Introduction

This chapter outlines the approach that is typically taken in epidemiology to the study of disease risk, gives examples of studies that currently serve as paradigms for epidemiological thinking, and compares these traditional approaches with what is needed to understand health problems associated with global environmental change. It is argued that in order to meet this new challenge, epidemiology must encompass three things: models of extended causation, multiple levels of analysis, and differential vulnerability to disease and injury. This chapter examines in detail the concept of vulnerability, the condition of individuals or groups that modifies the effect of environmental exposures on disease outcomes.

Modern epidemiology is both useful and wrong. By this I mean that epidemiology is an asset to public health because it points out ways to prevent disease and injury, but the foundations of the enterprise are shaky. “Wrong” is putting it strongly – “limited” may be closer to the mark. When you look closely you can see that the methods used by epidemiologists assume a world that does not exist. This is not a criticism of epidemiology, rather an observation on the way that science in general works. Like all scientific disciplines, epidemiology is a tool for creating knowledge. All tools provide partial and incomplete access to the external world because they themselves are imperfect reflections of the world. (Hammers assume there is not a plank that cannot be nailed; telescopes presuppose a universe in which all that matters can be seen.)

Introduction

According to the World Health Organization (WHO, 1996) 30 infectious diseases new to medicine emerged between 1976 and 1996. Included are HIV/AIDS, Ebola, Lyme disease, Legionnaires' disease, toxic Escherichia coli, a new hantavirus, a new strain of cholera and a rash of rapidly evolving antibiotic-resistant organisms. In addition, there has been a resurgence and redistribution of several old diseases on a global scale; for example, malaria and dengue (“breakbone”) fever carried by (vectored by) mosquitoes. The factors influencing this lability of infectious diseases are many and varied. They include urbanization, increased human mobility, long-distance trade, changing land-use patterns, drug abuse and sexual behaviours, the rise of antibiotic resistance, the decline of public health infrastructure in many countries, and a quarter century of predominantly anthropogenic climate change. This complex mix of potential influences means, of course, that the scientific task of attributing causation is difficult. This chapter discusses the types of evidence relevant to the detection of changes in infectious disease occurrence in response to climatic variations and trends.

Arthropods such as mosquitoes and ticks are extremely sensitive to climate. Throughout the past century public health researchers have understood that climate circumscribes the distribution of mosquito-borne diseases, while weather affects the timing and intensity of outbreaks (Gill, 1920, 1921; Dobson& Carper, 1993). Paleoclimatic data (Elias, 1994) demonstrate that geographical shifts of beetles have been closely associated with changes in climate.

Introduction

Future global environmental exposures may be significantly different from those experienced in the past. Forecasting and preparing for the resultant potential ecological, social and population health impacts requires innovative and interdisciplinary research approaches, both to advance global change/health science and to contribute to informed policy decisions. These approaches include empirical analyses and scenario-based exposure modelling to achieve meaningful risk assessments of the potential impacts of climate and ecological changes. This chapter focuses on the application of epidemiology (an empirically based discipline) to understanding the potential health consequences of global environmental change. The empirical knowledge gained from epidemiological studies should be used iteratively with model development to strengthen the foundation of predictive models.

Epidemiological research can be used in the three domains introduced in Chapter 1: first, historical analogue studies to help understand current vulnerability to climate-sensitive diseases (including contributions to understanding the mechanisms of effects) and to forecast the health effects of exposures similar to those in the analogue situation; second, studies seeking early evidence of changes in health risk indicators or health status occurring in response to actual environmental change; and third, using existing empirical knowledge and theory to develop empirical-statistical or biophysical models of future health outcomes in relation to defined scenarios of change. This chapter discusses some standard epidemiological methods used to generate quantitative estimates of exposure–disease associations for studies in these three domains. The examples focus primarily on climate variability and change to maintain consistency throughout the discussion.

Traditional Western environmental medicine acquired renewed significance during the 1990s. Significant global climate change is likely to occur during the twenty-first century, and will alter the needs for population health maintenance as well as the resources available for the management of disease crises. In the past, environmental medicine held that human health and disease could not be assessed independently of climate and place. Interactions between changing climate and human health were thus assumed. Those who hope to recover a measure of this more ancient stance towards medicine question the utility of framing future epidemiology in narrow clinical paradigms. Advocates of a more global epidemiology turn away from the study of risk factors and therapy, in favour of larger environmental models of health and disease.

