SEARCH

SEARCH BY CITATION

Keywords:

  • ESAs;
  • heather moorland;
  • Molinietum;
  • moorland management;
  • palaeoecological techniques

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results and interpretation
  6. Discussion
  7. Acknowledgements
  8. References

1. A characteristic of some heath and moorland areas in maritime north-west Europe is the widespread dominance of Molinia caerulea (purple moor grass). The overwhelming local supremacy of this species concerns farmers, owing to its relatively low palatability for grazing stock, and conservationists, owing to the monotonous, species-poor landscapes that often result under Molinietum.

2. In some environmentally sensitive areas (ESAs) in England and Wales, Molinietum is believed to have ousted Callunetum in recent decades; experiments sponsored to control the species have predicated its infiltration and replacement of heather-dominated stands.

3. Experimental control of Molinia in ESAs on Exmoor, England, was paralleled by palaeoecological studies to verify its recent rise, assess its status in moorland, and test the utility of the techniques for such research.

4. Peat profiles from two localities on Exmoor were sampled and subjected to recently developed techniques of plant macrofossil counting and to conventional pollen analysis. One locality was ‘white moor’, clearly dominated by Molinia; the other was ‘grey moor’ (an admixture of ericaceous shrubs) that had become invaded (allegedly recently) by Molinia.

5. Dating of profiles employed a range of methods, including conventional radiocarbon dating, Accelerator Mass Spectrometry (AMS) dating and the counting of spheroidal carbonaceous particles, to attempt to delimit horizons of recent peat growth.

6. The pollen and macrofossil data confirmed the recent ousting of Calluna and rise to dominance of Molinia in the grey moor, but also provided evidence of an earlier unsuspected (pre-Callunetum) presence of Molinia. The overwhelming dominance of Molinia in the white moor was also a recent phenomenon, but was only partly at the expense of Calluna. The palaeoecological data indicated a greater antiquity and former abundance of Molinia than is often appreciated and suggested that, over the past millennium, vegetation dominance has alternated between Callunetum and grass moor containing at least some Molinia, while the former Calluna-dominated grey moor itself developed originally from grass moor.

7. These findings have implications for conservation management and for restoration targets in ‘degraded’ moorland. Similar palaeoecological studies have since been adopted in Wales, directly to inform conservation and management policy.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results and interpretation
  6. Discussion
  7. Acknowledgements
  8. References

Globally, heaths are a relatively scarce resource; they are naturally found either in particular high-altitude areas above the tree line, or in some high latitudes in cool environments where trees are sparse or absent, as for example in sub-Arctic or sub-Antarctic regions (Gimingham & de Smidt 1983). However, extensive heaths are also found in western Europe as part of the pastoral cultural landscape, in which trees are kept at bay through a combination of grazing and burning. This ‘heath region’ of north-west Europe extends from the west of southern Norway, through southern Scania, Denmark, the Low Countries, the maritime north-west and west of France, and encompasses the British Isles (Gimingham 1975).

A major heather-dominated (as opposed to grass-dominated) manifestation of upland heathland is heather moorland (as Callunetum and variants), which in Britain has decreased substantially over the past 50 years and has been identified by conservation agencies as a threatened habitat (NCC 1984; Thompson et al. 1995). It is regarded as being increasingly under threat from a combination of biotic and human pressures, particularly from over- (or under-) grazing and inappropriate burning (Mallik, Gimingham & Rahman 1984; Bardgett, Marsden & Howard 1995), although climatic change since the ‘Little Ice Age’ and, in certain areas, fallout of atmospheric pollution may also be factors involved (Chambers, Dresser & Smith 1979).

Spread of molinia

In parts of the heath region of north-west Europe, Molinia caerulea (L.) Moench (purple moor grass or flying bent) has allegedly been increasing this century at the expense of heather in a range of heath, mire and moor habitats, including Danish and Dutch heaths (Hansen 1976; Berendse, Schmitz & de Visser 1994), disturbed mires in Belgium (Damblon 1992) and moorlands in western Britain. The invasion of heather stands by Molinia is a phenomenon of concern to landowners, farmers and government farming agencies, owing to the species’ relatively low palatability for grazing stock. Its supposed recent rise to dominance in some environmentally sensitive areas (ESAs) in Britain, and the resulting monotonous, species-poor landscapes, are also of concern for nature conservationists and officials in several of the national parks in England and Wales. In some of these regions, Molinietum is believed to have ousted Callunetum in recent decades (Todd 1995).

So dominant had M. caerulea become in some areas of England that, in 1995, the Ministry of Agriculture, Fisheries and Foods (MAFF) commissioned 3-year experiments to ascertain the best control methods. These management experiments were conducted to find methods to limit Molinia encroachment in ESAs, to encourage heather-dominated communities, and to promote the regeneration of heather in areas dominated by Molinia (Phillips & Marrs 1995; Todd 1995). Various management treatments were applied and monitored at sites in the Pennine Dales, in the North Peak and on Exmoor; these experiments involved the treatment and monitoring of replicate bi-plots, with combinations of burning (or not), summer grazing and existing grazing pressure, and the application of herbicide, and were designed to evaluate the most effective combination of controls (Todd 1995, 1996, 1997).

Rationale and aims

Straker & Crabtree (1995) emphasized that in order to understand the nature of the development of the heather- and Molinia-dominated landscapes, and to be able to ‘respond to proposals to alter land use and provide the background for management plans’, palaeoenvironmental research was required that should involve strategically targeted pollen and plant macrofossil analysis. The present study on Exmoor was conceived partly as a direct response to this call for further palaeoenvironmental research, but recognizing also that it could have implications and application elsewhere.

