Abstract:
The catchability of scampi is known to vary on a daily basis related to light levels, on seasonal basis due to moulting or reproductive activities, and also in relation to the lunar tidal cycle. Additionally, previous research studies suggested that their relation to catchability may not be linear. The main purpose of this project is to investigate different ways of estimating these patterns in catchability of scampi from New Zealand commercial catch and effort records, and, particularly, applies approaches to understand the daily patterns by fitting an hourly pattern for tows which are typically 5-7 hours duration. The semi-parametric models - generalized additive mixed models (GAMMs) are proposed here as a flexible alternative to the preliminary models - periodic regression models and generalized additive models (GAMs). Final GAMM models suggest that there is evidence of cyclic daily and lunar patterns in catchability, and also support the suggestion of nonlinear relationships between catchability and other explanatory variables. The GAMMs developed on New Zealand data are then applied to overseas catch data.