19 M22 ( subnodulosus– Cirsium palustre) fen meadow

19.1 Context

M22 is present in a number of SSSIs, but is not usually the main designation feature. In some regions where fens are scarce, such as the South Midlands, most sites with M22 have not been designated as SSSIs, though it sometimes occurs in SSSIs designated for other features (such as Pilch Fields, Buckinghamshire). The community has apparently been used as a basis for SAC habitat designation under the category ‘calcareous fens with Cladium mariscus and of the CARICION DAVALLIANAE’ (see Table 3.1and Table 3.3), but this is exceptional and has dubious legitimacy.

19.1.1 Concept and status

The vegetation encompassed by M22 was recognised and described first by Wheeler (1980c) as ‘rich-fen meadow’, and subsequently became incorporated into the National Vegetation Classification as M22 (Rodwell, 1991b). One reason why Wheeler referred to it as ‘rich-fen meadow’ rather than as a formal syntaxon was because it was a particularly variable unit which was difficult to define and characterise. It was also difficult to identify clear subdivisions within the community, and repeat analyses indicated several alternative subdivision solutions, none of which was entirely satisfactory. The variability of the unit doubtless reflects the range of edaphic and topographical circumstances in which this type of vegetation occurs, as well as the vagaries of individual management regimes, both past and present. This is coupled with effects of fragmentation and the identity of adjoining vegetation types, and is doubtless enhanced by the large number of examples of the community. Despite these caveats, the unit forms, as Rodwell (1991b) recognised, a broadly coherent unit referable to the Calthion alliance. Nonetheless, the variability of the unit means that specification of environmental thresholds is particularly difficult. This is manifest in the WETMEC analyses which show that the community can be equally at home on consistently wet spring mounds and in intermittent seepages. Moreover, its relationship to water regimes appears to be partly conditional on the nutrient status of the substratum: examples of M22 can be found in intermittent seepages across much of the observed fertility range, but it usually occurs in permanent seepages only in mesotrophic or eutrophic conditions. Whereas versions that occur in drier conditions are often floristically distinct from those in permanent seepages, these differences do not necessarily correspond to the sub-communities of M22 recognised by Rodwell (1991b). A more exact assessment of environmental thresholds may therefore require a more precise intra-community classification, though experience suggests it may be difficult to arrive at a universally acceptable solution. Stands of M22 can be transitional to those of other communities, both in concept and in the field, in time as well as in space. Spatial relationships between tall herb derivatives from M22 often appear abrupt and clear in the field, because of management boundaries, but this can mask a more gradual temporal change in species composition with plenty of overlap between M22 and tall herb communities, especially S25. M22 also shows some overlap with M13, and the status of the transitional Juncus–Carex lepidocarpa nodum identified by Wheeler (1980c) is discussed under M13 (see 15.1.1). However, in general the overlap is not great. A much bigger problem, as recognised by

Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales 607

Rodwell (1991b), is the relationship between M22 and M24, as many transitional examples occur (Figure 15.2).

19.1.2 Floristic composition

M22 fen meadow is typically dominated by sedges and rushes of medium height. Juncus subnodulosus is the most characteristic rush, but it is not always present, and in some sites Juncus acutiflorus and, occasionally, J. inflexus may predominate. Carex acutiformis and C. disticha are particularly characteristic sedges and on occasion can be strongly dominant. Although distinctive physiognomically, this vegetation type is not easy to define because of its floristic variety and lack of good positive characterisation. The species that are particularly distinctive are essentially (wet) meadow ; these not only occur in wet meadows but many, such as Juncus subnodulosus, Cirsium palustre, Filipendula ulmaria, Lotus uliginosus, Calliergon cuspidatum, can also be found in examples of M13 (though usually with lower frequency and constancy than in M22) and other communities. Juncus subnodulosus, a frequent dominant of M22, can also be dominant in M13, but whilst other M22 dominants such as Carex acutiformis and C. disticha can occur in M13, they are not usually as dominants. The community is variable, but can be very species-rich (mean of 26, range 3–66 spp per sample) (Table 19.1). A total of 31 rare species were recorded from M22 stands, but the mean number of rare species per sample is less than one. Rodwell (1991b) recognises four sub-communities of M22: typical sub-community (M22a), Briza media–Trifolium spp. sub-community (M22b), Carex elata sub-community (M22c), Iris pseudacorus sub-community (M22d).

