Carbon (C) was also quantified in the N flux experiment using an elemental analyser to provide a percentage of carbon as dry weight (OEA laboratory Ltd.). The internal N and C (see Table S2) content is reported as grams per 100 g dry weight (% dw). To quantify changes in amino acid profiles with varying internal N content, all cultures were analysed for amino acids. All cultures in both experiments were analysed for aspartic acid, asparagine, glutamic acid, glutamine, serine, histidine, glycine, threonine,

alanine, arginine, tyrosine, valine, methionine, phenylalanine, isoleucine, leucine, lysine, and proline (Tables S1 and S2). As asparagine is hydrolysed to aspartic acid and glutamine to glutamic CHIR-99021 purchase acid during analysis, the sum of these amino acids were reported as asparagine/aspartic acid or glutamic acid/glutamine. For the stocking density experiment cysteine and taurine were also analysed, but not thereafter as they were minor constituents (cysteine <0.36% and taurine <0.04% of total Z-VAD-FMK price amino acids, see Table S1). Amino acids were analysed after 24 h liquid hydrolysis in 6 M HCl at 110°C using a precolumn derivitiz6ed HPLC at the ChemCentre (stocking density experiment) and a Waters ACQUITY UPLC at the Australian Proteome Analysis Facility, Macquarie University, Sydney (N flux experiment) using procedures based on the Waters AccQTag amino

acid methodology (Cohen 2001, Bosch et al. 2006). Internal N content (% dw) and SGR (% d−1) were plotted against N flux for both experiments. Curves of best fit were applied for both relationships using SigmaPlot 10.0 (Systat Software Inc., San Jose, CA, USA), r2 values reported. ANCOVAs were used to test the effect of stocking density on internal N content and SGR (two separate analyses), using data from the linear portion of curves (Systat10; Systat Software Inc). Amino acid quality of biomass in both experiments was analysed using nonmetric multidimensional scaling (MDS) using the statistical software PRIMER (PRIMER-E Ltd., Lutton, UK). A similarity matrix was calculated from the 4th root transformed MCE with individual amino acids contents (as a percentage of total amino acid content), N% and SGR

as variables in the MDS cluster diagram and vector plot. For the N flux experiment; total amino acid, methionine, lysine, and glutamine/glutamic acid contents (g · 100 g−1 dw) were plotted against internal N content for each water N concentration treatment. Correlations were made for internal nitrogen content versus total amino acids, methionine, lysine, and glutamic acid/glutamine contents (SigmaPlot 10.0, r values reported). A linear correlation was also made for SGR versus glutamic acid/glutamine contents. Internal N content increased rapidly for both stocking densities from 0.86% at the lowest water nitrogen flux (2 μM · h−1) to a maximum of 2.4% for 1 g · L−1 stocking density at 95.9 μM N · h−1 and 2.6% for 4 g · L−1 at 85.2 μM · h−1 (Fig. 2A).