As animals we have been a remarkably successful species; but also as animals we are vulnerable to environmental, in particular climate change. Such change is accelerating as a result of human activity, and global warming may already be taking place. Although we can foresee the trends, we cannot yet be specific about the results. Change usually proceeds by steps rather than gradients. But warming would probably include new risks to human health and contribute to an increase in human displacement. Of course climate change is only one among other complex problems facing human society, but it is closely related to them all, including population increase, environmental degradation and loss of biodiversity. We cannot prevent global warming but we can anticipate and mitigate some of its worst effects. Peoples and governments still need persuading of the need for action and of the magnitude of the issue at stake.

The atmosphere plays a key role in plant disease, but only recently has it become understood that atmospheric pollutants can influence the response of plants to attack by pests and pathogens. This paper reviews the evidence for this phenomenon, considering impacts of sulphur dioxide, nitrogen dioxide and ozone, mainly on fungal pathogens and aphid pests. Field observations in polluted areas have indicated changes in abundance of pests and pathogens and in some cases a causal link has been demonstrated in controlled experiments. A major study is described in which consistent marked positive impacts of SO2 and NO2 have been shown on a range of British agricultural aphid pests, using four different approaches: fumigations, nitration studies, exposure along air pollution gradients and a nation-wide field survey. Ozone, in contrast, produces a more complex range of responses. These effects are apparently mediated via chemical changes in the plant. Fungal pathogens show both positive and negative responses to air pollutants. A study is described in which these opposite responses in two different fungal species were observed in a field SO2–fumigation system and confirmed in controlled laboratory fumigations. Models are presented to describe the complex pathways by which air pollutants could influence host plant performance via impacts on pests and pathogens.

As there is little evidence of pollution affecting the health of fish and shellfish on a global scale, this paper attempts to put into perspective the pollution/fish disease relationship by reviewing examples of studies and reports in the historic and current literature. Although there is no dispute that pollution can affect the health of aquatic organisms under laboratory conditions and may be responsible for the decline of populations of such animals in some inland waters and some estuaries, most of the evidence for pollution causing or increasing disease in fish in open waters is circumstantial. Historical data proves that almost all fish and shellfish diseases known today have been described since the end of the last century. However, it is also known that water pollution, especially in inland waters, has for the past 400-500 years been the result of urbanization and industrialization. This has resulted in some major rivers becoming devoid of or deficient in fish stocks. The concern that pollution may influence the health status of fish and shellfish stocks has increased over the past 20 years. Initial attention was paid to epidermal diseases, including fin-rot in demersal fish, and protozoan diseases in molluscs in the heavily polluted bays and estuaries in North America. As the interest in this subject spread, it became political, and often controversial, especially amongst the North Sea countries. The disagreements have largely been settled amongst scientists because international bodies, such as the International Council for Exploration of the Sea (ICES), established workshops to investigate sampling methods and disease-reporting techniques. Recommendations from those workshops have contributed to some form of standardization for field work and the subject, although largely subjective, has some objective approaches which are described. As there are variable, interacting biological and physical influences in the aquatic environment, it is difficult to establish the background prevalences of diseases in populations offish and shellfish. Examples of the influences of climatic changes are presented, and these show that short-term catastrophes can be directly related. However, a more long-term problem is water acidification resulting largely from anthropogenic activities. In parts of Scandinavia this has, and is, leading to decimation offish stocks in inland waters. In general, diseases in fish and shellfish are very localized, but there is concern amongst scientists that certain cancers, especially liver tumours, occurring in demersal fish inhabiting polluted estuarine and coastal waters, are related to the release of chemicals, e.g. hydrocarbons, pesticides and heavy metals. This subject is discussed in detail, with examples of the author's own studies in North Sea fish. It is concluded that cancers in fish are of extremely low prevalence, and only present in a very few species, and then only in the oldest animals. Though changes in disease pattern may well be an indication of adverse environmental effects, further research is necessary for conclusive evidence.

Stratospheric ozone depletion threatens to increase exposure to ultraviolet (UV) radiation which is known to be a factor in a number of diseases. There is little doubt that cumulative exposure to UV radiation is important in the aetiology of non-melanoma skin cancers. Evidence is also strong for a link with cutaneous malignant melanoma, although here it appears to be intermittent intense exposure that is most damaging. More controversial is the view that exposure to solar radiation is a significant factor in ocular damage, particularly in the formation of cataracts. Earlier studies pointing to such an effect have been criticized and alternative aetiological hypotheses have been proposed. However, other studies do show an effect of UV exposure on cortical cataract. Concern is also growing that UV may be capable of activating viruses and have immunological effects that might exacerbate infectious disease. Very worrying is the possibility that UV exposure can activate the human immunodeficiency virus which might accelerate the onset of AIDS. Any such health effects that have been observed in human populations are the result of exposure to existing, naturally occurring levels of UV radiation. There is, therefore, great concern about the possible exacerbation of these impacts as a result of increased exposure to UV radiation associated with stratospheric ozone depletion. However, any assessment of the nature and scale of such impacts on human health has to deal with several major problems and these are the focus of this paper. There are uncertainties about recent trends in stratospheric ozone and problems in the prediction of future changes. Following on from this are the difficulties of estimating what effects these changes will have on UV flux at ground level in populated areas. Further problems arise in the prediction of changes in biologically significant doses to humans which might be affected by changes in behaviour as well as by changes in the environment. Finally, the limitations of existing epidemiological knowledge of the effects of UV exposure are a constraint on our ability to predict what the health effects of any changed UV doses might be.

