APPENDICESCONTENTS

A1Description of survey sites and summary of results 1994-71

A1.1Site selection and general considerations1

A1.2Southlake Moor, Somerset2

A1.3Walton Moor, Somerset2

A1.4Heigham Holmes, Norfolk3

A1.5Berney Marshes, Norfolk4

A1.6Other Broads Sites5

A1.6.1Belaugh5

A1.6.2Reedbush Carr5

A1.6.3Tatterford Common5

A1.6.4Twyford6

A1.6.5.Carlton Marshes6

Acknowledgements6

A2Results for 19987

A.2.1Figures7

A.2.2Tables12

A3Agro-economic data and methods for financial analysis of Tier 3 adoption28

A4Publications arising wholly or partly from research funded under BD130132

Note:Appendix 1 provides a summary of information (and background references) previously presented in the three earlier annual reports (Treweek et al., 1995; Mountford et al., 1996 and 1997b), and represents a shortened version of the corresponding appendix in Mountford et al (1998).

APPENDIX 1DESCRIPTION of SURVEY SITES and SUMMARY of RESULTS 1994-7.

A1.1Site selection and general considerations

Possible sites were suggested by ADAS (now FRCA) regional staff and selected after discussion between ADAS, Silsoe and the ITE. Four sites were chosen in 1994, another in 1995, and four more in 1996. Sites were selected on the basis of; a) geographical spread; b) soil type and c) Tiers represented. Ideally, each site had to possess fields both under unaltered and raised water-levels, though this proved possible only at Heigham Holmes, Southlake and Walton Moors. It was decided to cover more sites with varied characteristics, and use ordination techniques to help analyse variation between sites in relation to water-management. Site files may be consulted by MAFF officers at ITE Monks Wood or ADAS Gleadthorpe.

A common core of data was recorded at each site, and Tables 1 and 2 (Appendix 2) give a summary of the data gathered. The National Vegetation Classification (NVC - Rodwell, 1991, 1992, 1995, MS) was used to identify vegetation types, some of which have been targeted for conservation and restoration within ESAs (for a summary of the main types see Treweek et al., 1996). The site accounts (A1.2-A1.6) provide:

a standard context within which to evaluate both species data and community variables;

a standard comparison of vegetation between sites; and

a baseline from which to assess change; and

summaries of the key results for 1994-97.

A1.2Southlake Moor, Somerset

The site

Southlake Moor lies by the R. Parrett 9 km south-east of Bridgwater, at ca. 4 m A.O.D.. Soils are of the Midelney series: thin layers of alluvial clay over peat which is within 90 cm of the surface. Such soils develop where subject to regular flooding, and warping was practised on this site in the past. Embankments were built around the moor in the early fourteenth century, and the basic network of rhynes was excavated. The peat and lower levels of alluvium are permanently waterlogged, and surface soils are moderately acidic to neutral in reaction. Southlake Moor is under permanent grass, managed by a summer hay-cut with aftermath grazing. The moor is a Site of Special Scientific Interest (SSSI), and was subject to detailed grassland survey by English Nature. Several NVC grassland communities were recorded, and although some fields were improved in the past (MG7Lolium perenne leys and related grasslands), most of the moor is a mosaic of MG8 (Cynosurus cristatus-Caltha palustris grassland), MG9 (Holcus lanatus-Deschampsia cespitosa grassland) and MG6 (Lolium perenne-Cynosurus cristatus grassland), with MG5 (Cynosurus cristatus-Centaurea nigra grassland) in raised parts, and MG13 (Agrostis stolonifera-Alopecurus geniculatus inundation grassland) in depressions. Drainage channels (rhynes and ditches) were yet more diverse, and represented a major factor in the scheduling of the moor.

Water-management

Parts of the moor were included within Tier 1 from the declaration of the ESA in 1986, but achievement of Tier 3 conditions was not practical until a comprehensive scheme was introduced in 1993. Splash-flooding was implemented for the first time during the winter of 1993/4, and the 1994 field-season thus represented the first year under the new hydrological regime.

