This report has been prepared by Carter Ecological Limited , with all reasonable skill, care and diligence within the terms of the Contract with the client. We disclaim any responsibility to the client and others in respect of any matters outside the scope of the above. This report is confidential to the client and we accept no responsibility of whatsoever nature to third parties to whom this report, or any part thereof, is made known. Any such party relies on the report at their own risk. English Nature Contract Number 11 T4 00/01. English Nature is grateful for the assistance of the Environment Agency and the Medway and Swale Estuary Partnership in funding this report. 1/15 1 Introduction............................................................................................................ 3 1.1 Purpose of the report .......................................................................................... 3 1.2 Aims and objectives of the scoping survey ....................................................... 3 1.3 Survey approach ................................................................................................ 3 1.4 Areas surveyed................................................................................................... 4 2 Methods ................................................................................................................. 5 2.1 General............................................................................................................... 5 2.2 Variables recorded ............................................................................................. 6 2.2.1 Division across the sea-wall ...................................................................... 6 2.2.2 Salt Marsh .................................................................................................. 6 2.2.3 Structural features ...................................................................................... 6 2.2.4 Management .............................................................................................. 7 2.2.5 Plant species............................................................................................... 7 2.3 Data processing.................................................................................................. 8 2.3.1 Network and survey files ........................................................................... 8 3 Results ................................................................................................................... 9 3.1 GIS product........................................................................................................ 9 3.2 Species occurrence............................................................................................. 9 3.2.1 Species groupings ...................................................................................... 9 3.2.2 Widespread species .................................................................................. 10 3.2.3 Local species ............................................................................................ 10 3.2.4 Rare species ............................................................................................. 10 3.3 Species distribution in relation to habitat ........................................................ 11 3.4 Categorizing walls according to species richness ............................................ 11 4 Discussion and recommendations for further survey........................................... 13 4.1 Confirming the scoping survey data ................................................................ 13 4.2 Vegetation ........................................................................................................ 13 4.3 Habitat niches .................................................................................................. 14 4.3.1 Physical niche .......................................................................................... 14 4.3.2 Community niche..................................................................................... 14 2/15 1 INTRODUCTION 1.1 Purpose of the report This report provides background information on a botanical scoping survey of sea walls in the North Kent marshes, and discusses a recommended sampling strategy for the second stage of the project. The survey area comprised the sea walls of three SSSIs -The Medway Estuary and Marshes SSSI, The South Thames Estuary and Marshes SSSI, and The Swale SSSI. Fieldwork for the scoping survey was carried out by Carter Ecological Limited in August and September 2000 under English Nature contract 11T400/01. The principle output required was a report of the survey data as tables in MapInfo format, and this GIS product was supplied in final form to English Nature on 24/11/2000. The other output required was a written report on the recommended sampling strategy for the second stage of the project, and this is addressed here. This report also provides explanation of the methodology used in the scoping survey. 1.2 Aims and objectives of the scoping survey The scoping survey aimed to relate the occurrence of scarce plant species to readily observable aspects of the sea-wall environment using rapid surveying techniques. Specific objectives may be summarized as follows. 