application of bacterial indicator phylotypes to predict shrimp health status

• Published in: Applied microbiology and biotechnology Vol. 98; no. 19; pp. 8291 - 8299
• Main Authors: Xiong, Jinbo, Zhu, Jianlin, Zhang, Demin
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer-Verlag 01.10.2014; Springer Berlin Heidelberg; Springer; Springer Nature B.V
• Subjects: Analysis; Animals; Aquaculture; Bacteria; Bacteria - classification; Bacteria - genetics; Bacteria - isolation & purification; bacterial communities; bacterioplankton; Biomedical and Life Sciences; Biotechnology; Decapoda; Decapoda (Crustacea) - immunology; Decapoda (Crustacea) - microbiology; Environmental monitoring; Epidemics; Health aspects; health status; Independent variables; Life Sciences; Microbial Genetics and Genomics; Microbiology; Models, Biological; mortality; Multivariate analysis; Original Paper; Plankton; Ponds; prediction; Regression analysis; Shellfish; shrimp; shrimp culture; Shrimps; Studies; Variance analysis; Water quality; Indicator phylotypes; Incidence of disease; Logistic regression model; Bacterioplankton community
• ISSN: 0175-7598; 1432-0614
• DOI: 10.1007/s00253-014-5941-y

The incidence of shrimp disease is closely associated with the microbial composition in surrounding water, but it remains uncertain whether microbial indicator phylotypes are predictive for shrimp health status (healthy or diseased). To test this idea, we combined the data from our previous works, to investigate the feasibility of indicator phylotypes as independent variables to predict the health status during a shrimp culture procedure. The results showed linearly increased dissimilarities (P < 0.001) of the bacterioplankton community over time, while the communities dramatically deviated from this defined trend when disease occurred. This sudden shift in the bacterial community appears to cause severe mass mortality of the shrimps. In particular, we created a model to identify indicators that discriminated ponds with diseased shrimp populations from these with healthy shrimp populations. As a result, 13 indicative families were screened, in which seven are healthy indicator and six are diseased indictor. An improved logistic regression model additionally revealed that the occurrences of these indicator families could be predictive of the shrimp health status with a high degree of accuracy (>79 %). Overall, this study provides solid evidences that indicator phylotypes could be served as independent variables for predicting the incidences of shrimp disease.