Workshop on Mackerel, Horse Mackerel and Hake Eggs Identification and Staging 2 (WKMACHIS2; outputs from 2024 meeting)

The WKMACHIS2 workshop was held for plankton analysts involved in the 2025 ICES-coordinated mackerel and horse mackerel egg survey (MEGS), with the aim of standardizing and calibrating the identification and staging of eggs from the target species: mackerel (Scomber scombrus), horse mackerel (Trachurus trachurus), hake (Merluccius merluccius), and ling (Molva molva).

To calculate the number of spawning female fish in a stock using the Annual Egg Production Method (AEPM; Lockwood et al., 1981, Armstrong et al., 2001) or Daily Egg Production Method (DEPM; Lasker, 1985) correct identification of freshly spawned target species eggs (eggs in development stages 1a and 1b) is essential. It is therefore vital that MEGS analysts are able to correctly distinguish these eggs from those in later development stages, or eggs of other species at the same stage.

Participants analysed the same 513 eggs across two SmartDots events (one pre-workshop) and one physical event. Results were collated, processed and quality checked, and entered into standard Excel evaluation sheets. These sheets were utilised to identify species, stages, and individual eggs with low agreement scores. These eggs were presented on screen to demonstrate key diagnostic indicators and identify causes of disagreement and misidentification. In this way, the egg staging criteria (ICES, 2019) were reviewed based on shared expertise and group contributions.

These demonstrations resulted in increased agreement in staging overall and within individual species between SmartDots events 1 & 2, including between results filtered to only include expert readers. A decrease in agreement in stages 4 or 5 was observed in some species, however due to the very low numbers of these eggs in each event, disagreement on even one egg markedly distorted the percentage agreement.

In event 2 participants correctly staged 98% of all stage 1 (1a and 1b combined) eggs, the critical stage for DEPM and AEPM calculations for mackerel and the western component of horse mackerel. However, agreement with the ‘correct’ stage was notably lower when stage 1 was separated into stages 1a (84%) and 1b (65%). Agreement with the ‘correct’ stage was also poorer for the later developmental stages 2,3,4, and 5 (76%, 79%, 82%, and 73% respectively)

Overall agreement in event 2 was 75.7% and 83.2% for mackerel and horse mackerel respectively across all stages, and 99% and 96% for stage 1. Expert readers’ agreement was 77.6% and 84.3% across all stages, and 99% and 96% for stage 1.

WKMACHIS2 was the first workshop to include both SmartDots and physical egg sample exercises. Analysts previously expressed concerns that SmartDots exercises would produce poorer agreement in stage and species allocation due to the inability to manipulate eggs and to vary lighting or magnification. However, agreement in all species except hake was higher in SmartDots exercises. Less experienced readers found the precise measurements of egg dimensions in SmartDots increased their confidence. However, some species have identifying features that require physical manipulation of the egg to discern. As such, it was agreed that both image analysis and practical trials have a vital role in egg identification workshops.