Abstract
Aquaculture holds immense promise for addressing the food needs of our growing global population. Yet, a quantitative understanding of the factors that control its efficiency and productivity has remained elusive. In this study, we address this knowledge gap by focusing on the microbiome determinants of productivity, more specifically animal survival and growth, for one of the most predominant animal species in global aquaculture, whiteleg shrimp (Penaeus vannamei). Through analysis of the shrimp-associated microbiome from previous studies across Asia and Latin America, we established the presence of core phylogenetic groups, widely prevalent across aquaculture conditions in disparate geographic locations and including both pathogenic and beneficial microbes. Focusing on the early stages of growth (larval hatcheries), we showed that the composition of the microbiome alone can predict a remarkable fraction of the variation in shrimp larvae survival rates (ca. 50%). Taxa associated with high survival rates share recently acquired genes that appear to be specific to aquaculture conditions. These genes are involved in the biosynthesis of growth factors and protein degradation, underscoring the potential role of beneficial microorganisms in nutrient assimilation. By contrast, the predictability of the microbiome on the adult shrimp weight in grow-out farms is weaker (10%-20%), akin to observations in the context of livestock. In conclusion, our study unveils a novel avenue for predicting productivity in shrimp aquaculture based on microbiome analysis. This paves the way for targeted manipulation of the microbiome as a strategic approach to enhance aquaculture efficiency from the earliest developmental stages. IMPORTANCE: Aquaculture is a rapidly growing industry essential for global food security, yet its productivity is often constrained by high mortality rates and inefficient growth. While the microbiome is known to influence host health and nutrient assimilation, its broader role in animal production remains poorly understood. Here, we take a data-driven approach to address this gap by systematically analyzing shrimp-associated microbiomes across hatcheries and farms. By integrating microbiome data with machine learning, we demonstrate that microbial communities are powerful predictors of key production outcomes, shaping shrimp survival and growth. Our findings suggest that the microbiome could serve as a diagnostic tool for assessing production conditions and optimizing management strategies. In addition, machine learning techniques offer a promising avenue for identifying beneficial microbes and developing targeted microbiome therapies to enhance aquaculture sustainability and efficiency.