EN PRENSA Predicting semi-intensive production of Penaeus vannamei using simple linear regression models: A sensitivity analysis using environmental and management variables. EN PRENSA

Abstract

An analysis was conducted on semi-intensive production of a commercial farm of white shrimp (Litopenaeus vannamei) to establish recommendations for production improvement, considering the influence of environmental and management variables on growth and survival. For this purpose the records obtained of a production cycle. A variance analysis was conducted to confirm significant differences in environmental variables among ponds. The following relationships were determined: 1) environmental variables with management variables and 2) environmental and management variables with the production parameters (final weight and final survival). Subsequently, simple linear regression models were estimated between the variables related to the production parameters to conduct sensibility analysis by simulation. The analysis of variance showed statistically significant differences (P < 0.05) in temperature and dissolved oxygen concentrations among ponds. The correlation analyses between environmental and management variables showed that the final weight was positively related to temperature and dissolved oxygen and inversely proportional to the stocking density and duration of cultivation (P < 0.05). Survival was negatively correlated with pond size (P < 0.05). Sensibility analysis by simulation showed that variability of dissolved oxygen affected the final production more than other variables, as it increased from 1.521 kg ha^-1 to 2.429 kg ha^-1. The remainder variables in order of importance were: temperature, duration of cultivation, amount of feed, stocking density and pond size, respectively. Within the ranges tested, higher levels of dissolved oxygen, appropriate stocking densities and small pond size could substantially improve the production in semi-intensive Litopenaeus vannamei shrimp farms.