Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
18 Restoring Derelict Lands in Great Britain Reclamation of derelict lands is of increasing importance for industrial- ized societies, both because quantities of lands so affected are increasing and because the value attached to restoring natural vegetation on these lands is increasing. Early reclamation attempts tried to produce vegetation that looked pleasing and that reduced soil erosion. More recently, society has demanded the creation or restoration of more complex natural com- munities of plants and animals, even though this goal is more difficult and expensive. The task requires knowledge of the tolerances of plants for the environmental conditions; of the ways in which plants compete with one another for access to water, light, and mineral resources; and of the course of succession of plant species after establishment of pioneering vegetation. Less attention has been paid to restoration of animal com- munities, in part because animals often invade these sites naturally once suitable vegetation has been established, and in part because less is known about members of animal communities and how to manipulate them. The Longstone Edge reclamation project in Derbyshire, England, rep- resents the current state of the art of land reclamation. The project drew on existing ecological knowledge about plant competition and succession, but also included experiments to help develop new theories more applicable to the specific conditions of reclamation with which the investigators had to work. A particular part of the challenge one that is generally a com- ponent of such projects was to produce vegetation that would be self- regenerating once established, so that costly maintenance would not need to be continued for long periods. In Britain, most derelict sites are on 248 .
RESTORING DERELICT LANDS IN GREAT BRITAIN 249 very poor soils; the specifics of the Longstone Edge results are, accord- ingly, applicable to many other sites in that country. In North America, however, many sites being restored are on richer soils, and techniques adapted to those conditions, as well as to the more continental climates prevailing in most of North America, will need to be developed to meet challenges there. Case Study PETER WATHERN, Department of Botany and Microbiology, University College of Wales, Aberystwyth INTRODUCTION Bradshaw and Chadwick (1980) estimated the amount of derelict land in the United Kingdom in April 1971 at over 55,000 hectares (ha), and the consensus is that the amount has increased steadily in recent years. Much of this dereliction is the result of mining activities, and pressures on resources in the United Kingdom are such that mineral deposits are likely to be worked with little or no concern for their location. Thus, mining operations often occur in some of the most scenically attractive and environmentally sensitive areas, such as national parks. Examples include Dartmoor (kaolin), Yorkshire Dales (heavy metals and fluorspar), and the Peak District (heavy metals, fluorspar, and limestone). Many areas, including western and northeastern Wales and Cornwall, carry a legacy of dereliction from mines long defunct. In the United Kingdom, reclamation of despoiled land has been a fertile area for research for more than 2 decades, since the pioneering work in the Lower Swansea Valley of Wales reported by Hilton (19671. Recla- mation research has been geared to developing techniques for achieving three main objectives: to restore land so that it can be used productively, to remove local sources of environmental pollution associated with toxic materials in discarded wastes, and to make a site visually attractive again. Reclamation has generally been achieved by establishing a vegetation cover that prevents further weathering and allows such productive uses as agriculture, forestry, housing, manufacturing, and recreation. Thus, rein- statement can be regarded as an essential component of the recycling of land through a succession of uses. In some instances, derelict land detracts markedly from an otherwise pleasing landscape and must be reclaimed in the interests of visual amenity, irrespective of a readily apparent productive after-use. The major concern
250 SELECTED CASE STUDIES in the reinstatement of such sites is the creation of an area that, at worst, blends with the surrounding landscape or, preferably, is pleasing in its own right. This case study describes experimental work undertaken to develop techniques for dealing with such situations. Longstone Edge, Derbyshire, is a limestone scarp situated in a prom- inent position in the heart of the Peak District National Park. It visually dominates an area intensively used for recreation. Fluorspar deposits are found in a vertical vein (locally known as a rake) running for about 1.5 km along its crest. The deposits are worked by a combination of open- cast and deep mining techniques. The area has a long history of lead mining, and the steep scarp face of Longstone Edge is covered with waste heaps tipped to a depth of some 5 m by medieval miners. These waste heaps of gangue minerals, such as fluorspar and barytes, are being re- worked by the mining company. Much of the site, including the medieval waste heaps, is covered with species-rich calcareous grassland, and heath dominated by heather (Calluna vulgaris) has developed on superficial loessic drift (Balme, 1953; Pigott, 19621. The limestone grassland contains several rare species, including Epipactis atrorubens, Minuartia verna, Orchis apifera, and the terricolous lichen Cetraria islandica. As a result of mining activities, seminatural vegetation is being destroyed over much of the site. The general objective of the experimental work described here was to learn whether it would be possible to recreate the seminatural grasslands being destroyed by mining operations. The experimental areas were flat plots beside the worked-out rake. In addition, plots were established on the steep south-facing slopes of the Edge itself. The mining company had attempted reinstatement of some areas that provided a valuable contrast with the experimental plots. The experiments were set up in February- September 1972 and covered about 2 ha. The experiments are continuing, in an attempt to establish the long-term fate of the sown vegetation the only real criterion of success in reinstatement. ENVIRONMENTAL PROBLEMS IN REINSTATEMENT Derelict land reclamation has been the subject of intensive research in the United Kingdom so long that few waste materials or mine sites present insurmountable problems for reinstatement. The existence of large amounts of derelict land reflects the limited resources that have been made available for such work, rather than a lack of appropriate techniques. As a result of the work mainly of A. D. Bradshaw and his co-workers at Liverpool University since the early 1960s, even seemingly intractable wastes con- taminated by toxic substances can be revegetated with ecotypes able to
RESTORING DERELICT LANDS IN GREAT BRITAIN 251 tolerate the presence of these materials. Toxic metals present such a special problem that there is probably no alternative to the use of this approach. However, the treatment of nontoxic wastes has been much debated. The primary consideration has been to develop a recipe that can be used to produce a quick green cover, irrespective of its appropriateness as a long- term solution. This search for universal recipes that eliminate, rather than accommodate, constraints on vegetation development has colored much of the research on derelict land reclamation. The major constraints on plant growth on derelict land, apart from possible toxic effects, are low nutrient status, low soil organic content, drought stress, and, if the material has been mechanically sorted, uniform particle size. Abandoned mineral workings and waste materials such as alkali wastes (Lee and Greenwood, 1976), quarries (Davies, 1976; Hodgson, 1982), and wastes contaminated with heavy metals (Holliday and Johnson, 1979; Johnson, 1978; Johnson et al., 1978) often develop into important pre- serves for seminatural vegetation. The evidence suggests that there is no need to reclaim derelict or abandoned mineral workings, because an in- teresting vegetation will develop naturally, given sufficient time. But that stance neglects important changes that have rendered old sites unreliable as models. The changes affect all three main phases of the diversification of sown vegetation (Wathern, 19771: · Transfer. Propagules of the species must be dispersed from an es- tablished stand and arrive at the reinstated site. · Establishment. Propagules must become established and develop into mature plants. · Periodic recruitment. From time to time, new individuals derived from the original source or from the new population must become estab- lished, and the rate of recruitment should at least equal the death rate, if the population is to be sustained. The abandoned mineral workings that are now biologically so interesting began their development in the early decades of this century. The landscape over much of the United Kingdom was different from that seen today. Seminatural vegetation was the basis of much of the land use. Wood products were derived largely from broad-leaved woodlands, rather than softwood plantations. Pastoral agriculture relied on the use of seminatural grassland for grazing and hay. Thus, the countryside had a basic matrix of seminatural vegetation that could serve as a source of propagules for the diversification of abandoned mine workings. Transfer distances were short and within the capabilities of most species. The preponderance of wind-dispersed species on the alkali wastes highlighted by Lee and Green- wood (1976) indicates the importance of isolation, in that these waste
