Flora of North America species comparison

Of conservation concern.

Geum schofieldii is one of the more attractive yet seldom seen species of avens in North America; it is known from the rocky wet mountains of the Queen Charlotte Islands and northern Vancouver Island.

Because Geum schofieldii appears to be intermediate morphologically between G. calthifolium and G. rossii, some botanists have considered it to be a hybrid and included it in synonymy with G. ×macranthum [G. calthifolium × G. rossii]. The 16-ploid chromosome number (2n = 112) of G. schofieldii suggests an allopolyploid origin. Geum rossii is not known to occur in the Queen Charlotte Islands or on Vancouver Island; the plants cannot be of recent hybrid origin. Geum ×macranthum is found in the Aleutian Islands and the Alaska Peninsula where the two parental species overlap. Compared to G. schofieldii, its leaves are more pinnate than lyrate-pinnate, with 9 to 13 major leaflets instead of 5 to 7. Also, G. ×macranthum appears to be sterile; G. schofieldii appears to be fertile, suggesting that the latter is a valid species. The distinctness of G. schofieldii pollen also supports its specific status (R. J. Hebda and C. C. Chinnappa 1990).

(Discussion copyrighted by Flora of North America; reprinted with permission.)

Species ca. 45 (16 in the flora).

The most distinctive feature of many members of Geum is their geniculate-jointed style. The style has two parts, a persistent proximal segment and a deciduous distal segment. Where the two segments join, each bends to form a hook, with the hook of the proximal segment joined to the hook of the distal one. Seven of the 16 North American species of Geum lack this distinctive feature. Differences in style structure have led to differences in opinion as to generic delimitation and subgeneric classification of Geum.

During the early twentieth century, species traditionally included in Geum were divided among an increasing number of segregate genera. F. Bolle (1933), in his worldwide monograph, recognized five genera. After conducting extensive cytogenetic studies on Geum for two decades, W. Gajewski (1957) concluded that the most reasonable approach was to classify all the species in one genus. Most flora writers since Gajewski have maintained the species in a single heteromorphic genus. J. E. E. Smedmark and her collaborators have utilized DNA sequences from ITS, chloroplast trnL-trnF region, and nuclear low-copy gene GBSSI (Smedmark and T. Eriksson 2002; Smedmark et al. 2003), along with scanning electron microscopy of gynoecial development (Smedmark and Eriksson 2006), to test hypotheses by Gajewski of fruit evolution and allopolyploidy in Geum and to explore the phylogenetic relationships among the species. Their data support none of the previous circumscriptions of Geum as a monophyletic group, nor most of the suggested segregate genera favored by taxonomists such as Bolle, E. L. Greene (1906), and P. A. Rydberg (1913b). The taxonomic solution chosen here for North American species is traditional yet consistent with the molecular data and recognizes two genera: Geum for the polyploid (2n = 28–116), herbaceous species and Sieversia for the diploid (2n = 14), suffrutescent species. Smedmark (2006) found that Waldsteinia might best be treated within Geum. It is also possible that Coluria R. Brown, Taihangia T. T. Yu & C. L. Li, and Waldsteinia are sister to Geum (Smedmark and Eriksson 2002). Waldsteinia (2n = 14) is retained here as a separate genus, with its traditional circumscription.

Although most of Geum species in the flora area are reasonably distinct morphologically, their distinctive characteristics are not well correlated so as to be combined in a key using several characters per couplet, as was pointed out by E. G. Voss (1972–1996, vol. 2), whose key to Geum in Michigan served as a model for the key below. In particular, the characters that are useful in identifying specimens in flower do not correlate well with characters helpful in the identification of fruiting specimens. Specimens with both flowers and fruits are the easiest to identify with certainty, and both leads at couplet 11 in the key should be followed if possible. Petal color in Geum fades upon drying. Both white and yellow petals ultimately turn cream to pale yellow, thus rendering an important key character for fresh plants unreliable with most herbarium specimens. Collectors should note petal color on their labels.

For taxonomists familiar with Geum, leaves provide valuable characters for distinguishing the species, especially when flower color is not known. The complex morphology of leaves and their variability do not suit them for wide use in keys. Variability occurs both among plants of one species and on one plant. In many of the species with geniculate-jointed styles, the basal leaf form changes over the growing season. Spring-developing leaves are often pinnate followed by lyrate-pinnate or 3-foliolate leaves, and in late summer by simple leaves that are winter-persistent. For instance, the basal leaves of G. laciniatum may appear like those of typical G. aleppicum in late spring, but look more like those of G. macrophyllum var. macrophyllum in late summer. Cauline leaves may vary from pinnate proximally to 3-foliolate at mid stem to simple just proximal to the inflorescence. Because variability of leaves within a single species is nearly as great as between some species, leaf descriptions are kept short. Measurements of leaf length include the petiole, if present.

(Discussion copyrighted by Flora of North America; reprinted with permission.)