The influence of dietary fat concentration on fatty acid signatures and stable carbon isotopes of fatty acids in liver of Atlantic pollock (Pollachius virens)

Abstract

A number of biomarker techniques have been employed to investigate animal diets, including analyses of fatty acid (FA) proportions and stable carbon isotopic signatures. Here, the influence of dietary FA concentrations on fish FA signatures and δ13C values of FA in fish liver was evaluated by feeding Atlantic pollock (Pollachius virens; n=108) three diets with different FA concentrations for 20 weeks. Calibration coefficients (CC) and discrimination factors were calculated to compare the FA profile and δ13C values of FA in liver to those of diet. Most CC showed significant variation across the three diets; however, a number of FA present in smaller proportions had CC that did not vary among the diets. Overall, the two CC sets derived from experiments utilizing feeds with the higher FA concentrations were very similar to each other and showed substantial variation compared to CC derived from feeding the diet with the lowest FA concentrations. Using Quantitative Fatty Acid Signature Analysis (QFASA) and the initial samples of pollock, the three sets of CC derived from fish fed the three treatment diets were applied to determine the influence of these three CC on diet estimates. The results indicated that QFASA may be less sensitive to variation in CC than previously assumed. Discrimination factors calculated using liver and diet δ13C values varied with dietary FA concentration for five of ten FA. Most discrimination factors for saturated and monounsaturated FA (MUFA) showed significant deviations from zero; in contrast, few discrimination factors for polyunsaturated FA (PUFA), specifically 20:5n-3 and 22:6n-3, were significantly different than zero. These results suggest that δ13C values for PUFA sampled from a wild fish would closely mirror that of the prey consumed. Combined biomarker techniques, employing both FA proportions and δ13C values, may prove to be the most reliable in estimating diets of animals in the wild at this time.