252 SELECTED CASE STUDIES areas are many miles from potential seed sources. Species with other types of dispersal mechanism usually move only a short distance with each generation and can move long distances only incrementally through a chain of suitable habitats. At present, however, species are becoming concentrated in fewer and fewer refuges. It is likely, therefore, that a quarry site abandoned today will be many miles from potential sources of propagules and that transfer distances will be too great for many species. In addition, because mining and quarry operations have changed over the years, it is unlikely that modern sites will acquire the same ecological interest as the old sites. Large-scale workings are preferred, and mineral extraction produces uniform conditions over extensive areas. A modern limestone quarry lacks the physical diversity of an old quarry with its uneven face, rubble-strewn floor, and scattered waste heaps, which ac- count for much of the biological interest. Such diversity can be created only by a program of treatment after quarrying operations have ceased. Many of the old skills have been lost, but Humphries (1977) describes how, for example, a quarry face could be treated. The final objection to a laissez-faire attitude to site rehabilitation is not biological. Society will no longer tolerate the visual intrusion of a derelict mineral working or quarry over a long period in the hope that interesting habitats will eventually develop. Industry must be seen to be rehabilitating environments degraded by its activities. The challenge in derelict land reclamation is to reinstate a site so that its visual impact is reduced, while maintaining its potential for development into a biologically interesting area. The traditional approach to reclamation of derelict land for visual amen- ity has been based on agricultural techniques. In the period 1967-1972, the mining company made considerable efforts to revegetate areas on Longstone Edge where extraction was complete or where there was a temporary halt in operations. The approach adopted was to sow agricultural seed mixtures (Lolium perenne, Dactylis glomerata, and Phleum berto- lonii) on raw subsoils with heavy fertilizer applications. It soon became apparent that this agricultural approach would provide no long-term so- lution. Whenever such grass-seed mixtures were sown, a green cover developed briefly. After a short period of vigorous growth, swards became moribund, the grasses showed signs of nutrient deficiency, and standing crops declined. On these free-draining soils lacking organic matter, nu- trients, particularly nitrogen, were leached out. Periodic applications of fertilizers became the main form of management used by the mining company. Clearly, there would be considerable advantages in developing some form of single-application treatment that would lead to stable veg- etation. It would be more visually acceptable to the public and to the
RESTORING DERELICT LANDS IN GREAT BRITAIN 253 controlling authority, the Peak Park Planning Board, and it would have obvious economic benefit. Either of two approaches could be adopted: constraints on the produc- tivity of high-yield species could be eliminated, or constraints on the system could be accommodated by creating some form of low-productivity system whose demands for nutrients would be better attuned to the ability of the system to supply them. The first approach has been used successfully in the revegetation of kaolin wastes in Cornwall. Kaolin wastes are nu- trient-deficient, drought-stressed environments (Bradshaw et al., 1978), and fertilizers must be added regularly to maintain vegetation cover (Brad- shaw et al., 19751. To overcome this problem, a biological source of nitrogen was sought. Rhizosphere fixation was inadequate to achieve the target of about 700 kg/ha estimated by Bradshaw et al. (1975) to be necessary to produce a self-sustaining system. The potential of nitrogen fixation by nodule bacteria in legumes was investigated in a series of trials. The results showed the use of legumes to be a means of overcoming nutrient limitations (Jefferies et al., 19811; Lupinus arborous, the tree lupine, was the most effective species (Palaniappan et al., 1979), but is not native to the United Kingdom. Although sowing legumes is useful for reinstating some nutrient-deficient substrates, it was rejected for Longstone Edge, because the seminatural vegetation of Longstone Edge comprises calcareous grassland and Calluna heath, both of which are naturally nu- trient-deficient systems. A productive agricultural sward created on the mineral working would be out of context with the surrounding vegetation. Consequently, it was decided to attempt to produce a low-productivity vegetation in harmony with the existing nutrient-deficient system. The potential of these swards appeared great. Grime (1973a) has argued that environments moderately stressed by drought, nutrient deficiency, or graz- ing support diverse communities. In contrast, high-productivity systems are often species-poor. These low-productivity grasslands could be ex- pected to develop into diverse swards akin to the existing seminatural vegetation. Local populations of plants could not be used as the sole source of propagules in this research, because of the difficulty of interpreting the experimental results. Because the natural arrival of propagules, mainly seeds, in the experimental plots would be a chance occurrence, it would be impossible to determine whether the failure of a particular species to invade an area resulted from a lack of propagules or from an inability to establish under the prevailing conditions. The ability of a sward to ac- commodate a new species is termed its receptivity (Wathern, 19761. Nat- ural seed rain provides a flow of only a small irregular number of propagules into experimental areas; that makes it difficult to measure the receptivity
254 SELECTED CASE STUDIES of contrasted swards. Therefore, we diversified the swards artificially by introducing known quantities of seeds of a range of native species, hoping to create a visually pleasing community rapidly. The decision to introduce native species deliberately raised an important issue, which has since achieved greater prominence. Deliberate introduc- tions are discouraged by scientists, mainly because they disrupt natural distribution patterns, particularly of local ecotypes. However, because plant distribution in the United Kingdom has been modified drastically by human activity over a long period, that objection has little validity. In these experiments, only commonly occurring species were used, and they were collected from local seed sources whenever possible. This is an ethical consideration that requires more thorough discussion, because fail- ure to transfer species artificially might doom them to local extinction. The ethics of introductions in the construction of natural ecosystems are discussed by van der Hock (1982~. APPROACHES TO RECREATING SEMINATURAL GRASSLAND Experiments were established to investigate whether the yield of species on the Longstone Edge subsoils was related to relative growth rate (RGR). Monocultures of Lolium perenne (RGR~`, 1.30), Festuca rubra (1.18), and F. ovina (1.00) showed an inverse relationship between yield and RGRmaX after 3 years. Thus, the species with the greatest potential for growth, L. perenne, performed least well. Other trial areas were sown with two contrasting mixtures one consisting of F. ovina and F. rubra and the other consisting of L. perenne, Dactylis glomerata (1.31), and Poa trivialis (1.40) and monitored for 5 years. After 1 year, the Lolium sward had almost twice the standing crop of the Festuca sward. After 2 years, both grasslands had increased in yield, but they did not differ significantly from one another. After 5 years, the Lolium sward had the same standing crop as it had after 2 years; in the intervening 3 years, it had become moribund. In contrast, the Festuca sward continued to increase its standing crop throughout the experiment. The results of this experiment, described more fully by Wathern and Gilbert (1979), confirmed that nutrient-deficient subsoils could be rein- stated successfully with low-RGR grasses. Grasses, however, provide only the general matrix of the community in seminatural grasslands. In the United Kingdom, many of the distinctive elements, and certainly the visually dominant species, are dicotyledons. Indeed, the seminatural grasslands of the carboniferous limestone are renowned for their floral richness. Consequently, in parallel with the trials
RESTORING DERELICT LANDS IN GREAT BRITAIN 255 on low-productivity swards, the feasibility of recreating seminatural grass- lands was investigated. Experiments were established as long-term trials in September 1972 and still cannot be considered complete. For the first 5 years, the plots were monitored fairly intensively, but they have since been assessed only infrequently. One area of Longstone Edge provided an opportunity to investigate the effect of a grass seed mixture on the rate of invasion by native species under natural conditions. Three years before experiments on Longstone Edge began, the company had regraded a large area. Half this area had been sown with a grass seed mixture, and the remainder had been inex- plicably left fallow. The vegetation of the two sections was surveyed, but no important differences in the number of adventive species could be found. Initial seeding with a mixture producing an open sward did not appear to affect the rate of diversification. This suggested that the two- part objective of reinstatement reducing visual impact without signifi- cantly affecting sward receptivity could be achieved. The second major conclusion from this work was that, although diversification occurred naturally, most of the adventives were species that produced large numbers of highly dispersable seeds. It appeared that, without deliberate introduc- tion, the more characteristic species of climax limestone grasslands with larger, less readily dispersed seeds might not become established in the short term. In the light of these observations, experiments were set up to establish the most appropriate means of diversifying swards and to determine the receptivity of contrasted swards to native species. The Festuca- and Lol- ium-based mixtures used in the low-productivity sward experiment de- scribed above were used. Various sources of propagules of native species were investigated. Hand-collected seed was broadcast at known rates in the swards and fallow areas. After 2 years the Festuca sward contained 22 sown species; six species showed an establishment rate of over 10%. The Lolium and fallow areas contained 15 and 17 species, respectively, at low frequency. After 5 years, although the frequency of native species had increased in all areas as a result of additional recruitment, self-set seed, and vegetative spread, the differences among the Festuca, Lolium, and fallow plots were even more marked than they had been after 2 years. Thus, not only was the Festuca sward initially more receptive to native dicotyledons, but this increased receptivity was sustained over a long period. Adventive species also showed higher frequencies in Festuca areas and after 12 years included important elements of the limestone grassland flora. In a separate experiment, sowings along a soil depth gradient showed that sward performance determined establishment rates for native species.
