Grazing Marsh for Birds and Livestock: Development of Management Guidelines

BD1317 Technical Appendices

BD1317

Grazing Marsh for Birds and Livestock: Development of Management Guidelines

Technical Appendices to the Final Report

CONTENTS

Appendix I Modelling the relationships between field selection by 1

Breeding birds on grazing marshes and the grazing

marsh habitat in different areas

Appendix II Effects of grazing management on the sward structure of 23

grazing marsh in the Norfolk Broads Environmentally

Sensitive Area

Appendix III Development of birdhabitat models as an aid to targeting 49

measures for the rehabilitation of grazing marsh (BGS

conference paper)

Appendix IV Habitat models of bird species distribution on coastal grazing 53

marshes: a multi-site comparison (AAB conference paper)

Appendix V Predictions of lapwing (Vanellus vanellus) habitat occupancy 63

using rule-based models (IALE conference paper)

i

BD1317 Technical Appendix I 1

APPENDIX I

i

BD1317 Technical Appendix I 13

i

BD1317 Technical Appendix I 13

MODELLING THE RELATIONSHIPS BETWEEN FIELD SELECTION BY

BREEDING BIRDS ON GRAZING MARSHES AND THE GRAZING MARSH HABITAT IN DIFFERENT AREAS

1. INTRODUCTION

An empirically based modelling approach has been used to identify common and site-specific bird-habitat relationships on grazing marshes and to consider the implications for the formulation and targeting of conservation management prescriptions for ground-nesting birds on grazing marshes. This work has been carried out in support of the UK Biodiversity Action Plan for Coastal and Floodplain Grazing Marsh (Anon., 1995).

Models of field selection by groups and individual species of ground-nesting birds and the grazing marsh habitat on the North Kent Marshes were constructed in a previous DEFRA-funded project (BD1405). In this project, the models from North Kent have been used to predict field selection by these birds in three geographically separate areas of grazing marsh in England.

This appendix describes the model testing procedure and a meta-analysis of the models from the North Kent Marshes and site-specific models from the test areas. The results of these two approaches have been used to distinguish between bird-habitat relationships that are common to several areas and those that are site-specific. The implications for the preparation and application of conservation management prescriptions are discussed.

2. METHODS

2.1 Study areas

Three study areas were chosen through consultation with JNCC and English Nature (Table 1). The initial list comprised coastal grazing marshes at Halvergate in Norfolk Broadland and in Northwest England, and floodplain grazing marshes in the Lower Derwent Valley in Yorkshire. However, an inspection of possible sites on the Solway Firth, Morecambe Bay and Ribble Estuary revealed that the coastal grazing marshes there were small, highly fragmented and few in number. Moreover, many of those on the Ribble had been converted to arable land. As the sample of sites in NW England was too small and too uniform for bird-habitat modelling, an alternative was sought. The Arun Valley, in West Sussex, was finally chosen. When combined with the study area on the North Kent Marshes, the sample of study areas in the present project comprised a good balance of inland and coastal sites. Their geographical distribution complemented that of the study areas selected by CEH and ADAS for the DEFRA funded projects BD1309 and BD1301 (Caldow et al., 1999). With the exception of Higham Holmes and Belaugh in the Norfolk Broads, the CEH and ADAS study areas were situated in south-western England, on the Somerset Levels and Moors, Upper Thames Tributaries and near Wareham, Dorset.

Table 1. Study areas
Study area / Number of land-holdings / Number of marshes/fields / Area (ha)
Halvergate Marshes, Norfolk Broadland / 7 / 180 / 800
Lower Derwent Valley, Yorkshire / 37 / 152 / 657
Arun Valley,
West Sussex / 9 / 204 / 645

A detailed account of the Halvergate Marshes is given by George (1992). The Lower Derwent Valley and parts of the other study areas, Berney Arms on Halvergate and Pulborough Brooks in the Arun Valley, are described in the case studies presented in the Wet Grassland Guide (Treweek et al., 1997).

