CHAPTER 3 Species Diversity in Grasslands Jonathan Mitchley CONTENTS How Does Grasslands Diversity Vary in Space and Time? . . . . . . . . . . . . . . 46 What Determines the Number of Species That May Coexist and Makes Some Grasslands Richer Than Others?. . . . . . . . . . . . . . . . . . . . . . 47 Models of Species Coexistence Assuming Equilibrium Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Models of Coexistence Assuming Nonequilibrium Conditions . . . . 48 Are Diverse Grasslands More Stable, and, If So, Which Components of Diversity Are More Important? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 How Can We Conserve Species Rich Grasslands and Restore Diversity to Degraded Grasslands?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Restoration of Grassland Biodiversity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Grasslands are a major vegetation type covering about 30% of the terrestrial globe (Henwood, 1998) and providing a critical resource for pastoral and arable agriculture. Despite their significance, there remain fundamental the- oretical and practical questions to be answered concerning species diversity in grasslands. This chapter is concerned with seminatural grasslands, i.e., human-influenced grasslands which occur in a wide range of geographical locations and climates when the natural climax vegetation (e.g., forest) has been cut down and vegetation is grazed or cut for hay, thus preventing trees from re-establishing. This chapter addresses four themes: 45 0-8493-0904-2/01/$0.00+$.50 © 2001 by CRC Press LLC 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 45 46 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS DESIGN AND MANAGEMENT • How does grassland diversity vary in space and time? • What determines the number of species that may coexist and makes some grasslands richer than others? • Are diverse grasslands more stable, and, if so, which components of diversity are most important? • How can we conserve species-rich grasslands and restore diversity to degraded grasslands? HOW DOES GRASSLAND DIVERSITY VARY IN SPACE AND TIME? Species richness may be defined quite simply as the number of species in a given area. Species diversity, on the other hand, is a more complex concept which combines species richness with the relative abundances of the species present (Magurran, 1988). Grasslands differ in the relative abundance and spatial patterns of the constitutent species. Some communities may be char- acterized by one or two dominant species with a number of associated species occurring at low frequency in the sward, e.g., agricultural pastures dominated by Lolium perenne and Trifolium repens with low frequency and ground cover of associated species such as Taraxacum officinale and Bellis perennis (Rodwell, 1992). Other communities are characterized by a more even distribution of relative abundances among species, e.g., grazed calcare- ous grasslands in which twenty or so species of grasses and forbs may occur with higher frequency and cover in any particular sample of vegetation in an area (Rodwell 1992; Grubb et al., 1982). It is usual to consider the former types of grasslands of high dominance and low diversity (and evenness), and the latter grasslands of high diversity (and evenness) and low dominance (Magurran, 1988). Whittaker (1975) has defined three different kinds of diversity — alpha, beta and gamma—based on the scale of observation. Alpha diversity is the number of species in a defined area (e.g., a single quadrat). But diversity relates both to the number of species in an area and also to the difference in species composition between different areas (thus diversity is the “biology of number and of difference,” Gaston, 1996). Beta diversity is the difference in diversity between two sample areas, and gamma diversity is the regional dif- ference in species diversity. Very monotonous landscapes, such as those dominated by intensive arable agriculture, have low levels of diversity at all scales, while more var- ied landscapes such as those of mixed farming systems with meadows, per- manent pasture, woodland, etc. have a greater (but landscape specific) degree of alpha, beta, and gamma diversity. Differences in beta and gamma diversity can often be straightforwardly ascribed to differences in soil, topog- raphy, and climate or microclimate. Explaining differences in alpha diversity presents the greater challenge, however, as discussed below. 