Introduction: the challenges of agricultural change and the involvement of ecology

With the global human population now propelled towards seven billion and probably over 800 million people undernourished, there is an irrefutable case that our shared needs for agricultural production have never been greater. Already, almost 40% of the world's land area is under some form of agriculture (Table 1). These facts alone, and the sheer magnitude of the numbers involved, mean that the involvement of ecology in agriculture is one of the most crucial of all uses of our subject. But consider also the complexity of the circumstances and choices that surround food production: how do we balance the needs to ensure global food security, optimize production and make appropriate use of changing technology against the needs to preserve environmental function and halt current trends in the erosion of biodiversity? The information of ecologists should clearly be pivotal, particularly at the local and national levels where agricultural policies and their implementation have their greatest direct impact.

In Europe, two areas that are particularly topical for ecologists are the reform of the European Union (EU) Common Agricultural Policy (CAP) and the advent of GM crops. The CAP has been with us in some form since 1962, initially aimed at increasing domestic food production while reducing reliance on imports. The story of the overproduction and intensification that followed, along with major consequences for farmland ecosystems, is now widely known (Chamberlain et al. 2000). More latterly, however, reforms to the CAP have promoted modest spending on styles of farming considered less damaging environmentally, with further developments in this direction agreed by EU agriculture ministers on 26 June 2003. For the first time, at least some of the €40 + billion currently spent in the CAP per year will now be completely decoupled from production and linked to the need for farmers to comply with requirements to maintain farmland in ‘good environmental condition’. Modulation will also improve funding for rural development.The exact sums involved are currently unclear and they will be moderated locally. Nevertheless, environmental farm schemes will develop throughout the current and enlarged EU further than ever before, with real potential to change environmental quality. At the same time, adverse changes in grazing, further intensification and specialisation in some sectors might have negative environmental impacts. While the ramifications for ecologists should be clear – for example through our engagement in advising on new schemes and assessing their effectiveness – the evidence about our involvement appears to be otherwise. In this issue of the Journal of Applied Ecology, Kleijn & Sutherland (2003) show, for example, that very few ecological assessments of agri-environment schemes have ever been made and, where assessments exist, there are often basic problems with their design. With the development of the CAP now at a critical phase, not only should there be a more robust assessment of the true value of agri-environment schemes, but also ecologists should be far more fully engaged with those who administer agri-environment measures at local, national and European levels.

The issue of genetically modified (GM) crops in the UK could not provide a more contrasting example of the involvement of ecologists in a major topic of agricultural development. While much of the public debate about GM technology has centred on the risks of gene transfer (Raybould & Gray 1993; Barton & Dracup 2000; Beringer 2000; Pretty 2001; Thomson 2001; Desplanque et al. 2002), the UK's recently reported farm-scale trials of genetically modified crops involved what, in effect, was a classical ecological enquiry: factorial experiments in split fields were coupled with conventional sampling to assess how arable plants and invertebrates responded to broad-spectrum herbicides applied to tolerant crops (Firbank et al. 2003; Perry et al. 2003). In other words, irrespective of the genetic design and constitution of the crop plants, ecological effects were at the heart of the trials and were expected to arise as a consequence of the conditions in which each crop was cultured (Squire et al. 2003). This interpretation is borne out by the central results: herbicide application to two of the GM crops (sugar beet and spring oilseed rape) negatively affected some components of biodiversity, but in the third case (GM maize) effects were less marked, perhaps because the conventional reference crop received its own characteristic herbicide dressing (Heard et al. 2003; Brooks et al. 2003). It can now be more confidently surmised that any future ecological consequences from genetic modification will be at least as likely to reflect the diversifying opportunities, methods and conditions for crop growth as any other factor in the brave new GM landscape. Moreover, these same consequences, and in particular whether they are negative or positive, should, from here on, be judged on ecological criteria relative to ‘conventional’ methods of agricultural production (Dewar et al. 2003; Squire et al. 2003).