2.2 Sampling

As the aim was to describe the range of conditions accepted by ground-nesting species, as well as the optimum combination, each study area comprised a nature reserve, landholdings under Environmentally Sensitive Area, Countryside Stewardship or English Nature agreements and, where possible, non-agreement land. Selection of land-holdings and individual fields/marshes followed the rationale used in the North Kent Marshes study (Milsom et al., 2000).

At Halvergate and the Arun Valley, study areas were selected by landholding and all marshes within a landholding were then sampled. However, sampling was not grouped by landholding in the Lower Derwent Valley because the pattern of land tenure there was very different to that at other study areas. Many landowners only owned a single field or had mowing strips across part of a field. The number of landholdings was, therefore, much higher in the Lower Derwent Valley than in the other study areas (Table 1).

Fieldwork was carried out at Halvergate in 1998 and 1999, in the Lower Derwent Valley in 1999 and 2000, and in the Arun Valley in 2000 only. A second year of fieldwork had been planned for the Arun in 2001 but it was cancelled due to Foot and Mouth Disease restrictions.

Table 2. Habitat measurements
Variable /

Measurement details

SWARD / Mean sward height (cm)
TUSS / 0 = no tussocks; 1 = tussocks sparse (<5% of ground cover); 2= tussocks occasional (5-15% cover); 3 = tussocks frequent (15-35% cover); 4= tussocks abundant (>35% cover); 5 = tussocks localised; absent or sparse in most of field but locally frequent or abundant.
TUSSIZE / 0= no tussocks; 1 = small tussocks (5-10cm above surrounding sward); 2 = medium tussocks (10-20cm above sward); 3=large tussocks (>20cm above sward)
HOLLOWS / 0=none; 1=few; 2=many; 3=abundant
RILLS / Density of rills per field/marsh (m ha-1)
WETMUD / Density of rills that contain floodwater or wet mud per field/marsh (m ha-1)
DRAINS / Density of footdrains per field/marsh (m ha-1)
WET / 0=surface of all hollows/rills dry and hard; 1=all dry on top but some soft; 2=many dry on top, some moist; 3= many moist; 4= many wet but no standing water; 5= many wet, some standing water; 6=many wet, most with standing water.
FLOODS / 0=no floodwater; trace=1-4% by area; 1=5-15%; 2=15-25%; 3= 25-35%; 4=35-45%; 5=45-55%; 6=55-65%; 7=65-75%; 8=65-75%; 9=85-95%.
AREA / Area of marsh/field to nearest 0.1 ha
BOUND / Index of degree of field enclosure; scale varies from 1=open boundary to 6=field/marsh completely enclosed by hedgerows, trees or walls >2m high.
POWER / Presence/absence of National Grid pylons in, or powerlines over, each field/marsh
ARABLE / Presence/absence of arable field on boundary of field/marsh
FPATH / Presence/absence of public footpath in field/marsh
ROAD1 / Shortest distance (m) from field centre to nearest public road
ROAD2 / Shortest distance (m) from field centre to nearest A class road or motorway
ROAD3 / Length of route (m) from field centre via access points on field boundaries to nearest public road
Further details of measurements are given in Milsom et al. (2000).

2.3 Habitat measurements

Each marsh or field was visited to make habitat measurements. Details of the variables and measurement methods are shown in Table 2.

Visits were made in April, May and June of each year to estimate mean sward height (cm) (SWARD), to record frequency and size of tussocks (TUSS, TUSSIZE), and to measure surface wetness of rills or footdrains (WET), or percentage cover of flood water (FLOODS). The presence or absence of National Grid power lines (POWER), the presence or absence of arable crops in adjacent fields (ARABLE), and the frequency of surface features (HOLLOWS) and boundary habitats were recorded in the field. An index of the degree of marsh enclosure by hedges, trees or man-made structures (BOUND) was calculated using the method described in Milsom et al. (2000). Lengths of rills and footdrains in each marsh were measured from aerial photographs and converted to densities (m ha-1) of wet and muddy rills (WETMUD), all rills (RILLS) and footdrains (DRAINS). Areas (ha) of individual marshes (AREA) were taken from farm maps. Distances (m) of each marsh to roads (ROAD1, ROAD2, ROAD3) and presence or absence of footpaths in each marsh (FPATH), were measured from the latest revisions of 1:25000 scale Ordnance Survey maps. Further information is given in Milsom et al. (2000).