Summary of results 1994-7

The ITE and Silsoe College surveyed seven fields in 1994, but no further fieldwork was undertaken under the present project until 1998. In the intervening time the site was monitored by English Nature (Cox, 1997). The 1994 survey suggested that Tier 3 management had had rather little impact on the vegetation, though those eight species which showed significantly different cover under either Tier 1 or Tier 3 were typical of respectively moist-damp and wet situations (Ellenberg, 1988; Hill et al. submitted). Bare ground was more extensive under Tier 3, possibly as a result of extensive sward death due to the flood. In subsequent years, the impact of raised water-levels became more apparent (Cox, 1997):

Tier 1: Clear changes in the composition were observed, where species that increased were typical of moist (but not waterlogged) conditions e.g. Centaurea nigra, Cerastium fontanum and Plantago lanceolata. In contrast, species of wetland situations (Carex nigra, Cardamine pratensis, Glyceria fluitans and Ranunculus flammula) declined under Tier 1. These changes in “control” fields may reflect the exceptionally dry springs and summers of 1995-6.

Tier 3: Raised water-levels led to an increase in wetland species , notably sedges (Carex disticha and C. nigra), Cardamine pratensis, Lysimachia nummularia, Oenanthe fistulosa and Senecio aquaticus, as well as plants more typical of damp (but not wet) pastures e.g. Deschampsia cespitosa. During the drought years, Holcus lanatus and other plants of drier pastures increased in Tier 3 fields. Major declines were observed in Cerastium fontanum, Ranunculus acris, Rumex acetosa and Trifolium pratense, all more typical of MG5 grasslands. Although changes were very pronounced at first, the trend lessened after the dry years of 1995 and 1996, and then increased once again under more typical weather patterns.

A1.3Walton Moor, Somerset

The site

Lying at ca. 4 m A.O.D. between the Poldens and High Ham uplands, the site is in the Cary catchment on earthy eufibrous peat soils (Altcar 1 series). Most of the moor is under long-term grass and grazed by cattle (or locally sheep), though some arable remains and a few fields were sown to grass in the 1980s (including a Tier 3 field selected for study). Large parts have been improved agriculturally (MG7), and swards of conservation interest (MG8) are poorly developed, though some areas approach MG9 or MG10 (Holcus lanatus-Juncus effusus rush-pasture), and inundation grasslands (MG13) locally occur. Isolated fields appeared under-grazed, with frequent couch and false oat-grass. A few fields had a sward allied to U1 (Festuca ovina-Agrostis capillaris-Rumex acetosella).

Water-management

A raised water-level scheme was introduced in 1994/5, following engineering in the previous summer, enabling two blocks to meet Tier 3 prescriptions, as well as areas of Tiers 1 and 2. Field-work in 1994 was completed before water-levels were raised, providing a baseline assessment of the grassland.

Summary of results 1994-97

Prior to the raising of water-levels, vegetation was essentially similar in fields subsequently allocated to the two different tiers except that Lolium perenne was dominant in Tier 1, and Holcus lanatus in Tier 3. It may be assumed that later differences resulted from altered water-regime. Sward composition in the two tiers diverged from 1995-7:

Tier 1: The vegetation type present was MG7d (Lolium perenne-Alopecurus pratensis sub-community). Over the period Cirsium arvense markedly increased, though coarser grasses (Elytrigia repens and Festuca pratensis) declined. Despite some between-year variation L. perenne remained significantly commoner and dominant in Tier 1, indicating fertile conditions (more intensive agriculture) as also did C. arvense, Stellaria media and Taraxacum.

Tier 3: The sward had abundant Agrostis stolonifera, with sub-dominant H. lanatus and Ranunculus repens, as well as frequent Anthoxanthum odoratum, Carex hirta, L. perenne, Poa trivialis, Ranunculus acris and Rumex acetosa in a community allied to MG10. The cover of A. stolonifera in 1997 was three times its 1994 level, and R. acris and R. acetosa also appeared to increase, whilst Cynosurus cristatus, Juncus effusus, Phleum pratense, R. repens and T. repens declined.

A1.4Heigham Holmes, Norfolk

The site

The Holmes are at ca. 2 m A.O.D. by the R. Thurne near Martham Ferry and on pelo-calcareous alluvial gley soils (Newchurch 2 series). The grassland is cattle-grazed and considered of minor interest, compared to the ditches and birds which led to its designation as a SSSI and purchase by the National Trust. The preponderance of MG7 indicated past intensive agriculture, with liming before 1992, and spot treatment of C. arvense. Some fields are mown prior to cattle grazing. Where farming intensity was less, a Festuca rubra-Agrostis stolonifera-Potentilla anserina grassland (MG11) occurred – the typical form in Tier 3 and the Lolium perenne sub-community (MG11a) in Tier 1. There were large areas of coarse grass (Elytrigia repens and Festuca arundinacea).