Only green plants can convert the single carbon units of atmospheric carbon dioxide into the multi-carbon organic molecules on which all forms of life depend. Only green plants can provide the oxygen required by man and other aerobic organisms. In addition to his basic need for preformed organic molecules and oxygen, man also depends on plants to provide him, directly or indirectly, with an array of specific compounds such as vitamins and essential amino acids. Inadequate supplies of these may hinder growth and development or give rise to well defined deficiency diseases. At the present time information concerning the distribution and concentrations of such essential nutrients in plants is largely restricted to those plants that are already used as human foods. Nothing or virtually nothing is known about the chemical composition of approximately 250000 wild and little-used species. Amongst these there may be many that could provide us with cheap and plentiful new sources of essential nutrients that could be of enormous benefit to those suffering not only from full-blown deficiency diseases but also suffering sub-normal health due to partial deficiencies. The destruction of much of the world's wilderness areas has already deprived us of the opportunity to evaluate the contributions that a great many plant species might have made towards the elimination of deficiency diseases.

Many plants used as human foods contain compounds that are toxic to man. If intake is sufficiently high, these toxins may cause disease. Breeding programmes designed to eliminate toxins from crops species or reduce their concentrations to acceptable levels depend on genetic variability within the species or the possibility of producing hybrids with the desired characteristics. The motor neurone disease, lathyrism, which affects populations in the Indian sub-continent, Africa and China is caused by a toxin in the seeds of Lathyrus sativus. Surveys of cultivated plant populations have shown great variability in toxin levels and such genetic differences make it possible to select and breed toxin-low varieties. The existence of toxin-free species within the same genus has led to research aimed at producing toxin-free hybrids suitable for agricultural use. Approaches, designed to reduce or eliminate diet-related diseases, depend on the maintenance of the greatest possible diversity among both wild and cultivated plant populations. Such diversity is under threat.

Some 250 plant species are used as sources of drugs in western medicine. Most of these drugs are obtained from plants whose therapeutic value was recognized long before the compounds were isolated. In the developing countries of the world, it is estimated that 25000 plant species may be used in medicinal preparations. Few of these plants have been studied systematically to determine whether their reputations are justified and if so, the nature of the drugs they contain. If only 0.1 % were to yield useful drugs, they could make a major contribution to human health and welfare. The plants and the indigenous populations who understand their uses are both disappearing and it is a matter of great regret that much that could be of value has been and will be lost.

This review addresses' changes in the ecology of vectors and epidemiology of vector-borne diseases which result from deforestation. Selected examples are considered from viral and parasitic infections (arboviruses, malaria, the leishmaniases, nlariases, Chagas Disease and schistosomiasis) where disease patterns have been directly or indirectly influenced by loss of natural tropical forests. A wide range of activities have resulted in deforestation. These include colonisation and settlement, transmigrant programmes, logging, agricultural activities to provide for cash crops, mining, hydropower development and fuelwood collection. Each activity influences the prevalence, incidence and distribution of vector-borne disease. Three main regions are considered – South America, West & Central Africa and South-East Asia. In each, documented changes in vector ecology and behaviour and disease pattern have occurred. Such changes result from human activity at the forest interface and within the forest. They include both deforestation and reafforestation programmes. Deforestation, or activities associated with it, have produced new habitats for Anopheles darlingi mosquitoes and have caused malaria epidemics in South America. The different species complexes in South-East Asia (A. dirus, A. minimus, A. balabacensis) have been affected in different ways by forest clearance with different impacts on malaria incidence. The ability of zoophilic vectors to adapt to human blood as an alternative source of food and to become associated with human dwellings (peridomestic behaviour) have influenced the distribution of the leishmaniases in South America. Certain species of sandflies (Lutzomyia intermedia, Lu. longipalpis, Lu. whitmani), which were originally zoophilic and sylvatic, have adapted to feeding on humans in peridomestic and even periurban situations. The changes in behaviour of reservoir hosts and the ability of pathogens to adapt to new reservoir hosts in the newly-created habitats also influence the patterns of disease. In anthroponotic infections, such as Plasmodium, Onchocerca and Wuchereria, changes in disease patterns and vector ecology may be more difficult to detect. Detailed knowledge of vector species and species complexes is needed in relation to changing climate associated with deforestation. The distributions of the Anopheles gambiae and Simulium damnosum species complexes in West Africa are examples. There have been detailed longitudinal studies of Anopheles gambiae populations in different ecological zones of West Africa. Studies on Simulium damnosum cytoforms (using chromosome identification methods) in the Onchocerciasis Control Programme were necessary to detect changes in distribution of species in relation to changed habitats. These examples underline the need for studies on the taxonomy of medically-important insects in parallel with long-term observations on changing habitats. In some circumstances, destruction of the forest has reduced or even removed disease transmission (e.g. S. neavei-transmitted Onchocerca in Kenya). Whilst the process of deforestation can be expected to continue, hopefully at a decreased rate, it is expected that unpredictable and sometimes rapid changes in disease patterns will pose problems for the public health services.