Supplementary data on the species composition of the field vegetation were collected in the Derwent and Arun Valleys in 2000. The presence or absence of five key species/genera (Glyceria maxima, Juncus spp., Carex spp., Phalaris arundinacea, Caltha palustris) were recorded from each of 20 1m ´ 1m quadrats which were evenly spaced along the diagonal of each field. These frequency data were not used directly. Correspondence analyses (Digby & Kempton, 1987) were carried out on data from each study area to identify the major axes of variation in species composition between fields. The scores from each of the four site axes (VEG1 …VEG4) were included in the variable lists for the Derwent and Arun models.

2.4 Bird counts

Counts of ground-nesting birds in each marsh or field were made using a modified field-by-field method (Milsom et al., 2000). They were carried out in April and repeated in May and June of each year.

The assemblages of ground-nesting bird species on grazing marshes at Halvergate were similar to those on the North Kent Marshes (Table 23). Skylark, meadow pipit, lapwing and redshank were the most widespread species in both areas. However, there were some differences between areas. Halvergate supported fewer species than the North Kent marshes, and the very scarce breeders in North Kent Marshes study area, black-tailed godwit, ruff and short-eared owl, were absent from Halvergate. Meadow pipits and yellow wagtails occupied proportionately fewer fields at Halvergate than in the North Kent Marshes whereas shelduck, shoveler and oystercatcher occupied proportionately more fields at Halvergate than in North Kent.

The four most widespread species in the coastal study areas - lapwing, redshank, skylark and meadow pipit - were also widespread in the inland study areas (Table 34). However, the species assemblages in the inland study areas differed from those at the coast in several respects. Ten species, including yellow wagtail, were not recorded from either of the inland study areas whereas curlew, spotted crake and black-headed gull occurred only in the Lower Derwent Valley. Snipe were widespread in the Lower Derwent Valley but were very scarce in the other areas.