256 SELECTED CASE STUDIES Establishment was greatest in swards of intermediate density. In such swards, competition is not severe, but there are enough plants to ameliorate the microclimate of developing seedlings. Dense swards eliminate native species by competitive exclusion, whereas failure rates of seedlings in open situations are high because of harsh environmental conditions (Wath- ern and Gilbert, 19781. These experiments suggest that low-productivity swards increase the rate of diversification by native species and that sowing seed of native species by hand can create visually pleasing grasslands rapidly. Other sources of native species were investigated. From casual obser- vations on the site, it was noted that fragments of old turf had become established, particularly in compacted bulldozer tracks, providing small islands of seminatural grassland. Therefore, turf stripped from an area about to be reworked by the mining company was used in the diversifi- cation trials with an oversewing of the Festuca mixture (Gilbert and Wathern, 19801. The use of seed-rich topsoil on this site was not suc- cessful, but the approach should not be dismissed, inasmuch as it has been successful in other experimental areas. Seed-rich Calluna litter (Gim- ingham, 1960) has been used to produce a Calluna heath community on an acidic subsoil road verge on highway A57 west of Sheffield (Gilbert and Wathern, 1976~. Similarly, Farmer et al. (1982) and Tacey and Glos- sop ( 1980) have since used seed-rich topsoils for the reinstatement of mine areas. To obtain additional information on the fate of man-made swards, a chronosequence of grasslands up to 150 years old was analyzed. A sample of 69 quadrats covering the complete range of grassland age was recorded. Soils were taken for laboratory analysis to determine whether edaphic factors were controlling the composition of the vegetation. For grasslands less than 25 years old, specifications for the original seed mixtures were known. The detailed procedures, particularly the seed mixtures, used to create the older grasslands could not be discovered. The major finding of this study was that soil nutrient status was the most important factor in controlling succession in urban grasslands (Wathern, 1976, 19801. USES OF KNOWLEDGE Of the ecosystem modifications listed by Wathern (1984), the most important changes resulting from development that influence reinstatement appear to be simplification, the opening of previously closed nutrient cycles, and the loss of vegetation types. The objective of revegetation
RESTORING DERELICT LANDS IN GREAT BRITAIN 257 should be to redress that progression. Many of the objections to devel- opment are based on amenity, so it seems appropriate for reinstatement to aim at creating vegetation that has or can rapidly acquire amenity value. The application of agricultural technology to this problem, however, has often failed to achieve this objective. In the case reported here, ecological knowledge was used to develop a more subtle approach to reinstatement. The revegetation of mineral workings on Longstone Edge was designed to satisfy five criteria: reinstated swards should be inexpensive to produce, result from a single simple treatment, require no intensive management, be visually pleasing, and resemble surrounding seminatural vegetation. The research involved field observations and theoretical considerations aimed at understanding the relationships between reinstated grasslands and the environment and then application of this information in devising a reclamation scheme. The application of ecological considerations pervaded the study. However, ecological knowledge figured most prominently in four ways: production of low-productivity swards, sward diversification, assessment of the potential of soil seed banks, and urban grassland succes- sion. Low-Product~vity Swards The most important concept used in this phase of the work involved the ecological importance of relative growth rate. Hunt (1970) and Grime and Hunt (1975) measured the RGR of 115 species in a uniform high- nutrient environment. The RGRma,, of each species was determined. This study showed that low-productivity environments are characterized by species with low RGRma,~, whereas productive systems contain species with high RGRmax. The agricultural species used previously in landscape reinstatement have high RGRm~. When sown on Longstone Edge, they initially produced a green cover, because artificial fertilizers provided the high concentrations of nutrients required for their growth. These skeletal soils, however, have a low organic content. Under these conditions, nutrients are not held within the system, but rapidly leach out. For the highly productive species to flourish, fertilizers must be added regularly. From Hunt's observations, it can be predicted that such soils would naturally carry species with low RGR. If those species were sown during reinstatement, the results should be more satisfactory than those obtained with agricultural species. Thus, the work on RGR provides the theoretical basis of the reinstatement of low-productivity swards. In turn, the experiments provide a field trial of the importance of growth rate in explaining plant distribution.
