use of beneficial farming practices. The scheme targets a range of habitats which include wet grassland, dry grassland (acid, calcareous and neutral), uplands (hay meadows and moorland), lowland heathland and arable margins. In general the aim of the ESAs is to maintain and enhance the wildlife, landscape, historical and access value of the area through the adoption of traditional management practices.

Research Requirements

1.5The study was carried out over a period of two years; the start up period included project planning and design and devising sampling strategies and method protocols. It was intended that the sample should be representative of land being managed in the ESAs as a whole, covering the range of habitats under agreement and reflecting the differences between ESAs and tiers.

1.6The methods were based on those used in the Countryside Stewardship Monitoring Programme (Monitoring and Evaluation of Countryside Stewardship Scheme Methodology Report, November 2000).

1.7The study:

provides national estimates of the extent of Biodiversity Action Plan (BAP) Broad and Priority Habitats under ESA agreement;

provides national estimates of vegetation character, and hence ecological ‘quality’, of ESA agreement land, with particular emphasis on Priority Habitats;

analysed the spatial distribution and vegetation characteristics of the habitats surveyed (with special reference to Priority Habitats) with regard to geographic location and tier; and

establishes a baseline for future evaluation of changes in ecological ‘quality’.

1.8A secondary objective of the project was to develop and test digital data capture methods in the field using PDA computers. CEH carried out this objective in conjunction with Position Resources Limited.

2THE SAMPLE

2.1The sampling rationale was based on two of the objectives of this study that state that the project should:

provide national estimates of the extent of Biodiversity Action Plan (BAP) Broad and Priority Habitats under ESA agreement; and
provide national estimates of vegetation character, and hence ecological ‘quality’, of ESA agreement land, with particular emphasis on BAP Priority Habitats

2.2Unfortunately Priority Habitats would be unlikely to be found frequently enough in a random sample of all agreements to fulfil these objectives. This is because the Priority Habitats are generally not common and some are rare e.g. upland calcareous grassland. This problem was alleviated by creating a stratified random sample whereby the management tiers were grouped into habitat divisions (strata) based on the likelihood of them containing target Priority Habitats.

2.3Unlike CSS, each ESA has individual objectives and suites of management tiers (Appendix 8). Hence a sample based on a simple stratification by tier or management prescriptions would have proved very unwieldy, if not impossible. As the objective was to report by Broad Habitat we stratified the sample by habitat following the rationale outlined below and shown diagrammatically in Figure 2.1.

Rationale

2.4The rationale behind this sampling strategy was to make sure that as many habitats as possible were sampled with an adequate replication.

2.5An over-riding assumption was made, based on resource availability, that approximately 500 agreements could be sampled.

2.6A list of 10,915 extant agreements (up to year 2000) was supplied by Defra for use in the stratification process.

2.7The first level of stratification was by habitat. Divisions of habitat type were identified to reflect Broad Habitats and obvious divisions of them that could also be linked to the objectives of individual ESAs, where priority habitats are targeted. As Defra had funded a separate survey of upland hay meadows in the Pennine Dales these were excluded from the stratification. Eleven habitat divisions were chosen.

1)Arable and arable reversion (to sample cereal field margins within arable)

2)Wet grassland (to sample coastal & floodplain grazing marsh within improved grassland)

3)Upland acid grassland

4)Lowland acid grassland

5)Pasture/in-bye (neutral) (to sample semi-improved, grazed neutral grassland)

6)Hay meadows (neutral) (to sample lowland and upland hay meadow priority habitats)

7)Upland calcareous grassland

8)Lowland calcareous grassland

9)Dwarf shrub heath (upland heath)

10)Dwarf shrub heath (lowland heath)

11)Fen/marsh/swamp

2.8The second level of stratification was to assign ESA tiers to these habitat divisions (Table 2.1). This table was updated based on expert knowledge from Defra staff and uses a generalised set of management tier codes from the document “Agri-Environment Scheme Measures That Contribute To BAP Priority Habitat Targets” (Ovenden et al. 1999) provided by them. These codes will be unfamiliar to those used to the tier codes used within individual ESAs and so a translation is given in Appendix 8.