[en] A rearing protocol for the edible European sea urchin Paracentrotus lividus in a closed cycle (control of the whole life cycle of the echinoid) and in a recirculating system (control of the environment around the echinoid) is set up and tested at a pilot scale. This protocol is used to experiment on growing postmetamorphics whose age and genetic origin are perfectly known. Among the various measurements of size, we determined that the test diameter is both rapid and accurate for quantifying somatic growth. Causes and mechanisms of asymmetrical, or even sometimes multimodal, size distributions among previously homogeneous cohorts are studied. Results evidence the existence of a size-based intraspecific competition, causing a reversible growth inhibition of smaller individuals. A new growth model (called 'fuzzy-remanent'), including a component of intraspecific competition, is elaborated by defuzzifying a fuzzy model. Traditional least-square regression is abandoned in favor of quantile regression to fit it. Both the model and the regression method are adapted to include individual variations (we call this an 'envelope model'). This envelope model has functionally interpretable parameters. One of them quantifies the degree of inhibition caused by intraspecific competition. Since many similar fuzzy-remanent functions can be designed and fitted with this method, this approach is promising to model growth of other organisms in a functional way. This model rehabilitates von Bertalanffy's theory on individual growth. Moreover, the latter theory is now verified for Paracentrotus lividus, despite the observation of an initial lag phase in growth. A functional classification of growth curves is proposed.