Table 19.1 Number of species recorded in stands of M22 Total Mean SE Minimum Maximum All species (spp 4 m–2) 403 25.8 0.07 3 66 Mire species (spp 4 m–2) 152 14.9 0.06 2 46 Rare mire species (spp 4 m–2) 31 0.7 0.04 0 10

* These include: Blysmus compressus, Calamagrostis canescens, Calliergon giganteum, Campylium elodes, Carex acuta, Carex appropinquata, Carex diandra, Carex elata, Carex lasiocarpa, Carex viridula ssp viridula, Cladium mariscus, Dactylorhiza praetermissa, Dactylorhiza traunsteineri, Eleocharis uniglumis, Epipactis palustris, Erica ciliaris, Eriophorum latifolium, Hypericum undulatum, Juncus alpinoarticulatus, Lathyrus palustris, Oenanthe lachenalii, Osmunda regalis, Peucedanum palustre, Philonotis calcarea, Plagiomnium elatum, , Ranunculus lingua, Sphagnum teres, Stellaria palustris, Thalictrum flavum, Thelypteris palustris.

19.1.3 Distribution

The distribution of the community in England and Wales is shown in Figure 19.1. M22 has been recorded from 331 sites and is the most widespread community of base-rich fens in England and Wales. In some regions (such as much of the South Midlands) it provides the only real representative of herbaceous fen, but it is also widespread in East Anglia, often as a derivative of other, more distinctive, vegetation types. Its main distribution is in Central and Eastern England, but this is probably due to the presence of suitable substratum conditions (wet, base-rich, mesotrophic soils) rather than a direct influence of climate (Rodwell, 1991b). Similar vegetation occurs in some base-rich fen meadows in parts of Scotland (such as Ardblair and Myreside SSSI (Perth and Kinross)), but usually with neither Juncus subnodulosus nor Carex acutiformis and, because of the sparsity of data for comparable

608 Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales

vegetation in Scotland, it is not clear if this is best considered a northern variant of M22, or another community (perhaps M26).

(data from FenBASE database)

Figure 19.1 Distribution of M22 in England and Wales

19.1.4 Landscape situation and topography

M22 stands are particularly a feature of lowland valleyhead fens, although this has less to do with the hydrological characteristics of these systems than the fact that many of them are (or, until recently, were) grazed, whereas this is less true of many topogenous systems. When grazed, base-rich floodplain sites can support M22 vegetation (such as Burgh Common (Norfolk), Woodwalton Fen (Cambridgeshire)); if grazed (or annually mown litter) fens were more widespread in floodplain systems, M22 would be more widespread within them. The community also occurs in some partly drained grazing levels. The majority of stands occupy more or less flat situations or hollows, but a large number occur on seepage slopes (of varying grades of steepness) and in certain (base-rich mesotrophic-eutrophic) circumstances, they can cover spring mounds.

19.1.5 Substratum

M22 stands can occur on very shallow peaty soils, sometimes organic gleys, but also on deep (more than 1.5 m) peats in floodplains or basins. About 75 per cent of the stands were recorded in valleyheads, and these typically have very shallow peat (less than 0.5 m). Some 20 per cent of samples were recorded from floodplains, with a mean peat depth of 1.47 m. Only eight per cent occupied peat deposits deeper than 1.5 m, all in basins or floodplains. The substratum and irrigating water are typically of high (circumneutral) pH, though there are examples of lower pH in some floodplain peats, especially where these have been partly drained or are near the upland margins of some sites (values between 4.5 and 5.0 were recorded along part of the margin of Catfield Fen, Norfolk). Lower pH values are also found in the few sites associated with less base-rich bedrocks (5.5 on Lower Greensand, such as

Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales 609

Nares Gladley Marsh, Bedfordshire). There is considerable variation in the fertility of the deposits, but the majority are clearly mesotrophic. M22 stands can be associated with a wide variety of bedrocks. Many of these are obviously strongly calcareous (Chalk, Jurassic and Carboniferous Limestone), but the community is also sometimes associated with other types, such as Old Red Sandstone (Pont y Spig Monmouth), Upper Greensand (such as Stowell Meadow, Somerset) and Lower Greensand. Examples on the Lower Greensand tend to be more acidic than others, are often dominated by Juncus acutiflorus and represent the base-poor extreme of M22 (which is perhaps transitional to M23). Many examples of M22 are located upon various superficial deposits, particularly glacial sand and gravel (such as Cors Hirdre, Caernarfon), sometimes expressed as sandy lenses within Boulder Clay (such as Clack Fen, Buckinghamshire) and may have little interaction with the bedrock. A few are located over non-sedimentary bedrocks.

