海州湾鹰爪虾栖息地适宜性研究

根据2011年及2013?2017年春季和秋季在海州湾进行的底拖网调查数据，结合同步测定的底层水温、底层盐度、水深和资源量等数据，开展鹰爪虾(Trachypenaeus curvirostris)栖息地适宜性的研究，先利用广义加性模型对环境因子进行筛选，再应用提升回归树模型确定各环境因子的权重，然后分别采用算术平均法和几何平均法建立栖息地适宜性指数模型，并通过交叉验证选择最优模型。结果表明:春季鹰爪虾的栖息地适宜性指数模型采用算术平均法构建，选择水深和底层盐度作为变量，具有最小的拟合；秋季鹰爪虾的栖息地适宜性指数模型采用几何平均法构建，选择底层水温和底层盐度作为变量，具有最小的拟合。对春季栖息地适宜性指数模型总偏差贡献率最大的是水深(76.23%)，其次是底层盐度(23.77%)；对秋季栖息地适宜性指数模型总偏差贡献率最大的是底层水温(82.56%)，其次是底层盐度(17.44%)。海州湾春季鹰爪虾的最适栖息水深为24 m以内，底层盐度为29.7～31.8；秋季的最适栖息底层水温为18～24℃，底层盐度为29.2～31.5。本研究表明，环境因子的优化有助于改进栖息地适宜性指数模型，并提升其预测能力。

Study on the habitat suitability of Trachypenaeus curvirostris in the Haizhou Bay

Based on the bottom-trawl survey data collected from the Haizhou Bay in spring and autumn of 2011 and 2013?2017, the habitat suitability of Trachypenaeus curvirostris was analyzed based on environmental factors such as bottom temperature, bottom salinity and water depth. Generalized additive model (GAM) was used to determine the optimal combination of environmental factors. Boosted regression tree (BRT) was used to evaluate the weight of each environmental factor in the habitat suitability index (HSI) model. The arithmetic mean model (AMM) and geometic mean model (GMM) were used to build HSI model, and the best model was selected by cross validations. Results showed that HSI model built with depth and bottom salinity in spring had the minimum AIC value, while HSI model constructed with bottom temperature and bottom salinity in autumn had the minimum AIC value. BRT model showed that the weight of depth and bottom salinity were 76.23% and 23.77% in spring, and the weight of bottom temperature and bottom salinity were 82.56% and 17.44% in autumn. The optimal range of depth and bottom salinity for T. curvirostris in spring were within 24 m and 29.7?31.8, respectively. In autumn, the optimal range of bottom temperature and bottom salinity were 18?24℃ and 29.2?31.5, respectively. This study suggested that the optimization of environmental factors was proved to be able to improve the performance of HSI models.