Gagea (Wolfe, 1998)

Zarrei, Mehdi, Wilkin, Paul, Ingrouille, Martin J. & Chase, And Mark W., 2011, A revised infrageneric classification for Gagea Salisb. (Tulipeae; Liliaceae): insights from DNA sequence and morphological data, Phytotaxa 15, pp. 44-56 : 46

Gagea sensu stricto and subgenera

Classifying Gagea (excluding Lloydia ) into two subgenera, as was originally suggested by Pascher (1907), is supported by some morphological synapomorphies. Apart from seed characteristics, which are frequently cited as being of taxonomic importance, tunic texture and geography are also valuable in establishing subgenera ( Zarrei et al. 2009). Flattened seeds and a fibrous, reticulate to fibrous-papery tunic are synapomorphies for G. subgenus Hornungia (Bernh.) Pascher; these taxa mostly occur in dry habitats. Pyriform to terete seeds and a leathery tunic are shared characters among G. subgenus Gagea members, which mostly occur in more humid and boreal habitats. There are a few exceptions, such as G. alexeenkoana Mishchenko (1908: 76) in G. subgenus Hornungia, which extends into humid areas, and G. kunawurensis as well as G. dschungarica Regel (1879: 513) in G. subgenus Gagea , which grow in drier areas.

However molecular (nrITS and plastid) studies do not support the existing subgeneric classification for Gagea ( Peterson et al. 2008, Zarrei et al. 2009). Neither G. subgenus Gagea nor G. subgenus Hornungia is monophyletic. Analysis of sequence data from the low-copy nuclear gene ncpGS also does not support the monophyly of either subgenus (Zarrei et al., unpubl. data).

Micromorphological characters do not display congruence with the previous subgeneric classification either. Apart from tepalar characters that are to some extent uniform in all species studied ( Zarrei et al. 2010d), characters associated with the different types of basal leaves and pedicels have been shown to be dispersed throughout both subgenera; widespread convergent evolution for these traits is indicated.

Palynomorphological characters do not support the existence of two subgenera ( Zarrei & Zarre 2005) because, although most species of G. subgenus Hornungia were classified under the reticulate and foveolate pollen type, microreticulate and perforate species are dispersed throughout the genus. The perforate pollen type included mostly G. subgenus Gagea species , but some species have reticulate and micro-reticulate palynomorphs. Evidently these palynomorphs have evolved repeatedly.

In summary, although some macromorphological characters support the monophyly of subgenera in Gagea sensu stricto (as explained earlier in this section), neither micromorphological traits nor molecular analyses confirm this classification, which should therefore be revised.