The experiments on Exmoor were therefore paralleled by palaeoecological studies to ascertain whether the dominance of Molinia is genuinely recent and to assess Molinia's status in moorland. This project was linked geographically to, but was distinct in aims from, the monitoring study. The research sites were located in two extensive and distinct areas of moorland: (i) ‘grey moor’, a mixture of Calluna vulgaris (L.) Hull, Vaccinium myrtillus L. and frequent Molinia[equivalent to a variant of the National Vegetation Classification (NVC) type M15, ‘wet heath’; Rodwell et al. 1991], as represented at Larkbarrow; and (ii) ‘white moor’, Molinia-dominated moorland (approximating the NVC type M25; Rodwell et al. 1991) adjacent to areas of heather moorland and believed to be derived from it, in that ericaceous shrubs are present occasionally, but whose past ecology was unknown, as represented at Lanacombe (Fig. 1).

image

Figure 1. Location of Lanacombe and Larkbarrow study areas on Exmoor. For accounts of core locations (x), see text. Squares show experimental bi-plots of Todd (1996, 1997). &ãgr; Crown copyright ED0267A.

Download figure to PowerPoint

This paper presents the palaeoecological data generated in projects that were conducted (i) to ascertain the sequence of vegetational changes leading to the dominance of Molinia in these moorlands; (ii) to assess the suitability of these techniques for further research into the status and spread of Molinia in moorland elsewhere; and (iii) to inform conservation and management policy in Molinia-dominated moorland in ESAs.

Study area

The study sites are located centrally in the Exmoor National Park, in south-west England. Exmoor is largely underlain by Devonian sedimentary rocks, and is dominated by a high plateau area, which in winter experiences a harsh climate. Whilst having a varied and fragmented character (Burton 1975), with altitudes ranging from sea level to 520 m a.s.l., the Park is principally renowned for its tracts of treeless moorland. Burton (1975) describes a generalized vegetational transition from western grass moor, through swampy middle moor (as near the study sites), east to heather moors. Heather-dominated moorland is regarded as being endangered, now covering approximately 19 000 ha compared with an estimated 34 400 ha at the turn of the century (White 1994). The Park (together with some adjacent areas of similar landscape character) has recently been declared an ESA.

Moorland on Exmoor has been categorized into coastal heaths, northern heather moors, central grass moors, southern heather moors and Brendon heaths (Miller & Miles 1984). The heather and grass moors are grazed by horses, cattle and sheep. Although red grouse Lagopus lagopus scoticus was introduced to Exmoor last century and became established after 1900, its population has declined markedly in the last 30 years, with extinction now likely, and so is no longer of significant influence here (Miller & Miles 1984; White 1994). Much of central Exmoor was protected from Saxon times under Forest Law, and was used for summer agistment of sheep and cattle (MacDermot 1911); it is believed that in central areas a vegetation dominated by M. caerulea and Trichophorum cespitosum (syn. Scirpus cespitosus) developed widely under this summer grazing regime (Binding 1995; Todd 1996). In recent decades, Molinia has been identified as a ‘problem’ species (Phillips & Marrs 1995) that is in danger of expanding further and is believed to be replacing C. vulgaris, particularly in ageing unburnt stands or where vegetation is burnt but then heavily browsed in winter (Todd 1996).

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results and interpretation
  6. Discussion
  7. Acknowledgements
  8. References

Field sampling

The locations of the study areas are given in Fig. 1 (latitude 51°10′ N; longitude 3°45′ W). These areas were investigated systematically for peat-depth variability, using a 1·8-m steel pin. Two cores of peat from the white moor site near Lanacombe were then collected using a Russian-pattern sampler (Jowsey 1966) from locations representing the range of peat depths, whilst a monolith of peat was extracted from the grey moor Larkbarrow site using a peat cutter, following the procedures detailed by Lageard, Chambers & Grant (1994). The Lanacombe cores were from (i) a location at National Grid Reference SS 766 425 adjacent to the bi-plots, but unaffected by the experiments; and (ii) an area of deeper peat at National Grid Reference SS 767 424, adjacent to rheotrophic flushed mire, approximately 100 m distant from the experiments. The Larkbarrow monolith was taken close to the bi-plots at National Grid Reference SS 825 418, a few metres outside the fenced plots, and unaffected by them.

Plant macrofossil analysis

Identification and quantification of vegetative macrofossils was undertaken at 2-cm intervals for Core 2, taken from the Lanacombe study site. The semi-quantitative Quadrat and Leaf Count macrofossil analysis technique (QLCMA), developed by Haslam (1987) and Clarke (1988), was used. This experimental methodology has been adopted by Stoneman (1993) and by Barber et al. (1994), and allows an assessment of the volume percentages of floral macrofossils to be made. The methodology was originally devised to reconstruct former vegetational assemblages on raised mires, where it continues to be applied (Mauquoy & Barber 1999), although it has been used on deep blanket mire in northern England at Moor House, Northern Pennines (McTiernan et al. 1998). It was used for the first time on shallower peats in the present Exmoor study.

In the laboratory, a slice of peat measuring 1 cm × 1 cm × 4 cm was cut from each level with a surgical scalpel. The samples were placed on a 125-µm sieve and sprayed with a high pressure jet of water from a narrow plastic pipe connected to a tap, which served to force the fine, highly humified material through the sieve and so disaggregate the sample. Macrofossils retained on the 125-µm sieve were placed in a glass Petri dish and 70 ml of water added to create a monolayer of remains over a 10 × 10 square-grid graticule. The remains were scanned using a Meiji EMZ-TR stereozoom microscope at ×10 magnification (supplied by Prior Scientific, Fulbourn, Cambs, U.K.), and the abundance of unidentified organic matter (UOM), Ericales roots, monocotyledons, identifiable Sphagnum and other bryophytes was estimated using 15 averaged quadrat counts. Where a component covered more than half each individual square, a score of 1% was given to it. Calluna vulgaris seeds were recorded on a 1–5 scale (where 1 = rare, 2 = occasional, 3 = frequent, 4 = abundant and 5 = very abundant). Charcoal fragments in the glass Petri dish were counted and assigned to five size classes.