2S8 SELECTED CASE STUDIES Sward Diversification High RGR, which reflects the ability of a species to use resources that are freely available, is only one of the characteristics by which one plant can gain a competitive advantage over another. Structure and height affect competition for light, and the ability to produce litter that persists from one growing season to the next is an aid to occupying space when a plant is not actively growing. Grime (1973a) argued that these four attributes can be amalgamated into an index of competitive ability. Species with high competitive index ("competitors" in Grime's terminology) grow in productive environments where resources, such as nutrients and water, are freely available. Such environments are species-poor, because competitors eliminate other spe- cies that are less able to exploit the resources. Environments where re- sources or physical and chemical constraints severely limit plant production, described as "stressed," are also species-poor. Only a few species are adapted to tolerate extreme conditions of, for example, drought or heavy- metal contamination. Such species are characterized by very low RGRm~,. Composition of a particular community, however, is dictated not only by the growth rates of the constituent species and by the physical and chemical characteristics that can limit the full realization of this potential. The degree of disturbance of the environment, such as grazing, is a third factor. Poverty of species is associated with a high incidence of distur bance. Grime advocated a triangular ordination as the means of analyzing community composition, the axes being the relative importance of com- petition, stress, and disturbance. At the extremes, species-poor commu- nities are present. Diverse communities are found where competition, stress, and disturbance are all moderate, and they are characterized by species that exhibit the "C-S-R" strategy in Grime's terminology. C, S. and ~ stand for competitive, stress-resistant, and ruderal (weedy). Such plants are "confined to habitats in which competition is restricted to moderate intensities by the combined effects of stress and disturb- ance . . ." (Grimes 19771. The Longstone Edge reinstatement areas have this combination of features, so they should ultimately develop into rich communities of C-S-R fortes. The major representatives of C-S-R strategies in the U.K. flora are small tussock grasses, small deep-rooted fortes, small stoloniferous species, fortes with short rhizomes, legumes, and small sedges and rushes (Grime, 19791. The major grasses used successfully to form the low-productivity sward on Longstone Edge were two small tussock-formers, Festuca ovina and
RESTORING DERELICT LANDS IN GREAT BRITAIN 259 a nonstoloniferous variety of F. rubra (var. fallax). The grasslands were diversified by using the seeds of wildflower species that could be collected in the area. Many achieved high rates of seedling establishment and con- version to mature plants. Several of the species that Grime (1979) lists among a short illustrative group of U.K. C-S-R strategists established successfully from seed in the Longstone Edge trials. In contrast, several tall herb species included for comparative purposes showed initial high rates of seedling establishment, but then failed to survive. Thus, Grime's concepts of community ecology are directly relevant to derelict land reclamation. Although the concepts were known in general terms at the start of this research, the detailed design of the experiments was dictated by pragmatic considerations, such as seed availability. Before the trials, it was known that diverse natural grasslands were characterized not only by species with low growth rates, but also by environmental constraints on plant productivity. Such considerations influenced the basic suppositions underlying the diversification trials. The substantial theoret- ical developments that took place during the project permitted a more thorough and systematic analysis of the results and set the conclusions in a broader context. Soil Seed Banks Established ecological theory was used to appraise the potential of soil as a source of propagules for diversification. Initially, only the general characteristics of soil seed banks had been determined. Seed banks from beneath particular types of seminatural vegetation have now been described (Brown and Oosterhuis, 1981; Donelan and Thompson, 1980; Thompson and Grime, 1979), but when our experiments were being formulated, most of the literature was related to the weed populations of arable systems (see Harper, 19771. Spray (1970) and Grime (1978) characterized the soil seed bank of tall herb communities. On the basis of high numbers of some native species, Spray suggested that seed-rich topsoil could be used to diversify reinstated vegetation. Experiments were undertaken to establish whether his tech- niques could be used on a large scale. Soil from beneath limestone turf was taken from a part of the site about to be worked for minerals. In laboratory germination tests, this soil was found to contain seeds of various native species. The field trials were not successful, but the reasons for the failure have never been adequately explained.
260 SELECTED CASE STUDIES Succession in Urban Grassland Sown vegetation is not static. Over the course of time, it changes as a result of a complex array of interacting factors. Succession is a phenom- enon that must be accommodated by any vegetational scientist involved in derelict land reinstatement. In the United Kingdom, few data on grass- land successions are available. Wells (1965) monitored the changing com- position of a sown agricultural grassland. Of particular interest, however, is the description of a complex of grassland types that have developed on formerly arable land on Salisbury Plain after various periods of abandon- ment (Wells et al., 1976~. Changing composition was associated with nutrient accumulation within the ecosystem. Analysis of a chronosequence of sown urban grasslands provided some insight into the fate of vegetation created for visual amenity. In Sheffield, extensive areas of grassland have been produced for public open space around municipal housing in recent years. The age of the grasslands, up to 25 years, and the seed mixtures originally sown were known from the records of the local authority. Older grasslands were discovered around large country houses in the vicinity. In the case of the latter grasslands, the composition of the sowing mixtures could not be determined, but the chronosequence could be extended over a 150-year period with estate records. The results of this survey are discussed by Wathern (1976, 19801. Diversification is the most important successional phenomenon for re- vegetation studies. The process in natural successions is accompanied by the accumulation of materials, particularly nutrients, within the system (Wells et al., 19761. In the urban grassland chronosequence, no increased diversity with age could be detected; older swards did not contain more species. However, the soils beneath these grasslands are subject to leach- ing. Grime (1973a) has shown that the number of species in seminatural grassland in the Sheffield region is inversely related to the acidity of the soil. Inasmuch as few species tolerate extremely acidic soils, a chrono- sequence in which there is a decrease in pH is likely to be one in which the number of potential recruits also decreases. Grime's data make it possible to separate the effects of diversification from pH changes. For each quadrat in the urban grassland survey, the number of species present was divided by the number of species anticipated in a soil with the pH of that quadrat. Values of a relative diversity index (Wathern, 1976), when plotted against age, showed that as a grassland gets older it acquires more of the species that are capable of growing at the appropriate pH. The results confirm the conclusion that succession is important in molding the composition of sown grasslands and indicate that many species invade
RESTORING DERELICT LANDS IN GREAT BRITAIN 261 only after long periods. Invasion of Calluna into these grasslands under conditions of natural seedfall, for example, takes about a century. The second major conclusion with important implications for the di- versification trials concerned the significance of soil nutrient status in determining composition. This was assessed with the Rumex acetosa seed- ling bioassay developed by Rorison (1967), a standard technique used in Sheffield. Increasing relative diversity was associated with decreasing nutrient status. These results are at variance with the observations of Wells et al. (1976), but it should be noted that soils in the urban grassland chronosequence initially had high nutrient concentrations as a result of fertilizer applications to topsoils. The implications with respect to the triangular ordination of community characteristics are discussed below. The observations suggested that nutrient deficiency in the limestone subsoil of the mineral workings on Longstone Edge should not be regarded as a major constraint on vegetation development, but potentially as an asset that should lead ultimately to high diversity if an appropriate method can be developed. SOURCES OF ECOLOGICAL INFORMATION The overall objective of the Longstone Edge trials was to create semi- natural vegetation resembling the surrounding diverse grasslands on the nutrient-deficient subsoils of the reinstated mineral workings. The research had two basic assumptions: sward production depends on the relation of soil nutrient status to the growth rate of constituent species, and diversity depends on sward productivity. Each experiment was established to ad- dress these factors, with the intention of combining the results in the development of an approach for the reinstatement of this and similar sites. The important ecological sources in this case, therefore, were those which either aided identification of soil nutrient status and relative growth rate as key issues in the research or influenced experimental design. As an illustration of the application of ecological theory to the resolution of ecological problems, the Longstone Edge revegetation trials are un- usual. The development of the experimental work preceded the publication of much of the ecological theory to which it is most closely allied. This apparent anomaly can be explained only with a thorough understanding of the situation in which the work was conducted. The evolution of an experimental research project is a complex process and one in which it is exceedingly difficult to unravel the detailed history of an idea. In a research environment like a university, many important innovations begin in a few chance remarks, in discussions with colleagues