1.3 To map boundary features and sea-wall types. To map readily observable aspects of sea-wall vegetation and management. To map readily observable aspects of salt-marsh condition outside the sea-wall. To produce a preliminary map of the occurrence of scarce species characteristic of the sea-wall habitat on the basis of conspicuous colonies only (the project brief specified that time should not be spent on detailed searching for plants). Survey approach The survey was carried out by walking the sea-walls with a surveying wheel to provide extremely accurate information on location. A range of variables were selected for observation, including appropriate plant species, various aspects of sea wall structure and management, and various aspects of salt-marsh condition. Wherever the state of any variable altered (e.g. a scarce plant population ended, or mowing management started) the chainage was recorded from the surveying wheel, and the information was either entered directly into a hand-held computer or noted more conventionally on paper. The computer software for field data entry required the sea-walls to be divided into more or less straight sections up to 2-km long, so that positional information from the surveying wheel could be correctly interpreted. 3/15 1.4 Areas surveyed All sea walls in the three SSSIs were surveyed as follows. South Thames Estuary and Marshes SSSI, 28 sections totalling 33km and 68m. Medway Estuary and Marshes SSSI, 49 sections totalling 41km and 596m. Swale SSSI, 54 sections totalling 54km and 898m. The total length of wall surveyed was therefore 129km and 562m. Sea frontages within the SSSIs that were clearly not sea walls were omitted as follows. South Thames SSSI, frontage of Cliffe Fort and Jetties, frontage of Allhallows Caravan Site and Leisure Park, and frontage of Grain village. Medway SSSI, frontage of Kent Oil Refinery and east side of Colemouth Creek, frontage of Stoke Creek Wharf from Bee Ness Jetty to western boundary of Kingsnorth Power Station, frontage of Otterham Quay Wharf to Horsham Farm Wharf, the western side of the Barksore Peninsula from Lower Halstow to Barksore marshes, the southern Shore of Bedlams Bottom including Raspberry Hill, Chetney Moat and Chetney Hill, and the northern end of Rushenden Marshes. Swale SSSI, frontage of Ridham Dock, frontage of Kemsley Marshes and Landfill area, the Isle of Harty, and Elmley Hills. Most of these are frontages modified by man, but a few (including Chetney Hill, Isle of Harty and Elmley Hills) are natural frontages above sea level having intrinsic nature conservation interest. 4/15 2 METHODS 2.1 General Fieldwork was carried out between 30/8/2000 and 6/10/2000 by Andrew Henderson, Richard Carter, Andrew Cross and Elaine Dromey. Surveyors worked in pairs, though at the beginning of the survey all four surveyors worked together to develop and harmonize the methodology and set common standards for recording. Where methodology was adjusted on the basis of this experience the trial stretches of sea-wall were subsequently re-surveyed to the finally agreed standard. The survey covered the flood defence embankment (i.e. the sea-wall proper) and the habitats on either side of it, including the nearer part of the saltmarsh (if any) on the outer side and the berm on the inner side. The landward boundary was set at the seaward side of the borrow dyke or other boundary where there was no borrow dyke. The width of the belt surveyed accordingly varied, depending on the width and multiplicity of the various habitat features (e.g. trackways, second banks and berms). It seldom exceeded 50m. In addition the character of land adjacent to the survey belt on the landward side was recorded (e.g. pasture, arable). The sea walls were walked in the direction that placed the sea on the left (i.e. chainage increased in this direction). Besides avoiding possible confusion, this was necessitated by the computer software. Recording sections were delimited on pragmatic grounds including the following. A requirement for easily recognizable start and end points determined by recognizable physical features (allowing location on maps and facilitating future relocation of the sections). A software constraints setting the maximum permissible section length at around 2km. A software constraint disallowing more than a very few sharp bends in any one section (this would disrupt registration of data against base maps). Owing to seasonality there were disparities in the ease with which different species could be recorded. To find the less conspicuous species it was necessary to have one recorder walking along the crest of the wall, while another ranged across the inner face and berm, working in those places that appeared most promising for scarce plants. This mode of working was adopted throughout the survey. Reasonably comprehensive recording of the target species along the wall could be achieved, though it was impossible to examine the whole width closely in places where there were wide berms or multiple berms. The majority of data were collected on a Husky field computer using specially written software adapted by Data Collection Ltd from their highways surveying package. Additional ecological data were recorded on paper. The sections were measured using a Trumeter wheel with chainages recorded from accurately located start points. The wheel is accurate to approximately 1%. 5/15 2.2 Variables recorded 2.2.1 Division across the sea-wall For data recording purposes the survey belt was divided into six parallel strips as follows (from seaward side to landward side): the salt marsh; the outer face of the sea-wall; the crest of the sea-wall; the inner face of the sea-wall; the berm; and the edge of the borrow dyke. The landward habitat use constituted a seventh recording location. The strips were treated as recording locations numbered 1 to 7 from seaward to landward. At any given chainage (or over any chainage interval) variables could be recorded in any or all of these seven recording locations. 2.2.2 Salt Marsh Characteristics of the salt marsh abutting the sea wall were recorded on the field computer in recording location 1 as follows. 