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 46 SPECIES DIVERSITY IN GRASSLANDS 47 Ecological systems are dynamic, and species coexisting at one point in time may or may not persist together through time; local extinction may result in changes in diversity through time (Ricklefs and Schluter, 1993). The pollen record provides evidence for the historical origins of grassland as well as for significant changes in diversity over time scales measured in millennia (Godwin, 1984). Many seminatural grasslands originated from forest clear- ance with species characteristic of open habitats colonising from an available species pool within the forest or from other areas (Rackham, 1986). Changes in diversity over more recent time-scales of tens or hundreds of years is read- ily confirmed through documentary evidence, for example repeated surveys at time intervals in the same area (Fischer and Stocklin, 1999). Habitat frag- mentation combined with changes in grassland management and the land- scape context have resulted in significant increases in local species extinction and loss of grassland diversity at a variety of spatial scales. WHAT DETERMINES THE NUMBER OF SPECIES THAT MAY COEXIST AND MAKES SOME GRASSLANDS RICHER THAN OTHERS? Gause’s competitive exclusion principle states that in uniform and con- stant conditions the most competitive species will come to dominate an area to the exclusion of all other species. The fact that grasslands composed of a single species rarely occur is testimony to the failure of the competitive exclu- sion principle. Species coexist and the exacting assumptions of the competi- tive exclusion principle rarely fit in nature. Environments are not homogenous in space nor in time; they are seasonal, spatially patchy, period- ically disturbed, and the plants regularly subject to competition from other species and to impacts from herbivores, pathogens, pollinators, and disper- sal agents (Crawley, 1997a). Explaining species coexistence in grasslands has been the subject of much research in recent decades (see Tokeshi, 1999, for a comprehensive review). Models explaining species diversity either assume equilibrium con- ditions where coexistence is possible even in uniform environments or non- equilibrium (stochastic) conditions where competitive exclusion is prevented by environmental or biotic fluctuations. A critical review of models of species diversity has been provided by Crawley (1997a) and Tilman and Pacala (1993); a summary follows. Models of Species Coexistence Assuming Equilibrium Conditions Niche separation and resource partitioning—This model states that the niches of coexisting species are sufficiently different that competitive exclu- sion simply does not occur. Species rich communities may be composed of 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 47 48 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS DESIGN AND MANAGEMENT species with narrow niches, species with broadly overlapping niches or habi- tats providing longer niche axes (Roughgarden, 1976). This is an interesting theoretical idea, but to test it in the field we must be able to define niche breadth for each coexisting species. This can be problematic, especially for plants, which share the same basic resources of light, water, and nutrients. Models of Coexistence Assuming Nonequilibrium Conditions Spatial heterogeneity—This model states that environments are het- erogenous and patchy and contain sufficient microsites/microhabitats for all coexisting species (Grubb et al., 1982). All environments are heterogenous and patchy, so this model is potentially helpful. Unfortunately, the model explains very little since we must know exactly what each niche is in order to predict species richness. In reality, very rarely can we define the niche of even a single species, let alone all those coexisting in a grassland. Temporal heterogeneity—This model stems from the observation that populations go up in good years and down in bad years. Good and bad years differ for different species, and, for coexistence to occur, good years must exceed bad years for all coexisting species (Chesson, 1986). In addition, there must be a storage effect to counter against bad years, e.g., long-lived adults or a seed bank. Also, the inferior competitors must do disproportionately well in good years to make up for poor years; likewise, there must be bad years for the best competitors. This is a model that can at least be tested, pro- vided we can define what constitutes good and bad years for each coexisting species. Competition and dispersal—This model suggests that coexistence of similar species is possible in a patchy environment when the dispersal abil- ity of the inferior competitor is sufficiently great that its rate of migration between patches is higher than the rate of extinction of populations within patches (Levin, 1974). In effect the poor competitor stays one jump ahead of the better competitor through a trade-off between competitive ability and dispersal. Herbivory—Models show that selective herbivory can enhance coexis- tence and species richness. All grazing animals are selective in their grazing, and even a mowing machine is selective since it defoliates