BD1317 Technical Appendix I 13

Table 3. Distribution of ground-nesting birds on coastal grazing marshes in the North Kent Marshes and Broads ESAs. Conservation status of species Red (R) or Amber (A) listing for breeding and UK Biodiversity Action Plan (B) shown.
Species / Status / North Kent Marshes ESA / Halvergate, Broads ESA
1995
(n=429) / 1996
(n=433) / 1998
(n=174) / 1999
(n=177)
Fields: n (%) occupied / Pairs per marsh: min & max / Fields: n (%) occupied / Pairs per marsh: min & max / Fields: n (%) occupied / Pairs per marsh: min & max / Fields: n (%) occupied / Pairs per marsh: min & max
Mute Swan / Cygnus olor / 0 (0) / - / 2 (0.5) / 0-1 / 4(2..3) / 0-1 / 6 (3.4) / 0-1
Greylag Goose / Anser anser / 3 (0.7) / 0-1 / 5 (1.2) / 0-3 / 2(1.2) / 0-3 / 0(0) / -
Canada Goose / Branta canadensis / 31 (7.2) / 0-13 / 33 (7.6) / 0-7 / 7 (4.0) / 0-2 / 4 (2.3) / 0-1
Shelduck / Tadorna tadorna / A / 13 (3) / 0-3 / 37 (8.6) / 0-15 / 33 (19.0) / 0-13 / 32 (18.1) / 0-7
Gadwall / Anas strepera / 2 (0.4) / 0-2 / 7 (1.6) / 0-2 / 1 (0.6) / 0-2 / 1 (0.6) / 0-1
Teal / Anas crecca / 0 (0) / - / 3 (0.7) / 0-1 / 2(1.2) / 0-4 / 2 (1.1) / 0-1
Mallard / Anas platyrhynchos / 49 (11.4) / 0-12 / 59 (13.6) / 0-8 / 40 (22.9) / 0-10 / 41 (23.2) / 0-2
Garganey / Anas querquedula / A / 1 (0.2) / 0-1 / 0 (0) / - / 0(0) / - / 0(0) / -
Shoveler / Anas clypeata / 20 (4.7) / 0-6 / 12 (2.8) / 0-6 / 14 (8.1) / 0-3 / 15 (8..5) / 0-2
Pochard / Aythya ferina / 2 (0.5) / 0-3 / 5 (1.2) / 0-3 / 1(0.6) / 0-1 / 3 (1.7) / 0-2
Tufted Duck / Aythya fuligula / 9 (2.1) / 0-7 / 11 (2.5) / 0-9 / 0(0) / - / 3 (1.7) / 0-1
Grey Partridge / Perdix perdix / R, B / 40 (9.3) / 0-1 / 44 (10.2) / 0-2 / 0(0) / - / 2 (1.1) / 0-1
Oystercatcher / Haematopus ostralegus / 66 (15.4) / 0-8 / 69 (15.9) / 0-14 / 48 (27.6) / 0- / 39 (22.0) / 0-2
Avocet / Recuvirostra avosetta / 3 (0.7) / 0-1 / 4 (0.9) / 0-52 / 3 (1.7) / 0-6 / 8 (4.5) / 0-4
Ringed Plover / Charadrius hiatacula / 15 (3.5) / 0-4 / 18 (4.2) / 0-6 / 1 (0.6) / 0-1 / 0(0) / -
Lapwing / Vanellus vanellus / 165 (38.5) / 0-25 / 165 (38.1) / 0-25 / 65 (37.4) / 0-12 / 63 (35.6) / 0-8
Ruff / Philomachus pugnax / A / 1 (0.2) / 0-2 / 0 (0) / - / 0(0) / - / 0 (0) / -
Snipe / Gallinago gallinago / A / 8 (1.9) / 0-1 / 8 (1.8) / 0-1 / 1 (0.6) / 0-1 / 1(0.6) / 0-1
Black-tailed Godwit / Limosa l. limosa / R / 6 (1.4) / 0-2 / 4 (0.9) / 0-2 / 0 (0) / - / 0 (0) / -
Redshank / Tringa totanus / 119 (27.7) / 0-18 / 145 (33.5) / 0-28 / 35 (20.1) / 0-4 / 50 (28.2) / 0-6
Short-eared Owl / Asio flammeus / 1 (0.2) / 0-1 / 0 (0) / - / 0(0) / - / 0(0) / -
Skylark / Alauda arvensis / R, B / 335 (78.1) / 0-23 / 312 (72.1) / 0-19 / 128 (73.6) / 0-5 / 111(62.7) / 0-5
Meadow Pipit / Anthus pratensis / 269 (62.7) / 0-13 / 264 (61) / 0-12 / 62 (35.6) / 0-3 / 68 (38.4) / 0-6
Yellow Wagtail / Motacilla flava sp. / 64 (14.9) / 0-10 / 71 (16.4) / 0-10 / 14 (8.1) / 0-2 / 5(2.8) / 0-2
Reed Bunting /

Emberiza schoeniclus

/ R, B / 22 (5.1) / 0-3 / 11 (2.5) / 0-3 / 0(0) / - / 0(0) / -
Corn Bunting / Miliaria calandra / R, B / 15 (3.5) / 0-3 / 31 (7.2) / 0-2 / 0(0) / - / 0(0) / -
Notes: Sources: Gibbons et al. (1996), Anon (1995), UK Biodiversity Group (1998)

BD1317 Technical Appendix I 13