The study of the history of disease and biometeorology during the last century carried the expansive environmental perspective far more than did clinical and community epidemiology. History can have relevance now for those crafting a new global epidemiological vision. Medicine's former interest in weather and climate directed investigation and intervention towards population health maintenance. Withdrawing from grand and costly goals, western medicine increasingly focused on individuals and local environmental hazards, even in the arena of public health. The heroes and often-told stories of medical history relayed by medical and scientific practitioners accentuate this narrowed perspective. By questioning accepted and self-congratulatory historical constructions of the past, epidemiologists excavate new foundations for the future.

Introduction

There is a long history of man's awareness of both climate and environmental influences on his health and well-being. Hippocrates wrote his “On Airs, Waters and Places” to educate his students on the climate and geographical risk factors that could aid in the prediction and diagnosis of diseases. Early climatologists defined climate in terms of the effects on the organs of the human body (such as von Humboldt in the early nineteenth century), and weather prediction has long been based on aching corns and squeaky joints.

Extensive literature has accumulated over the last few hundred years on the associations between climate, environment and disease. However, the consistency of these associations was often found to depend on geographical location, and seldom resulted in a scientific consensus on the causative pathway. Malaria obtained its name from its presumed cause – the bad odours emanating from marshes in Italy – but this olfactorial connection is absent from many other malarious regions. Our biological and ecological knowledge regarding the dynamics of malaria has greatly improved over the last century. However, the plethora of factors that determine the outcome of disease still causes major disputes over assessing the contribution of a single factor.

Medical geography and medical climatology have existed as scientific disciplines for a long time. However, they are largely descriptive and have been of little practical significance. The current need to assess the impact of global environmental change on disease intensity and distribution has given a new relevance to these disciplines.

Introduction

The study of global environmental change and its current and potential health impacts encounters uncertainties at many levels. These have profound implications for the form and content of science, scientific communication and policy-making. Indeed, it was in the course of the first substantive formal international discussion of these large-scale environmental issues, at the 1992 UN Conference on Environment and Development (the “Rio Conference”), that the Precautionary Principle was clearly enunciated and endorsed. That principle states, in essence, that scientific uncertainty in relation to a phenomenon with potentially serious, perhaps irreversible, consequences does not justify lack of preventive action.

During the past decade, the important role of uncertainty has become especially evident in the scientific and public debate around global climate change and its impacts upon human societies and population health. In each such discourse, pertaining to global environmental changes, there is continuing debate about the nature and quality of the scientific evidence for both the process and its impacts; about the relevance and legitimacy of multi-decadal-length modelling predictions; about the assumptions that can or should be made about future human societies and their capacity to handle changed circumstances; about the extent to which the medium to distant future should be discounted; about moral obligations between rich and poor nations and between present and future generations; about the extent to which decisions can be made within an orthodox economic framework by assigning money values to all present and future variables in the equations; and about the decision-making structures and forms of international governance appropriate to this type of large-scale phenomenon.

Introduction

The meaning of the word “environment” is elastic. Conventionally it refers to the various external factors that impinge on human health through exposures common to members of groups, communities or whole populations, and that are typically not under the control of individuals (i.e. the exposures are predominantly involuntary). Thus, “environmental exposures” are usually thought of as physical, chemical and microbiological agents that impinge on us from the immediately surrounding (ambient) environment.

The “environmental” roles of socioeconomic status in the determination of disease patterns, including aspects such as housing quality and material circumstances, have also claimed increasing attention from health researchers. This, however, requires a more inclusive definition of “environment” – one that embraces social and economic relations, the built environment and the associated patterns of living.

Note also that we typically view the environment as being “out there”. It surrounds us, it impinges on us – but it is not us. This implied separateness reflects the great philosophical tradition that arose in seventeenth-century Europe as the foundations of modern empirical western science were being laid by Bacon, Descartes, Newton and their contemporaries. For several centuries this view helped us to manage, exploit and reshape the natural world in order to advance the material interests of industrializing and modernizing western society. In recent times, however, the magnitude of that environmental impact by human societies has increased exponentially.

Introduction

In order to better understand the potential impact of global environmental change on human health it is necessary to accomplish two contrasting tasks. First, it is important to widen the conceptual structure by which health risks are evaluated to include a broader array of more distal risk factors than has been common in public health in recent years. In addition, to focus the analysis in terms that facilitate meaningful comparisons with other important risks to health, there is need to structure the analysis in absolute measures of ill-health and in terms of standard and emerging decision-making tools. Progress in both these arenas will be needed to effectively guide intervention policies.