taller plants dis- proportionately over shorter plants (Crawley, 1983). Selective grazing by her- bivores alters the competitive relationships between plant species and coexistence may be facilitated, for example, if there is a positive correlation between palatability and competitive ability. Similarly, frequency-dependent selection by herbivores promotes coexistence since the scarce species escape herbivory (Crawley, 1997b). Another way in which large herbivores promote coexistence is through the formation of regeneration sites; e.g., trampling creates gaps in the sward. Grazing by large herbivores and the associated 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 48 SPECIES DIVERSITY IN GRASSLANDS 49 activities such as trampling are all forms of disturbance, which is discussed immediately below. Disturbance—Here coexistence is facilitated through destruction of dominant species and creation of regeneration microsites. Important distur- bance factors in grasslands include grazing, trampling, cutting or mowing, and fire. Because species richness will be reduced at low and high levels of disturbance (by competitive exclusion and environmental stress respec- tively) species richness will be potentially greatest at intermediate levels of disturbance, and of productivity (Grime, 1973; Connell, 1978; Huston, 1979). Arguably, the search for a general hypothesis of species richness may never succeed simply because such an hypothesis by explaining everything will explain nothing (Crawley, 1997a). Specific models may be more success- ful, especially as the complexity, precision, and testability of mathematical models of community dynamics continues to increase (Pacala, 1997). Models are only one element of the species diversity story; carefully designed experiments, especially long-term experiments, to test the models are the other essential element. Commencing in 1856, the longest running ecological experiment in the world is the Park Grass experiment at Rothamsted, U.K. There, initially for agricultural purposes, a lowland per- manent pasture was subjected to various fertilizer and manure treatments to assess the impact on hay yields (Johnston, 1994). The subsequent changes in productivity and especially in species composition have provided a rich source of data for plant ecologists interested in species diversity, and the experiment is now considered a classic in ecology. Despite the lack of formal replication, the large plots (0.2 ha) have provided some important conclu- sions regarding long-term trends in grassland species diversity (Tilman et al., 1994): 1. In the absence of perturbation, grassland diversity is quasistable but influenced by climatic variation. Some usually rare species have brief periods of dominance and a few usually dominant species have brief periods of rarity. These observations have impli- cations for prediction of the potential impacts of climate change on grassland diversity. Some of these climate induced shifts in relative abundance may be predictable from plant life history attributes (see also Sternberg et al., 1999; Grime, 1997) while other shifts, espe- cially the potential increase in sparse species, may be less pre- dictable. 2. Species diversity is dependent on the rate and ratio of limiting soil nutrients, nitrogen, phosphorus, and potassium. All experimental inputs of nutrients resulted in a decline in species richness. To this day, species richness remains greatest on the unfertilized, unlimed control plots. It seems likely that with increased nutrient availabil- ity nutrient limitation weakens, and competition for light becomes 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 49 50 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS DESIGN AND MANAGEMENT the decisive factor (Newman, 1973). Competition for light appears to be more asymmetric than competition for nutrients, and, conse- quently, it is more likely to drive inferior species to extinction. This is, together with seedling recruitment limitations, the most impor- tant cause of a decrease in species richness high nutrient levels (Tilman et al., 1994). The practical lesson here is that conservation of species-rich grasslands requires low or preferably no inputs of fertilizers. There can be no easy compromise between productivity and diversity—grasslands can be productive in agricultural terms, or diverse in wildlife terms, but not both (Haggar and Peel, 1994). 