No one can doubt the profound and growing importance that scientific understanding of the environment and its resources will have on human activity over the next 20 or 50 years.

The six key issues on the agenda in the United Kingdom for the next 20 years will be: the identification, management and sustainable exploitation of renewable and non-renewable resources (including earth and marine resources, fresh water, land and soils); environmental risks and hazards; pollution of the atmosphere, freshwater, oceans and land; disposal of waste; global change; biodiversity. Scientific knowledge, understanding and prediction of these issues is required by the public, by government as proxy for the public and society in making policy at a national and international level, and by industry and commerce wishing to meet environmental and legislative standards and benefit from the burgeoning $200 billion per annum market in environmental technologies.

11 = /The research agenda of the scientific community in generating basic understanding can be to a large extent related to these six environmental and resource issues. The following paragraphs touch briefly on examples of the first four issues before going into more detail on the last two.

Water is a resource in the news. The drought of the past few weeks has re-emphasised that even in a wet country such as the United Kingdom, supplies cannot be taken for granted. With global demand for water doubling every 21 years, and water shortages affecting 40 per cent of the world's population, the World Bank has identified water as the likeliest cause of new wars. Scientific understanding of water quantity and quality is increasingly important.

³ÉÈËVRÊÓƵ

The traditional environmental risks and hazards include flooding, storms (both of particular relevance to the UK), earthquakes and volcanoes. Here the aim of research is to generate further understanding of the basic processes in order to enhance the power of predictive models. Often, understanding of local risks and hazards depends on global understanding, as illustrated by the recent demonstration that annual variation in the severity of monsoon events in India depends on the ocean currents (El Ni$o) off the Pacific Coast of South America.

A new kind of environmental hazard will emerge in the next few years as a result of developments in biotechnology. Already 28 species of crop plant in the UK have been genetically modified with the potential for commercial agricultural use. Some of these (for example, sugar beet, clover, ryegrass, lucerne and cabbage) have a high probability of exchanging genetic material with native plants, including, for example, genes for herbicide resistance. Perhaps even more of a challenge is the use of genetically modified micro-organisms in agriculture, medicine and industry.

³ÉÈËVRÊÓƵ

Pollution is of growing concern for four reasons. First, it may pose a direct threat to human health. Two examples are the 40 per cent decline in male sperm counts over the past 25 years (together with comparable effects in many non-human vertebrate species) which may be related to man-made oestrogenic compounds in the environment, and the emerging possibility of links between respiratory disease and air pollution. Second, pollutants may enter the food chain and thus indirectly threaten human health. Third, natural species or habitat may be threatened, as in the forest dieback of the 1970s in many European countries, originally attributed to sulphur dioxide but probably more to do with oxides of nitrogen. Fourth, industry is increasingly under pressure to meet legislative standards and therefore needs to know about the origin and transformations of pollutants and their impacts. All new technologies will be subject to close scrutiny in this regard: future competitiveness has to be environmental as well as economic. Legislation is not a threat: it also represents an opportunity for the environmental industries to make "brass" out of "muck".

The current and emerging EU policy for waste disposal in landfill prohibits both co-disposal of domestic and difficult industrial waste and attenuated, as opposed to contained, disposal in a sealed site. On the other hand, the preferred UK option is for attenuated co-disposal. Is the UK solution better from the point of view of the environment? Knowledge and understanding of the movement of pollutants through soils and rocks, of the breakdown products of landfill sites and of the impacts on groundwater are re-quired to answer this question: a recently launched joint EPSRC-NERC programme will provide opportunities for research on these questions.

The term global change encompasses things that are already, incontrovertibly, happening such as the 25 per cent increase (from 280 to 350 parts per million) in carbon dioxide in the earth's atmosphere over the past 200 years and changes in the ozone layer over the polar regions. These include not only the well known "ozone hole" over Antarctica which, according to recent British results is still deepening (with springtime values at Halley Research Station in Antarctica now at less than 40 per cent of values seen in the 1960s) and also now extending into the summertime, but also the 7 per cent reduction in ozone in the Arctic vortex over the past 15 years.

Here, the question is not whether there are changes, but what their impact will be on living systems. For example, the effect of increasing carbon dioxide levels on natural and agricultural ecosystems is almost totally unknown. Nor are the effects necessarily ones that can be predicted on the simples physiological hypothesis that more carbon dioxide means more raw material for photosynthesis and therefore more plant growth. In fact, results of field experiments at the NERC Unit of Comparative Plant Ecology at Sheffield show that experimentally increasing the level of carbon dioxide around individual plants in limestone plant communities can actually decrease plant growth, perhaps because the enrichment of carbon dioxide increases the metabolic activity of soil micro-organisms which, as a result deplete nutrients essential for plant growth. The general point, applicable to many environmental problems, is that unexpected interactions in a system as a whole can produce results that run counter to expectation from the analysis of the system components.