These and other areas of agricultural ecology have many key lessons that reach out across ecology and its application. In all stages from their design, execution and reporting, the farm-scale trials reflected the skills and experience of ecologists or ecological agronomists: aspects of experimental design, statistical power, experimental scale, data collection, expected effects and likely targets were all fundamentally ecological in conception (Firbank et al. 2003; Perry et al. 2003). Over the latter part of the 20th century, and certainly over the 40-year life of the Journal of Applied Ecology, our science has become sufficiently quantitative, experimental and multidisciplinary to address the challenges provided by this, and other, large-scale ecological issues. Where problems are beyond the scale of empirical experimentation, we have an array of heuristic modelling tools to generate hypotheses or even guide policy (Watkinson et al. 2000; Squire et al. 2003; Stephens et al. 2003). Although the difficulties in extrapolating ecological results and processes across spatio-temporal scales are undiminished (Manel et al. 2000; Schneider 2001), the ability of the applied ecological community to respond to the needs of environmental management at these large scales has never been greater (Ormerod et al. 2002). With around 20% of papers published each year in the Journal of Applied Ecology related in some way to agricultural themes, agro-ecology has not only contributed considerably to this strength, but it also continues to represent one of the key areas in which applied ecologists are active (Table 2). Given that such expertise is readily available, the contrast between the involvement of ecologists in the UK farm-scale trials and European agri-environment schemes is all the more striking: these two issues illustrate contrastingly good and bad examples of knowledge transfer between ecological researchers and policy makers.

The farm-scale trials of GM crops also provide the clearest evidence so far that the assessments of risk or benefit of GM technology must go beyond issues of gene transfer and food safety to address also those effects that might arise ecologically: as plants are modified to grow in new ways, new circumstances or with new properties, new ecological questions are bound to arise. It is germane that the experience of ecologists allows them to consider any such changes alongside other technologies that have altered the intensity of land management. The resulting trends in agriculture, particularly in the developed world, have been exemplified in this and other journals (Chamberlain et al. 2000; Squire et al. 2003). We are increasingly well-informed not only about the processes that affect associated ecological systems (McLaughlin & Mineau 1995; Vickery et al. 2001; Robinson & Sutherland 2002; Dalton & Brand-Hardy 2003), but also those through which organisms, populations and communities respond (Fuller et al. 1995; Siriwardena et al. 1998; Brickle et al. 2000; Ambrosini et al. 2002; Benton et al. 2002; Colling et al. 2002; Park et al. 2002). And while our knowledge has grown in depth, so it has broadened in extent. Although conspicuous taxa such as birds have dominated agro-ecological research over the last decade (Chamberlain et al. 2000; Ormerod & Watkinson 2000; Stephens et al. 2003), several recent reviews and individual contributions have made assessments of effects on other taxa a clear feature of the Journal of Applied Ecology (Benton et al. 2002; Robinson & Sutherland 2002; Hartley et al. 2003; Hutton & Giller 2003; Ormerod 2003a; Peveling et al. 2003). So, too, have the Journal's authors considered how the effects from agricultural land can affect adjacent ecosystems; for example, in bordering riparian habitat or in downstream wetlands and rivers (Jansen & Robertson 2001; Green & Galatowitsch 2002; Clément et al. 2003; Middleton 2003; Pretty et al. 2003; Wickramasinghe et al. 2003). Not only in addressing the case of new GM technology, but also in appraising the effects of conventional agricultural developments, the ability to integrate across communities and processes is a key strength of ecology. To a large extent, ecologists can understand, recognize and sometimes even predict, at least qualitatively, the effects of pesticide applications, patterns of fertilizer use, altered drainage, new crop choices and rotations, the effects of habitat modifications, the communities of organisms expected to occupy agricultural land, the effects of demographic change upon them, and the consequences of altering predator–prey interactions (Table 2). When linked with the knowledge of agronomists, pedobiologists and those from other key disciplines, the strength of this ecological knowledge base is further multiplied.

There will, of course, be unexpected and unpredictable consequences from the advent of GM and other new technologies as there have been already from other forms of agronomic development. Already, papers on the consequences of conventional agriculture illustrate how management problems can be compounded, or can develop unexpected twists, when they are linked to other pressures. Excess nitrogen additions, for example, alter the risks of alien plant invasion or survival (Brooks 2003; Green & Galatowitsch 2002), mediate the consequences of grazing (Hartley et al. 2003), affect the outcomes of attempts at ecological restoration (Pywell et al. 2003) and might change communities of plant symbionts (Kiers et al. 2002). Pesticide use can give rise to unexpected resistance in target organisms (Twigg et al. 2002), or have effects that vary between habitats (Lee et al. 2001). Large-animal grazing can affect the conditions for seed survival in scarce plants – both directly or indirectly by altering populations of seed predators (Meeson et al. 2002; Noy-Meir & Briske 2002). Habitat simplification might alter the populations of organisms that have disproportionately important ecosystem functions (Jones et al. 2003). Increasingly, also, the consequences of agricultural change must be determined in uncertain climates that will not only alter the way that land is used, crops are grown and pests develop, but also the way that ecosystems respond to these and any existing stresses (Moss et al. 2003). In all these respects, the skills and experience of ecologists in interpreting and investigating ecological systems will be even more crucial than in those instances where the ecological effects of agriculture are straightforward and predictable.