3. Soil pH modified via liming and fertilization greatly affects plant species diversity. In the Rothamsted experiment, species diversity was greatest at high pH; there were more species in high pH plots than in low pH plots. This result emphasizes the role of historical origins of the species pool in determining regional species diver- sity, and it reflects the greater species pool of calcicoles compared with calcifuges in the region of this experimental system. In other geographical regions, such as the Atlantic coastal plain of North America, calcifuges may outnumber calcicoles (largely because acidic substrates were more apparent here than in Europe during the Quaternary period), and the most species-rich grasslands occur in fire-disturbed savanna grasslands on acidic soils (Grubb, 1986; Walker and Peet, 1984). 4. Different rates of nutrient supply act as a selective force on plant populations, causing measurable evolutionary changes (Snaydon and Davies, 1972). The implication here is that explanations for species diversity not only must acknowledge both current (or recent) conditions, such as soil nutrient status or pH, but also his- torical characteristics, such as land use history and evolutionary factors. Historical factors are rarely considered in models of species diversity but are often of great significance. For example, Partel and Zobel (1999) report a study of variation in small-scale plant species richness between areas of calcareous alvar grasslands in western Estonia. The community type is very species rich at the small-scale (high alpha diversity). The pattern of diversity is often explained by variation in ecological factors that makes it possible to avoid competitive exclusion (see above). However, species richness here was found to be positively correlated with the size of the species pool, with community age, and with vegetation density. The authors conclude that historical processes, on both regional and local levels, determine the arrival of any particular species to a tar- get community, and it is historical factors, not environmental het- erogeneity, which are responsible for the variations in species richness between communities of this type. 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 50 SPECIES DIVERSITY IN GRASSLANDS 51 ARE DIVERSE GRASSLANDS MORE STABLE, AND, IF SO, WHICH COMPONENTS OF DIVERSITY ARE MORE IMPORTANT? Stability can be defined both in terms of resistance to change (e.g., resist- ance of native grasslands to invasion by exotic species) and in terms of resilience to perturbations (e.g., the ability of native grassland to rebound fol- lowing perturbations, such as fire or overgrazing). The idea that diversity influences critical ecosystem processes such as stability and resistance to invasion has been widely discussed (Chapin et al., 1998). In particular, the relationship between diversity and stability is far from clear cut (e.g., May, 1973), and various hypotheses have been proposed. For example the rivet hypothesis likens species in an ecosystem to the rivets in an airplane (Ehrlich and Ehrlich, 1981). The loss of a few rivets may go unnoticed since they are redundant; however, beyond a certain threshold, losses will bring about cat- astrophic collapse. Field and laboratory studies have attempted to disentangle some of these issues. For example, Tilman (1996, 1999) examined the relationships between biodiversity and stability at both the population and the ecosystem level in a long-term study of grassland plots. Results demonstrated that biodiversity stabilized community and ecosystem processes, but not population processes—hence, diversity does result in stability but at the ecosystem not the population level. During drought years, the change in total plant com- munity biomass from before a drought to the peak of the drought was highly dependent on species richness. The chance of drought resistant species occur- ring in a grassland was greater in more diverse grasslands (i.e., a “sampling effect”) providing community recovery insurance against climatic extremes. In contrast, year-to-year variability in species abundances was not stabilized by plant species richness for either all years or nondrought years. Tilman explains this difference between species and community biomass in terms of interspecific competition. When climatic variations harm some species, unharmed competitors increase. Such compensation stabilizes total commu- nity biomass but causes species abundances to be more variable. A corollary of the rivet hypothesis is the concept of functional groups that suggests species per se may not be the most significant influence on ecosys- tem function. Instead, groups of ecologically equivalent species, functional groups, may be the critical element, e.g., grasses and legumes (Tilman, 1997) or dominants, subordinates, and transients (Grime, 1998). Ecosystem func- tion may be little impaired by species losses if these represent all functional groups; however, big impacts may result if most or all the representatives of a particular group are lost, e.g., all the legumes or all the dominants. There is considerable scope for further research into the role of diversity both within and between functional groups in key grassland ecosystem processes such as productivity, stability, and invasiveness. 