The genus has an amphitropical distribution in North and South American deserts, and its phylogeny remains unresolved. This genus is conspicuous and specious within the Monte Desert, the largest, although understudied, southern South American dryland. presents an interesting case for phylogenetic studies due to its paternally inherited chloroplasts, its species hybridize in nature, and although nominal species are morphologically distinct, hybrids might be cryptic. We analysed ITS2 nuclear (nDNA) and rbcL chloroplast (cpDNA) sequences of the bifoliolate section, , including () from North America, and its South American congeners: () and (), and sequences of the multifoliolate section: (), (), and a morphological hybrid swarm. We aligned and analysed sequences from 111 individuals collected at 31 populations sampled along the range of each species. The nDNA revealed 56 haplotypes, and median-joining and maximum likelihood reconstructions provided clear separation among species and suggested hybridization between -. The nuclear phylogeny showed that the section diverged earlier than within which diverged first, meanwhile, consistent with previous studies, forms a monophyletic group sister to . Comparatively, cpDNA was less variable, with only six haplotypes shared between and -, and rarely between -. Our results emphasize the significance of separately considering nuclear and plastid evolutionary signals when reconstructing unresolved relationships. While nuclear markers clarified phylogenetic relationships and cryptic hybridization among species, the chloroplast revealed the retention of widespread ancient polymorphisms, which were conserved in populations of distinct species. Each marker provided insights into particular evolutionary patterns, highlighting that genetic variation may be more influenced by hybridization and mode of chloroplast inheritance than previously recognized.