MEPS 177:27-35 (1999)  -  doi:10.3354/meps177027

ABSTRACT: Using data assimilation techniques, we tested coherency between different data sets of parameters measured during a biodegradation experiment performed on natural seawater. Parameters were dissolved organic carbon (DOC), dissolved free amino acids (DFAA), bacterial biomass and ectoproteolytic activity measured with the fluorogenic substrate L-leucine-4-methyl-7-coumarinylamide (MCA-Leu). The model, which considers dissolved organic matter as 3 polymeric pools (Mh1 very labile, Mh2 semi-labile and Mh3 refractory), was used to investigate consistency in the use of MCA-Leu to trace protein hydrolysis. Three assimilations were performed: (1) the first assimilation was on the set of data excluding the ectoproteolytic activity, (2) in the second assimilation, data from the MCA-Leu technique were compared to the hydrolysis rate of the labile pool of proteins (Mh1) given by the model, and (3) the third assimilation was performed assuming that the measured proteolytic activity was representative of the slow hydrolysis rate of the semi-labile pool (Mh2). The short-term sampling frequency permitted detection of a peak of DFAA during the first 50 h of experiment. However, in the model, MCA-Leu measurements were not compatible with the important fluxes of protein hydrolysis needed to reproduce this peak. MCA-Leu hydrolysis rates appeared to be more coherent with the slow hydrolysis rates of less labile organic matter (Mh2). We suggest that the high concentration of proteins in our sample was responsible for the induction of fast hydrolyzing enzymes which were not able to hydrolyze MCA-Leu. The V_max per cell ratio being relatively constant showed that MCA-Leu hydrolysis was likely to be due to constitutive ectoproteases. In the biodegradation experiment, proteolytic fluxes were due to a combination of inducible (high activity with respect to the Mh1 pool during a first phase) and constitutive enzymes (lower activity with respect to the Mh2 pool, second phase). The assimilation technique, applied to the model, showed that the MCA-Leu poorly traced the hydrolysis of labile proteins.

KEY WORDS: Data assimilation · Modelling · Biodegradation · Ectoproteolytic activity · MCA-Leu