Do rising temperatures matter?

Relationships between the environment and population dynamics in ecology have a long history of polarized debates usually involving two alternative and opposite explanations that exclude each other. One of the best known examples is found in the early question of what regulates populations and the role of extrinsic vs. intrinsic factors, opposing the importance of the environment (or density-independent factors) to that of feedbacks within the nonlinear dynamics of the population itself (or density-dependent factors). Since then, population ecology has recognized the importance of the interplay between external environmental forcing and intrinsic nonlinearities in determining population dynamics. With some delay, the same polarization appeared, and is still present, in debates on the role of climate in the dynamics of infectious diseases at different temporal scales, from interannual fluctuations to longer trends. Although opposite views serve to organize our thinking and the testing of hypotheses, taken to extremes such polarization is unproductive, and delays understanding. From this perspective, Lafferty’s (2009) review is a welcome addition to the existing literature on the importance of considering climate in the context of other aspects of change (e.g., Lindblade et al. 2000, Chaves et al. 2008). However, Lafferty’s description of the state of the science as a ‘‘crisis discipline, reminiscent of the early phases of conservation biology’’ is already outdated, applying for example to the earlier scenarios for the global resurgence of malaria. An earlier reaction to such a ‘‘crisis discipline’’ has contributed to a polarization of the field and it would be unfortunate if conclusions in this review, on the lack of a net effect of climate change on the geographic range of infectious diseases at large spatial scales, were mistakenly taken as further support for the view that climate does not matter. Lafferty argues that climate driven range shifts are more likely than range expansions (with no net positive change) in vector-transmitted diseases such as malaria. We focus first on this point to emphasize that range shifts will matter and should be better understood, regardless of net effects at the large scales of continents or the globe. Importantly, assessments of net effects at too large a spatial scale miss fundamental considerations, as emphasized here with a focus on epidemic malaria in tropical regions. Furthermore, assessments based on current indices of climate suitability are problematic given the limitations of the current indices themselves. Similarly problematic are extrapolations from the history of malaria in Europe to epidemic regions that are climate sensitive. At shorter temporal scales, the relevance of climate has been addressed with seasonal and interannual patterns. Whether and how climate variability drives seasonality and cycles of longer period is important for the mechanisms by which infectious diseases will respond to climate change, for example in the length and timing of the seasons and in the amplitude of phenomena such as the El Nino Southern Oscillation (ENSO). The second part of our commentary concerns Lafferty’s review of interannual climate effects. We argue that the conclusions of the re-analysis of historical yellow fever and ENSO in the US do not extrapolate to yellow fever in general, and that more recent results on epidemic malaria in African highlands support a role of climate variability. With the case study of cholera on the evidence for a role of ENSO, we illustrate current approaches that consider climate drivers in the context of nonlinear disease dynamics. The challenge of considering climate in the context of other aspects of change goes beyond socioeconomic factors and includes pathogen evolution. Importantly, however, this is not an argument against the importance of climate itself, as we illustrate with the open question of the possible synergy of climate and drug resistance.