11 = /Some other aspects of global change, for example change in the earth's climate, are surrounded by greater uncertainty. It is tempting to be seduced by individual events such as the recent hot spell in England and last winter's severe flooding in Holland, but sober statistical analysis such as that recently published by the European Climate Support Network is needed to evaluate the significance of long-term trends. The same publication compares predictions of the three major European climate models (from France, Germany and the UK): the current best bet is that by the middle of the next century, the UK will be warmer and (at least in winter) wetter. The current models include many components: atmospheric physics, ocean dynamics (with their vast bulk and enormous heat capacity these are the principal drivers of climate change), land surface and hydrological effects, clouds and aerosols. However, the full coupling together of all these components, together with the inclusion of biological feedbacks and changes in the chemistry of the atmosphere, still remain as challenges for the future.

³ÉÈËVRÊÓƵ

From the point of view of the policymakers, not only is it important to know the likely impacts of global change on the human population and on other living systems, but also whether changes in human activity (costly in social and economic terms) could influence the present trends and if so on what timescale. Understanding these aspects of global change is a huge scientific challenge which has stimulated a new level of international collaboration through programmes such as the World Climate Research Programme and the International Geosphere Biosphere Programme. It has underlined the need for integrating the work of physicists, chemists, geologists, biologists and others. Global change is also potentially the most serious and long-term of the "environmental crises" facing us today.

The other major environmental crisis that has become apparent in the past 15 years is the global loss of biodiversity. In the three-billion-year history of life on earth there have been five mass extinction events in which a significant proportion of the biodiversity on earth disappeared. There is now virtually universal acceptance by evolutionary biologists and ecologists that the sixth mass extinction is in train at this moment, with a rate of species loss, estimated in a recent survey by May and Lawton of 1,000 to 10,000 times the "background rate". Man is playing a major role in destruction of biodiversity, as a result of a mixture of activities including habitat destruction and fragmentation, pollution, hunting, introduction of pests, diseases and competitors.

The UK has the best-studied fauna and flora of any country in the world, which puts us in a unique position of opportunity and responsibility to unravel the causes and consequences of loss of biodiversity. The number of species that have gone extinct in the UK over the past 100 years is modest, probably of the order of 200, but a far greater number of species are declining dramatically in range and abundance. Although the likely cause is to do with intensification of agriculture, the exact causes are not yet understood for the majority of species.

³ÉÈËVRÊÓƵ

For some taxonomic groups (for example, bacteria) the basic task of documenting diversity is still in its infancy, so that the current estimates of the world's total number of species range from 1.5 to 30 million. Beyond refining our taxonomic knowledge, an understanding of the influence of habitat fragmentation on population persistence, of past rates of speciation and extinction (increasingly accessible from molecular phylogenies) and of the spatial scales at which ecological processes operate are all central to the challenge of understanding bio-diversity.

Let me end with three general comments. First, advances in our understanding of these environmental issues will often come in quite unexpected ways and from purely curiosity-driven basic research. The discovery of the Antarctic ozone hole by the British Antarctic Survey was an unexpected by-product of long-term, curiosity-motivated study of the upper atmosphere: knowledge of the taxonomy and ecology of obscure deep-sea organisms is crucial to appraisal of the option of waste disposal (including Brent Spar) in the deep oceans; the kinds of fluid dynamic models that have been developed for understanding how volcanoes erupt can be applied to enhance the efficiency of production of steel castings.

11 = /Second, many of the scientific problems in the environment have been intractable because of measurement problems. But in the past few years new technologies, ranging from satellite observation for measuring minute details of the earth's vegetation, to molecular probes for measuring the structure of plant and animal populations and for characterising bacterial types in the environment, have been developed. These are opening up new possibilities for "measuring the unmeasurable": equivalent developments for measuring properties of the marine environment are a high priority.

Third, understanding and predicting of the environment rely to a substantial extent on modelling complex sytems with properties that are either genuinely emergent (cannot be understood within the theoretical framework of a lower reductionist level) or are counter-intuitive properties of interactions between elements of the system. A range of modelling techniques is used. At one extreme are simply phenomenological models that give an impressionist insight into the behaviour of a system and extract underlying principles. At the other are detailed, large-scale simulations, such as those used in the analysis and preduction of global change. Although the UK has world-class expertise across this range of approaches, it does not yet have a focus, analogous to the Sante Fe Institute, specifically dedicated to the abstract problems of analysing complexity.

³ÉÈËVRÊÓƵ

John Krebs Chief executive of NERC, Royal Society research professor in the department of zoology, Oxford University and a fellow of Pembroke College. He runs a research group investigating population change in farmland birds.