2.2.3 Grazing levels – high, low or none. Poaching levels – high, low or none. Eroding edge – present or absent. Spartina anglica – if present abundance was noted using the DAFOR scale. Structural features Structural features of the sea wall were noted as follows. Outer wall (location 2 only) – construction type (e.g. concrete, concrete block, other). Outer wall (location 2 only) – vegetation cover (high, low, none). Tracks (locations 3 and 5 only) – surface type (mud, gravel, asphalt etc.). Saline areas (location 5 only) – rutted or other. Landward boundary (location 7) – borrow dyke, fence or other. 6/15 2.2.4 Management Walls were unmanaged, mown or grazed. Areas were recorded as grazed or mown, not both. Management features of the sea wall were noted as follows. 2.2.5 Grazing (locations 3 to 5) – grass length to nearest 5cm. Grazing (locations 3 to 5) – poaching level (high, low, none). Mowing (locations 3 to 5) – grass length to nearest 5cm. Mowing (locations 3 to 5) – cuttings left or not. Weed control (locations 2 to 6) – if any, topping or chemical. Plant species Species were recorded as present or absent in 100m stretches of sea-wall. Precise chainages were noted only for rare species (e.g. Lactuca saligna). The following plant species were recorded wherever they were seen. Astragalus glycyphyllos; X Agropogon littoralis; Bupleurum tenuissimum; Carex divisa; Hordeum marinum; Inula crithmoides; Lactuca saligna; Lepidium latifolium; Petroselinum segetum; Peucedanum officinale; Polypogon monspeliensis; Puccinellia fasciculata; Puccinellia rupestris; Torilis nodosa; Trifolium squamosum; and Vicia bithynica. The survey was carried out at an optimal time for the following species: Astragalus glycyphyllos, Bupleurum tenuissimum, Hordeum marinum, Inula crithmoides, Lactuca saligna, Lepidium latifolium, Petroselinum segetum, Peucedanum officinale and Polypogon monspeliensis. There was no difficulty in spotting or recognizing these species, though some are not conspicuous at the best of times (especially Bupleurum tenuissimum). The sedge Carex divisa was less conspicuous than in June when the flowering heads tend to stand erect, but was otherwise intact. The remaining species were dead, but could mostly be recognized from dead remains (especially the characteristic fruiting heads of Torilis nodosa and Trifolium squamosum). There is however a danger that shattered material of these species was being overlooked. Species not encountered included the following. 7/15 Alopecurus bulbosus which is conspicuous only at flowering in June. It can be located at other times of the year, but could have been missed in rapid surveys since close searching of turf would be required. Trifolium ornithopodioides which grows on sea-wall paths and tracks, and is virtually indistinguishable from trampled material of Trifolium dubium once flowering is over. Chenopodium chenopodioides which could have been overlooked among depauperate plants of Chenopodium rubrum. 2.3 Data processing 2.3.1 Network and survey files For each section, the start point and survey line of the field survey were drawn onto OS digital maps using Fastmap GIS. The survey data were then laid over the centreline and linked to its chainage. This process allocated a grid reference to all data points. All data was collected according to centreline chainages, and displayed at the appropriate chainages but offset to the left or right. In addition, each of the habitat variables were offset to arbitrarily determined distances to enable clarity of display. The offset distances are uniform throughout the survey, despite the real variation in width of wall and berm. 8/15 3 RESULTS 3.1 GIS product The survey produced MAPINFO overlays of the species and habitat variables on maps of the sea-walls produced from OS line-data. Thus for any section of sea-wall it is possible to call up a visual display of the incidence of plant species present, and further to display this information in relation to the habitat variables recorded. This GIS tool is the principle output of the survey. It is lodged with the Kent office of English Nature to whom enquiries should be addressed. It allows interactive probing of the data set, and in this respect cannot be summarized in a written report. 3.2 Species occurrence 3.2.1 Species groupings Scarce species tend to occur in distinct parts of the sea-wall habitat complex as follows. General short grassland assemblage – short open grassland, often grazed and poached, or else at the edge of trampled areas, supports an assemblage of scarce species comprising Bupleurum tenuissimum, Torilis nodosa and Trifolium squamosum. In one location Vicia bithynica was present. Several modestly uncommon species not recorded in the scoping survey have their headquarters here including Lathyrus nissolia and Lotus glaber. Rutted saline areas – rutted saline areas on the berm support an assemblage of grasses including Hordeum marinum, Polypogon monspeliensis, Puccinellia fasciculata and Puccinellia rupestris. The very rare X Agropogon littoralis was found in this assemblage in one place. More frequently flooded sites with Polypogon monspeliensis may be the locus for Chenopodium chenopodioides not found in the scoping survey. Grassy berms – the only species generally confined to this habitat appears to be Carex divisa, though Bupleurum tenuissimum and Trifolium squamosum are often present as well. In one location Astragalus glycyphyllos was present. This may be the locus for Alopecurus bulbosus not found in the scoping survey. Outer faces of walls – The swash-bank at the top of walls and cracks in concrete facings generally support few scarce species, though a number of the more uncommon species occur mainly in this habitat, including Lactuca saligna and Lepidium latifolium. Species found in other sea wall habitat components may occur in small quantity, especially Hordeum marinum and Torilis nodosa. Several modestly uncommon species not recorded in the scoping survey have their headquarters here including Anthriscus caucalis and Carduus tenuiflorus. Paths on wall crests – trampled paths on the crest of the wall often support Puccinellia fasciculata and Puccinellia rupestris, while the edges of such paths 9/15 often support Torilis nodosa and Trifolium squamosum. This is also likely to be the locus for Trifolium ornithopodioides not recorded in the scoping survey. 