Water-management

Following a Water Level Management Plan, raised water-levels were introduced in 1993/4. Since the soils have poor hydraulic conductivity, the National Trust renovated the “foot-drains” in order to convey water from high-level carriers into the field centre. The ITE baseline survey of 1994 was undertaken in the first year after implementation of a Broads ESA Tier 3 water-regime.

Summary of results 1994-97

In 1994, Tier 1 fields were dominated by L. perenne and Tier 3 by E. repens, a pattern which probably reflected selection of less productive fields for raised water-levels. Some changes were observed in both Tier 1 and 3 fields, following extensification of farming under the Trust and ESA e.g. increasing cover of Festuca rubra, H. lanatus and Trifolium repens The sward in all fields was very species-poor in 1994, and both regimes witnessed species-recruitment from 1994-7 (from 3.5 up to 7 species m-1).

Tier 1: Fields dominated by couch and rye-grass in a species-poor sward, with sub-dominant Festuca arundinacea and H. lanatus. E. repens spread after 1994, whilst A. capillaris, Alopecurus geniculatus and L. perenne declined.

Tier 3: Significantly more species-rich, the sward was dominated by A. stolonifera, E. repens and H. lanatus, with L. perenne, F. rubra, A. pratensis, Poa trivialis and T. repens occasional. A. stolonifera had greatly increased , whilst couch had declined, and there was evidence that the original coarse sward was changing toward a mixed grassland with some relationships to MG9, MG10 and MG11.

A1.5Berney Marshes, Norfolk

The site

Part of Halvergate Marshes SSSI, the site is between sea-level and 1m. A.O.D., by the R. Yare. Although mostly on ground-water gleys of the Newchurch 2 series, the local pattern is complex. The entire study area is under cattle-grazed permanent grass, with local spot treatment of thistle. In 1995, two species-poor grassland types predominated, with MG6 mainly under Tier 2 and MG11/MG11a in the Tier 3. Some lower-lying parts of Tier 3 had MG12 or MG13, and the scrapes possessed S21 (Bolboschoenus maritimus swamp), or salt-marsh (SM16Juncus gerardii, SM20Eleocharis uniglumis and SM23Spergularia marina-Puccinellia distans; Rodwell MS). Where water stood late in the Tier 3 scrapes Tier 3, A21Ranunculus baudotii community occurred (Rodwell, 1995).

Water-management

On the RSPB reserve, raised water-levels, bunding and other management for conservation were first implemented from 1986-91, and the site now comprises Tiers 2 and 3 fields, the latter area generally lying slightly lower than the Tier 2. Shallow scrapes were dug in some fields which later qualified for Tier 3, but data-gathering in 1995-8 focused on the grassland. Hydrology has been studied and modelled by ADAS for >6 years, deriving mean water-table depth for all fields.

Summary of results 1995-7

Data were gathered from a large number of field, reflecting two situations:

  • Tier 2 for 5 (usually 9-10) years, but water-regime not markedly altered during that time.
  • Tier 3-type regime from at least 1991 (some since 1986) - controlled raising of water-level.

Berney was the only site within the present study where high salinity was important, though this additional factor did not mask clear effects of raised water-levels on the vegetation.

Tier 2: Dominated by A. stolonifera and L. perenne and T. repens, with C. cristatus, H. lanatus and Hordeum secalinum frequent. Apart from A. stolonifera (which declined over the period) and H. secalinum, these species were significantly commoner in a sward which was clearly more species-rich than the Tier 3.

Tier 3: A. stolonifera often dominant, with areas of open water and bare soil, but quite varied with other species locally important. The mesotrophic grasses of Tier 2 were generally uncommon and often declining, but annuals, species of saline and wet or disturbed sites were all better represented in the Tier 3 fields. Since raised water-levels had been longer-established on Berney Marshes, it was not surprising that a greater number of species showed differences in abundance between the tiers than at other study sites.

