Hierarchical clustering was applied in order to get co-occurrence assemblages. These sets were compared with diagnostic groups defined a priori, life forms and their status in South America (introduced vs native species). In order to detect geographic and climatic distribution patterns of the assemblages, we extracted climatic summary variables from the WorldClim database and performed a principal component analysis (PCA). We detected seven species assemblages with a significant correspondence to diagnostic groups defined a priori (P < 0.01), though their association is relatively low (Cramér's V = 0.49). Native species dominate over adventive ones in each assemblage suggesting low risk of outcompetition by introduced species adapted to these habitats. The annual life form is over represented in most of the cooccurrence groups. Despite the huge temperature and precipitation gradient in the studied transect, plant assemblages are occurring almost along the whole range, supporting the hypothesis of azonality (low dependence on macro-climate). Assemblages defined in this work can be potentially used for further supervised classification strategies in order to define plant community types within vernal pool vegetation.