3.2.2 Salt marsh margins – the bottom of the wall is the main locus for Inula crithmoides which may extend a short distance onto the salt-marsh where elevations are high. Miscellaneous – other rarities including Petroselinum segetum and Peucedanum officinale occur rather indiscriminately across habitat components in the few places where they occur. Widespread species The following species were widespread throughout the three SSSIs – Bupleurum tenuissimum, Carex divisa, Hordeum marinum, Inula crithmoides, Puccinellia fasciculata, Puccinellia rupestris, Torilis nodosa and Trifolium squamosum. 3.2.3 Local species Polypogon monspeliensis was locally common in disturbed saline ground (and to a limited extent at the edge of adjacent arable fields) east of Gravesend, on the Barksore Marshes, and especially on the Chetney Marshes. Petroselinum segetum was encountered in very few places. It was not anticipated that this species would be so uncommon. 3.2.4 Rare species X Agropogon littoralis was encountered only on the Barksore Marshes. It was not found near Kingsnorth where there are old records, though it could easily have been missed in a rapid survey. Astragalus glycyphyllos was recorded from mown grass on the berm on the eastern side of the Motney Hill peninsula. Lactuca saligna is a Red Data Book species that in recent years has been known only from two sites at Fobbing in Essex (sea-walls) and Rye Harbour in Sussex (semiruderal habitats). Unconfirmed rumour has it that the plant is no longer to be found at Rye Harbour, so this species is precarious in the extreme. It is reported that in summer 2000 plants were recorded by Francis Rose from the Isle of Grain. In our survey two populations, one large and one small, were recorded in sites over a kilometre apart on the Isle of Grain, and a third population was located east of Kingsferry Bridge on the Isle of Sheppey. This brings the count of UK populations from one or two to four of five (depending on whether the Rye Harbour population is extant). In this survey the plant was mainly found on the outer face of the sea-wall. Lepidium latifolium is not generally rare, but was only encountered once in our survey on the outer face of the sea-wall on the eastern side of the Motney Hill peninsula. 10/15 A single patch of Vicia bithynica was found among a sward consisting almost entirely of annual legumes on the inner face of the sea-wall west of Seasalter. 3.3 Species distribution in relation to habitat The data broadly support the impression gained by surveyors in the field that the scarce species of sea walls are less frequent in two situations as follows. Recently reconstructed walls, where recolonization appears to be slow even where there are short open swards maintained by grazing. Completely unmanaged walls, where swards of the tall grasses Arrhenatherum elatius and Elytrigia repens exclude most other species. A factor in the case of reconstructed walls may be the lack of saline areas on the berm behind the wall (since topover and seepage are reduced) – this excludes the complement of species associated with the Hordeum marinum community. Beyond this definite patterns of occurrence in relation to observable current management are obscure. This may be because factors such as grazing are routinely varied from year to year according to the availability of grass and other accidents of the agricultural economy. It may also relate to past periods of neglect or adverse management leading to floristic impoverishment in some areas that appear suitable, while elsewhere plants may persist in areas where management is currently adverse. All of this may make simple relationships between species occurrence and current management hard to discern. 3.4 Categorizing walls according to species richness The tables in Appendices show the occurrence of species in 100m stretches of sea wall throughout the areas surveyed. Walls may be categorized according to speciesrichness in various ways as follows. The total number of species in each 100m stretch can be found, and parts of the sea walls with high numbers of species in consecutive 100m stretches can be identified by inspection. The same approach may be refined by using the species groupings presented in Section 3.2.1 to identify parts of the sea walls with high numbers of species in particular habitat components, e.g. picking out areas that are important for rutted berm species where short turf species are lacking, so that the areas do not stand out on the basis of total numbers of species. A rarity-weighted index may be computed for a section of sea wall using the formula ‘number of records for species (all species or those in a specified species group as in Section 3.2.1) divided by the number of 100m sections’. On this basis the following wall sections (see Appendices) are exceptionally speciesrich across a range of species groups, often with five to seven scarce species per 100m stretch: Elmley 03, Yantlet 05, Blacketts 01, Allhallows 01 and Allhallows 02. The 11/15 following are only marginally less species-rich: Yantlet 02, Horsham 02 and Murston 02. The following are moderately species rich across a range of species groups, often with three to five scarce species per 100m stretch : Chetney 04, Minster 01, Yantlet 01, Yantlet 04, Spitend 03, Spitend 04, Horsham 01, Conyer 01 (part) and Teynham 02. Other wall sections may be important for particular species groups. Some of the rarest species occur in the less species-rich sections, e.g. large Polypogon monspeliensis populations near Gravesend and on Chetney Marshes, Peucedanum officinale on Faversham Creek, X Agropogon littoralis on Barksore Marshes, Vicia bithynica at Seasalter, and both Astragalus glycyphyllos and Lepidium latifolium on the Motney Hill peninsula. 