262 SELECTED CASE STUDIES working in related fields, or in questions raised in research seminars. It is often difficult later to apportion progress among these causal factors. Chance occurrences often change the course of research or provide valuable information that permits a more comprehensive analysis of re- sults. For example, the soil-gradient experiment described earlier provided much information on the effect of sward performance on the establishment of native species. The experiment was not part of the original scheme; in some ways, its establishment was accidental. The company requested the use of the low-productivity sward mixture on this part of the site and, because surplus seeds of native species were available, this was added almost as an afterthought. Therefore, although planning is an essential component of good research, no program should be so intensely planned that there is no room for exploiting chance occurrences. The factors that influenced the evolution of approaches to the reinstate- ment of disturbed land can be considered in four categories: ecological facts, ecological theory, analog studies, and expert judgment. Ecological Facts The major impetus in experimental design came from the field obser- vations of research workers on site, who analyzed unsuccessful attempts at reclamation undertaken by the company, and from a detailed appraisal of the likely causes of failure in the light of ecological knowledge. A number of published and unpublished sources aided the study of low- productivity swards and grassland diversification. The experimental grass sowings on Longstone Edge were designed to assess the performance of different species on limestone subsoil and to investigate the feasibility of creating low-productivity swards. Initial spe- cies selection therefore had to take account of the commercial availability of seed, as well as of potential productivity. Hunt (1970) compiled data on the relative growth rate of a range of species found in different habitats in the Sheffield region. These data confirmed the differences in growth rates of the species selected for the grass monoculture and the low-pro- ductivity sward experiments and allowed these differences to be quantified. The urban grassland survey described above was also important in species selection. The results showed that the younger grasslands on soils with high nutrient status were dominated by Lolium perenne swards, whereas older nutrient-poor soils carried grasslands dominated by fine-leaved Fes tuca species. Although the older urban grasslands were more interesting, there was no evidence from an initial analysis of the data that they were more diverse than younger grasslands. Not until after the publication of the Grime
RESTORING DERELICT LANDS IN GREAT BRITAIN 263 (1973a) paper, after the Longstone Edge trials were established, were the data reinterpreted to take account of the pH effects discussed above. The effect of modifications of soil nutrients on species composition in the United Kingdom, however, has been known for many years, with respect to upland grasslands (Milton, 1940), lowland meadows (Brenchley and Warington, 1958), dunes (Willis, 1963), and limestone grasslands (Jef- frey, 1971; Jeffrey and Pigott, 19731. The major conclusion from these experiments is that grassland com- position can be manipulated by changing nutrient status with fertilizer applications. The park grass experiment described by Brenchley and War- ington (1958) resulted in the production of grassland types totally different from a common sward over the course of a century of contrasted fertilizer applications to trial plots. Most important, these investigations showed the high diversity of swards on soils of low nutrient status. This information motivated our expectation that rich grassland would develop on the nu- trient-poor subsoils of the reinstated areas. Data on the relative growth rate and competitive ability of different species proved to be the most valuable unpublished materials available to the study. The interrelationships among environmental characteristics, competitive ability, and plant strategies were already being investigated by members of the Unit of Comparative Ecology at Sheffield University. Grime proposed that nutrient-rich communities were species-poor because of the competitive dominance of the constituent species. A competitive index, quantifying this dominance, had been developed and calculated for 441 species present in the Sheffield region. The index was based on maximal height, structure, and maximal accumulation of persistent litter. The data collected by Hunt (1970) on relative growth rates were also included. Only incomplete data related to ~ 15 species were available, so growth rate was not included formally in the index. The index confirmed the high competitive ability of agricultural grasses compared with that of the low-productivity sward mixture. In more recent discussions of com- petition (Grime, 1973a), growth rate has been incorporated into the index as a result of the availability of additional data. The Unit of Comparative Ecology at Sheffield has been collecting sys- tematic data on the characteristics of grasslands in the Sheffield region for many years. These data were used to produce a computerized data base on grasslands and their constituent species, but the first analysis of grasslands in the area (Lloyd et al., 1971) was too general to be valuable during this study. The data were also used to produce an ecological atlas of common grassland plants (Grime and Lloyd, 1973), including detailed information on the environmental characteristics of the habitats in which each of 94
264 SELECTED CASE STUDIES species occurs. This atlas has been described as a practical aid for land- scape managers involved in, for example, revegetation, but that appears to be a post hoc rationalization of its function. Selection of species for landscape reinstatement is determined by pragmatic considerations, such as seed availability, ease of establishment, and likelihood of success. These considerations are not addressed in the atlas. For example, productivity of the systems in which a species occurs, acknowledged elsewhere by Grime as a major factor in determining distribution, is not considered. The atlas therefore has the wrong emphasis to be valuable in revegetation. Consequently, the information on distribution was not used in the re- vegetation trials. Species from a wide range of habitats were selected for the diversification experiments. It was felt that this approach would be better, because it would provide information on the conditions necessary not only for success, but also for failure. Ecological Theory The results of the seral study on urban grasslands and the subsequent diversification trials illustrate well the implications of the theory of plant strategies and community ecology for reinstatement studies. There were important changes in the status of this theoretical background over the course of the work described here. At the outset, the theory was still being developed by Grime in Sheffield. It was not published until most of the experiments had been set up, although the general tenor of the ideas was known. Thus, during formulation of the reinstatement work, the urban- grassland study and papers on the relationship between diversity and nu- trient status were more important. Preliminary analysis of the experimental work began in 1974, after publication of the basic theoretical concepts (Grime, 1973a,b, 19741. The theory has been further developed by Grime (1977, 1979), but not all aspects have been accepted without dissent (see, for example, Newman, 19734. With this more comprehensive theoretical base, the results of the experimental work became more important and could be set within a broader context. The results demonstrate that diverse vegetation can be created artificially under conditions of moderate stress, as the theory pre- dicts, and that the techniques used constitute a worthy strategy for rein- statement of disturbed land for visual amenity. Analog Studies No reports of previous attempts to reinstate mineral workings to semi- natural vegetation were discovered during a systematic search. Work on
RESTORING DERELICT LANDS IN GREAT BRITAIN 265 diversification of road verges, however, is relevant to the study. Additional diversification trials were established on road verges in the Sheffield area, because they have the potential to make a major contribution to the reserves of seminatural habitats in the United Kingdom (Mellanby, 19741. The results were reported by Wathern and Gilbert (19781. A number of previous attempts at establishing native species in verges have been reported in the conservation press, and these were investigated in detail. Most cases were in the form of "rescue" operations, in which popu- lations of rare species were threatened by highway construction. Generally, mature plants were returned to the verge after a period in a nursery to bridge the construction phase. The most spectacular success has been the mass planting of native Narcissus pseudonarcissus (daffodil) on the M40. There were a few attempts at more general diversification. Calluna turf was used successfully for reinstatement at the Devil's Punch Bowl, Surrey, after roadwork and at Cumbernaud New Town. Commercially available herbs have been used on road verges in southeastern England. The results of this work have never been published. Significantly, some members of voluntary organizations had suggested more general diversification of highway verges in their areas, but such proposals were invariably aban- doned after objections on the grounds that distribution patterns would be disrupted. The main value of the analog studies was in confirming that artificial diversification could be achieved. In general, the labor-intensive approaches were rejected as being inappropriate for large-scale projects. Experimental work in the Netherlands on landscape reinstatement was of particular interest, but it was not seen until the Longstone Edge trials had been established. The Dutch experience was useful in interpretation of the experimental results, because a similar approach had been adopted independently in many instances. For example, the problem of low-pro- ductivity swards on sandy soils was investigated early (Hoogerkamp, 1971), and trials on the establishment of Calluna on highway verges had shown the importance of a nurse crop in seedling establishment, as was found in the United Kingdom (Gilbert and Wathern, 19761. Expert Judgment It should be clear from this discussion that a university provides a unique atmosphere for research. The interplay of personnel leads to a rapid evo- lution of ideas. The Longstone Edge study was undertaken in the De- partment of Landscape Architecture of Sheffield University as a doctoral research project. The supervisor was an experienced ecologist, and the research student a botany graduate. There were also important links with the Botany Department, in particular its Unit of Comparative Ecology.