A1.6Other Broads Sites

A1.6.1Belaugh

The site: By the R. Bure west of Wroxham, the site comprised cattle-grazed fields on deep well-humified peat of the Adventurers'-Prickwillow complex, though riverine mud had been spread to produce an area raised above the general altitude of <2 m A.O.D. Though derived from tall-herb rich-fen, the site was modified by variable grazing intensity producing a species-rich and diverse vegetation. Grassland on the dumped mud was the Juncus inflexus sub-community of MG10 (i.e. MG10b), whereas the peat supported a vegetation allied to Phragmites australis-Peucedanum palustre tall-herb fen (S24), or tending toward a Juncus subnodulosus-Cirsium palustre fen-meadow (M22) where grazing was more intense.

Water-management: The site was accepted as meeting Tier 3 criteria in 1992, but since the drainage network was directly connected to the river, this required no raising of water-levels. The ditches were cleaned in winter 1993/4 to facilitate water movement through the site and to enhance field wetness.

Summary of results 1994-97: No clear dominant occurred, and Thelypteris palustris was the only species with >10% mean cover. Most of the site is a mosaic of areas where sedges, rushes and grasses (including locally Phragmites) are important. The fields became apparently more species-rich from 1994-7 with Carex species spreading together with Hydrocotyle vulgaris and Calliergon cuspidatum. However, there was evidence of a decline in Equisetum palustre, Juncus articulatus and Phleum pratense.

A.1.6.2Reedbush Carr

The site: At the west edge of the Wickhampton Marshes and <1 m A.O.D., the fields were permanent grassland, summer grazed by, but left for long periods without stock. The site covered a variety of soils including pelo-alluvial and typical humic-alluvial gleys, and an area around the carr itself with the same peat soil as at Belaugh - the boundary between peat and gleys visible as a clear change in elevation and vegetation. On the higher-lying alluvial soils was species-poor MG11a, with A. stolonifera, H. lanatus and L. perenne dominant, and prominent R. repens and T. repens. The lower-lying peat had species-rich M22 with a higher proportion of mesotrophic grasses than at Belaugh. Some areas approached the Briza media-Trifolium spp. sub-community (M22b), and possessed a number of uncommon mire species.

Summary of results 1995-97: There was no change in water-management. A. stolonifera, rye-grass and Ranunculus acris declined, whilst Anthoxanthum odoratum, Cirsium arvense, Dactylis glomerata, Festuca spp., and Holcus lanatus spread with a concomitant increase in species-richness.

A1.6.3Tatterford Common

The site and water-management: On the south bank of the R. Wensum, near Tatterford village and at ca. 40m A.O.D., the soils are typical humic-sandy gleys (Isleham 2 association), but varied with much sand and gravel in the subsoil, and often a markedly humose topsoil. Rather under-grazed in 1996, with abundant C. arvense, by 1997 cattle grazing had increased. The site had a coarse grassland with some areas of MG11a and MG13, others of MG10b, and patches of the Poa trivialis sub-community of MG9a. Where the topsoil was peaty a somewhat richer flora occurred with Dactylorhiza and Carex species. The entire common was entered for Broads Tier 3 in 1996, when water-control structures were installed, including bunds within the channels draining the site, and a sluice.

Results 1996-7: The rather species-poor sward was dominated by A. stolonifera, with patches of C. arvense, and subsidiary H. lanatus, Ranunculus repens and Potentilla anserina. Trends in the year following implementation of Tier 3 reflected either decline through weed control (Cirsium arvense and Senecio jacobaea) or spread due to more intensive defoliation.

A1.6.4Twyford

The site and water-management: On both banks of a canalised tributary of the Wensum between Guist and Twyford at ca. 27m A.O.D, soils were apparently like those at Tatterford, though the pattern was more complex, and peaty areas were local. In early 1996, the fields had been too wet for cattle, and were thus under-grazed. The sandier (drier) Tier 1 land had MG6a or MG11a, depending on past improvement. The wetter Tier 2 land had peatier soil, and a more species-rich vegetation, with some areas resembling MG9a and MG10b, and richer parts approaching M22 fen-meadow (with much Carex and Juncus species). In 1996, the fields were allocated to Tiers 1 and 2, and the ditches cleaned to aid water-movement. There were both extensive foot-drains, and springs draining high ground to the north. Such variation in water-regime influenced both the vegetation and allocation to ESA tiers.