Wendt separated these latter two subgenera based on morphological characters. However, this division was later rejected by the same author (Wendt, pers. comm. 1989).
Marques joined the two subgenera because both have a carinated, trilobed posterior petal, a thick style, and an apically-free staminal sheath. Although the first two characters are constant in the species for these two groups, most of species of subg. Pterocarya have auriculate styles.
The decision to put the two subgenera together was primarily based on the morphological study of the Brazilian species. For example, the author also characterized subg. Monnina as having only free posterior sepals; however, 10 Ecuadorian species and 23 Peruvian species (Ferreyra 1946, 1953) have these sepals connate.
This author divided the genus based on a morphological and anatomical cladistic analysis using 15 out of the ca. 180 species of Monnina. According to her strict consensus tree, the subg. Monnina comes in a different clade than the other subgenera (12% of bootstrap suppport), subg. Pterocarya comes in a different clade from the last subgenus (13% of bootstrap support), and subg. Monninopsis comes together with Polygala and other genera (12% of bootstrap support).
More species should have been studied, especially for coding the characters. For example, although most of the species of subg. Pterocarya are herbaceous plants, Monnina martiana is a shrubby plant of up to 4 m high (Freire-Fierro 1992). Or, while most species of subgenus Pterocarya and subgenus Monnina have an apically free staminal sheath, in some species it is partially connate. Although most of species of subg. Monnina have drupes, the Ecuadorian Monnina sodiroana has samaroid fruits (Freire-Fierro 1989), from subg. Pterocarya while most of species have samaroid fruits, Monnina leptostachya, M. exalata and M. oblongifolia (Freire-Fierro 1992) have true drupes, or from subg. Monninopsis, while Monnina insignis has samaroid fruits, Monnina malmeana has drupes (Freire-Fierro 1992). This character is of crucial importance since it was the only morphological character that separated the three subgenera in the cladistic analysis.