Lack of birth seasonality in a nineteenth-century agricultural population: Escazú, Costa Rica.

Demographic studies repeatedly indicate that human births tend to fluctuate in a seasonal pattern, particularly in agricultural groups. Interestingly, lack of seasonal variation has been infrequently reported. Recently, Arcury et al. (1990) indicated that only one published study identified a population in which no seasonality of births was demonstrated. Here, I report the results of a historical demographic study that investigates whether Escazú, a nineteenth-century population from Costa Rica, experienced birth seasonality. The data set consists of certificates of baptisms from 1851 to 1901 collected at the Parish of San Miguel de Escazú, Costa Rica. The data are an accurate reflection of the actual number of births in the population because even newborn infants who appeared to be in imminent danger of dying (and who subsequently did) were baptized. The baptism series was analyzed by fitting an ARIMA model (0, 1, 1) and cross-correlating it with two climatic series: rainfall and temperature. The autocorrelation analysis of the baptism data fails to demonstrate any cyclical pattern of births in Escazú. Moreover, there is no indication that temperature or rainfall influences the frequency of births. This lack of seasonality is unexpected in an agrarian population such as Ecazú. The results of this analysis indicate that human fertility cycles might not be as prevalent as previously thought.