Morpho-anatomical traits explain the effects of bacterial-feeding nematodes on soil bacterial community composition and plant growth and nutrition

• ● Bacterivorous nematode species differentially promoted plant growth and nutrition. • ● Predation by bacterial-feeding nematodes altered the soil bacterial communities . • ● Nematode effects were predicted better by traits than by cp classes or species identity. • ● Body and egg sizes as well as mouth and procorpus width were likely effect traits. • ● The trait approach could explain the variable effect of nematodes on ecosystem functions. Bacterial-feeding nematodes are abundant soil animals regulating microbial communities and enhancing plant nutrition and growth. However, the biological metrics driving the variable within-trophic group effects of these organisms on soil and plant functions are not yet identified. In this study, we determined the effects of eight bacterial-feeding nematode genera ( Acrobeloides butschlii , Brevibbuca sp., Cephalobus sp., Diploscapter sp., Oscheius akosreti , Panagrellus redivivus , Pristionchus pacificus , Zeldia sp.) on soil bacterial communities and Pinus pinaster growth and nutrition from organic N (glutamate) and P (phytate) supplied in a high P-adsorbing soil. Because nematode species harbored contrasting morpho-anatomical characteristics, we hypothesized that traits, in comparison to species identity or life strategy groups (cp classes), could explain the variable effects of the nematode species. After 50 days of growth, the inoculation of bacterial-feeding nematodes caused variable but never negative effect on plant growth and net plant N and P uptake depending on nematode species. We also observed significant shifts in root traits and bacterial community composition. For both soil bacterial composition and plant growth and nutrition, a selected panel of traits ( e.g. body size, mouth width, procorpus width, demanian ratios b and c, relative egg size) explained soil and plant function more than did species identity or life strategy groups. For instance, traits explained 77 % of plant N and P amounts while cp classes and species-identity explained 33 % and 46 %, respectively, according to PLS models. The identity of traits varied according to the functions. Structural equation modeling with latent variables emphasized the direct effects of nematodes on plant nutrition through excretion in comparison to indirect effects through shifts in bacterial composition and improved root ramification. We discuss the relevance of bacterial-feeding nematode traits as promising biological metrics to explain the effect of these worms within the soil microbial loop framework.