To approach these tasks, we follow a temporal progression. First, we briefly examine historical views of human health and the environment to show that the challenges now created by global environmental change actually extend contemporary public health's scope into realms previously embraced by the field. We next offer an analytical structure for addressing the linkages and pathways between multiple social and ecological processes acting at various scales, ultimately influencing health. Issues of causality and capability lead us to examine how a disease-based, resource-effectiveness paradigm might be expanded to understand environment–health connections today. We quantify the current contribution of environmental risk factors to ill-health, and from this basis peer into the future. We discuss how attributable and avoidable risk calculations relate to public health planning, and the implications of incorporating considerations of net present value and sustainability.

Over the past two decades there has been a rapid evolution of research concepts and methods in relation to global environmental changes – their processes, impacts and the response options. The scale and complexity of these environmental problems are, in general, greater than those that individual scientists or their disciplines usually address. This is particularly so for those components of the topic that are furthest “downstream” from the pressures, or their drivers, that initiate the processes of global environmental change.

Indeed, in seeking to detect or forecast the population health impacts of global environmental changes there is an additional difficulty. Not only is the impact of research contingent on various assumptions, simplifications and projections made by scientists working “upstream” on the environmental change process per se, but the category of outcome – a change in the rate of disease or death – is one that usually has multiple contending explanations. If a glacier melts, then temperature increase is a very plausible explanation. Likewise, if birds, bees and buds exhibit their springtime behaviours a little earlier as background temperatures rise, that too is reasonably attributable to climatic change. However, if malaria ascends in the highlands of eastern Africa, regional climate change is just one contending explanation – along with changes in patterns of land use, population movement, increased urban poverty, a decline in the use of pesticides for mosquito control, or the rise of resistance to antimalarial drugs by the parasite.

Acronyms

1. AATSR Advanced Along Track Scanning Radiometer

2. ADEOS Advanced Earth Observation Satellite

3. ALOS Advanced Land Observing Satellite

4. ARIES Australian Resource Information & Environment Satellite

5. ASTER Advanced Spaceborne Thermal Emission & Reflection Radiometer

6. AVHRR Advanced Very High Resolution Radiometer

7. AVIRIS dvanced Visible and Infrared Imaging Spectrometer

8. AVNIR Advanced Visible & Near Infrared Radiometer

9. CBERS China Brazil Earth Resources Satellite

10. ENVISAT Environmental Satellite

11. EO-1 Earth Orbiter-1

12. ERS ESA (European Space Agency) Remote Sensing

13. ETM+ Enhanced Thematic Mapper (Landsat)

14. GLI Global Imager

15. IRS Indian Remote Sensing Satellite

16. LISS Linear Imaging Self Scanning System

17. MARA Mapping Malarial Risk in Africa

18. MODIS Moderate Resolution Imaging Spectro Radiometer

19. MSS Multispectral Scanner

20. NOAA National Oceanographic & Atmospheric Administration

21. PAN Panchromatic

22. RS Remote Sensing

23. SAR Synthetic Aperture Radar

24. SPOT Système Pour l'Observation de la Terre

25. TM Thematic Mapper (Landsat)

26. WiFS Wide Field Scanner

27. XS Multispectral (SPOT)

Introduction

There has been much speculation about the potential impacts of climate change on the map of human health, particularly on the patterns of vector-borne diseases (e.g. Longstreth & Wiseman, 1989; WHO, 1990; Dobson & Carper, 1992; Shope, 1992; Kovats, 2000; Chapters 7 & 8). The impact of climate change on the transmission patterns of these diseases can be both direct (e.g. effect of changes in precipitation on populations of arthropod vectors) and indirect (e.g. human population dynamics and their effects on exposure risk, changes in vegetation, hydrology and other landscape features).

Introduction

Although landscape ecology is generally considered a terrestrial discipline, ecologists working on streams and rivers have long been interested in the spatial relations and geographical distribution of aquatic organisms and their habitats. In fact, if a landscape is defined as a spatially heterogeneous area, and landscape ecology as the study of its structure, function, and change (Turner, 1989), then landscape ecology has the potential to be an important force in the appropriate management of streams and rivers. The application of landscape ecology to riverine management is particularly well suited to three classes of activities: conservation of biodiversity, fisheries management, and restoration/ rehabilitation of biological integrity.

