13. Viola sagittata Aiton (arrow-leaved violet, arrowhead violet)

V. sagittata var. subsagittata (Greene) Pollard

V. arkansana Greene

V. emarginata (Nutt.) Leconte

Pl. 576 j, k; Map 2704

Plants perennial, to 30 cm tall, with a prostrate to ascending rhizome 4–6 mm thick. Aerial stems not produced. Leaves in a basal rosette, long-petiolate (except sometimes in early-season leaves), the petiole glabrous or finely hairy. Stipules conspicuous, free from the petiole, membranous to somewhat herbaceous, narrowly lanceolate, the margins toothed and glandular-hairy. Leaf blades 1.5–8.5 cm long, all except sometimes those of the first leaves of the growing season lobed (occasionally appearing only coarsely toothed), the lobes produced only near the blade base and less than 1/3 the length of the midvein, the blade much longer than wide, elliptic to narrowly triangular, narrowly angled to a rounded to sharply pointed tip, broadly rounded to truncate or cordate at the base, the margins otherwise bluntly toothed below an often more or less entire apical portion, the surfaces glabrous or finely hairy. Cleistogamous flowers produced. Flower stalks not or only slightly overtopping the leaves (those of the cleistogamous flowers erect or ascending). Sepals 4–8 mm long, lanceolate, narrowly angled to a sharply pointed tip, the margins sometimes minutely hairy, the basal auricles well-developed, elongating as the fruits develop. Corollas 8–14 mm long (except in cleistogamous flowers), the petals oriented forward with arched or outward-curved apical portions, longer than the sepals, bluish purple to purple (with a greenish white throat) with darker veins (rarely all white elsewhere), the lateral and lower petals bearded on the upper surface with slender hairs, the spur conspicuous, well-exserted beyond the sepal auricles, stout and often somewhat hemispheric in shape. Stamens not exserted, typically not visible without dissection of the flower. Style slender, slightly expanded into a narrow, more or less scoop-shaped, hollow, truncate tip. Fruits 11–18 mm long, narrowly ellipsoid to ellipsoid, green, drying to tan, the surface glabrous. Seeds 1.2–1.5 mm long, tan to brown. 2n=54. April–June.

Scattered nearly throughout the state, but uncommon or absent from the Mississippi Lowlands Division and the western half of the Glaciated Plains (eastern U.S. west to Minnesota and Texas; Canada). Dry prairies, sandstone glades and edges or clearings of open oak forests, on dry to seasonally somewhat moist sand or gravelly sand.

This species is relatively widespread in Missouri, but apparently is nowhere very abundant in the state. Steyermark (1963) noted the extreme seasonal variation in leaf morphology within V. sagittata, with early-season leaves small and tending to have relatively short petioles and relatively entire blades. Blades of late-season leaves tend to be much shorter relative to their width than those produced earlier in the growing season. Most botanists recognize two varieties within the species (McKinney, 1992; Ballard, 1994). Missouri plants generally correspond to the widespread var. sagittata, with leaves becoming erect and long-petiolate at fruiting and with the blades of all but early-season leaves lobed. The var. ovata (Nutt.) Torr. and A. Gray (V. fimbriatula Sm.) comprises populations in the Great Lakes region and portions of the Appalachians that have the leaves spreading and relatively short-petiolate at fruiting, as well as blades that are at most coarsely toothed, rather than lobed. Yost (1987) studied plants of this complex that were clonally produced or grown from seed planted into shaded and sunny plots in a forest and adjacent field in New York. She concluded that leaf size and petiole length were strongly influenced by environmental factors and were not good characters for taxonomic distinctions in the group, but did not address whether other potential distinguishing features might be under genetic control. Gil-Ad (1997), who treated the two taxa as separate species, tabulated a number of additional (often subtle or overlapping) morphological characters said to separate them and also noted apparent differences in habitat preferences.

Viola sagittata produces putative hybrids with several other species of stemless blue violets. Russell and Risser (1960) studied plants that had been called V. emarginata and concluded that these represented a series of such interspecific hybrids, but McKinney (1992) noted that the type specimen of this name appears to be referable to V. sagittata. Occasional hybrids with V. sororia were studied by Russell and Cooperrider (1955) and have been called V. ×greenei House. Steyermark (1963) also reported rare putative hybrids between V. sagittata and V. pedatifida, but these have not been confirmed by later specialists.