First steps towards developing an open access software designed for the recognition of daily microstructures used in ageing fish

Ageing methods for most of the exploited fish species are not validated, and consequently age data series often have important gaps and uncertainty manifesting its consequences on age-based fisheries assessment. Amongst the most recognizable otolith microstructures are those corresponding to daily growth increments producing reliable estimates of larval and juvenile growth in an ample number of fish species. From the temporal perspective, otoliths are the only calcified structures that show growth marks at two different scales, the seasonal and daily markings. Seasonal rings or otolith macrostructure is mostly used to age juvenile and adult individuals while microstructure is typically used in larvae, because the DGI in adult otoliths are too numerous and cumbersome to count. Nonetheless, in several hake species DGI have been counted and measured in juveniles by interpretating daily increments through thin transversal sections of the otolith (50-100 μ). Since these sections can only be obtained after a process of otolith resin embedding, section precision cutting and final polishing, the main handicap lacking today is the appropriate digital tool for counting and measuring otolith DGI widths. In this communication, the research multidisciplinary team of researchers from the research fields of fisheries and engineering plans to develop a new tool designed to add efficiency in the weighty task of estimating the age of fishes based on DGI analysis. The technique is based on new computer algorithms that interpret DGI patterns, whether it be in a semi-automatic way or in such a manner that otolith readers are allowed to interact with the system or run automatically when the technique finds determined types of otolith structural patterns considered consistent and reliable. The technique, when registered, will be made available freely to the international scientific community in the open access mode. Although some image analysis software is available in the market nowadays, including some centered in otolith image processing, they all show at least two main restrictions: they do not manage ‘live’ images which allow to specifically focus determined microstructures at great magnification, and moreover, the high costs of the software. We present here the first results on the development of this forthcoming software which is expected to resolve the before mentioned limitations. Larval and adult otoliths measure from some hundred μm to several mm, respectively, while DGI widths fluctuates from slightly less than 1 μm increments close to the nucleus in larval otoliths to more than 15 μm in juveniles and adults. In order to discriminate DGI, light microscope magnification varies from x1000 in larvae to x200 in juveniles and adults. As a consequence, most of the times the whole otolith or section cannot be represented on a single image and thus, several images are needed to construct a sole image to get the vision of the total otolith size that can be measured where most DGI counts can be included. In these cases, the image series should be tiled as a panoramic view or mosaic, where each section superposes with adjacent ones. The first step in the software development has been a superposing routine, identifying automatically common areas between adjacent images. Once a single high quality image is set by otolith at an acceptable magnification some arrangements on contrast, brightness, etc. can be implemented when desired. The next stage is the controlled or semi-automatic counting of DGI along a growth trajectory. Expert readers particular interpretations could be matched to those made au- tomatically and eventually adjust the software reading standards. Finally, DGI counts and measurement data are recorded and also exported to work files, where further analysis could be performed.As soon as having ready an appropriate and affordable tool to count and measure daily growth incre- ments in an automatic or semi-automatic way, otolith microstructure analysis can have a reliable age estimation method that can be extended and applicable to age adult fish. Although, in its present form the technique is focused in otolith microstructure analysis, further developments of the initiated software development intends to extend its application to other calcified structures, namely bivalve shells and cephalopod peaks.Universidad de Málaga. Campus de Excelencia Internacional Andalucía Tech