Determination of protein for studies of marine herbivory: a comparison of methods

In this study we compared techniques for the estimation of protein for stud ies of marine herbivory. As an example we use gut fluid from the temperate marine herbivorous fish Kyphosus sydneyanus. There were two components to t his work: (1) to compare the ability of spectrophotometric methods to estim ate protein in gut fluid; and (2) to determine the effect of freezing diges ts samples on the extraction of gut fluid for the measurement of total hydr olysable amino acids (THAA). The Bradford, Lowry, deoxycholate/trichloroace tic acid (DOC/TCA) Lowry and the bicinchoninic acid (BCA) spectrophotometri c protein assays were compared for their ability to estimate protein in poo led gut fluid and compared with protein determined by quantitative amino ac id analysis using an automated procedure. The spectrophotometric assays all gave differing estimates of protein content and none correlated well with quantitative amino acid analysis. The Lowry and BCA methods gave similar es timates of protein, which were higher than those measured in the same sampl es by amino acid analysis. Both the DOC/TCA Lowry and the Bradford assay su bstantially underestimated protein in pooled gut fluid. Internal standard r ecovery experiments for the protein assays showed interference of colour fo rmation in the Bradford method but not the other spectrophotometric protein assays. This suggested the low values in the Bradford assay were due to gu t fluid interfering with colour formation. Therefore, five algal species kn own to occur in the diet of Kyphosus sydneyanus were selected to test for t heir interference in the Bradford assay. The algae included three species o f phaeophytes, one rhodophyte and one chlorophyte. Extracts from all five a lgal species interfered with this protein assay to significant but varying degrees. Two methods of extracting gut fluid were compared: (i) method A, w hich involved removing the gut fluid from the digesta immediately post capt ure; and (ii) method B, in which samples of digesta were frozen and gut flu id removed later. The fluid from method B gave significantly higher levels of THAA in some gut segments. Freeze-thawing of digesta in method B was tho ught to be the basis of the increased THAA. This finding suggests that meth od A more accurately reflects in vivo gut fluid THAA concentrations. The ab ove methodology provides a framework to compare the ability of different he rbivorous species at extracting THAA from their diet. These results highlig ht that the estimation of protein can be affected by methodology and by sam ple composition, and shows that methods must be optimized for a particular application. (C) 2000 Elsevier Science B.V. All rights reserved.