In difficult times of rising input costs and declining selling prices, shrimp farmers are looking more than ever to reduce their production costs. Feed consumption represents their highest cost of production, and their productivity is generally summed up in their yield in pounds/kilos per hectare, partially linked to the weekly growth and survival achieved in each pond.
Precision aquaculture has been one of the recently promoted developments. It consists of controlling as precisely as possible the different production parameters that influence costs and productivity. The increase in connected electrical and electronic equipment in recent years has made it possible to generate a large amount of data and many applications. D igital platforms with intelligent data analysis use them. However, even with artificial intelligence, it is difficult to take advantage of this large volume of data if such data is not reliable, and even worse, manipulated or modified for whatever purposes.
A first example is the non-taking into account of the % of “free” postlarvae at the time of stocking, not only for the reason that they were not invoiced and therefore cannot appear in the "inventories" but also for the objective of presenting an apparent false good survival at the end of the cycle. The perverse effects of this behavior are the risks of (a) underfeeding, because they do not appear in the calculations of biomass to be fed, (b) reaching a higher real biomass that the culture system can support, preventing the best development of the shrimp in the right good conditions. They may represent a significant loss of economic efficiency and accuracy, or relevance, of automated production models.
The second example is the manual modification of certain figures of weekly weight samples, just to be able to present a "linear" weekly growth and/or close to the targets to please the hierarchy. It doesn’t make much sense since , as everyone knows, due to the moulting stages with variable frequency, the increase in the weight of a shrimp is not the same every day, or every 3 or 7 days. The perverse effects of this behavior are the risks of (a) wrongly estimating the biomass to be fed in the following days and even the biomass to be harvested, (b) wrong feeding growth models and everything related to those data!
The current situation of modern shrimp farming requires a modernization of spirits to keep improving the efficiency of the sector. This will only be possible with as much sincerity as possible from the data generated, especially from data controlled by people.