19.1.6 Zonation and succession

Numerous examples of M22 do not display clear zonations with other mire communities. In some cases they occur as small fragments, in others they occupy entire (usually valleyhead) sites and their limits are bounded by transitions into drier ground or watercourses. In some instances (such as where M22 covers seepage slopes), their transition into drier habitats is often abrupt and determined by the topographical disposition of the site and controls upon the emergence of groundwater. In some floodplain locations, M22 can occupy large areas and in some instances entire compartments, bounded by dykes, are covered by the community. Such expanses of M22 are not necessarily uniform, but floristic variation within them is expressed in terms of different versions of M22 rather than different communities (often because of the strong selective pressures imposed by grazing management). In many instances M22 occurs in juxtaposition with other mire communities. M22 is essentially maintained by grazing or regular mowing and where these occur differentially, the community may adjoin dereliction derivatives such as S24 or S25. The boundary between M22 and other communities may be abrupt (for example, along the line of a fence). The community also occurs in more natural zonations. In some seepage systems it forms a zone flanking the main seepage communities (such as M13), in conditions that may or may not be drier but which are often more fertile (when the main seepage is also quite fertile, the whole system tends to be blanketed by forms of M22). Some stands of M22 contain a number of typical Molinion species, and these may grade out into examples of M24 in drier conditions. However, other examples of M22 can be as dry as examples of M24, and the consistent difference between these two communities is that M22 is more fertile than M24. M22 frequently forms a zone in wet hollows, surrounding wetter forms of fen or swamp and grading out into wet or dry grassland, as is seen clearly in some of the West Norfolk pingo fields. In many instances, M22 is not obviously part of the terrestrialisation sequence of the hollows, but occurs on shallow peat or mineral ground around them. Nonetheless, examples of M22 do occur on surfaces which have originated by terrestrialisation, but the community mainly occurs as a grazing-maintained secondary feature (plagioclimax), derived by scrub clearance and encouraged by partial drainage. This seems to be the status of M22 in the topogenous basin at Great Cressingham Fen, where the natural herbaceous vegetation appears to have been a form of M9. Likewise, examples on floodplains may be a product of scrub clearance or of grazing of tall herb fen (S24, S25), again often – but not always – enhanced by drainage. A corollary is that M22 can disappear as a result of dereliction, though the process can be slow and is not always complete: in a number of locations in Broadland, patches of strong Juncus subnodulosus-dominance within S24 are probably the relicts of former M22 litter fens, where mowing seems to have been abandoned well over fifty years ago. Lambert (1948) observed in the Yare valley that replacement of former litter fen by tall herb fen as a consequence of dereliction occurred most rapidly alongside the dykes

610 Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales

and least rapidly in the centres of compartments. M22 has now virtually disappeared from unmanaged examples of these mires, but at Wheatfen small patches of Juncus subnodulosus dominance still persist in locations distant from the dykes.

19.2 Water supply mechanisms and conceptual model

Samples of M22 vegetation were recorded from a wide range of WETMECs: from 5 through to 17. Most are from areas with permanent or intermittent seepages or where groundwater tables are shallowly sub-surface year round, sometimes peripheral to permanent seepages: 30 per cent were from WETMEC 11 (Intermittent and Part-Drained Seepages, such as Booton Common (Norfolk), Cors Hirdre (Caernarfon)) and 22 per cent from WETMEC 10 (Permanent Seepage Slopes, such as Cors Hirdre, Buxton Heath (Norfolk)). M22 stands can be irrigated by surface water and groundwater, depending on their situation. It is estimated that groundwater provides the main component of water supply to the rooting zone of about 70 per cent of sites and surface water, about 10 per cent. The remaining 20 per cent are either mixtures of groundwater and surface water (around five per cent), or sites with low summer water tables (where the surface is mostly exclusively rain-fed). Examples on river floodplains and in other valley bottoms mostly appear to be dependent upon surface water inputs, whereas those in valleyhead situations are mostly groundwater-fed. However, in some topogenous situations the surface water may be largely derived from proximate groundwater sources, whilst in some valleyhead systems, where the community occupies intermittent seepages, rain-generated run-off may have greater importance in contributing to the summer water supply than is the case with permanent seepage faces. Wheeler and Shaw (2000a) found that there appears to be an interaction between soil nutrient status and water conditions occupied by M22 in spring-fed fens. For example, oligotrophic spring mounds fed by Chalk water rarely support M22: M13 is usually the main vegetation type, with M22 occurring only in peripheral locations (Boyer and Wheeler, 1989). However, mesotrophic spring mounds fed by water from glacial sands and gravels may have the entire surface and surroundings covered by M22 (Wheeler, 1983). Wheeler and Shaw (2000a) provide further discussion on water sources and supply to stands of M22.