266 SELECTED CASE STUDIES These contacts occurred during a formative and productive period for the Unit, when ideas concerning plant strategies and community ecology were being developed. The work on the comparative ecology of individual species and the composition of plant communities might seem esoteric, but the revegetation work on Longstone Edge has shown its relevance to reinstatement. Although not published, information was freely available to the study, and informal discussions around this work were useful in the development of ideas concerning reinstatement. Consultations with other organizations were not generally important in the evolution of the work. However, discussions with experts outside the university system, particularly government scientists in the Institute of Terrestrial Ecology (ITE) and the Nature Conservancy Council (NCC), took place. In addition, a British Council travel fellowship enabled the work undertaken in the Netherlands to be discussed with government scientists at IBS Wageningen in 1974. The responses of research scientists differed. Most research workers were concerned primarily with investigating the effects of management on grassland composition. A few research scientists at ITE, however, were also interested in creating seminatural grassland; indeed, they set up experimental work in 1973 and 1978 (Wells, 19831. A guidance manual on diversification has also been prepared (Wells et al., 19811. The dis- cussions in 1971 generally concerned the need to develop techniques, rather than how this might be done. Sowings of wildflower species became quite common in the Netherlands at one time (Cole and Keen, 19761. During the discussions with Dutch scientists, however, the ethics of the approach were called into question. This hinged not on theoretical considerations of disrupted distributions of different ecotypes, but on a more fundamental point: whether people should interfere in the development of an ecosystem. The attitude of conservation officers in the United Kingdom was de- fensive and antagonistic. Many saw within the techniques a potential diminution in their ability to object to development proposals on conser- vation grounds, because they believed that any suggestion that plant com- munities could be recreated after development would undermine their position. This attitude reflected a failure to understand the concepts un- derlying the techniques. There is no suggestion of constructing commu- nities that contain all manner of rare species. The approach seeks merely to redress somewhat the losses of seminatural vegetation that are occurring. The alternative is that yet more highly productive grazing land will be created in reinstatement, irrespective of location. This misconception con- cerning the overall approach, unfortunately, continues.
RESTORING DERELICT LANDS IN GREAT BRITAIN CONTRIBUTION OF ECOLOGICAL KNOWLEDGE 267 A Peak Park Planning Board official has described the Longstone Edge trials discussed here as "the most exciting piece of reinstatement work which has been undertaken within the area." The excitement had two sources. First, the resulting sward is diverse and contains many of the elements of seminatural calcareous grassland. The presence of these char- acteristic species is the result of both deliberate introductions of propagules and chance invasions. Even 12 years after establishment, the grasslands are still acquiring additional species. Second, development of these tech- niques marks an important departure from previous approaches to rein- statement. Formerly, reclamation was based on agricultural technology with re- medial measures, so the environment was modified primarily by fertil- ization to support the species sown. The scheme adopted on Longstone Edge was based on a more objective ecological approach. In the simplest terms, it can be described as an attempt to sow species capable of surviving under the environmental regime of the limestone subsoils. The change in emphasis was possible because the objectives of reinstatement were re- considered and ecological theories concerning the control of community composition were reappraised. Development of theory was particularly useful in two respects: it could be used to explain the composition of seminatural communities in the area and the grassland chronosequence studied early in the research (it was on the basis of these field observations and the literature on the effects of fertilizers on grassland diversity that the more intuitive experiments were established), and, during interpre- tation of the experimental results, it enabled success and failure to be explained, rather than merely noted. There is one major criticism of the work. Covering no more than 2 ha, the plots are at best large trials. With one exception, they have yet to be scaled up. Although increased scale should not affect the outcome of reinstatement, it would establish whether the approach is usable on an industrial scale. The possibility of developing a "wildflower" mixture for reinstatement schemes based on commercial sources of native species has been considered and rejected. Commercial seed is prohibitively ex- pensive, but there are more serious ecological objections to such a mixture. The research was developed in an attempt to avoid the previous approach- namely, producing a recipe to be applied in all situations even if it could guarantee a visually attractive grassland. The species that are available in commercial quantities constitute inappropriate ecotypes. For example,
268 SELECTED CASE STUDIES native populations of Lotus corniculatus figured prominently and suc- cessfully in the experiments, and this species seems to be useful in rein- statement, whereas commercial material is a robust, erect ecotype from central Europe that is selected for its high yield and ease of harvesting and is quite unlike the calcareous grassland ecotype found on Longstone Edge. Hand collection of local ecotypes is far more satisfactory, but its feasibility and, more important, its impact on local populations are not known. Some form of bulking up might be required, because large quan- tities of seed could be involved. Wells (1983), for example, recommends up to 2 kg/ha for some species. These practical considerations must be resolved. Work on Calluna has been scaled up by other workers, and the creation of Calluna heathland has generated considerable interest. Large areas of moorland and heath over much of the United Kingdom are degraded and in need of reinstatement. In addition, serious visual intrusion has resulted from the use of traditional methods of reinstatement in the revegetation of such developments as pipeline rights of way. Reinstatement of Calluna heathland is now feasible on a large scale as a result of further field trials (Meaden, 1984; Moorland Restoration Project, 1983; Putwain et al., 19821. The Longstone Edge research has stressed the necessity for a long-term perspective in analyzing the results of reinstatement schemes. Results over short periods can have misleading implications for small-scale pilot schemes. Differences that were encountered in these trials over time could be ex- plained by applying theories of community ecology. The main value of the case described here concerns the interrelationship between theoretical ecology and applied ecology. The study has provided experimental support for theories of community ecology, and theoretical considerations, particularly of plant strategy and community ecology, help to weld the experimental work into a coherent whole. This unity converts what might have been regarded merely as an interesting set of empirical observations on a set of reinstatement trials on a small site in Derbyshire into a radical alternative strategy for the revegetation of derelict and des- poiled land. ACKNOWLEDGMENTS The work described here was undertaken in collaboration with my research supervisor, Oliver Gilbert, and formed the basis of a doctoral thesis submitted to the University of Sheffield in 1976. The work was funded by a research studentship from the Natural Environment Research Council. The assistance of the mining company, Laporte Industries Ltd.,
