NCCOS Assessment: Groundfish biodiversity hotspots off the Pacific Coast of Oregon from 1971-09-05 to 2010-09-20 (NCEI Accession 0156467)

This data set comprises maps of predicted long-term groundfish biodiversity hotspot probabilities off the Pacific Coast of Oregon. Predicted hotspot probabilities are given for four biodiversity indices: 1) relative abundance, 2) relative biomass, and 3) species number for all groundfishes, and 4) relative abundance for only nearshore groundfishes. The nearshore species assemblage includes: Sand Sole, English Sole, Pacific Sanddab, Speckled Sanddab, Petrale Sole, Starry Flounder, and Butter Sole. Hotspots were defined as areas with predictions in the top 10% of values of at-sea observations.

Areas of relatively high biodiversity were predicted using at-sea groundfish observations collected from 1971 to 2010 and associative models linking species observations with environmental covariates. Groundfish observations were taken from existing fishery-independent trawl data collected by NOAA’s National Marine Fisheries Service and flatfish trawl data collected by the Oregon Department of Fish and Wildlife. Environmental predictors included position on the shelf, sea floor habitat, depth, slope, aspect, slope of slope, and oceanographic productivity were used to predict areas with relatively high groundfish biodiversity. Binary logistic regression trees were used to associate species observations with environmental covariates and predict categorical results (Hotspot/Low classes). Mapped values indicate the probability of a raster cell belonging in the hotspot class.

本数据集包含美国俄勒冈州太平洋沿岸海域的长期底栖鱼类（groundfish）生物多样性热点区域预测概率分布图。该数据集针对四项生物多样性指数提供了热点区域预测概率：1）所有底栖鱼类的相对丰度、2）相对生物量、3）全部底栖鱼类的物种数量，以及4）仅近岸底栖鱼类的相对丰度。近岸底栖鱼类群落包括：沙鳎（Sand Sole）、英国鳎（English Sole）、太平洋沙鲽（Pacific Sanddab）、斑点沙鲽（Speckled Sanddab）、佩特拉勒鳎（Petrale Sole）、星鲽（Starry Flounder）以及黄油鳎（Butter Sole）。热点区域被定义为海上观测值排名前10%的对应预测区域。
本研究依托1971年至2010年采集的海上底栖鱼类观测数据，以及关联物种观测与环境协变量的预测模型，完成了生物多样性相对较高海域的预测。底栖鱼类观测数据来源于两类公开数据集：一是美国国家海洋和大气管理局（National Oceanic and Atmospheric Administration, NOAA）下属国家海洋渔业服务局采集的现有渔业独立拖网调查数据，二是俄勒冈州鱼类与野生动物局收集的比目鱼拖网调查数据。本研究选用的环境预测因子包括陆架位置、海底栖息地类型、水深、坡度、坡向、坡度曲率以及海洋学生产力，以此预测底栖鱼类生物多样性相对较高的海域。研究采用二元逻辑回归树模型，将物种观测数据与环境协变量进行关联，以预测栅格单元的分类类别（热点/非热点类别）。最终绘制的地图栅格值代表单个栅格单元被归类为热点区域的概率。