Aims: To explore the relationship between plant communities and key environmental drivers in the Andes of northwest Patagonia, Argentina, and to evaluate the applicability of traditional phytosociological definitions to this region. Methods: We conducted 141 vegetation samples using a stratified systematic sampling design. This was done along two steep gradients of aridity and temperature, which are further modified by local factors that include successional changes due to fire and soil variation. We employed a series of multivariate approaches, including hierarchical clustering, Indicator Species Analysis (ISA), restricted Monte Carlo permutation tests, and both constrained and unconstrained ordinations, to identify (a) the main plant community types, (b) their representative species, and (c) the primary drivers of variation in species composition. Finally, we compared the obtained groups and species to associations described by earlier expert-based classifications. Results: From the set of analyses, we identified six different plant community types with 241 recorded species. We found significant differences across communities’ species composition and their environmental indicators. Among the considered environmental variables, elevation and annual precipitation had the strongest effect on species composition. Additionally, variation in composition was significantly related to forest structure, land use and soil characteristics. We further outlined the influence of locally biased classifications based on a predominance of sampling in areas western to the Andes in classification systems developed in the region. Conclusion: Our analysis allowed us to identify the most relevant environmental drivers and indicator species of the six classified plant communities based on numerical methods. The findings highlight the importance of considering full ecological gradients and communities’ responses for developing stable classification approaches.<br> Taxonomic reference: Anton and Zuloaga (2023).<br> Abbreviations: db-RDA = distance-based Redundancy Analysis; ISA = Indicator Species Analysis; NMDS = Non-metric Multi-Dimensional Scaling.
