Predicting Pyrodinium bahamense occurrence using weather pattern data in Eastern Visayas, Philippines.

Pyrodinium bahamense, a harmful alga which causes paralytic shellfish poisoning (PSP), has been found in seawater samples collected along bays in Eastern Visayas in the Philippines. Due to its negative impacts and uncertainty in its occurrence, there is a need to develop a real-time monitoring device of the harmful algal bloom (HAB) occurrence. This study aims to determine whether there is significant relationship between the dependent variable, the P. bahamense cell density, and the independent variables, namely, temperature, rainfall and humidity, bays in Eastern Visayas, and sampling month. This was done through data mining technique of P. bahamense cell density and weather patterns and statistically analyzing the datasets, with P. bahamense cell density as the dependent variable and weather patterns, bays and sampling month as independent variables using M5P regression analysis. Through the regression analysis, the usefulness of the predictive model as an initial development can be assessed. Results of the study showed that the correlation coefficient of P. bahamense with the five independent variables is 0.725 which signified a moderate correlation between P. bahamense cell density and bay, sampling month, temperature, rainfall and humidity. Maximum cell density, which was 20 cells/L occurred during months of wet season in the bays in Eastern Visayas, except in Cancabato bay while the minimum cell density, which was 0 cell/L occurred during months of dry season. It can be concluded from the results that there is enough data to say that there is a significant relationship between P. bahamense occurrence and temperature, rainfall and humidity, so the model is meaningful and useful at alpha = 0.05 or 95% and 0.01 or 99% confidence interval. The data can be used to generate a predictive model for P. bahamense occurrence relative to its cell density.