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 51 52 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS DESIGN AND MANAGEMENT In conclusion it should be noted that higher plants, whatever their func- tional type, are only one component of the grassland ecosystem. Recent work on plant mycorrhizae suggests that below-ground diversity of arbuscular mycorrhizal fungi is a major factor contributing to the maintenance of plant biodiversity and to ecosystem functioning (van der Heijden et al., 1998). Thus, the key element to ecosystem functioning may not be in front of our eyes and counted in quadrats but in the soil itself. HOW CAN WE CONSERVE SPECIES RICH GRASSLANDS AND RESTORE DIVERSITY TO DEGRADED GRASSLANDS? For hundreds of years, agriculture supported the enrichment and diver- sification of the vegetation in central Europe including its grasslands. Now many landscape elements are no more than relics of historical land use. Today’s intensive agriculture is considered to be the main agent responsible for the decline of plant species (Stanners and Bourdeau, 1995). The remaining fragments of species-rich grasslands are considered important elements of the landscape, and they feature significantly in lists of protected areas. For example, there are more grassland biotopes listed in the European Habitats Directive than any other single vegetation type (Stanners and Bourdeau, 1995). However, the extent of species-rich grassland in Europe has declined dramatically in recent decades due both to agricultural development and to neglect of traditional management practices. For example a recent survey of seminatural grasslands in lowland England and Wales showed that unim- proved seminatural grassland accounted for only 1–2% of the total cover of permanent lowland grassland (Blackstock et al., 1999). These figures reflect losses of up to 97% of the extent of these grasslands since the 1940s (Ratcliffe, 1984). Even existing seminatural grassland sites, such as protected areas, may be threatened as landscape ecological processes become disrupted. For example, Fischer and Stocklin (1999) used botanical survey data collected in 1950 and in 1985 to investigate rates of local extinction in remnants of extensively grazed calcareous grassland in Switzerland. The results showed higher rates of local extinction for smaller populations, for species with shorter life cycles, for species with short-lived seed bank, and for species with higher habitat speci- ficity. Thus, most characteristic species of calcareous grassland do not persist in the seed bank and cannot rely on this mechanism as a buffer against local extinction even in intact grassland remnants. One important consequence is that effective conservation of species-rich grasslands requires not only an understanding of the site-specific factors allowing coexistence of species but also an understanding of the landscape level processes that have given rise to the grasslands. Maintenance of ade- quate management through disturbance regimes involving grazing, mowing, 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 52 SPECIES DIVERSITY IN GRASSLANDS 53 or burning needs to be combined with landscape processes such as provision of buffer zones to reduce edge effects, maintenance of transhumance of livestock to maintain landscape scale seed dispersal, and provision of networks of reserves in a more sympathetic landscape matrix. Many of these principles have been included in the Pan European Biodiversity and Landscape Strategy, adopted in 1995, which provides a landscape scale vision for integrated grassland conservation with ecologically sustainable agriculture (Goriup, 1998). Grasslands exist within a landscape matrix very often including agricul- ture of one type or another. Potentially more sustainable systems of agricul- ture, such as organic farming, can promote species diversity of arable fields and grassland. A higher total number of species and also more endangered “red list” species can be found in organic fields than conventional fields (van Elsen, 2000). However, economic pressure leads to technological advances, such as mechanical weed control and undersowing, which reduce diversity. Thus, the aims of biodiversity conservation will not necessarily be achieved by converting wholesale to organic farming. An integrated approach is needed to avoid the polarization of the landscape into the minority set aside for nature conservation and the majority for intensive agricultural produc- tion. Biodiversity conservation certainly needs to be integrated directly into agricultural policy for organic farming, but also we should not ignore the potential benefits from integrating environmental concerns into the more intensive forms of conventional agriculture