Wherever monocotyledon epidermal tissues were visible in the Petri dish, these were removed using watchmaker's forceps, then mounted in Aquamount (BDH Laboratory Supplies, Poole, UK) and examined at ×100–400 magnification to confirm the taxa present. Each monocotyledonous taxon is expressed as a mean percentage value of the peat matrix scanned while performing the 15 quadrat counts. The drawings in Grosse-Braukmann (1972), Katz, Katz & Skobeyeva (1977) and Clarke (1988), in addition to a reference collection of type material, aided the identification of sedges encountered in the sample peat matrices. In many instances, the Cyperaceous remains could not be identified to species level, particularly where macrofossils were largely derived from rhizome material; in this case, and where there was no epidermal material present in subfossil samples, monocotyledonous remains were simply classified as monocotyledons undifferentiated (monocots undiff.), and expressed as a percentage value.

Seeds of C. vulgaris and spindles from Eriophorum vaginatum were identified using a type reference collection and the drawings and photographs in Grosse-Braukmann (1972, 1974) and Grosse-Braukmann & Streitz (1992). Most graminoid subfossils were identifiable only as monocot. remains, but the epidermal tissues of M. caerulea are very distinctive (van Geel 1978). Subfossil epidermal tissues were compared to modern reference type material of M. caerulea that had been boiled in 10% NaOH to age it artificially. The drawings in Grosse-Braukmann (1972) and the type collection of Professor Bas van Geel, at the University of Amsterdam, the Netherlands, were also consulted to aid the identification of this grass species.

Core 1, taken from Lanacombe, and the monolith taken from Larkbarrow, contained highly humified peat matrices, which militated against detailed macrofossil analysis. Despite this, a coarse analysis of plant macrofossils was undertaken at 8-cm intervals for the Lanacombe core and at 4-cm intervals for the Larkbarrow monolith, specifically to search for remains of Molinia.

Spheroidal carbonaceous particle analysis

Spheroidal carbonaceous particles (SCPs) are formed from the high-temperature combustion of fossil fuels. Analyses of their concentrations in lake sediments by Griffin & Goldberg (1981), Renberg & Wik (1984, 1985) and Rose et al. (1995), and in peat stratigraphy (Barber et al. 1999), have served to indicate the historical development of coal and oil burning in each of the countries studied. The analysis of SCPs might offer an extra degree of dating control, as in areas adjacent to industrial activity it allows peat from the very recent past to be dated. Hence, SCP analysis could be a useful technique to date very recent peat and lake stratigraphy, provided the spherules are present in sufficient quantities. As the Exmoor ESAs are some distance from heavy industry, analysis was thought unlikely to yield large quantities of SCPs; analysis was applied to evaluate the utility of the technique for rural areas remote from pollution sources.

The methodology adopted was based on Rose (1990, 1994). Differences from that experimental procedure, which principally uses concentrated nitric acid to remove organic material, were (i) the omission of hydrofluoric and hydrochloric acid digests, because these are designed to remove siliceous and carbonate material and these components were largely absent from the peat stratigraphy investigated; and (ii) a higher temperature setting of 180 °C rather than 80 °C, during the acid digestion phase. A Gerhardt scrubber unit (Gerhardt, Bonn, Germany; via Gerhardt UK Ltd, Brackley, UK), which incorporates a pump to draw acidic fumes through a 10% NaOH solution, was used to neutralize acidic vapour evolved during the acid digestion. SCP concentrations are expressed as per gram dry mass of sediment (g DM–1). The diameters of SCPs ranged from 5 to 30 µm.

Pollen analysis

Samples of 0·5-cm vertical thickness were prepared for pollen analysis after Barber (1976), but with an extra stage involving micromesh sieving. After treatment with 10% NaOH, samples were initially washed through a 180-µm sieve, and then through a 10-µm micromesh sieve to remove any clay particles and to concentrate the pollen grains in the sample. Pollen grains were identified with the aid of an extensive pollen reference collection and with pollen keys in Faegri & Iversen (1989). Plant nomenclature follows Clapham, Tutin & Moore (1987), as this was found to be more easily compatible with the pollen key than the nomenclature of Stace (1991). Charcoal on the pollen microslides was recorded semi-quantitatively using a 5-point scale.

In the pollen diagrams, the arboreal taxa are arranged in conventional order, whilst the non-arboreal types are arranged in broad ecological groups, from woodland, through woodland edge, ruderal, arable, grassland and mire habitats. All taxa are represented as percentages of total land pollen (TLP), but with spore taxa excluded from the TLP sum (these are marked in parentheses in Figs 2, 3 and 5).

image

Figure 2. Pollen diagram from the upper section of Lanacombe Core 1. Pollen and spore data are expressed as percentages of total land pollen; taxa in parentheses were excluded from the pollen sum. The position of radiocarbon samples is shown (date in year bp). Charcoal concentration was assessed on a 5-point scale.

Download figure to PowerPoint

image

Figure 3. Macrofossil diagram from Lanacombe Core 2. With the exception of Calluna vulgaris seeds, which were assessed on a 5-point abundance scale, all plant macrofossil components are expressed as percentages of the Petri dish sample. Charcoal particles were counted and assigned to five size classes.

Download figure to PowerPoint

image

Figure 5. Pollen diagram from Larkbarrow. Pollen and spore data are expressed as percentages of total land pollen; taxa in parentheses were excluded from the pollen sum. The position of radiocarbon samples is shown (date in year bp). Charcoal concentration was assessed on a 5-point scale.

Download figure to PowerPoint

Radiocarbon dating

After the plant macrofossil and pollen analyses had been conducted, critical horizons were identified in the diagrams, and slim samples of peat > 200 g wet weight were taken from the Larkbarrow monolith for conventional bulk radiocarbon dating. It was not intended that the radiocarbon assays should provide either a precise or a detailed time scale; rather, the purpose was to provide an estimate of the onset of peat accumulation at the Larkbarrow site, to provide an overall time scale for the vegetational history shown in the pollen diagram and to supplement relative dating indications from the pollen and SCP data. Small samples of core material were taken from the Lanacombe cores for radiocarbon dating by accelerator mass spectrometry (AMS) to provide an outline time scale for the vegetational changes shown in the macrofossil and pollen diagrams. All radiocarbon dating was conducted by Beta Analytic Inc. (Miami, FL, USA).

