The Impact of Particulate Matter on Daily Mortality in Bangkok,Thailand

ABSTRACT

Several studies conducted in U.S. cities report an association between acute exposures to particulate matter (PM), usually measured as PM_10, and mortality. Evidence of high concentrations of PM₁₀ in Eastern Europe and in large metropolitan areas outside of the United States, such as Mexico City and Bangkok, underscores the need to determine whether these same associations occur outside of the United States. In addition, conducting studies of mortality and air pollution in regions that have distinctly different seasonal patterns than those of the United States provides an effective opportunity to assess the potentially confounding aspects of seasonality. Over the last few years, daily measures of ambient PM₁₀ have been collected in Bangkok, a tropical city of over 6 million people. In this metropolitan area, PM₁₀ consists largely of fine particles generated from diesel- and gasoline-powered automobiles, and from two-stroke motorcycle engines. Our analysis involved the examination of the relationship between PM₁₀ and daily mortality for 1992 through 1995. In addition to counts of daily natural mortality (total mortality net of accidents, homicides, and suicides), the data were compiled to assess both cardiovascular and respiratory mortality, and natural mortality by age group. A multivariate Poisson regression model was used to explain daily mortality while controlling for several covariates including temperature, humidity, day of the week, season, and time. The analysis indicated a statistically significant association between PM10 and all of the alternative measures of mortality. The results suggest a 10-µg/m³ change in daily PM₁₀ is associated with a 1–2% increase in natural mortality, a 1–2% increase in cardiovascular mortality, and a 3–6% increase in respiratory mortality. These relative risks are generally consistent with or greater than those reported in most studies undertaken in the United States.     

Vocal dialect and genetic subdivisions along a geographic gradient in the orange-tufted sunbird

At least four hypotheses have been suggested to explain the formation and maintenance of song dialects among birds: historic processes (epiphenomenon), genetic or local adaptation, acoustic adaptation, and social adaptation. We studied spatial and temporal distribution of dialect in the orange-tufted sunbird (Nectarinia osea), a small nectarivorous bird that expanded its breeding range in Israel during the past 100 years from the southern part of Rift Valley to the entire country. Sunbird range expansion was concurrent with the establishment of many small settlements with an ethos of gardening, which introduced many ornithophilous plants. We recorded songs and genetically screened individual sunbirds in 29 settlements distributed across a 380 km north–south gradient along the Rift Valley. We show that dialects cluster together into geographical regions in 70% of cases, a moderate concurrence to geography. Settlement establishment date, geographical position, and genetic distance between local populations (i.e., settlements) were all poor predictors for the variance among song dialects. The specific effect of habitat was not tested because all sampled localities were similar in their physical and acoustic properties. Using a network analysis, we show that dialects seem to aggregate into several network communities, which clustered settlement populations from several regions. Our results are best explained by either the epiphenomenon hypothesis or the social adaptation hypothesis, but at present our data cannot state unequivocally which of these hypotheses is better supported. Last, we discovered a negative association between network centrality and genetic diversity, a pattern that requires further examination in other systems.