We used generalized additive models (GAM) to analyze the relationship between spatiotemporal factors and catch, and to estimate the monthly marine fishery yield of single otter trawls in Putuo district of Zhoushan, Ch...We used generalized additive models (GAM) to analyze the relationship between spatiotemporal factors and catch, and to estimate the monthly marine fishery yield of single otter trawls in Putuo district of Zhoushan, China. We used logbooks from five commercial fishing boats and data in government's monthly statistical reports. We developed two GAM models: one included temporal variables (month and hauling time) and spatial variables (longitude and latitude), and another included just two variables, month and the number of fishing boats. Our results suggest that temporal factors explained more of the variability in catch than spatial factors. Furthermore, month explained the majority of variation in catch. Change in spatial distribution of fleet had a temporal component as the boats fished within a relatively small area within the same month, but the area varied among months. The number of boats fishing in each month also explained a large proportion of the variation in catch. Engine power had no effect on catch. The pseudo-coefficients (PCf) of the two GAMs were 0.13 and 0.29 respectively, indicating the both had good fits. The model yielded estimates that were very similar to those in the governmental reports between January to September, with relative estimate errors (REE) of <18%. However, the yields in October and November were significantly underestimated, with REEs of 36% and 27%, respectively.展开更多
基金Supported by the National Natural Science Foundation for Young Scientists of China (No. 40801225)the Natural Science Foundation of Zhejiang Province (No. Y3090038)
文摘We used generalized additive models (GAM) to analyze the relationship between spatiotemporal factors and catch, and to estimate the monthly marine fishery yield of single otter trawls in Putuo district of Zhoushan, China. We used logbooks from five commercial fishing boats and data in government's monthly statistical reports. We developed two GAM models: one included temporal variables (month and hauling time) and spatial variables (longitude and latitude), and another included just two variables, month and the number of fishing boats. Our results suggest that temporal factors explained more of the variability in catch than spatial factors. Furthermore, month explained the majority of variation in catch. Change in spatial distribution of fleet had a temporal component as the boats fished within a relatively small area within the same month, but the area varied among months. The number of boats fishing in each month also explained a large proportion of the variation in catch. Engine power had no effect on catch. The pseudo-coefficients (PCf) of the two GAMs were 0.13 and 0.29 respectively, indicating the both had good fits. The model yielded estimates that were very similar to those in the governmental reports between January to September, with relative estimate errors (REE) of <18%. However, the yields in October and November were significantly underestimated, with REEs of 36% and 27%, respectively.