Results and interpretation

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results and interpretation
  6. Discussion
  7. Acknowledgements
  8. References

Lanacombe, core 1

This core, taken from close by the experimental bi-plots, was 50 cm deep. The highly humified nature of the peat allowed only a coarse analysis; nevertheless this yielded evidence of Molinia epidermis at 8 cm depth in the core.

Pollen samples were prepared and analysed every 2 cm, down to 20 cm depth (Fig. 2). The pollen data showed an earlier period of Sphagnum spore abundance, which was succeeded by a horizon with markedly increased Calluna representation. This implied a drying out of the mire surface. There then followed a decline in Calluna, principally accounted for by both an initial rise in Cyperaceae and by a gradual rise for Gramineae (sensuFaegri & Iversen 1989; equiv. Poaceae) from 40% TLP at 16 cm depth to approximately 55% TLP at 8 cm depth. Above this, Gramineae showed a marked increase to approximately 70% TLP, whilst Cyperaceae values were reduced.

The AMS radiocarbon dates suggested a relatively rapid rate of peat accumulation, and confirmed that the rise in Gramineae pollen was a modern feature.

Lanacombe, core 2

The data for Lanacombe Core 2 are presented in Fig. 3 (plant macrofossils, 0–50 cm) and Fig. 4 (pollen, 0–20 cm). The macrofossil data showed distinct changes in local vegetational dominance. At the base of the analysed core, the local vegetation was clearly dominated by Sphagnum, principally S. section cuspidata– probably S. cuspidatum– a pool-dwelling species, indicating a wet local environment. The Sphagnum dominance was temporarily interrupted by a phase in which epidermis of E. vaginatum (cotton sedge) was recorded. Subsequently, Sphagnum dominance was reasserted, but this time including S. papillosum, in a horizon (at 24 cm) containing the first identified remains of Molinia. The representation of Sphagnum then declined, as representation of epidermis of E. vaginatum increased. The phase of Eriophorum dominance then ceased abruptly as representation of UOM increased. A rise in UOM is usually taken to indicate a drier local environment, which generates a more humified peat with fewer identifiable plant macrofossils. The peak of UOM representation coincided with an increase in the representation of epidermis of M. caerulea. The uppermost phase then contained the highest representation of Molinia in the profile. This phase also contained the highest representation of charcoal fragments in the profile.

image

Figure 4. Pollen diagram from the upper section of Lanacombe Core 2. Pollen and spore data are expressed as percentages of total land pollen; taxa in parentheses were excluded from the pollen sum. The position of radiocarbon samples is shown (date in year bp). Charcoal concentration was assessed on a 5-point scale.

Download figure to PowerPoint

As the intention was to ascertain the recent vegetational history, pollen data were generated down to 20 cm depth only, to encompass the changes in the macrofossil diagram from Sphagnum dominance to dominance by graminoid communities. In contrast to the plant macrofossil data, which showed distinct changes in the local vegetation, the pollen data principally showed broad trends, which reflected the integration of changes in the wider vegetational community. At 20 cm depth, Sphagnum spore representation is higher than in the rest of the diagram, and presumably reflects the high local representation of Sphagnum in the plant macrofosssils. The striking aspect of the pollen diagram is the high representation of Gramineae (grass) pollen, which increased from 50% of TLP at 20 cm depth to over 70% TLP at the surface, compared with the lower values for Cyperaceae (sedges), which varied from < 5% TLP at 20 cm depth to approximately 15% TLP at 16 cm and 4 cm depth. These data imply that the undifferentiated Monocot. remains in Fig. 3 may well be of grass rather than sedge.

A salient feature of this pollen diagram is the slightly increased representation of Calluna pollen in the depth interval 6–8 cm. This is in a period when the UOM content was highest, which indicates a drier environment. The implication is that a decrease in mire-surface wetness gave rise to better growing conditions for Calluna. Calluna pollen representation does not exceed 10% TLP in this profile and so it may be inferred that heather was more abundant in areas of shallower peat (compare Fig. 2). The elevated Calluna representation in Fig. 4 is at a time when records of Molinia macrofossils started to increase; a decline in Calluna pollen then took place as the values for Gramineae pollen rose, eventually to reach their highest percentages in the uppermost horizon. This increase may be interpreted as recording the rise to overwhelming dominance by Molinia in this locality.

Peat at 50 cm depth dates to 360 ± 40 bp (cal. ad 1460 to cal. ad 1630, to 1 σ), which indicates that the peat immediately above this horizon accumulated during the so-called Little Ice Age (i.e. after ad 1420). It is notable that this phase was dominated by Sphagna characteristic of wet conditions, which might be attributable to a cooler, wetter climate. As at Lanacombe Core 1, the top 20 cm of peat appeared to have accumulated rapidly. The calibrated age (to 1 σ) for peat at 20 cm depth was between cal. ad 1690 and cal. ad 1930; and a modern age is indicated for the major rise in Gramineae at 6 cm (Fig. 4). This immediately succeeds the rise in Molinia epidermis, but also marks the major rise in charcoal records in the macrofossil diagram (Fig. 3).

Larkbarrow

Although the peat here was highly humified, and so prevented detailed macrofossil analysis, fragments of M. caerulea epidermis were identified in samples at 4, 8, 16 and 24 cm depth. Of the samples examined, only those from 12 and 20 cm depth had no identifiable Molinia, but this absence might not be significant and could be attributable to the highly humified nature of the peat.

