Analyses for bulk delta-15N and delta-13C:

To ensure accuracy, international reference standards, USGS40 and USGS41a, were analysed at the beginning, middle and end of each run. An internal standard of ground prawn Penaeus vannamei with well characterized delta-13C and delta-15N values (-22.6 and 6.8 per mille, respectively) was analysed every 10 samples to monitor precision, which was <0.2 per mille for both delta-13C and delta-15N.

Analyses for delta-15N of amino acids (delta-15NAA)

Each sample was analyzed in duplicate and a triplicate measurement was made if the mean delta-15NAA values fell outside the expected measurement error (<1.0 per mille). Precision and accuracy were determined using a mixed amino acid standard prepared from 7 amino acids (alanine, valine, leucine, aspartic acid, glutamic acid, glycine and phenylalanine) with known delta-15N values (University of Indiana, USA and SI Science Japan). The mixed standard was analyzed every 4 injections. The mean precisions and accuracies were ± 0.9 per mille and ± 0.7 per mille (1 sigma, n = 29), respectively. Raw delta-15NAA sample values were corrected following the methods of McCarthy et al. (2013). This method takes into consideration the response of individual amino acids to the stationary phase of the column and is based on the offset between the measured delta-15NAA values in the nearest mixed standard and their known delta-15NAA values (Eq. 1).

Eq. 1: delta-15N-Sample-reported= Avg delta-15N-Sample-measured (delta-15N-Standard-measured - delta-15N-known),

where Avg delta-15N-Sample-measured is the average delta-15N for an amino acid in a sample (n = 2), delta-15N-Standard-measured is the delta-15N for the AA in the nearest mixed standard and delta-15Nknown the known elemental analysed offline value for the same standard.