Abstract. This paper presents two metazoan zooplankton datasets obtained by imaging samples collected on the Bay of Biscay continental shelf in spring during the PELGAS integrated surveys, over the 2004–2019 period. The samples were collected at night, with a WP2 200 µm mesh size fitted with a Hydrobios (back-run stop) mechanical flowmeter, hauled vertically from the sea floor to the surface with a maximum depth set at 100 m when the bathymetry is deeper. The first dataset originates from samples collected from 2004 to 2016, imaged on land with the ZooScan and is composed of 1,153,507 imaged and measured objects. The second dataset originates from samples collected from 2016 to 2019, imaged on board the R/V Thalassa with the ZooCAM and is composed of 702,111 imaged and measured objects. The imaged objects is composed of zooplankton individuals, zooplankton pieces, non-living particles and imaging artefacts, ranging from 300 µm to 3.39 mm Equivalent Spherical Diameter, individually imaged, measured and identified. Each imaged object is geolocated, associated to a station, a survey, a year and other metadata. Each object is described by a set of morphological and grey level based features (8 bits encoding, 0 = black, 255 = white), including size, automatically extracted on each individual image. Each object was taxonomically identified using the web based application Ecotaxa with built-in, random forest and CNN based, semi-automatic sorting tools followed by expert validation or correction. The objects were sorted in 172 taxonomic and morphological groups. Each dataset features a table combining metadata and data, at the individual object granularity, from which one can easily derive quantitative population and communities descriptors such as abundances, mean sizes, biovolumes, biomasses, and size structure. Each object’s individual image is provided along with the data. These two datasets can be used combined together for ecological studies as the two instruments are interoperable, or as training sets for ZooScan and ZooCAM users.

中文翻译：

---

比斯开湾的后生动物浮游动物：16 年来 ZooScan 和 ZooCAM 成像系统的个体大小和丰度

摘要。本文介绍了 2004-2019 年期间 PELGAS 综合调查期间春季在比斯开湾大陆架上收集的成像样本获得的两个后生动物浮游动物数据集。样本是在夜间收集的，采用 WP2 200 µm 网眼尺寸，配备 Hydrobios（逆止器）机械流量计，当水深测量更深时，从海底垂直拖到表面，最大深度设置为 100 m。第一个数据集源自 2004 年至 2016 年收集的样本，使用 ZooScan 在陆地上成像，由 1,153,507 个成像和测量对象组成。第二个数据集来自 2016 年至 2019 年收集的样本，在 R/V Thalassa船上成像使用 ZooCAM，由 702,111 个成像和测量对象组成。成像对象由浮游动物个体、浮游动物碎片、非生命颗粒和成像人工制品组成，范围从 300 µm 到 3.39 mm 等效球直径，单独成像、测量和识别。每个成像对象都经过地理定位，与站点、调查、年份和其他元数据相关联。每个对象都由一组基于形态和灰度级的特征（8 位编码，0 = 黑色，255 = 白色）描述，包括大小，在每个单独的图像上自动提取。每个对象都使用基于网络的应用程序 Ecotaxa 进行分类识别，该应用程序具有内置的随机森林和基于 CNN 的半自动分类工具，然后进行专家验证或校正。这些对象被分类为 172 个分类学和形态学组。每个数据集都有一个以单个对象粒度结合元数据和数据的表格，从中可以轻松导出定量的种群和群落描述符，例如丰度、平均大小、生物量、生物量和大小结构。每个对象的单独图像与数据一起提供。这两个数据集可以结合起来用于生态学研究，因为这两个仪器可以互操作，或者作为 ZooScan 和 ZooCAM 用户的训练集。