19.3 Regimes 19.3.1 Water regime

Mean values for annual rainfall and potential evaporation for the sites examined are given in Table 19.2, together with mean recorded values for summer water table associated with stands of M22.

Table 19.2 Mean rainfall and potential evaporation for M22 stands Mean Minimum Maximum Rainfall (mm a–1) 651 539 1,050 Potential evaporation (mm a–1) 601 435 638 Mean summer water table (cm) –10.8 –175 12

Water conditions associated with M22 are variable (Table 19.2). Consequently, mean water table values have limited value, are potentially misleading and should be interpreted with caution. A very low value of 175 cm bgl has been measured at Cornard Mere in a

Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales 611

drought period, but this is exceptional. Typically, water conditions range from being rather dry to just above the surface, the latter being associated with permanent seepages. Much of the variation in species composition can be attributed to differences in the kind and degree of waterlogging (Rodwell, 1991b). For example, species such as Carex acutiformis, Carex paniculata and Carex disticha tend to be associated with wetter conditions, whilst species such as Carex hirta and Deschampsia cespitosa are more typical of summer-dry conditions. Specific time-series data for stands of M22 are not available for the majority of sites. It is therefore not possible to specify precise water regimes, or tolerance to change, but the following comments can be made: Optimal water levels x Most examples of M22 are characterised by summer water tables that are below the surface (–5 to –18 cm). x The M22 stands with the highest summer water tables are mostly those with the strongest groundwater inputs. x The most species-rich stands are found at water levels of between about –5 and –20 cm. Sub-optimal or damaging water levels x Very wet sites (summer water table usually above-surface between tussocks) tend to be less species-rich. Prolonged, deep inundation, particularly in the summer, is likely to be damaging. x Moderate reduction in water levels may actually increase species richness (Shaw and Wheeler, 1991), but a long-term reduction of the summer water table beneath high quality stands of M22 can be expected to result in some loss of botanical interest. More discussion of the relationships between hydrological conditions and floristic variation within M22 stands (and comparison with M13 stands) can be found in Wheeler and Shaw (2000a).

19.3.2 Nutrients/hydrochemistry

Typically found in base-rich conditions and of moderate fertility, although M22 can span a wide range (Table 19.3). The community tends to occupy more fertile situations than M24 or M13 (Shaw and Wheeler, 1991). In general, some of the least fertile examples were the most species-rich, although Shaw and Wheeler (1991) found no relationship between substratum fertility and species richness, indicating that other variables may be more important in regulating this. Low fertility conditions may help to retard invasion by tall herb fen and scrub into unmanaged stands. Shaw and Wheeler (1991) reported a decrease in species richness associated with an increase in base status, suggesting an avoidance of some species of particularly base-rich conditions. Wheeler and Shaw (1991a) report a mean increment (April to September) in dry weight of above-ground standing crop of 547 g dry wt m–2, which is significantly higher than that of M13.

Table 19.3 pH, conductivity and substratum fertility measured in stands of M22

612 Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales

Variable Mean SE Min Max Water pH 6.6 0.02 4.5 8.1 Soil pH 6.9 0.03 4.9 7.6 Water conductivity (Kcorr, μS cm–1) 612 1.2 113 1,780 Substratum fertility1 (mg phytometer) 13.9 0.25 2 49

19.3.3 Management

This community owes its origin to management (mowing or grazing) and depends on this for its maintenance. It once formed extensive areas in the mowing and grazing marshes of Broadland and some other floodplain systems. As far as is known this vegetation has no natural analogues, but is a product of the clearance of wet woodland followed by management. A corollary of this is that many of the plant species typically found in this vegetation are shade-tolerant and grow readily in wet woodland. Variations in management regime (including timing, frequency and intensity) and their histories are reflected in the variations in species composition.