as well (Burch et al., in press). RESTORATION OF GRASSLAND BIODIVERSITY Increasingly conservation strategy looks to ecology for the techniques for putting biodiversity back into the landscape through ecological restoration. For example, the U.K. Biodiversity Action Plan includes quantitative targets for restoration as well as for conservation of grasslands (U.K. Steering Group, 1995). Grassland restoration may involve reintroduction of appropriate man- agement practices to neglected grassland, the recreation of species-rich grass- lands on arable land, or the diversification of species-poor improved grassland (Muller et al., 1998). Species-rich grassland communities still occur in low-intensity farming systems throughout Europe. Gradually, such systems have either been aban- doned or more intensively exploited, with a subsequent decline in species. Until recently, it was believed that restoration of these communities would be relatively straightforward. However, abiotic constraints (especially eutroph- ication and acidification) have prevented restoration (Bakker and Berendse, 1999). Moreover, biotic constraints may be limiting; many plant species are not present in the soil seed bank (Bakker and Berendse, 1999; Stamfli and Zeiter, 1999) and, even if present, may not be recruited unless appropriate disturbance regimes are applied (Edwards and Crawley, 1999). Furthermore, 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 53 54 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS DESIGN AND MANAGEMENT dispersal of many species may be limited in the current fragmented land- scape where certain pathways of seed dispersal, such as attachment to the fleece or coats of livestock during traditional livestock transhumance, have been lost (Poschlod et al., 1998). Effective restoration requires techniques to address the critical habitat and environmental constraints of the site, including methods for reducing soil fertility, for introducing propagules of target species, and for reinstating appropriate disturbance regimes and management. Where the aim is to restore high species diversity to previously fertilized meadows, Kirkham and Kent (1999) have shown the importance both of reducing soil fertility and of encouraging seed production in those species that have declined. In other sit- uations, active introduction of wildflower seed or plants may be required to achieve rapid increases in diversity. Methods need to be developed and tested for the successful establishment of desirable species into existing species-poor swards through regional seed mixtures (Jones and Hayes, 1999), hay strewing (Jones et al., 1995), sod transplantation (Partel et al., 1998), and even complete habitat transference (Good et al., 1999). Once provision of propagules (naturally or artificially) has been made to suitable area and substrate, implementation of an appropriate management regime is the next key step in restoration. Most species-rich grasslands require defoliation, from large grazing animals or by mowing machine, for the maintenance of species richness. Without this disturbance, species rich- ness declines and natural succession proceeds to transform grasslands into scrublands and ultimately secondary woodland (Mitchley and Ispikoudis, 1999). Recreated grasslands often require rigorous defoliation regimes in the early stages of restoration to maximize recruitment from the seed bank and seed rain (Gibson et al., 1987). For established grazed pastures, the timing and intensity of grazing is critical not only to maintenance of plant species richness but also to maintenance of vegetation structure and habitat for the associated invertebrate fauna (Cherrill and Brown, 1990; Mitchley, 1994). For mown meadows, again timing of defoliation is the critical element. Economic considerations in modern agricultural management often result in earlier cutting dates than in traditional systems. Earlier cutting dates can result in shifts in species richness and composition; for example, late flowering species may not have set seed by the time of mowing (Smith and Jones, 1991). The loss and fragmentation of habitat is a major threat to the continued survival of many grassland habitats and species. Application of landscape ecological principles can help to devise integrated strategies to reverse these trends. Huxel and Hastings (1999) reported a “restoration lag” in simulations of species restoration when randomly selecting habitat for restora- tion. They found that nonrandom or targeted restoration practices, such as restoring only habitat that is adjacent to those occupied by the target species, can dramatically turn round any restoration lag. Many restora- tion efforts have limitations on both funds and available sites for restora- tion, necessitating high potential success on any restoration efforts. The 920103_CRC20_0904_CH03 1/13/01 10:44 AM Page 54 [...]