RESTORING DERELICT LANDS IN GREAT BRITAIN 269 and Jack Harwood in providing facilities for this work is gratefully ac- knowledged. The research should be regarded as the joint work of Oliver Gilbert and me, but the observations and interpretation of events included in this report are my responsibility. REFERENCES Balme, O. E. 1953. Edaphic and vegetational zoning on the carboniferous limestone of the Derbyshire Dales. J. Ecol. 41:331-344. Bradshaw, A. D., and M. J. Chadwick. 1980. The Restoration of Land. Blackwell, Oxford, Eng. Bradshaw, A. D., W. S. Dancer, J. F. Handley, and J. C. Sheldon. 1975. The biology of land revegetation and the reclamation of the china clay wastes in Cornwall. Pp. 363- 384 in M. J. Chadwick and G. T. Goodman, eds. The Ecology of Resource Degradation and Renewal. Br. Ecol. Soc. Symp. 15. Blackwell, Oxford, Eng. Bradshaw, A. D., R. N. Humphries, M. S. Johnson, and R. D. Roberts. 1978. The restoration of vegetation on derelict land produced by industrial activity. Pp. 249-274 in M. W. Holdgate and M. J. Woodmand, eds. The Breakdown and Restoration of Ecosystems. Plenum, New York. Brenchley, W. E., and K. Warington. 1958. The Park Grass Plots at Rothamsted, 1856- 1949. Rothamsted Experimental Station, Harpenden, Eng. Brown, A., and L. Oosterhuis. 1981. The role of buried seeds in coppiced woods. Biol. Conserv. 21:19-38. Cole, L., and C. Keen. 1976. Dutch techniques for the establishment of natural plant communities in urban areas. Landscape Design 116:31-34. Davies, B. N. K. 1976. Wildlife, urbanization and industry. Biol. Conserv. 10:249-291. Donelan, M., and K. Thompson. 1980. The distribution of buried and viable seeds along a successional series. Biol. Conserv. 17:297-312. Farmer, R. E., M. Cunningham, and M. A. Barnhill. 1982. First-year development of plant communities originating from forest topsoils placed on southern Appalachian minesoils. J. Appl. Ecol. 19:283-294. Gilbert, O. L., and P. Wathern. 1976. Towards the production of extensive Calluna swards. Landscape Design 114:35. Gilbert, O. L., and P. Wathern. 1980. The creation of flower-rich swards on mineral workings. Reclam. Rev. 3:217-221. Gimingham, C. H. 1960. Biological flora of the British Isles: Calluna vulgaris (L.) Hull. J. Ecol. 48:455-483. Grime, J. P. 1973a. Control of species density in herbaceous vegetation. J. Environ. Manage. 1:151-167. Grime, J. P. 1973b. Competitive exclusion in herbaceous vegetation. Nature 242:344-347. Grime, J. P. 1974. Vegetation classification by reference to strategies. Nature 250:26-31. Grime, J. P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. Am. Na. 111:1169-1194. Grime, J. P. 1978. Interpretation of small-scale patterns in the distribution of plant species in space and time. Pp. 101 - 124 in A. H. J. Freyson and J. W. Woldendorp, eds. Structure and Functioning of Plant Populations. Elsevier-North Holland, Amsterdam. Grime, J. P. 1979. Plant Strategies and Vegetation Processes. John Wiley & Sons, Chich- ester, Eng.
270 SELECTED CASE STUDIES Grime, J. P., and R. Hunt. 1975. Relative growth rate, its range and adaptive significance in a local flora. J. Ecol. 63:521-534. Grime, J. P., and P. S. Lloyd. 1973. An Ecological Atlas of Grassland Plants. Edward Arnold, London. Harper, J. L. 1977. Population Biology of Plants. Academic Press, London. Hilton, K. J. 1967. The Lower Swansea Valley Project. Longmans Green, London. Hodgson, J. G. 1982. The botanical interest and value of quarries. Pp. 3-11 in B. N. K. Davies, ed. Ecology of Quarries. ITE Symposium 11. Institute of Terrestrial Ecology, Cambridge, Eng. Holliday, R. J., and M. S. Johnson. 1979. The contribution of derelict mineral and industrial sites to the conservation of rare plants in the United Kingdom. Min. Environ. 1: 1-7. Hoogerkamp, M. 1971. Probleme bet Ansaaten an Strassenrandern. Rasen 2:85-86. Humphries, R. N. 1977. An ecological approach to the revegetation of limestone quarries in the United Kingdom. Pp. 2-39 in E. J. Perry and N. A. Richards, eds. Limestone Quarries. Allied Chemical Corporation, Jamesville, N.Y. Hunt, R. 1970. Relative Growth Rate: Its Range and Adaptive Significance in a Local Flora. Ph.D. thesis, University of Sheffield. Jefferies, R. A., A. D. Bradshaw, and P. D. Putwain. 1981. Growth, nitrogen accumulation and nitrogen transfer by legume species established on mine spoils. J. Appl. Ecol. 18 :945- 956. Jeffrey, D. W. 1971. The experimental alteration of a Kobresia-rich sward in Upper Teesdale. Pp. 79-89 in E. Duffey and A. S. Watt, eds. The Scientific Management of Animal and Plant Communities for Conservation. Br. Ecol. Soc. Symp. 11. Blackwell, Oxford, Eng. Jeffrey, D. W., and C. D. Pigott. 1973. The response of grasslands on sugar-limestone in Teesdale to application of phosphorus and nitrogen. J. Ecol. 61:85-92. Johnson, M. S. 1978. The botanical significance of derelict industrial sites in Britain. Environ. Conserv. 5:223-238. Johnson, M. S., P. D. Putwain, and R. J. Holliday. 1978. Wildlife conservation value of derelict metalliferous mine workings in Wales. Biol. Conserv. 14:131-148. Lee, J. A., and B. Greenwood. 1976. The colonisation by plants of calcareous wastes from the salt and alkali industry in Cheshire, England. Biol. Conserv. 10:131-149. Lloyd, P. S., J. P. Grime, and I. H. Rorison. 1971. The grassland vegetation of the Sheffield region. J. Ecol. 59:863-886. Meaden, D. P. 1984. The Restoration and Creation of Heather Moorland Vegetation. Ph.D. thesis, University of Liverpool. Mellanby, K. 1974. The changing environment. Pp. 1-6 in D. L. Hawksworth, ed. The Changing Flora and Fauna of Britain. Academic Press, London. Milton, W. E. J. 1940. The effects of manuring, grazing and cutting on the yield, botanical and chemical composition of natural hill pastures. I. Yield and botanical composition section. J. Ecol. 28:326-356. Moorland Restoration Project. 1983. Phase 2 Report: Re-vegetation Trials. Moorland Res- toration Project, Bakewell, Eng. Newman, E. I. 1973. Competition and diversity in herbaceous vegetation. Nature 244:310. Palaniappan, V. M., R. H. Marrs, and A. D. Bradshaw. 1979. The effect of Lupinus arboreus on the nitrogen status of china clay wastes. J. Appl. Ecol. 16:825-831. Pigott, C. D. 1962. Soil formation and development on the carboniferous limestone of Derbyshire. I. Parent materials. J. Ecol. 50:145-156. Putwain, P. D., D. A. Gillham, and R. J. Holliday. 1982. Restoration of heather moorland
