Spatial and Temporal Patterns of Zooplankton Species in the Gulf of Alaska as Revealed by Image Analysis

Kate Mayer, Research Experience for Undergraduates, University of Alaska Fairbanks, Fairbanks, AK, United States, Cheryl Clarke-Hopcroft, University of Alaska Fairbanks, Fairbanks, AK, United States and Russell R Hopcroft, University of Alaska Fairbanks, College of Fisheries and Ocean Sciences, Fairbanks, AK, United States
The time (and money) required to process samples remains a logistical challenge for zooplankton research. With the addition of several new transect lines to the new Northern Gulf of Alaska Long-Term Ecological Research program we assessed the ability of the ZooScan image analysis system to yield data of sufficient fidelity to reveal and extract major spatial and temporal patterns in this ecosystem. During 2018, three research cruises traversed four sampling lines, collecting zooplankton within the upper 200m of the water column using 505 µm Bongo or Multinets. At least 500 animals were scanned per sample, rough sorted by the ZooScan software, and then manually validated. Cluster and non-parametric multidimensional scaling analysis using Bray-Curtis similarity from all three cruises showed a clear distinction between communities found both between seasons and across habitats. The spring cruise was the most distinctive, having separate and dissimilar dominant species than those found on the summer and fall cruises. The distinctness of the spring cruise arose from the dominance of the copepod genera Neocalanus that was uncommon during the summer or fall cruises. On all three cruises, cross-shelf gradients emerged. It appears that the ZooScan is capable of capturing major community gradients in the NGA-LTER and will have a huge impact on making the analysis methodology more tangible and expedited for large-scale and long-term sampling.