... Agricultural and Ecological Sciences, 9 38 CABI, Wallingford Jones, A.T and Hayes, M.J 1999 Increasing floristic diversity in grassland: the effects of management regime and provenance on species introduction Biol Conserv., 87, 38 1 39 0 Jones, G.H., Trueman, I.C., and Millet, P 1995 The use of hay strewing to create species-rich grasslands (i) General principles and hay strewing versus seed mixes Land Contamination... Stability and Complexity in Model Ecosystems Princeton University Press, Princeton Mitchley, J 1994 Sward structure with regard to conservation In: Haggar, R.J and Peel, S., Eds Grassland Management and Nature Conservation, 43 53 British Grassland Society, Reading Mitchley, J and Ispikoudis, I 1999 Grassland and shrubland in Europe: biodiversity and conservation In: Papanastasis, V.P., Frame, K.J., and Nastis,... Grasslands and Woody Plants in Europe, 239 –251 European Grassland Federation, Thessaloniki 9201 03_ CRC20_0904_CH 03 58 1/ 13/ 01 10:44 AM Page 58 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS DESIGN AND MANAGEMENT Muller, S., Dutoit, T., Alard, D., and Grevilliot, F 1998 Restoration and rehabilitation of species-rich grassland ecosystems in France: a review Restoration Ecol., 6, 94 –101 Newman, E.I 19 73 Competition... Ecol., 7, 33 6 34 7 9201 03_ CRC20_0904_CH 03 1/ 13/ 01 10:44 AM SPECIES DIVERSITY IN GRASSLANDS Page 57 57 Goriup, P 1998 The Pan-European Biological and Landscape Diversity Strategy: integration of ecological agriculture and grassland conservation Parks, 8, 37 –46 Grime, J.P 19 73 Competitive exclusion in herbaceous vegetation Nature, 242, 34 4 34 7 Grime, J.P 1997 Climate change and vegetation In: Crawley,... AGROECOSYSTEMS DESIGN AND MANAGEMENT REFERENCES Bakker, J.P and Berendse, F 1999 Constraints in the restoration of ecological diversity in grassland and heathland communities Trends Ecol Evol., 14, 63 68 Blackstock, T.H., Rimes, C.A., Stevens, D.P., Jefferson, R.G., Robertson, H.J., Mackintosh, J., and Hopkins, J.J 1999 The extent of semi-natural grassland communities in lowland England and Wales: a... Where grassland diversity has been lost, there may be potential for restoration by reinstating management practices or by recreating grasslands on marginal arable and other land Effective restoration requires an understanding of the abiotic and biotic features of the site, including history, substrate, and the local management regimes, as well as landscape-level processes of species colonization and dispersal... Targeting conservation and restoration to the most appropriate areas and sites using landscape ecological principles can provide cost-effective strategies for the maintenance and enhancement of this important agroecological resource ACKNOWLEDGMENTS Grant Edwards provided helpful comments on the first draft of this paper 9201 03_ CRC20_0904_CH 03 56 1/ 13/ 01 10:44 AM Page 56 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS... attempt to explain how species coexist in grassland In general terms, spatial and temporal heterogeneity and disturbance, especially defoliation by large herbivores and/ or cutting and their impacts on species dynamics and recruitment (from seed bank or seed rain) are critical to maintenance of species richness in grassland 2 Diverse grasslands may be more stable than less diverse ones in terms of ecosystem...9201 03_ CRC20_0904_CH 03 1/ 13/ 01 10:44 AM Page 55 SPECIES DIVERSITY IN GRASSLANDS 55 incorporation of spatial analyses and targeting in restoration management may drastically improve the success rate (Connor et al., in preparation) Therefore, general principles that incorporate spatial processes and appropriate management are needed to guide restoration strategies... Harper and Row, New York Connell, J.H 1978 Diversity in tropical rain forests and coral reefs Science, 199, 130 2 – 131 0 Connor J., Simmons E.A and Mitchley J (in preparation) Targeting areas for restoration to chalk grassland using a GIS Crawley, M.J 1997a The structure of plant communities In: Crawley, M.J., Ed Plant Ecology, 2nd edition, 475 – 531 Blackwell Science, Oxford Crawley, M.J 1997b Plant-herbivore . grasslands. Maintenance of ade- quate management through disturbance regimes involving grazing, mowing, 9201 03_ CRC20_0904_CH 03 1/ 13/ 01 10:44 AM Page 52 SPECIES DIVERSITY IN GRASSLANDS 53 or burning. Furthermore, 9201 03_ CRC20_0904_CH 03 1/ 13/ 01 10:44 AM Page 53 54 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS DESIGN AND MANAGEMENT dispersal of many species may be limited in the current fragmented land- scape where certain pathways. such as productivity, stability, and invasiveness. 9201 03_ CRC20_0904_CH 03 1/ 13/ 01 10:44 AM Page 51 52 STRUCTURE AND FUNCTION IN AGROECOSYSTEMS DESIGN AND MANAGEMENT In conclusion it should be noted