Performance Report for Cooperative Agreement No: na06oar4810163 for the Period from September 1, 2006 to August 31, 2012 University of Maryland Eastern Shore

Project Description: The primary goals with this project are to assess the utility and accuracy of the In Situ Ichthyoplankton Imaging System (ISIIS) as a method for quantifying ichthyoplankton and other meso-zooplankton. To this end, efforts will first go towards validating the image analysis software being developed by the co-PI as well as validate the system with respect to net collections.

Three sets of comparisons are planned to validate the software and utility of the ISIIS: 1) software counts of fish larvae of digital data vs. manually analysis of digital data; 2) comparison of digital fish counts vs. plankton samples counts; 3) accuracy of digital fish identification (taxonomic) vs. plankton sample identifications. Digital imagery was clear (see Fig. 4 below – high latitude and low latitude examples provided) and preliminary review shows easy distinction of different fish taxa, as well as many other planktonic taxa. Manual segmentation (i.e. extraction of regions of interest or ROIs) is proceeding rapidly – taking approximately 10 weeks to complete (we are ~ 50% right now). Shortly after we receive the BONGO data we should be able to complete our comparison analyses. While not many fish are being found, a predominance of flatfish larvae can be seen– likely Paralichthys dentatus, and a second more elongate (Merluccius spp.?) taxon. The digital analysis software has been undergoing significant reworking to improve the segmentation step. It is noteworthy that the water column sampled was very rich in phyto- and zooplankton and marine aggregates (e.g. snow). Further, the desire was to enable the segmentation process to adjust for particles that are in poor focus or appear broken in two (i.e. something precluding a complete outline). This required the development of additional algorithms and software programming. Efforts have also been made to improve the speed (efficiency) of the segmentation process. This effort is completed now and a reanalyses of the full digital record for fish and other functional plankton groups is being done. A Ph.D. graduate student (Adam Greer) has been hired to begin his dissertation work in association with this cruise. He is responsible for examining all of the digital imagery for manual selection of target organisms for direct comparison with image analysis results and the plankton tow work. Analyses will be forthcoming this fall and initial presentation of results is planned for the 2010 Ocean Sciences meeting in Portland Oregon in February.