The pollen diagram (Fig. 5) shows five main pollen assemblage zones. The lowermost (zone a) is dominated by Gramineae, implying the presence of grass heath. The radiocarbon date of 1360 ± 60 bp calibrates to between cal. ad 610 and cal. ad 760 (at 1 σ) and suggests that this zone commenced in the Dark Ages (5th to 9th centuries ad). The succeeding zone (b) is dominated by Calluna, which reaches some 65% of TLP. This zone marks the establishment of Callunetum; its age is not known with certainty but is likely to be Medieval, and may well include the Medieval climatic optimum (c. ad 1100–1300), when overall conditions here could have been relatively dry. Zone c is dominated by Gramineae, and has very low Calluna values. Its age is not known with certainty; the radiocarbon age indicates Modern, which implies (but does not conclusively confirm) a late-20th century age, whilst elevated records of SCPs from 7 cm (Fig. 6a) suggest an age at or after the start of the Industrial Revolution (from c. ad 1760). The dominance of Gramineae in the pollen spectra is much reduced in zone d, when there is increased Calluna representation, implying a resurgence of heather. The current vegetation is represented in zone e, in which the higher Gramineae values and reduced Calluna representation seem to bear out the anecdotal evidence for Molinia displacing Calluna in recent decades.

image

Figure 6. Spheroidal carbonaceous particle (SCP) concentrations recorded in (a) the upper horizons of Larkbarrow peat, and (b) the upper horizons of blanket peat from Drygarn Fawr, mid-Wales. Note the scales are different; the pattern is similar, but much higher SCP concentrations were recorded in Wales.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results and interpretation
  6. Discussion
  7. Acknowledgements
  8. References

Long-term changes

The data presented here add to existing palaeoecological data from Exmoor, which hitherto have concentrated on deeper peats (Merryfield & Moore 1974; Merryfield 1977; Moore, Merryfield & Price 1984; Francis & Slater 1990, 1992). The previous studies showed that initiation of the deeper peats took place in prehistory, principally replacing woodland (Straker & Crabtree 1995), but that there was a range of peat-inception dates, from the earliest Neolithic to the late Iron Age. Pollen data from the deeper peat sites had indicated past fluctuations between heather- and grass-dominated moorland over millennial time scales (Francis & Slater 1990; Straker & Crabtree 1995). The present study shows that there have also been marked changes in species dominance in shallow peat sites, but over historical times. These changes were not previously known. Also, a later (Dark Age) peat inception date is indicated for the Larkbarrow site, which parallels the peat initiation date for some shallow heath peats in South Wales (Chambers 1983) and which may coincide with a Dark Age climatic deterioration (Blackford & Chambers 1991).

Whilst the results from this study provide evidence of former plant communities of the two moorland types, thereby directly suggesting potential plant communities for each type of moorland, the techniques and resulting data are more widely applicable in attempting to understand past and contemporary ecological change in a major threatened habitat.

Utility of the techniques

Palaeoecological methods, principally pollen analysis, were used by Stevenson & Thompson (1993) in an ambitious, albeit imprecise, attempt to assess long-term changes in the extent of heather moorland in upland Britain and Ireland. However, until now, neither pollen analysis nor plant macrofossil analysis of shallow, relatively highly humified peats would have been thought to have the potential to provide useful data to investigate the vegetational history of an alleged invasive species in moorland. Pollen analysis on its own could not have confirmed the past status of Molinia, owing to poor taxonomic separation, whilst macrofossil analysis itself is of limited use in shallow, highly humified heath peats; but when, as here, plant macrofossil analysis is applied in combination with pollen and other palaeoecological data then a fuller picture may emerge. This work reinforces the findings of Damblon (1992) in that regard, but the present study takes the techniques into a different habitat, upland moorland rather than the lowland fens and bogs of Damblon's study, for which the techniques could be expected to succeed. The combination of methods is a significant advance on the use of pollen analysis alone, employed in moorland by Stevenson & Thompson (1993).

Limitations regarding the dating of the peat profiles are acknowledged, but very precise dating was beyond the scope of this study. The use of SCPs has since been shown to be of assistance as a dating tool for Exmoor, principally as work in progress has demonstrated the utility of SCPs to help date the rise of Molinia elsewhere; for example, in Fig. 6b data from Wales show a similar pattern to those from Exmoor (Fig. 6a) in the upper peat horizons, but the scale of values is an order of magnitude higher in the Welsh peat. Nevertheless, the similar pattern produced allows us to be more confident that the rise in SCP records in the Exmoor peat does correspond with the early part of the Industrial Revolution. The combination of techniques employed here is being used in a study of blanket bog degradation in South Wales (F.M. Chambers & D. Mauquoy, unpublished data) directly to inform conservation management objectives of the Countryside Council for Wales. This study has therefore borne out Straker & Crabtree's (1995) recommendation that palaeoenvironmental research can be of assistance in devising management plans, and more specifically that the use of palaeoecological methods can inform conservation and management policy in Molinia-dominated moorland in ESAs.

Management implications

The invasion of heather moorland by Molinia is now a cause of concern in many areas of upland Britain, whilst the spread of Molinia has also been affecting cultural landscapes elsewhere in Europe, including heathlands in Denmark (Hansen 1976) and Norway (Prøsch-Danielsen & Øvstedal 1994) and disturbed mires in the Haute-Ardenne (Damblon 1992). The decline of heather and replacement by grasses in the Netherlands, particularly, has been well documented (Gimingham & de Smidt 1983; Heil & Diemont 1983; Berendse, Schmitz & de Visser 1994). Of the range of possible causes suggested by Chambers, Dresser & Smith (1979) for the decline in upland heather, Berendse, Schmitz & de Visser (1994), in experiments on lowland heaths, confirmed the hypothesis that an increased nutrient supply can cause a shift from heather to grass dominance, and that Molinia is favoured by nutrient enhancement.