19.4 Implications for decision making 19.4.1 Vulnerability

Conservation management typically involves ensuring relatively wet, mesotrophic and base- rich conditions. The main threats are from dereliction and drainage (or interception of supply). As the community does not normally define a SAC habitat, and because it is widespread, it is often not assigned a high priority for protection. However, in some districts it represents the only form of base-rich mire vegetation and repository for mire species, and can therefore have considerable local or regional significance. Perhaps the biggest threat to M22 vegetation is dereliction. This is likely to lead first to the development of tall herb vegetation (and associated species loss, particularly of small herbs) and then to development of some form of wet woodland (such as Salix cinerea–Betula pubescens–Phragmites australis woodland (W2)). The wet woodland may continue to support the majority of plant species that originally occurred in the M22 community, though in reduced numbers and probably without Juncus subnodulosus. A change in management regime (such as from mowing to grazing, or in timing or frequency) is also likely to result in a change in species composition. Management regimes can considerably affect the flowering performance of some of the less common species (such as Dactylorhiza spp.). Overgrazing can also be detrimental and may result in species loss, as well as poaching of the ground. The wide range of water table conditions under which stands of M22 occur make it difficult to make simple comments on vulnerability to drainage. Drying of M22 stands is likely to result in some changes in species composition, but the floristic impact of this does not depend just upon the magnitude of change, but also upon the wetness of the pre-drying starting point. In other cases, drying can lead to a change from one sub-community of M22 to another. The absence of a clear and consistent relationship between water levels and species richness of M22 vegetation, coupled with the fact that many of the species that distinguish M22 from other community types are essentially wet meadow species, means that some drying of M22

1 Experience has shown that N and P data derived from soil analysis has only limited use in assessing fertility of wetlands. Consequently the technique of phytometry (measuring the biomass of test species (phytometers) grown on soil samples) was developed. Typical phytometer yields (dry wt.); low fertility = <8mg, high fertility>18mg. Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales 613

stands may have little impact on species diversity per se, and in some cases could lead to a net increase in species richness. M22 can accommodate considerable eutrophication without change to its basic composition provided that active management continues, although there is likely to be some floristic change, and low-fertility stands may lose their distinctive features. Conversely, low-fertility conditions may help to retard invasion by tall herb fen and scrub if left unmanaged. Figure 19.2 shows the possible floristic impacts of changes to the stand environment. However, the concept of ‘vulnerability’ is complex and depends upon the starting conditions (including floristic composition), sensitivity of the stand and sensitivity of the site to the pressure of change. In such a context, the stand could be regarded as sensitive to change but not necessarily vulnerable. For this reason, accurate assessment of vulnerability is likely to require careful site-specific investigations.

614 Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales

Figure 19.2 Possible effects of environmental change on stands of M22 Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales 615

19.4.2 Restorability

As with all restoration measures, their likely success depends on the cause of the damage and how far the starting conditions are from the objective, both in time and conditions (such as numbers of species lost, damage to substratum, degree of enrichment). Limited information is available on the restoration of M22 stands, but the following observations can be made: x Where the community has been recently damaged, but this has not been intensive, corrective management may be sufficient to rehabilitate M22 in the short to medium term. x Scrub removal and re-instatement of vegetation management may help to restore M22 vegetation that has been left unmanaged for a while, provided that other conditions have not changed irreversibly. x Attempts to increase the wetness of examples of M22 by blocking outflows could be detrimental to the vegetation if they result in the establishment of prolonged periods with stagnant surface water and strongly reducing conditions.

19.4.3 Limitations of these guidelines and gaps in knowledge

The limitations of the information presented here include the following: x There are currently no data to better inform the temporal water table characteristics of M22 stands. Time series of dipwell measurements are required to fill this gap. x In order to make predictions on the vulnerability of M22 stands to water levels, models are required that can connect hydrogeological processes with hydrological conditions at the fen surface. This may require detailed ecohydrological investigations at representative sites. x Data on the spatial extent of M22 are lacking. x Possible differences in environmental conditions influencing the four sub- communities have not been explored here. x More information is needed on tolerance to nutrient enrichment and nutrient budgets. x More information is needed on appropriate restoration techniques.

616 Science Report A Wetland Framework for Impact Assessment at Statutory Sites in England and Wales