On the relationship of stock recruitment in Pacific Sardina 

population based on environmental factors

Abstract: Sardina is an economically and ecologically important pelagic species. They are warm water fishes, and only a few of them have lower optimum temperature. They feed mainly on plankton, and their optimum temperature of habitat is about 2030℃. In the Pacific, Sardina with age more than 1 live around the water of Kuroshio current for overwintering and spawning. In summer and autumn, they stay along the coast or migrate to the north with limited distance for growth. For presence of eggs, Sardina spawn from November to June of the following year. The peak of spawning occurs from February to April, and the best temperature for spawning activity of sardines is between 15℃ and 18℃. Previous research suggested that the recruitment of Sardina was very vulnerable to the marine environmental factors of habitat. However, many reasons including human activities, climate change and its induced local environmental variability have intensified the fluctuations in recruitment of sardines in the Pacific Ocean. Understanding and estimation of how those factors affect the spawningrecruitment (SR) relationship of Sardina is beneficial to sustainable utilization and management of the resource. Unfortunately, estimating parameters in SR relationship is one of the tough tasks in fisheries stock assessment. In addition, the SR relationship is greatly blurred by the abovementioned factors, and critical environmental factors are usually incorporated to explore the relationship between the two. Commonly, Ricker model with considering key environmental factors for spawning is a useful tool to derive the SR relationship. For Sardina in the Pacific Ocean, two validated factors, sea surface temperature (SST) on the spawning ground and Pacific decadal oscillation (PDO), are important for spawning and hatch, thus, we assume that SST and PDO can be the potential factors affecting the SR relationships of Sardina. We combine environmental data including SST on Sardina spawning habitat between 142°165° E and 35°40°N (from Columbia University Climate Center, http://iridl.Ldeo.columbia.edu/SOURCES/.IGOSS/.data_products.Html), PDO data ( form Japan Meteorological Agency, http: //www.jma.go.jp/jma/index. Html) from February to April (spawning season for sardine in the Pacific Ocean) and survey data obtained from the 2016 Pacific Fauna Resource Assessment Report to construct multiple models of the SR relationship of Sardina during 1996—2015. Specifically, we assume that SST and PDO  can affect the SR relationship separately and simultaneously, three alternative models (SST_Ricker, PDO_Ricker, and SST_PDO_Ricker) could be established for Sardina during 1996—2015. According to the survey data, the obvious peak and valley happened in spawning biomass and recruitment of Sardina. After 1996, the spawning biomass was decreasing to historical low level with fluctuation, however, the opposite situation of increasing in biomass reached 1.03 million tons in 2014 as the continuous good recruitment biomass of Sardina and the decreased fishing intensity. In addition, the extremely high recruitment was estimated to be 1.86 million tons in 2015.The performance and robustness of models were evaluated by the Akaike information criterion (AIC) values and leaveoneout cross validation, and the model with the lowest AIC value was the optimal model. Results showed that three alternative models were all better than the original Ricker model (without considering the environmental factor), the SST_PDO_Ricker model with the fewer outliers and the lowest AIC value of 49.99 was the final model to describe the SR relationships for the Pacific Sardina. It meant that the synchronous effect of SST and PDO could affect the mechanism of recruitment for Sardina. By examining the environmental coefficient of the optimal model, it was found that SST and PDO in the spawning period had a positive effect on the recruitment quantity of the Pacific Sardina. The SST and PDO on the spawning grounds could significantly reflect the dynamics of recruitment for the Pacific Sardina, and the SST had more influences on the Pacific Sardina compared with PDO. It was more conducive to the occurrence of recruitment when SST was greater of its mean value and rose steadily. Additionally, high PDO value could make a larger increase in recruitment of 2015. Commonly, PDO can be divided into cold and warm phases. When PDO has a warm phase, the surface water of the north American continent is abnormally warm, while the surface temperature of the North Pacific Ocean is abnormally cold, it can lead to an increase in Sardina recruitment. When PDO has a cold phase, the opposite is true. Compared with PDO, SST may have more ability to affect the recruitment of Sardina. When temperatures are relatively high and above average, there is a significant increase in recruitment. The rising of SST occurring in warm phase would produce more suitable habitat for Sardina in the North Pacific Ocean. This is validated by the observation that the number of 0yearold Sardina increassd significantly during warmer years, and the large and mediumsized Sardina are caught at lower surface salinity and temperatures than that of small Sardina in the Pacific Ocean. Each life stage of Sardina (as one of the poikilotherm fishes), such as spawning, hatching, growth is significantly correlated with water temperature positively or negatively. Understanding the effects of climate change induced environmental variability on fish recruitment on interannual and decadal time scales is fundamental to understand fish population dynamics. Our study suggests that the SR relationship incorporating SST and PDO should be considered for Sardina management. Further, other factors such as food sources and the relationship between other species need to be explored for better assessment and utilization of Sardina resource in the Pacific Ocean.