Phenotypic characterization of wild Myrciaria dubia (Kunth) McVaugh ex situ germplasm bank for breeding, conservation, and sustainable development in the Peruvian Amazon

Abstract

Introduction: Myrciaria dubia “camu-camu” is an economically important Amazonian fruit shrub known for its exceptionally high vitamin C content. Despite its commercial value, comprehensive phenotypic characterization of available genetic resources remains limited, hindering breeding programs and conservation strategies. This study aimed to characterize the phenotypic diversity of wild accessions maintained in a 36-year-old ex situ germplasm bank, one of the most comprehensive M. dubia collections globally, to provide baseline data for genetic improvement and conservation. Methods: We evaluated 43 wild accessions systematically collected from eight major hydrographic basins in the Loreto region using a stratified sampling approach to capture maximum ecological diversity and maintained in an ex situ germplasm bank established in 1988 at the National Institute of Agrarian Innovation in Loreto, Peru. Twenty-three quantitative and six qualitative morphological descriptors were assessed using standardized protocols specifically developed for M. dubia, emphasizing commercially important descriptors including fruit weight, pulp content, and seed characteristics. Statistical analyses encompassed univariate variability assessment, bivariate correlations, and multivariate classification through hierarchical clustering and principal component analysis. Results: Phenotypic characterization revealed moderate overall variability with coefficient of variation averaging 17.4%, with reproductive descriptors showing greater variation than vegetative traits. Fruit and seed descriptors exhibited the highest variability exceeding 20%, while qualitative descriptors showed limited diversity with Shannon Index of 0.823. Multivariate analysis identified four distinct phenetic groups with no significant correlation to geographic origin (Mantel test, p=0.4034). Principal component analysis revealed fruit-related descriptors as primary drivers of phenotypic differentiation, explaining 57.1% of observed variability. Three accessions from phenetic group 3 (PER1000416, PER1000423, and PER1000411) demonstrated superior trait combinations: fruit weight exceeding 13 g, pulp content above 75%, and reduced seed count below 2.5 seeds per fruit. Conclusion: The moderate phenotypic variability observed reflects the natural distribution patterns and limited domestication history of the species. This comprehensive characterization provides essential baseline data and a foundation for targeted breeding programs, conservation strategies, and sustainable production systems supporting development while preserving the genetic diversity in the Peruvian Amazon