Generalized Linear Models (GLMs) or Generalized Additive Models (GAMs) will be developed to determine predictors of the occurrence of selected billfish taxa through a model selection framework to rank the best models using information criteria. The model will consider predictors such as a proxy for market, accessibility, and population density of the nearest city, seasonal dynamics, and satellite derived sea surface temperature, chlorophyll, and productivity. Time and space are among the variables that will be considered for our predictive modeling. The analyses will demonstrate how gear type and selectivity in time and space affects the catch rates especially for recreational and artisanal fisheries. Time is defined in terms of hours, months and seasonal changes. Predictive models will be hierarchical, which means that factors associated with sampling (time, location, season, e.t.c.) that may impact the observation will be considered.