Plague dynamics are driven by climate variation

Abstract

The bacterium Yersinia pestis causes bubonic plague. In Central Asia, where human plague is still reported regularly, the bacterium is common in natural populations of great gerbils. By using field data from 1949-1995 and previously undescribed statistical techniques, we show that Y. pestis prevalence in gerbils increases with warmer springs and wetter summers: A 1 degrees C increase in spring is predicted to lead to a >50% increase in prevalence. Climatic conditions favoring plague apparently existed in this region at the onset of the Black Death as well as when the most recent plague pandemic arose in the same region, and they are expected to continue or become more favorable as a result of climate change. Threats of outbreaks may thus be increasing where humans live in close contact with rodents and fleas (or other wildlife) harboring endemic plague.

Fig. 1.

The field data used in this study were collected in a natural plague focus in Kazakhstan. The data are plague prevalence in great gerbils, counts of fleas collected from trapped gerbils, and meteorological observations. (Left Upper) Kazakhstan on a map of Central Asia with the PreBalkhash focus (between 74 and 78°E and 44 and 47°N) marked as a square. The historic climate (tree-ring) measurement sites are circles marked K (Karakorum) and T (Tien Shan). These sites are located ≈1,000 and 600 km from the research area, respectively. (Right Lower) The LSQ in the PreBalkhash focus from which we have prevalence. The four LSQs (40 × 40 km) circled in red, namely LSQs 78, 83, 91, and 105, represent key sites where collection of samples for testing the presence of plague was more regular and continuous. The Bakanas meteorological station is located in LSQ 117, marked by a red triangle. (Right Upper) The time-series plots of the observed prevalence per LSQ. Open and filled circles denote the observed prevalence during the spring and fall, respectively. The time series of the prevalence fitted by using the model defined by Eq. 1 is shown in red. Using the same model but without any climatic covariates gives the time series shown in gray. Note that owing to the presence of missing values in some covariates (occupancy) and prevalence data, the curves of the fitted values are discontinuous. The fitted values from the model provide a closer fit and reproduce the peaks in prevalence far better than the model without the climatic variables. (Left Lower) Time-series plots of the climate variables, spring rainfall, spring temperature, and summer rainfall (from left to right) in the model defined by Eq. 1.

Fig. 2.

The effect of changes in the environmental conditions on prevalence. (a) The effect of spring temperature on prevalence in the spring. (b) The effect of summer precipitation on prevalence in the fall. Note that the curves in a and b illustrate the mean effect of spring temperature and (log) summer rainfall, respectively, with other covariates and random effects set at their mean values. The unit of temperature is °C, and rainfall is on the log-mm scale (i.e., the untransformed rainfall data are in millimeters). Open circles are the partial residuals for spring temperature and summer precipitation, respectively. The partial residuals are defined as the mean effect of spring temperature (summer precipitation) plus Pearson residuals (i.e., raw residuals rescaled so that they have constant variance, and the constant variance equals the mean-squared deviations of the raw residuals about their mean). Another approach to assess the climate effects is to calculate the induced average changes in the prevalence, with the other covariates unchanged (and held at their historical values and the random effects equal to their estimates). Results of the latter approach, which are reported in the text, are broadly similar but nonidentical to those shown in this figure.