Pitcairn, Fowler & Grace (1995) evaluated the role of increased nitrogen deposition in Europe since 1950, which they considered potentially significant in vegetational change, including a decline of Calluna in Breckland heaths. Damblon (1992), however, pointed to changing management regimes as being responsible for initiating a shift from Calluna-dominance to Molinia-dominance in disturbed mires in Belgium, but argued that restoration and management could be facilitated by knowledge of the succession of plant communities leading to present-day vegetation.

Control of Molinia has in recent decades become an intractable problem in many moorland areas in Britain, with farm-scale trials being conducted in similar vegetation types in three national parks to identify successful methods for future management of Molinia (Todd 1996). A range of strategies has been adopted to try to control its alleged spread on Exmoor. The experiments to control the species were based on the premise that there had been infiltration and replacement of heather-dominated stands by Molinia in the recent past. Rodwell et al. (1991) lists M. caerulea as a possible member of a wide range of moorland communities and as a constant member of several (labelled in the UK as NVC types M13–M17, M21 and M24–M26). However, the opinion that Molinia needs to be controlled and prevented from spreading into significant areas of upland moors in ESAs seems to stem from a view that Molinia is acting, if not as an ‘alien’, then at least as an invasive and increasingly dominant (in reality domineering) species in such landscapes, almost analogous to the recent spread of Pteridium aquilinum (bracken) in heaths and moorland edges (cf. Smith & Taylor 1986; Lowday & Marrs 1992). In the absence of significant pre-war aerial photography, this perception is often based on anecdotal evidence, relying upon written or oral accounts of changes in the landscape. For example, in Elenydd, mid-Wales, a local landowner recounted how, according to local belief, the spread and dominance of Molinia dates from the turn of the century when cattle grazing was supplanted by sheep ranching (E.W. Cloutman, personal communication). This explanation (which remains to be confirmed) for the dominance of Molinia in those particular communities (M25a/M15; sensuRodwell et al. 1991) may not apply elsewhere, particularly in regions where there has long been a tradition of sheep grazing or where the date of the vegetational change is more recent.

Results of this palaeoecological study have shown that the rise of M. caerulea is indeed a recent feature at the studied sites in the Exmoor ESA, and that its rise has been partly at the expense of Calluna or Callunetum. This was particularly noticeable in the grey moor locality (Larkbarrow). Its major rise to dominance there (at 8 cm depth in Fig. 5) was accompanied by increased charcoal evidence for burning, implying more frequent and severe fires. Although the two Lanacombe cores were taken from locations whose contemporary quadrat data are different, the pollen data nevertheless exhibit similar time series. The dating evidence from both Lanacombe and Larkbarrow, when compared with dating evidence elsewhere (Fig. 6b), suggests the vegetational shift took place this century.

However, the vegetational history is more complex than this: former Callunetum (at Larkbarrow) itself originated from grass heath, probably in Medieval times. More recently, this Callunetum was supplanted by grass moor, probably Molinietum; but a temporary resurgence of heather immediately preceded the current invasion by Molinia. This evidence suggests that, during the past millennium, vegetation dominance has alternated between Callunetum and grass moor containing at least some Molinia.

The palaeoecological data suggest that Molinia is a long-standing member of the local plant communities, but that these have changed markedly over time. Current uncontrolled cultural methods, particularly annual (illegal) burning, are allegedly encouraging Molinia dominance, but the evidence for the antiquity of Molinia in moorland raises questions as to the ethics of applying chemical controls, and implies that controlled cultural methods might be successful in limiting its dominance, as they would appear to have been in the past. These cultural methods could involve different grazing regimes or different grazers, particularly cattle (cf. Grant et al. 1996) or ponies (Sutherland & Hill 1995). However, there is also circumstantial evidence from both localities to suggest that changes in vegetational assemblages may have coincided with climatic shifts, which implies that the vegetation type may be determined partly by prevailing climate.

The palaeoecological data emphasize that in ‘degraded’ moorland, management prescriptions based on a rigid adherence to present-day vegetational classifications (Rodwell et al. 1991) may exclude other, equally legitimate, vegetational assemblages. These palaeoecological data therefore raise questions as to future management practices for Molinia-dominated moorland elsewhere, and emphasize particularly the subjectivity involved in devising methods for land-use management and in identifying restoration targets for degraded moorland. In particular, the rigid use of contemporary vegetational classifications (in Britain the NVC) to determine management prescriptions ignores the ontogeny of the moorland vegetation and might exclude from consideration other possible vegetational assemblages, so limiting the potential restoration possibilities for these degraded cultural landscapes. In contrast, the use of palaeoecological data can broaden management vision for conservation objectives for moorland. In Wales, new palaeoecological studies have recently been commissioned to ascertain the vegetational changes that have given rise to Molinia-dominated moorland, directly to inform the management process (Yeo 1997).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results and interpretation
  6. Discussion
  7. Acknowledgements
  8. References

