First annulus formation in the European anchovy; a two-stage approach for robust validation

Abstract

The age determination in fast-growing short-living species, such as European anchovy (Engraulis encrasicolus), has been widely recognized as a difficult task and bias introduced by readers leads to bias in reconstructing the population age structure. In this context, it is worth to note that age structure of fish population represents key information in fishery ecology and for stock assessment models. The uncertainty in estimating the age of the European anchovy by otolith reading is linked to the number of false-growth increments (checks) laid down before the annulus formation. While direct validation methods (e.g. mark-recapture, rearing, radiochemical dating) are difficult to implement specially for this short living species, the use of different indirect methods, supported by a coherent statistical approach, represents a robust and easier validation tool. A statistical modeling approach has been here adopted to assess the coherence of two well-known methods, namely Edge Analysis and Marginal Increment Analysis, in order to validate the first annulus formation in European anchovy. Both methodologies in two different yearly cycles converged toward the same result, thus confirming the annulus identification for the first year class. In addition, the completion dates of the checks and the first annulus were computed in order to gain a better insight into otolith growth dynamic. According to the species spawning period, the completion date of the first annulus falls in the summer period, while the first and second checks completion dates were mostly found in summer and winter respectively. General additive models using marginal increments as dependent variable showed a significant effect of the month, highlighting the presence of only one clear minimum in July/August, as well as specific relationships with condition factor and gonadosomatic index. Modeling the otolith edge morphology, the probability to find a hyaline band displayed in both years a similar shape, characterized by a minimum in July/August and higher values between November and January. The obtained results evidenced temporally coherent patterns providing a better insight in the otolith growth dynamic as well as a more robust validation of the first annulus formation in the European anchovy.

Figure 1

Schematic representation of the European anchovy whole otolith sagitta: TR = otolith total radius, and annulus 1 is the combination of one opaque plus one true hyaline zone (The authors acknowledge Barbara Bottini for drawing this image).

Figure 2

Morphology of the otolith edge for each type: (a) opaque margin (O); (b) hyaline margin (H); (c) semihyaline margin (OH).

Figure 3

Age structure of European anchovy for each analyzed year from the GSA 16 stock: (a) proportion for age class; (b) proportion of age 1 fishes for length class in 2015 (n° 313) and 2016 (n° 757).

Figure 4

Boxplot of temporal trend (monthly) for fishes of age class 1: (a) gonadosomatic index (GSI); (b) condition factor (CF); (c) and otolith marginal increments (MI).

Figure 5

Monthly evolution (%) of edge morphology in 2015 (left panel) and 2016 (right panel) for fishes of age class 1: opaque plus semihyaline edges (O + OH, merged according literature, solid line); hyaline edge (H) (dotted line); opaque edge (O) (dashed line).

Figure 6

Plots of the fit (GAM) between the probability to find a hyaline margin and month. Shaded (light-grey) regions represent the confidence bands (±2 SE) for smooths. Model-predicted presence of translucent otolith edges in 2015 (left panel) and 2016 (right panel).

Figure 7

Completion date frequency distribution (TCA) of the analyzed translucent zones showing two cohorts per sampling year respectively in 2015 top panel, and 2016 bottom panel. The three different translucent zones were plotted respectively T1 in panel (a); T2 in panel (b); and T3 in panel (c).

Figure 8

Plots of fitted GAMs in 2015 (left panels) and 2016 (right panels) of the marginal increment ratio (MIR) considered as dependent variable and the three considered factors: gonadosomatic index (GSI), condition factor (CF), and month. Shaded (light-grey) regions represent the confidence bands (±2 SE) for smooths.