RESTORING DERELICT LANDS IN GREAT BRITAIN 271 and lowland heathland with special reference to heathlands. Environ. Conserv. 9:225- 235. Rorison, I. H. 1967. A seedling bioassay on some soils in the Sheffield area. J. Ecol. 55:725-741. Spray, M. 1970. Management of roadsides, hedgerows and grasslands for nature conser- vation and amenity. Pp. 28-33 in P. J. Grime, ed. People and Plants. Derbyshire Nat- uralists' Trust, Mattock, Eng. Tacey, W. H., and B. L. Glossop. 1980. Assessment of topsoil handling techniques for rehabilitation of sites mined for bauxite within the Jarrah Forest of Western Australia. J. Appl. E:col. 17: 195-201. Thompson, K., and J. P. Grime. 1979. Seasonal variation in herbaceous seed banks. J. Ecol. 67:893-922. Van der Hock, D. 1982. New man-made nature. Pp. 287-288 in S. P. Tjallingii and A. A. van der Veer, eds. Perspectives in Landscape Ecology. Centre for Agricultural Pub- lishing and Documentation, Wageningen, Neth. Wathern, P. 1976. The Ecology of Development Sites. Ph.D. thesis, University of Sheffield. Wathern, P. 1977. The ecology of herb establishment and survival in swards and its relevance to grassland reinstatement. Rasen 8:102-108. Wathern, P. 1980. The creation of natural vegetation after development. Pp. 5-20 in Proceedings of the Second International Conference on Environment Protection. Vol. II. Scientific Society for Building, Budapest. Wathern, P. 1984. Ecological impact assessment. Pp. 273-292 in B. D. Clark, A. Gilad, R. Bisset, and P. Tomlinson, eds. Perspectives on Environmental Impact Assessment. Reidel, Dordrecht, Neth. Wathern, P., and O. L. Gilbert. 1978. Artificial diversification of grassland with native herbs. J. Environ. Manage. 7:29-42. Wathern, P., and O. L. Gilbert. 1979. The production of grassland on subsoil. J. Environ. Manage. 8:269-275. Wells, T. C. E. 1965. Changes in the botanical composition of a sown pasture on the chalk in Kent, 1956-65. J. Br. Grassl. Soc. 22:277-281. Wells, T. C. E. 1983. The Creation of Species Rich Grasslands. Pp. 215-232 in A. Wa~Ten and F. B. Goldsmith, eds. Conservation in Perspective. John Wiley & Sons, Chichester, Eng. Wells, T. C. E., J. Sheail, D. F. Ball, and L. K. Ward. 1976. Ecological studies on the Porton Ranges. Relationships between vegetation, soils and land-use history. J. Ecol. 64:589-626. Wells, T. C. E., S. A. Bell, and A. Frost. 1981. Creating Attractive Grasslands Using Native Plant Species. Nature Conservancy Council, Shrewsbury, Eng. Willis, A. J. 1963. Braunton Burrows: The effects on the vegetation of addition of mineral nutrients to the dune soils. J. Ecol. 51:353-374.
272 SELECTED CASE STUDIES Committee Comment An increasingly important activity of environmental scientists is the reinstatement and reclamation of derelict land that has been modified for various reasons with various outcomes. Reclamation practices are most highly developed in the densely populated countries of western Europe, where human modifications affect virtually every piece of terrain. Much of this experience is summarized in a valuable book by A. D. Bradshaw and M. J. Chadwick, The Restoration of Land, published in 1980. In- creasing interest in this subject in the United States is reflected in the launching of a new journal, Restoration and Management Notes, published by the University of Wisconsin Arboretum. As in most fields of applied ecology, early attempts at restoration relied almost entirely on an empirical approach try a variety of approaches and see which ones work best. They were also highly constrained by cost, availability of materials for revegetation purposes, and the objectives expressed by society for rein- statement projects. Happily, the combination of several factors has resulted in important changes in the practice of restoration. Society is placing increasing value on land restoration, plant ecology has seen substantial recent development of relevant theory and data, and the goals of restoration have become much more sophisticated than merely getting something pretty and green. The project at Longstone Edge is an admirable example of applied ecology. It showed the importance of drawing on a variety of sources such as ecological facts, ecological theories, analog studies, and the reasoned judgment of persons familiar with the local areas if sound solutions are to be achieved. The Longstone Edge project took its direction from the recognition that most derelict land in Britain occurs on nutrient-poor soils and that attempts to make such areas fertile in the long run require repeated fertilization, which, because of its cost, is seldom feasible. But it was observed that diverse natural grasslands often occurred on nutrient-poor sites; appar- ently, therefore, the site could be ecologically interesting without being productive. These important observations played a role in the development of the powerful experimental design that was used in this project. Species-rich swards can be established on nutrient-poor soils when constituent species are poor competitors characterized by low maximal growth rates and when little competitive exclusion of species occurs. Under rich soil conditions, a few of the strongest competitors often come to dominate that is, interestingly, an empirical observation for which there is no satisfactory theoretical explanation. We do not understand compe- tition among plants in enough detail to be able to identify the mechanisms
RESTORING DERELICT LANDS IN GREAT BRITAIN 273 of interactions. Restoration projects might turn out to be especially val- uable for gaining further knowledge about these fundamental processes. They are carried out on a scale larger than is usually possible in purely academic research and, given current interest in them, are likely to be followed long enough to reveal aspects of the processes whose outcomes require decades and longer to be seen. Better understanding of competition among plants might offer new insights into the practical problems of reinstatement. For example, the Longstone Edge project used primarily information on relative growth rates to assess long-term success of species on the site. Other factors that influence interactions among plants, such as growth form and the relative allocation of resources to different tissues and organs, might also be used in the selection of species for seeding of experimental plots. Similarly, since the initiation of the Longstone Edge project, consid- erable progress has been made ire understanding the role of particular combinations of nutrients in determining the outcome of competition among plants, particularly algae. The experimental results and a theoretical inter- pretation of them have been provided by Tilman (19821. Because derelict lands are likely to be diverse in their nutrient status, even if they are mostly on poor soils, better understanding of the importance of particular combinations of soil nutrients for the success of competing plant species should enable both better selection of species for colonization and the cultivation of specific genotypes for use in particular circumstances. Restoration ecology might also contribute substantially to our knowl- edge and understanding of seed banks in the soils and their roles in the development of vegetation after disturbance. In the Longstone Edge project, important use was made of topsoil with seeds; this technology appears to have great potential. New advances in the understanding of the causes of seed longevity and the specific conditions that result in termi- nation of dormancy in various species might reveal which species have seeds that persist well in different circumstances. An important lesson of the Longstone Edge project is that in restoration ecology there are unlikely to be formulas of broad applicability that can be used in restoring most derelict lands. Solving problems for a particular site will inevitably require knowledge of its environmental conditions and of the genotypes of local plants that are potential colonizers of the site. Nonetheless, general concepts such as the relationships among soil nu- trients, growth rates, and competition are likely to be broadly applicable. In addition, society is likely to have multiple goals for reclaimed lands. Many sites in North America, for example, unlike those in Britain, are on potentially valuable agricultural land or are in highly productive sites that are to be restored to some seminatural state. Thus, apart from highly ;
274 SELECTED CASE STUDIES specific knowledge of local conditions, continued developments in the theory of competition, plant succession, seed ecology, and plant nutrition are likely to remain vital to restoration ecology far into the future. References Bradshaw, A. D., and M. J. Chadwick. 1980. The Restoration of Land. Blackwell, Oxford, Eng. Tilman, D. 1982. Resource Competition and Community Structure. Princeton University Press, Princeton, N.J.