In this chapter we review relevant stream-ecology concepts that encompass a landscape perspective and link these concepts to current management practices. We argue, with examples, that scale-dependent processes that are valuable to society underlie biotic patterns in streams. We show how recent evidence links certain land-use activities with altered stream condition, and attempt to present the underlying mechanisms responsible. To show the integration of landscape principles with practical management, we present examples related to stream restoration, recreational fishing, and a case study of an ongoing project to prioritize the conservation of aquatic biodiversity on a regional basis.

Landscape elements of stream ecology

Several categories of concepts have been developed to explain how streams function in the context of the landscape (Ward, 1989; Lorenz et al., 1997): concepts that focus on longitudinal changes of the biota; concepts emphasizing lateral interactions; concepts that integrate longitudinal, lateral, and vertical dimensions of streams; and concepts emphasizing spatial hierarchies and temporal changes.

Introduction

With ever-increasing loss and degradation of wildlife habitat, wildlife management decisions depend on a solid understanding of the influence of both patch characteristics and landscape structure on populations. Appropriately designed multi-scale ecological studies are becoming more and more important in determining how current and future land-use management decisions will affect the survival of natural populations. Effective management plans for populations and regions depend on clear and interpretable results from properly designed studies.

Historically, researchers designed studies to examine the effects of patch-scale characteristics on population dynamics. A patch is defined as a discrete area of contiguous and homogeneous habitat. Patch-based ecological studies address the relationship between the inherent characteristics of the individual patches (e.g., patch size, patch quality, patch isolation) and some ecological pattern (e.g., distribution and abundance of organisms) or process (e.g., dispersal, disturbance regimes, predation, or competition) (e.g., reviews by Andrén, 1994 and Bender et al., 1998).

Recently, researchers have begun to recognize the importance of considering the effect of the landscape context of the patch. A landscape-scale ecological study addresses one or more of (1) the effect of landscape structure on the distribution and/or abundance of organisms, (2) the effect of landscape structure on an ecological process(es) (e.g., animal movement), or (3) the effect of ecological process(es) (e.g., fire), or organisms (e.g., beavers; see Johnston, 1995), on landscape structure. Landscape structure implies spatial heterogeneity, which is described in terms of landscape composition and configuration. Landscape composition is the amount of the different landscape elements (e.g., habitat types, road cover) in the landscape. Landscape configuration describes the spatial arrangements of these elements.

Introduction

The challenges facing natural resource managers occur over entire landscapes and involve landscape components at many scales. Many resource managers are shifting their approach from managing resources such as fish, wildlife, and water separately to managing for the integrity of entire ecosystems (Christensen et al., 1996). Indeed, nearly all resource management agencies in the USA have recognized that informed management decisions cannot be made exclusively at the level of habitat units or local sites. It is generally accepted that ecological patterns and processes must be considered over large areas when biodiversity and ecological function must be maintained while the goods and services desired by the public are provided. For example, forest managers must determine the patterns and timing of tree harvesting while maintaining an amount and arrangement of habitats that will sustain many species. Managers of parks and nature reserves must be attentive to actions occurring on surrounding lands outside their jurisdiction. Aquatic resource managers must broaden their perspective to encompass the terrestrial and human landscape to manage stream and lake resources effectively (Hynes, 1975, widely regarded as the father of modern stream ecology and quoted above; Naiman et al., 1995). Landscape ecology also is implicit in the paradigm of ecosystem management (Grumbine,1994; Christensen et al., 1996).

Despite the acknowledged importance of a landscape perspective by both scientists and resource managers, determining how to implement management at broader scales is very much a work in progress.

Traditionally, natural resources have been often managed using information collected from small scales, resulting in variable and limited success. To improve these results, many scientists and natural resource managers have recognized the need to adopt a large-scale approach to natural resource management, using the concepts, principles, and methods of landscape ecology. At the same time, many landscape ecologists have also realized that further development of landscape ecology will benefit from better connections with resource management issues. However, as is often the case between academic and non-academic worlds, landscape ecologists and natural resource managers historically have not communicated well. Landscape ecologists often do research without regard to the needs for natural resource management, and managers often do not know how to apply landscape ecology to managing natural resources.

