Introduction
The world is stranger than we can imagine and surprises are inevitable in science; thus we found, for example, that pesticides increase pests, antibiotics can create pathogens, agricultural development creates hunger and flood control leads to flooding (Levins, 1995a, b). But some of these surprises could have been avoided if the problems had been posed so as to accommodate solutions in the context of The Whole, taking complexities into account. Predicting the impact of a changing world on human health is a hard task and requires an interdisciplinary approach drawn from the fields of evolution, biogeography, ecology and social sciences, and relies on various methodologies such as mathematical modelling and historical analysis (Awerbuch, 1994; Levins, 1995a, b; Awerbuch et al., 1996; McMichael, 1997). Indeed, integrated assessment modelling of human health has been recommended as a global methodology to develop prevention strategies, educate policy makers and assess the impact of interventions (Martens, 1998; see also Chapter 8).
When even a simple change occurs in the physical environment, its effects percolate through a complex network of physical, biological and social interactions that feed back and feed forwards. Along some pathways the effects are attenuated and may even disappear; along others they are amplified and can show up at points far removed from their original entry into the system; along still other pathways the effects may be reversed so that, for example, heating may lower the temperature or adding nitrogen to a lake may reduce the nitrogen level (Levins & Lane, 1977).