Evaluation of Sampling Methods for Maturation Stage Determination in the European Eel Anguilla anguilla

Abstract

Monitoring data is important in ecological research, but differences between and within areas or species in data collection methods could introduce bias in the analyses. Standardizing data collection is particularly important when monitoring migratory species that have a distribution that crosses several national borders. The European Eel Anguilla anguilla is an extreme example of such a species since it constitutes one stock across the entire distribution area. One important variable collected for the European Eel is maturation stage. This data is needed to monitor silver eel escapement to assess population trends. To determine maturation, data on length, weight, diameter of the eyes, and pectoral fin length are used to calculate Pankhurst eye index and Durif's silver index. In this study, we investigated effects of precision and interobserver variability on data collection relevant for maturation stage determination according to Pankhurst and Durif's indices. We found that eye diameter differed in size between the left and right eyes; however, the mean difference (0.19 mm) is probably an artifact of the large sample size (n = 16,977) and can be regarded as being within the measurement precision. Meanwhile, there was no significant difference in pectoral fin length. These results suggest that either side of the eel could be used without losing precision. Visually determined maturation stage classifications differed from those calculated with Pankhurst and Durif's indices but could still provide useful information; hence, it is recommended to collect this variable. Measurements performed using computer software generated greater precision than using calipers, which increased interobserver variability. Since the difference was relatively small and since computer analysis of images may not always be an option, measuring method can be decided based on the level of precision needed in each case. These suggested implementations can reduce observation bias and streamline the data collection used for stock assessments of the European Eel.