The ecological issues arising from agricultural development in general, and from genetic modification in particular, might be local in nuance, but they are very clearly global in extent. Irrespective of how the United Kingdom or European Union now makes policy on GM licensing, the first generations of genetically modified (GM) crops are already grown on over 50 million ha round the world (James 2001). The majority are in the developed world, but a fifth are in less developed countries. For comparison, this total area already approaches that of arable crops grown in the original 12 countries of the European Union (c. 66 million ha; UN Food and Agriculture Organization, http://www.fao.org/). In agriculturally active regions such as the expanded EU (i.e. with 15 member states) or parts of Asia, agriculture in some form now occupies over 40–50% of the land area, rising to over 60–70% in nations such as Denmark and the UK (Table 1). Even though smaller proportions of land are used outside these regions, the habitats involved are often highly sensitive to change or development – for example, in the case of the world's semi-natural rangelands and grasslands. Some of the prime examples recorded in the Journal of Applied Ecology concern grazing on rangelands in Australia, Africa and the Americas (see Table 2). Here, even small changes, for example the provision of stock-watering or stock-holding facilities, can have disproportionate ecological consequences (Landsberg et al. 2002, 2003; Riginos & Hoffman 2003). The message is clear: agriculture is a major global form of land management, and hence a major conduit through which genetic modifications or other agronomic developments might further affect the character of ecosystems. The global community of applied ecologists – our subscribers, readers and contributors – have an ideal opportunity in the pages of the Journal of Applied Ecology not only to trade experience, but also to compare subtle local variations from the overall pattern (Ambrosini et al. 2002; Table 2).

In the end, one of the over-riding lessons from the ecology of resource use is that ecological consequences and management challenges not only arise in the present, but also affect management options in the future. Increasingly, papers published in the Journal of Applied Ecology address the restoration, remediation or repair of a wide range of ecosystem types that have been changed through exploitation (Ormerod 2003b). Many such studies apply directly to systems previously altered by agriculture; for example, where extensification has now become an aim, or where moves to encourage or maintain biodiversity within farmed systems have been driven by changing economics or policy (Wolf et al. 2001; Calladine et al. 2002; Pywell et al. 2002, 2003; Middleton 2003; Pretty et al. 2003; Smith et al. 2002, 2003). Across those areas of Europe and the world where agri-environment schemes are developing, methods for restoring land after cultivation, for protecting adjacent systems such as wetlands, or for maintaining nature conservation interest during cultivation, must from here on respond to factors linked to GM and its associated new technologies. It is sobering to realize that the effects of even modest agricultural development in sensitive ecosystems can be particularly long-lasting (Augustine 2003).

A special profile: meeting the ecological challenges of agricultural change

In all these foregoing respects, the point of departure for applied ecologists meeting the challenge of agricultural change, whether from GM or any source, and whether through intensification or restoration, is bound to be in our existing knowledge. That this knowledge is extensive across the whole range of agriculture and its consequences is evident not only from the array of papers published recently in this Journal (Table 2), but also from the following special profile of seven papers linked by their relevance to this theme.

These papers illustrate the coverage of comparative agricultural ecology across global regions and ecosystems in the Journal of Applied Ecology, for example ranging from North American forest wetlands (Middleton 2003), through riparian wetlands in mainland Europe (Clément et al. 2003; Kleijn & Sutherland 2003), conventional farmlands in the British Islands (Hutton & Giller 2003; Wickramasinghe et al. 2003) to rangelands in Australia (Landsberg et al. 2003). These papers illustrate effects related to agriculture not only directly in the farmed landscape, but also in changes and processes involving hydrological run-off through riparian zones and into the open waters that receive agricultural drainage (Clément et al. 2003; Wickramasinghe et al. 2003). In so doing, they not only reveal the consequences of agriculture for organisms, but also for physico- and bio-chemical processes that can reflect and diagnose important processes in landscape management (Clément et al. 2003). Finally, as befitting the multi-taxonomic and botanical–zoological blend of papers in the Journal, they address organisms that range across wetland and rangeland plants, terrestrial invertebrates, birds and mammals.

Perhaps most of all, these seven papers illustrate very effectively the array of approaches used by applied ecologists in addressing agricultural questions including modelling (Stephens et al. 2003), meta-analysis (Kleijn & Sutherland 2003), surveys (Wickramasinghe et al. 2003), transect studies (Landsberg et al. 2003), experiments (Hutton & Giller 2003), seedbank assays (Middleton 2003) and process studies based on modern ecological methods (Clément et al. 2003). In so doing, the papers in this special profile reconfirm the role of the Journal of Applied Ecology in showcasing and developing some of the best methods currently available for problem solving in ecology.