A new model to forecast fishing ground of Scomber japonicus in the Yellow Sea and East China Sea

The pelagic species is closely related to the marine environmental factors, and establishment of forecasting model of fishing ground with high accuracy is an important content for pelagic fishery. The chub mackerel (Scomber japonicus) in the Yellow Sea and East China Sea is an important fishing target for Chinese lighting purse seine fishery. Based on the fishery data from China’s mainland large-type lighting purse seine fishery for chub mackerel during the period of 2003 to 2010 and the environmental data including sea surface temperature (SST), gradient of the sea surface temperature (GSST), sea surface height (SSH) and geostrophic velocity (GV), we attempt to establish one new forecasting model of fishing ground based on boosted regression trees. In this study, the fishing areas with fishing effort is considered as one fishing ground, and the areas with no fishing ground are randomly selected from a background field, in which the fishing areas have no records in the logbooks. The performance of the forecasting model of fishing ground is evaluated with the testing data from the actual fishing data in 2011. The results show that the forecasting model of fishing ground has a high prediction performance, and the area under receiver operating curve (AUC) attains 0.897. The predicted fishing grounds are coincided with the actual fishing locations in 2011, and the movement route is also the same as the shift of fishing vessels, which indicates that this forecasting model based on the boosted regression trees can be used to effectively forecast the fishing ground of chub mackerel in the Yellow Sea and East China Sea.

[1]  C. Graham,et al.  Selecting pseudo-absence data for presence-only distribution modeling: How far should you stray from what you know? , 2009 .

[2]  Jennifer A. Miller,et al.  Mapping Species Distributions: Spatial Inference and Prediction , 2010 .

[3]  Cui Xue,et al.  Fishing Ground Forecasting of Chilean Jack Mackerel(Trachurus murphyi) in the Southeast Pacific Ocean Based on CART Decision Tree , 2012 .

[4]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[5]  L. Gang.  Relationship between the resource and fishing ground of mackerel and environmental factors based on GAM and GLM models in the East China Sea and Yellow Sea , 2008 .

[6]  Xinjun Chen,et al.  Review of fishery forecasting technology and its models , 2013 .

[7]  Cheng Jia-hua,et al.  Study on the biological characteristics and status of common mackerel (Scomber japonicus Houttuyn) fishery in the East China Sea region , 2004 .

[8]  Gang Li,et al.  Habitat suitability index of Chub mackerel (Scomber japonicus) from July to September in the East China Sea , 2009 .

[9]  Greg Ridgeway,et al.  Generalized Boosted Models: A guide to the gbm package , 2006 .

[10]  H. Dewar,et al.  Prediction of fishing effort distributions using boosted regression trees. , 2014, Ecological applications : a publication of the Ecological Society of America.

[11]  J. Leathwick,et al.  Ontogenetic habitat associations of a demersal fish species, Pagrus auratus, identified using boosted regression trees , 2012 .

[12]  Leo Breiman,et al.  Classification and Regression Trees , 1984 .

[13]  J Elith,et al.  A working guide to boosted regression trees. , 2008, The Journal of animal ecology.

[14]  J A Swets,et al.  Measuring the accuracy of diagnostic systems. , 1988, Science.

[15]  F. Jiguet,et al.  Selecting pseudo‐absences for species distribution models: how, where and how many? , 2012 .

[16]  Xinjun Chen,et al.  Distribution of hotspots of chub mackerel based on remote-sensing data in coastal waters of China , 2014 .

[17]  Mark S. Boyce,et al.  Modelling distribution and abundance with presence‐only data , 2006 .

[18]  Wolf-Christian Lewin,et al.  The influence of anthropogenic shoreline changes on the littoral abundance of fish species in German lowland lakes varying in depth as determined by boosted regression trees , 2013, Hydrobiologia.

[19]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[20]  H. Andrade,et al.  The relationship between the skipjack tuna (Katsuwonus pelamis) fishery and seasonal temperature variability in the south-western Atlantic , 2003 .

[21]  Robert P. Anderson,et al.  Maximum entropy modeling of species geographic distributions , 2006 .

[22]  Dan Steinberg,et al.  Stochastic Gradient Boosting: An Introduction to TreeNet™ , 2002, AusDM.

[23]  G. Moisen,et al.  PresenceAbsence: An R Package for Presence Absence Analysis , 2008 .

[24]  G. Stunz,et al.  Spatiotemporal Predictive Models for Juvenile Southern Flounder in Texas Estuaries , 2013 .