Environmental structuring of mixoplankton functional types within marine protist communities: a global analysis

Mixoplankton, marine planktonic protists that combine photo-autotrophy and phago-heterotrophy, play vital roles in marine ecosystems as producers, consumers and nutrient recyclers. However, the environmental drivers of their global distribution remain poorly understood. Here, we utilised global DNA metabarcoding data sourced from the metaPR2 database and classified ~ 47,000 marine protist ASVs (amplicon sequence variants) into four mixoplankton functional types ‒ constitutive (CM), generalist non-constitutive (GNCM), endosymbiotic-specialist non-constitutive (eSNCM) and plastidic-specialist non-constitutive (pSNCM) ‒ and other functional groups (diatoms, non-diatom phytoplankton, protozooplankton, and parasites). We then applied a machine learning-based community clustering method to delineate assemblages, which were subsequently analysed using multivariate and statistical techniques to resolve the spatial distribution and environmental associations of mixoplankton within global protistan communities. Analysis of ASV richness and relative abundance confirmed that mixoplankton are ubiquitous components of protistan communities. Self-organising maps and distance-based redundancy analyses identified communities structured along environmental gradients aligned with classical oceanographic regions, and generalised additive models supported statistical associations with temperature, salinity and nitrate concentration. CM were broadly distributed in oligotrophic waters; eSNCM were restricted to warmer biomes; and GNCM and pSNCM showed niche-specific associations with nutrient regimes. Non-diatom phytoplankton had distributional patterns similar to CM, suggesting functional overlap or shared resource utilisation. In contrast, diatoms were predominantly associated with cold nitrate-rich waters, while protozooplankton and parasites displayed trends generally inverse to those of mixoplankton. Through the integration of metabarcoding data with machine learning, multivariate and statistical approaches, we demonstrate that distinct mixoplankton functional types occupy ecologically differentiated niches governed by temperature, salinity, and nutrient availability. These findings enhance our understanding of mixoplankton ecology and global distribution, helping to establish a foundation for their integration into predictive models of marine ecosystem dynamics.