Fig 1.
European eel (A. anguilla) larvae were reared in a common larval rearing tank during the pre-feeding (endogenous feeding) period (from 0 to 9 days post hatch (DPH)) (A) and in 8L acrylic Kreisel tanks (n = 3 for each diet) throughout and beyond the first-feeding window (B and C). Water from the RAS was sampled before (on 9 DPH) after (on 15 DPH) the initiation of the feeding for bacterial community composition analysis and larvae were sampled at different ages for bacterial community composition analysis and molecular analysis of immune and stress related genes, while larval survival was calculated through enumeration of dead larvae daily (D).
Table 1.
Composition of the three diets used during the experiment (see also [18]).
Table 2.
Oligos used for molecular analysis of immune- and stress-related gene expression.
Fig 2.
Effect of age for each diet (A-C) and effect of diet at each day-post-hatch on survival of European eel (A. anguilla) larvae. Values represent means (± SEM), while different lower-case letters represent significant differences (p < 0.05).
Fig 3.
Effects of diets and age (Days Post Hatch (DPH)) on richness and evenness of ASVs in European eel (A. anguilla) larvae (A) and inflowing water (B). Yellow and blue colours indicate the larval and water samples, respectively. In each box plot, the solid black line indicates mean alpha diversity measure, and the surrounding box indicates the third quantile.
Fig 4.
Relative abundances of the bacterial orders detected in European eel (A. anguilla) larvae as a function of feed and age, and in inflowing water from corresponding RAS units.
Each stacked bar represents the relative abundances of bacterial orders detected in each replicate sample (A) and the average of relative abundances in replicate samples (B). Unassigned is ASVs that could not be classified reliably at the order level. (D1 = Diet 1, D2 = Diet 2, D3 = Diet 3, PF = Pre-feeding).
Fig 5.
PCoA ordination plots based on Bray–Curtis (A) and Sørensen–Dice (B) dissimilarities for comparison of the bacterial communities of inflowing water and of European eel (A. anguilla) larvae feeding on different diets. Different colours indicate different samples. Bray-Curtis and Sørensen-Dice are both percent dissimilarities based on relative abundance and presence/absence, respectively.
Table 3.
PERMANOVA R2 and p-values based on Bray–Curtis and Sørensen–Dice indices for comparisons of bacterial communities between RAS water allocated for different diets, between European eel (A. anguilla) larvae and RAS water, and between larvae fed different diets at different developmental stages.
Significance level was set at <0.05 and significant p values were bolded.
Table 4.
PERMANOVA R2 and p-values based on Bray–Curtis and Sørensen–Dice indices for comparisons of bacterial communities in European eel (A. anguilla) larvae and RAS water allocated for different diets between different ages.
Significance level was set at <0.05 and significant p values were bolded.
Table 5.
ASVs of potentially harmful genera that were found by DESeq2 analysis to be significantly more abundant (p < 0.05) in larval bacterial communities on 22 dph compared to 15 dph in different experimental groups and nearest matches of the DNA sequences of those ASVs as found by the RDP SeqMatch tool.
Fig 6.
Expression of immune-related genes, il1b (A), tlr18 (B), c1qc (C) and stress-related gene, hsp90 (D) in European eel (A. anguilla) larvae fed three different diets. All y-axes display gene expression relative to the expression at hatch (0 dph). In each facet, the graphs, i, ii and iii show the effect of age on expression of the relevant gene for each diet, whereas the graph iv shows the effect of diets on gene expression at each age. Values represent means (± SEM), and different lower-case letters represent significant differences (P < 0.05).