To facilitate the communication between landscape ecologists and natural resource managers, we hosted the 13th annual conference of the US Regional Association of the International Association for Landscape Ecology (US-IALE) on the campus of Michigan State University in 1998. The conference's theme was “Applications of landscape ecology in natural resource management.” Clearly, this theme of linking landscape ecology with natural resource management reflected the desire of many others, as more than 500 landscape ecologists and natural resource managers from around the world participated in the conference (the largest number ever to attend a US-IALE annual meeting).The conference was a huge success, but we were urged by many attendees to produce a book expanding upon the ideas presented at the conference, reaching a larger audience, and promoting further communication and collaboration between the landscape ecology and natural resource management communities.

My theme is that when it comes to land-use research, planning, and management, there is a need to enlarge the frame of reference from the landscape to the region. Although the term “landscape” is often extended beyond the dictionary definition of “an expanse of scenery seen by the eye in one view” to include what can be distinguished in an aerial photo or satellite image, a landscape is also described by the interactions of different identifiable units (sometimes called ecotypes) on the land surface which are based upon ecological, social, and economic considerations (Turner, 1989; Turner et al., 1996). In terms of an absolute spatial scale, a landscape is a large geographic expanse encompassing anywhere from ten to several thousand square kilometers (Bailey, 1996). While the landscape perspective in ecology has enlarged the scale at which research is carried out, a more appropriate scale for addressing many land-use, land-tenure, and environmental problems is the region, which is the focus of this chapter.

In the 1930s, social scientists promoted the concept of regionalism in which social indicators were used to compare different geographical and political regions. This concept considered regions to be large geographic expanses (e.g., multiple counties, or multiple states) based primarily upon political or social boundaries (Odum, 1936). My father, Howard W. Odum, and his faculty and staff at the University of North Carolina, Chapel Hill, were leaders in developing this field. His books Southern Regions (1936) and American Regionalism (Odum and Moore, 1938) were very influential in shaping the political scene of North Carolina, and the southern region of the United States as a whole.

Introduction

Habitat fragmentation is the nearly inevitable result of contemporary land use. Beyond the biodiversity shadow cast by reserves and habitat protection plans, settlement patterns continue to extirpate remaining habitats. This reality should lead landscape ecologists and natural resource managers to vigorously investigate small patches of habitat. While accumulating knowledge of landscape structure and function leaves no doubt about the critical importance of large indigenous habitat patches, knowledge of the ecological function of small patches is comparatively meager. Yet opportunities for preserving or creating small patches characterize human land use. Informed by landscape ecology, conservation biology, ecosystem management, and restoration ecology, humans have only recently begun to preserve and reconnect the pieces of what were once large, continuous ecosystems. The undeniable trend of contemporary settlement patterns suggests that the small shall inherit the earth. How these small patches can serve ecological functions is a pragmatic question for landscape ecology.

A bird's-eye view of settled landscapes today presents a striking image of the impact of humans on the natural world. The landscape pattern observed is profoundly influenced by culture, created according to political systems, economic uses, aesthetic preferences, and social conventions (Nassauer, 1995). Culture, defined as “the sum total of ways of living built up by a group of human beings and transmitted from one generation to another” (Random House, 1987), not only influences landscape patterns, it can also suggest new landscapes designed to promote ecological function (Nassauer, 1995). New landscape patterns designed without consideration of the appearance of cultural values on the land are not culturally sustainable (Nassauer, 1992, 1997).

Introduction

Natural resource management has moved from a single disciplinary and one resource management approach to an interdisciplinary and ecosystem-based approach. Many conceptual models are being developed to understand and implement ecosystem management and forest certification initiatives that require an integration of data from both the social and natural systems (Vogt et al., 1997, 1999a,b). These changed approaches to natural resource management arose from a perception that variables critical in controlling the health and functioning of an ecosystem could only be determined by integrating information from both the social and the natural sciences (Vogt et al., 1997). However, it has been difficult to take many of the theoretical discussions and the frameworks or conceptual models that they have produced and to operationalize or put them into practice on the ground.

Despite these discussions and the recognition of their importance, social and natural science data have been ineffectively incorporated into the management and trade-off assessments of natural resources (Berry and Vogt, 1999).We hypothesize that some of this has occurred because of the distinct spatial scales being used by different disciplines which have not allowed for integration of information to occur at a causal level. The complexity and uncertainty of data needed to understand ecosystems by both social and natural scientists have also made it difficult for managers to recognize when the wrong indicators are being monitored or whether a system could degrade due to management (Larson et al., 1999; Vogt et al., 1999c). The need to link data causally from both disciplines as part of ecosystem management has given greater impetus to develop practical tools that would allow this integration to be accomplished.