12/15 4 DISCUSSION AND RECOMMENDATIONS FOR FURTHER SURVEY 4.1 Confirming the scoping survey data A first priority of the second stage of the survey should be to estimate how complete the species information from the scoping survey might be. To this end some areas should be surveyed closely, allowing estimation of the extent to which species may have been overlooked in the rapid scoping survey. Recommendations for technique include the following: Stretches of sea-wall selected should contain sufficient 100m stretches for adequate statistical comparison – 20 at least, perhaps 30, i.e. 2 or 3km stretches of wall. A sample of stretches should be selected on a stratified random basis, so that different sea-wall types and management types are equally represented in the sample. Sample stratification should also ensure that species-poor and species-rich sections of wall are included in the sample. Within selected areas, an exhaustive search for scarce species should be made, using smaller recording sections than in the scoping survey – perhaps 25m. A fuller description of the habitats present should also be attempted (see Section 4.3.1). In addition to recording data from a stratified random sample of walls, it would also be advisable to collect similar data from the places that were in fact richest, according to the findings of the scoping survey. There would be some advantage in carrying out the second stage surveys earlier in the year than the scoping surveys. Surveys conducted during June and July would be better for several species, especially the widespread species Puccinellia fasciculata, Puccinellia rupestris, Torilis nodosa and Trifolium squamosum, and worse for very few (though Bupleurum tenuissimum would be less conspicuous). There might be some loss of direct comparability with the scoping survey results, but there would be large gains in respect of the first priority of estimating how complete the species information from the scoping survey might be. However, if for any reason it were necessary to conduct the second stage surveys later it would not be a major problem. 4.2 Vegetation Vegetation may be a better indicator of where scarce plants are likely to occur than management type, even though management may ultimately control the vegetation. A possible objective of the second stage could therefore be to identify the vegetation types in which scarce species occur. It might equally be of value to know what types they do not occur in. It may well be easier to determine how to manage vegetation types that provide the locus for scarce species, than to determine what the species require individually. 13/15 Unfortunately sea wall vegetation is a recognized gap in the coverage of the National Vegetation Classification (Rodwell et al. 2000). It is probably beyond the scope of a study on these SSSIs to remedy this. It may however be possible to erect an ad hoc classification, perhaps identifying types by means of indicator species and physiognomy, e.g. tall MG1 Arrhenatherum elatius grassland with Anthriscus sylvestris , ditto without Anthriscus sylvestris, swash-bank community with Anthriscus caucalis etc. This would give an initial indication of where the scarce species occur in phytosociological terms. It is to be hoped that JNCC will urgently support a project to provide sea-wall relevees for consideration in relation to the National Vegetation Classification. The findings of the scoping survey could assist in providing justification for such a survey. 4.3 Habitat niches 4.3.1 Physical niche Another objective of the second stage could be to identify more closely the precise niche occupied by each scarce species in the sea-wall habitat complex, and to identify localities of importance for each species. This may in the first instance be possible from in-depth analysis of the scoping survey data, provided that further work suggests that species were not extensively overlooked. Beyond this it will be necessary to devise improved ways of characterizing the habitat conditions on the basis of experience from the scoping survey. Matters that especially need to be addressed include the following. 4.3.2 Characterization of grazing and mowing regimes which cannot always be inferred from a site visit. Consultation with site managers is required, and better ways of describing the conditions at the time of survey are also required – perhaps more objective indices of poaching and dunging and rutting. A wider range of wall types and structures could be recorded in more detailed surveys, e.g. double berms, unusual facing types, scrub. Community niche For the scarcest species it would be worth identifying the community type by recording species cover in quadrats targetted on the locations containing the scarce species. Appendices: Appendix 1 Excel spreadsheet of sections surveyed. 1. South Thames 14/15 2. Medway 3. Swale Appendix 2 Excel spreadsheet of species distribution. 1. South Thames 2. Medway 3. Swale 15/15