Monitoring water quality in aquaculture with a new sensor for microorganisms and toxicity

AnnaRosa Sprocati¹ Vincenzo Capuano¹ Giovanni Antonini² Pierdaniela Valenti³ Alessandra Aiello⁴

¹ENEA-Dip.Innovazione Div. Biotecnologie 
²Dip. Biologia di Base e Applicata - Università dell'Aquila 
³Istituto di Microbiologia II° Università di Napoli 
⁴MICROBO srl

A correct risk assessment in aquaculture activities must take into account both inflowing as outflowing waters, either in order to control risk factors for fish populations under stressful conditions or in order to avoid negative environmental impacts of the aquaculture plant. The development of innovative methods of control can positively contribute to a correct management of aquaculture activities. In the present work, sensor prototypes were installed on a mobile laboratory and utilised for monitoring faecal contamination and toxicity in a brackish aquaculture plant, situated on the Orbetello Lagoon, Italy. The final aim of the work was to verify the possibility to control the feacal contamination in real time, with reduced analysis costs, without analytical discontinuity, with automated data acquisition and employing non-specialised personnel. Thus the objectives were to conduct a methodological comparison between the sensor and the classical microbiological methods in brackish waters and to verify the efficiency of the prototypes over a long-time working period. The work has been split up into 5 campaigns during different seasons. Three different sampling points within the plant have been selected for which the following parameters were monitored: (i) faecal contamination index, (ii) nutrient content, (iii) pH and (iv) temperature. Sampling frequency was 2X a day during each campaign The sensor, patented by Microbo srl, is able to automatically measure the concentration of either total coliforms or faecal coliforms or faecal streptococci. The principle is based on the detection of bacterial metabolism, instead of replication (classical counting methods), through a specifically designed medium, added to the sample, in which products with oxido-reductive activity are contained and are metabolised by the microorganisms. The redox variation is converted into a change of electrical intensity or into a voltage change and compared to a reference electrode. The time required for the redox reaction is inversely proportional to the number of microorganisms present, varying e.g. for total coliforms, from 45 minutes to 12 hours, in case of sterility. Each prototype can process one sample at a time. The sensor can also be used for a toxicological test, which is based on the absence or reduction of microbial metabolism (a mutant strain of E. coli pva 45) in the presence of toxic substances. The lower range of sensibility is at a concentration of 10-4M. Validation has been done on phenol. 
Final results showed that the potentiality of this sensor is interesting and that the biological method is also very reliable for brackish waters. Nevertheless an improvement of the machinery is required to tolerate long-time working periods without specialised personnel and to carry out multiple samples at the same time. Indeed when the microbial concentration was lower than 104, only 39% of the foreseen analysis have been done within 24 hours, which is the time required by classical methods for T.C. For concentrations higher than 104, the total time of analysis has been shorter than classical methods. The toxicological test showed a light toxicity of the inflowing waters.
This work has been supported by the Ministero della Marina Mercantile in the Fourth National Plan for Agriculture and Fishery.