The Analysis On Trajectory Simulation Of Single Otter Trawl Vessels And Its Production Estimates As Well As Spatio-temporal Characteristics In Zhoushan Fishing Ground In Autumn And Winter

Researches on fisheries resources need a detailed understanding of fisheries about features and changes in the production and distribution, including statistical data such as spatial and temporal characteristics, fishing operations as well as catch and fishing effort. Currently, in China the main source of fishery catch data is fishing logbook. However, it still has some disadvantages:on the one hand, the logbook data recorded by fishermen is not comprehensive and accurate enough, generally it only records production of high-yield area; on the other hand, the logbook data is static, it often lacks the information on specific latitude and longitude where the vessel was fishing as well as its speed and course. Therefore, the scarcity of the fundamental data and the limitation of data sources in fisheries research in our country have affected the rapid development of the research on fisheries resources.In recent years, with the decline of the world’s major traditional fisheries resources, VMS (Vessel Monitoring System) has been applied in some developed countries for conservation and management of fisheries resources. The fishery information in China is relatively limited, if the VMS data can be combined with the logbook data, it cannot only give a new way to get the essential data but also help the development of fisheries stock assessment, therefore, it is a beneficial attempt in the research of fisheries resources.The research in this paper were based on the logbook data of6randomly selected single otter trawl vessels fishing in the Zhoushan fishing ground in2010and the logbook data and corresponding VMS data of above6vessels after the closed season from September2012to January2013. The logbook data included date, latitude, longitude, trawling time and speed. While the VMS data were including the identifying number, date, position, heading and speed. With these data, we analyzed the fishing characteristics of the sampled single otter trawl vessels. Generalized Additive Model (GAM) was built and it was used to estimate their catches from September2012to January2013.The spatio-temporal distribution of their yields in Zhoushan fishing ground was researched.We used cubic Hermite spline (cHs) interpolation method and linear interpolation method to simulate the real trawl trajectory under different time intervals. By comparing the result of both methods, we found that in a longer time interval, vessels are not likely to be travelling in a straight line and the straight line interpolation may give incorrect estimates of track simulation. The cHs method can make full use of additional information such as heading and speed, which can indicate the trend of vessel’s track. Hence, the cHs interpolation method is more useful than the linear method in estimating trajectories using data recorded by VMS. The results showed that the error of the simulated trajectory increased with the increase of time intervals of VMS data. In order to ensure the accuracy of the simulated trajectory, the time interval of VMS data should shorter than30minutes. We define the vessel’s behavior based on the combination of speed with the simulating trajectory; the result showed that when the vessel was fishing, its speed was between2.5Kn and5.0Kn.By using above combination method, we can determine the vessels’behavior with less error when just depend on speed.GAM was built to quantitatively analyze the relationships between single otter trawl catch and the spatio-temporal factors and environmental factors. Our results of step GAM analysis suggested thatthe factors which affect catch can be ordered following the importance as date, SSS (sea surface salinity), latitude, SST (sea surface temperature),SSH (sea surface height) and longitude. The AIC indicated the GAM with above6variables was optimal with best significance of the variances. The result showed that GAM can reflect the change of catch, and the results will be improved with the increases of the vessels and the VMS data.The analyses of the spatio-temporal distribution of catch per unit effort (CPUE) with sampled single otter trawl vessels showed that after the closed season, the CPUE of sampled vessels increased as time goes on from September to November2012. While in December2012and January2013, CPUE decline with time. The sampled vessel mainly distributed in the sea area near29°30’N～31°30’N,123°00’E～125°30’E. Additionally, the movement of the CPUE firstly had a tendency from south to north and then moved in the opposite direction on the latitude, but less changed on the longitude.