Thanks are due to the British Ecological Society for a small ecological project grant; The Heather Trust Ltd for funding initial sampling; and to Professor Bas van Geel for comparative plant macrofossil material. Figures 2–5 were produced using tilia.graph, supplied by Dr Eric Grimm. Comments from three referees helped to improve the paper.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results and interpretation
  6. Discussion
  7. Acknowledgements
  8. References
  • Barber, K.E. (1976) History of vegetation. Methods in Plant Ecology (ed. S.B.Chapman), pp. 5 83. Blackwell Scientific Publications, Oxford, UK.
  • Barber, K.E., Battarbee, R.W., Brooks, S.J., Eglinton, G., Haworth, E.Y., Oldfield, F., Stevenson, A.C., Thompson, R., Appleby, P., Austin, W.N., Cameron, N., Ficken, K.J., Golding, P., Harkness, D.D., Holmes, J., Hutchinson, R., Lishman, J.P., Maddy, D., Pinder, L.C.V., Rose, N., Stoneman, R. (1999) Proxy records of climate change in the UK over the last two millennia: documented change and sedimentary records from lakes and bogs. Journal of the Geological Society of London, 156, 369 380.
  • Barber, K.E., Chambers, F.M., Maddy, D., Stoneman, R., Brew, J.S. (1994) A sensitive high resolution record of late Holocene climatic change from a raised bog in northern England. Holocene, 4, 198 205.
  • Bardgett, R.D., Marsden, J.H., Howard, D.C. (1995) The extent and condition of heather on moorland in the uplands of England and Wales. Biological Conservation, 71, 155 161.
  • Berendse, F., Schmitz, M., De Visser, W. (1994) Experimental manipulation of succession in heathland ecosystems. Oecologia, 100, 38 44.
  • Binding, H. (1995) The Changing Face of Exmoor. Exmoor Books, Tiverton, UK.
  • Blackford, J.J. & Chambers, F.M. (1991) Blanket peat humification: evidence for a Dark Age (1400 bp) climatic deterioration in the British Isles. Holocene, 1, 63 67.
  • Burton, S.H. (1975) Exmoor. Britain's National Parks (ed. M.Bell), pp. 30 41. David and Charles, Newton Abbot, UK.
  • Chambers, F.M. (1983) Three radiocarbon-dated pollen diagrams from upland peats north-west of Merthyr Tydfil, South Wales. Journal of Ecology, 71, 475 487.
  • Chambers, F.M., Dresser, P.Q., Smith, A.G. (1979) Radiocarbon dating evidence on the impact of atmospheric pollution on upland peats. Nature, 282, 829 831.
  • Clapham, A.R., Tutin, T.G., Moore, D.M. (1987) Flora of the British Isles, 3rd edn. Cambridge University Press, Cambridge, UK.
  • Clarke, M.J. (1988) Past and present mire communities of the New Forest and their conservation. PhD Thesis. University of Southampton, Southampton, UK.
  • Damblon, F. (1992) Palaeobotanical analysis of Eriophorum and Molinia tussocks as a means of reconstructing recent history of disturbed mires in the Haute-Ardenne, Belgium. Review of Palaeobotany and Palynology, 75, 273 288.
  • Faegri, K. & Iversen, J. (1989) Textbook of Pollen Analysis, 4th edn(eds K.Faegri, P.E.Kaland & K.Krzywinski). Wiley, Chichester, UK.
  • Francis, P.D. & Slater, D.S. (1990) A record of vegetational and land use change from upland peat deposits on Exmoor. II. Hoar Moor. Proceedings of the Somerset Archaeological and Natural History Society, 134, 1 25.
  • Francis, P.D. & Slater, D.S. (1992) A record of vegetational and land use change from upland peat deposits on Exmoor. III. Codsend Moor. Proceedings of the Somerset Archaeological and Natural History Society, 136, 9 28.
  • Gimingham, C.H. (1975) An Introduction to Heathland Ecology. Oliver and Boyd, Edinburgh, UK.
  • Gimingham, C.H. & De Smidt, J.T. (1983) Heaths as natural and semi-natural vegetation. Man's Impact on Vegetation (eds W.Holzner, M.J.A.Werger & I.Ikusima), pp. 185 199. Dr W. Junk, The Hague, The Netherlands.
  • Grant, S.A., Torvell, L., Common, T.G., Sim, E.M., Small, J.L. (1996) Controlled grazing studies on Molinia grassland: effects of different seasonal patterns and levels of defoliation on Molinia growth and responses of swards to controlled grazing by cattle. Journal of Applied Ecology, 33, 1267 1280.
  • Griffin, J.J. & Goldberg, E.D. (1981) Sphericity as a characteristic of solids from fossil fuel burning in a Lake Michigan sediment. Geochimica et Cosmochimica Acta, 45, 763 769.
  • Grosse-Braukmann, G. (1972) Über pflanzliche Makrofossilien mitteleuropaischer Torfe. I. Gewebereste krautiger Pflanzen und ihre Merkmale. Telma, 2, 19 55.
  • Grosse-Braukmann, G. (1974) Über pflanzliche Makrofossilien mitteleuropaischer Torfe. II. Weitere Reste (Fruchte und Samen, Moose u. a.) und ihre Bestimmungsmoglichkeiten. Telma, 4, 51 117.
  • Grosse-Braukmann, G. & Streitz, B. (1992) Über pflanzliche Makrofossilien mitteleuropaischer Torfe. III. Fruchte, Samen und einige Gewebe (Fotos von fossilen Pflanzenresten). Telma, 22, 53 102.
  • Hansen, K. (1976) Ecological studies in Danish heath vegetation. Dansk Botanisk Archiv, 31, 7 116.
  • Haslam, C.J. (1987) Late Holocene peat stratigraphy and climatic change – a macrofossil investigation from the raised mires of Western Europe. PhD Thesis. University of Southampton, Southampton, UK
  • Heil, O.W. & Diemont, W.M. (1983) Raised nutrient levels change heathland into grassland. Vegetatio, 53, 113 120.
  • Jowsey, P.C. (1966) An improved peat sampler. New Phytologist, 65, 245 248.
  • Katz, N.J., Katz, S.V., Skobeyeva, E.I. (1977) Atlas of Plant Remains in Peats. Nauka, Moscow, Russia [in Russian].
  • Lageard, J.G.A., Chambers, F.M., Grant, M.E. (1994) Modified versions of a traditional peat cutting tool to improve field sampling of peat monoliths. Quaternary Newsletter, 74, 10 15.
