Population studies on Oncomelania quadrasi, the snail intermediate host of Schistosoma japonicum, in the Philippines. 3. Data transformation for significance test of snail density.

For the comparative study of densities of Oncomelania quadrasi, the snail intermediate host of Schistosoma japonicum in the Philippines, the survey data should be transformed to stabilize the variance in determining the significant difference of densities among populations and to evaluate control measures by the difference, since the snail is distributed unevenly in the field. Comparison was made for the efficiency of stabilizing variance among 7 transformation formulas so far reported, and 2 new additional formulas in the present study. As a result, a simple transformation by y=log (x+0.01) was found to be most reliable. This nonparametric transformation can be applied to any snail population irrespective of the type of distribution and the degree of clumping. By this formula, there is no need to discard snail-free samples which are important to evaluate control measures. Using this transformation, snail survey data were analyzed in the areas where control measures were undertaken during the past several years. And significant reduction of snail densities was proven using the t-test.

Population studies on Oncomelania quadrasi, the snail intermediate host of Schistosoma japonicum, in the Philippines. 2. Necessary sample size for snail density survey.

In order to save time and manpower for the density surveys of Oncomelania quadrasi, the snail host of Schistosoma japonicum in the Philippines, the minimal necessary sample size (q) was determined using a formula, q greater than t2/E2(1d - 1 + 1/-x). This formula is based on the Id dispersion index [2, 3] which varies with the degree of clumping of unevenly distributed animals. A nomograph was prepared for the ready determination of the necessary sample size for various degrees of clumping (Id) and mean density (-x) to be encountered in the field. The reliability of the sampling procedure with an interval of 5 m distance, which has been adopted to the routine snail survey in the Philippines, was examined using the relative error (E). The relative errors calculated ranged below or around 20% (E = 0.2). Therefore, the existing sampling method was proven to be highly reliable as the field survey. Comparison was made between the sample size actually taken and the minimum needed theoretically determined by this method, by adopting the Student's t = 1 and relative error E = 0.3 as a permissible reliability level in the field. The actual sample size was more than twice the minimum needed for 21 out of 30 populations surveyed and less than the needed in 5 of the 30 surveys.

Population studies on Oncomelania quadrasi, the snail intermediate host of Schistosoma japonicum, in the Philippines. I. Distribution pattern of the snail in the field.

Distribution of Oncomelania quadrasi, the snail intermediate host of Schistosoma japonicum, was studied in Leyte, Philippines for the type of distribution based upon a large number of survey data. Among four distribution models examined, three uneven distribution models, i.e., the double Poisson, Neyman type A and negative binomial fitted well in this order but the Poisson model showed the poorest fitness. This indicates that the snails are distributed unevenly over the field and that the population mostly consists of small clusters of individual snails. Elucidation of the snail distribution pattern is necessary for determination of sample size in snail survey and for data transformation in significance test of the snail density comparison.