The paper examines the possible contributions to be made by Geographic Information Systems (GIS) to studies on human and animal trypanosomiasis in Africa. The epidemiological characteristics of trypanosomiasis are reviewed in the light of the formula for the basic reproductive rate or number of vector-borne diseases. The paper then describes how important biological characteristics of the vectors of trypanosomiasis in West Africa may be monitored using data from the NOAA series of meteorological satellites. This will lead to an understanding of the spatial distribution of both vectors and disease. An alternative, statistical approach to understanding the spatial distribution of tsetse, based on linear discriminant analysis, is illustrated with the example of Glossina morsitans in Zimbabwe, Kenya and Tanzania. In the case of Zimbabwe, a single climatic variable, the maximum of the mean monthly temperature, correctly predicts the pre-rinderpest distribution of tsetse over 82% of the country; additional climatic and vegetation variables do not improve considerably on this figure. In the cases of Kenya and Tanzania, however, another variable, the maximum of the mean monthly Normalized Difference Vegetation Index, is the single most important variable, giving correct predictions over 69 % of the area; the other climatic and vegetation variables improve this to 82 % overall. Such statistical analyses can guide field work towards the correct biological interpretation of the distributional limits of vectors and may also be used to make predictions about the impact of global change on vector ranges. Examples are given of the areas of Zimbabwe which would become climatically suitable for tsetse given mean temperature increases of 1, 2 and 3 °Centigrade. Five possible causes for sleeping sickness outbreaks are given, illustrated by the analysis of field data or from the output of mathematical models. One cause is abiotic (variation in rainfall), three are biotic (variation in vectorial potential, host immunity, or parasite virulence) and one is historical (the impact of explorers, colonizers and dictators). The implications for disease monitoring, in order to anticipate sleeping sickness outbreaks, are briefly discussed. It is concluded that present data are inadequate to distinguish between these hypotheses. The idea that sleeping sickness outbreaks are periodic (i.e. cyclical) is only barely supported by hard data. Hence it is even difficult to conclude whether the major cause of sleeping sickness outbreaks is biotic (which, in model situations, tends to produce cyclical epidemics) or abiotic. The conclusions emphasize that until we understand more about the variation in space and time of tsetse and trypanosomiasis distribution and abundance we shall not be in a position to benefit from the advances made by GIS. The potential is there, however, to re-introduce the spatial and temporal elements into epidemiological studies that are currently often neglected.

Urbanization involves a physical change in which increasing proportions of populations live in urban settings, however defined. It also implies considerable changes in the ways in which these people live, how they earn their livelihoods, the food which they eat, and the wide range of environmental factors to which they are exposed. There is another underlying assumption that, increasingly, urban populations will be more healthy than their rural counterparts and that higher levels of urbanization will equate with better health status. This paper discusses some of the assumptions underlying this contention. It takes issue with certain of them, particularly the assumption that urbanization affects the health of all residents equally. It is manifestly evident that in many cities, particularly in the developing world, the poor are exposed to greater risks and have much lower health status than their richer neighbours. In addition, whilst urban residents may theoretically have a better access to health care and services than do residents in many rural areas, and whilst many indicators of health do appear better in more highly urbanized societies than ones less so, there are caveats. The paper introduces the concept of epidemiological transition, which suggests that, whilst life expectancy might be higher in many urbanized countries and in certain cities, the inhabitants are often merely suffering from different forms of ill-health, often chronic or degenerative, rather than infective. In certain cities in middle-income countries, residents, particularly the poor, are exposed to a double risk of both infection and chronic degenerative ailments. The paper concludes with a consideration of more general recent statements from the World Health Organization among others, on the impact of urbanization on health. The ‘Healthy Cities’ project is also discussed. WHO identifies a range of general determinants of urban health: physical, social, cultural and environmental. Many represent the by-products of modernization and especially industrialization. It is emphasized that urbanization, and the concentration of human beings into new areas in particular, can bring exposure to new risk factors for large numbers of people. The growth of infectious and parasitic disease in some urban settings must therefore be recognized, as must the emergence of chronic diseases, with the concomitant need for investment in new types of health and social care. However, a number of constraints militate against the achievement of improved urban health, especially in developing countries. The paper concludes by considering some important constraints: the very scale of urban health problems; the impacts of structural adjustment programmes which cut public expenditure on environmental health, health services and nutrition, thereby increasing the vulnerability of the poor; urban management problems; lack of political will; and the difficulties of measuring change in health and effects of policies.