  • Lowday, J.E. & Marrs, R.H. (1992) Control of bracken and the restoration of heathland. I. Control of bracken. Journal of Applied Ecology, 29, 195 203.
  • MacDermot, E.T. (1911) The History of the Forest of Exmoor. Exmoor Books, Taunton, UK.
  • McTiernan, K.B., Garnett, M.H., Mauquoy, D., Ineson, P., Coûteaux, M.-M. (1998) Use of near-infrared reflectance spectroscopy (NIRS) in palaeoecological studies of peat. Holocene, 8, 729 740.
  • Mallik, A.U., Gimingham, C.H., Rahman, A.A. (1984) Ecological effects of heather burning. I. Water infiltration, moisture retention and porosity of surface soil. Journal of Ecology, 72, 767 776.
  • Mauquoy, D. & Barber, K.E. (1999) A replicated 3000 yr proxy-climate record from Coom Rigg Moss and Felecia Moss, the Border Mires, northern England. Holocene, 14, 263 275.
  • Merryfield, D.L. (1977) Palynological and stratigraphical studies on Exmoor. PhD Thesis. University of London, London, UK.
  • Merryfield, D.L. & Moore, P.D. (1974) Prehistoric human activity and blanket peat initiation on Exmoor. Nature, 250, 439 441.
  • Miller, G.R. & Miles, J. (1984) Moorland Management: A Study of Exmoor. NERC Institute of Terrestrial Ecology, Cambridge, UK.
  • Moore, P.D., Merryfield, D.L., Price, M.D.R. (1984) The vegetation and development of blanket mires. European Mires (ed. P.D.Moore), pp. 203 235. Academic Press, London, UK.
  • NCC (1984) Nature Conservation in Great Britain. Nature Conservancy Council, Peterborough, UK.
  • Phillips, J. & Marrs, J. (1995) The MAFF Molinia study. The Heather Trust Annual Report (1995) (ed. J.Phillips A.Watson), pp. 39 40. Heather Trust, Kippen, UK.
  • Pitcairn, C.E.R., Fowler, D., Grace, J. (1995) Deposition of fixed atmospheric nitrogen and foliar nitrogen content of bryophytes and Calluna vulgaris (L.) Hull. Environmental Pollution, 88, 193 205.
  • Prøsch-Danielsen, L. & Øvstedal, D.O. (1994) Vegetation history of Molinia heaths in Tysvær, Rogaland, western Norway. Nordic Journal of Botany, 14, 557 568.
  • Renberg, I. & Wik, M. (1984) Dating of recent lake sediments by soot particle counting. Verhandlungen der Internationalen Vereinigung für Theoretische und Angewandte Limnologie, 22, 712 718.
  • Renberg, I. & Wik, M. (1985) Carbonaceous particles in lake sediments – pollutants from fossil fuel combustion. Ambio, 14, 161 163.
  • Rodwell, J.S., Pigott, C.D., Ratcliffe, D.A., Malloch, A.J.C., Birks, H.J.B., Proctor, M.C.F., Shimwell, D.W., Huntley, J.P., Radford, E., Wigginton, M.J., Wilkins, P. (1991) British Plant Communities. Mires and Heaths. Cambridge University Press, Cambridge, UK.
  • Rose, N.L. (1990) A method for the extraction of carbonaceous particles from lake sediment. Journal of Paleolimnology, 3, 45 53.
  • Rose, N.L. (1994) A note on further refinements to a procedure for the extraction of carbonaceous fly-ash particles from sediments. Journal of Paleolimnology, 11, 201 204.
  • Rose, N.L., Harlock, S., Appleby, P.G., Battarbee, R.W. (1995) Dating of recent lake sediments in the United Kingdom and Ireland using spheroidal carbonaceous particle (SCP) concentration profiles. Holocene, 5, 328 335.
  • Smith, R.T. & Taylor, J.A. (1986) Bracken, Ecology, Land Use and Control Technology. Parthenon Publishing Group, Carnforth, UK.
  • Stace, C. (1991) A New Flora for the British Isles. Cambridge University Press, Cambridge, UK.
  • Stevenson, A.C. & Thompson, D.B.A. (1993) Long-term changes in the extent of heather moorland in upland Britain and Ireland: palaeoecological evidence for the importance of grazing. Holocene, 3, 70 76.
  • Stoneman, R.E. (1993) Late Holocene peat stratigraphy and climate: extending and refining the model. PhD Thesis. University of Southampton, Southampton, UK.
  • Straker, V. & Crabtree, K. (1995) Palaeoenvironmental studies on Exmoor: past research and future potential. The Changing Face of Exmoor (ed. H.Binding), pp. 43 51. Exmoor Books, Tiverton, UK.
  • Sutherland, W.S. & Hill, D.A. (1995) Managing Habitats for Conservation, Cambridge University Press, Cambridge, UK.
  • Thompson, D.B.A., MacDonald, A.J., Marsden, J.H., Galbraith, C.A. (1995) Upland heather moorland in Great Britain: a review of international importance, vegetation changes and some objectives for nature conservation. Biological Conservation, 71, 163 178.
  • Todd, P.A. (1995) The MAFF Molinia study. The Heather Trust Annual Report (1995) (eds. J.Phillips & A.Watson), pp. 41 44. Heather Trust, Kippen, UK.
  • Todd, P.A. (1996) The MAFF Molinia study: control of purple moorgrass (Molinia caerulea (L.) Moench) in heather (Calluna vulgaris) moorland after reduction in sheep grazing. The Heather Trust Annual Report (1996) (ed. J.Phillips), pp. 13 16. Heather Trust, Kippen, UK.
  • Todd, P.A. (1997) The MAFF Molinia contract: Year 2. The Heather Trust Annual Report (1997) (ed. J.Phillips), pp. 16 17. Heather Trust, Kippen, UK
  • Van Geel, B. (1978) A palaeoecological study of Holocene peat bog sections in Germany and The Netherlands based on the analysis of pollen, spores and macro- and microscopic remains of fungi, algae, cormophytes and animals. Review of Palaeobotany and Palynology, 25, 1 120.
  • White, J. (1994) Exmoor Birds. Exmoor Books, Dulverton, UK.
  • Yeo, M. (1997) Blanket mire degradation in Wales. Blanket Mire Degradation: Causes, Consequences and Challenges (eds J.R.Tallis, R.Meade & P.D.Hulme), pp. 101 115. Macaulay Land Use Research Institute (for British Ecological Society), Aberdeen, UK.