Fast decision-making tool for monitoring recirculation aquaculture systems based on a multivariate statistical analysis

Abstract Aquaculture requires the monitoring of several parameters simultaneously such as temperature, salinity, ammonia, potential of hydrogen, nitrogen dioxide, bromine, among others to preserve the well-being of the fish, optimize production and meet the requirements of environmental ethics. To contribute to the fulfillment of these requirements, we propose here to use a data analysis technique called principal component analysis (PCA). This data processing technique is a mathematical procedure that performs a transformation on the data to convert possibly correlated parameters into a new set of parameters called principal components. Once the PCA performs an exploratory analysis on the data, we can use it to reveal the internal structure of the aquaculture data, to explain not only the variance in the data, but also the correlation between them. The choice of PCA in this study was based on the fact that it is a multivariate statistical technique of data that is very well known and widely applied in several scientific fields. Furthermore, the PCA allows to reduce the dimensionality of datasets, increasing their interpretability, but at the same time minimizing the loss of information. Results obtained with the use of PCA on the data related to the weaning and pre-fattening phases of Senegalese sole (species Solea senegalensis) collected during the years 2018 and 2019, respectively, show the feasibility of this proposal in the simultaneous monitoring of aquaculture parameters and take future critical conclusions to predict events in the aquaculture data analysis. Furthermore, there are no records to date of studies in the aquaculture industry in terms of data processing and analysis in detail. Currently, the scientific literature in the area of aquaculture exhibits a consistent body of knowledge based on the characterization of the physiology and environment of fish, with research on instrumentation still scarce. In this perspective, this study not only represents a substantial novelty for the field of aquaculture for performing the data processing directed at the aquaculture industry for the first time, but it will also allow the development of a new branch of research focused specifically in a support tool for aquaculture recirculating systems-based industry.