We continue my series of more and less exasperated criticisms of the Jepson eFlora and Flora of North America treatments of the genus Allium with an entry that is more. Here’s the FNA couplet separating our protagonists:

95 Leaf blade strongly falcate; umbel mostly 5–10-flowered . . . . Allium parvum
95 Leaf blade linear or weakly falcate; umbel 20–30-flowered . . . . Allium cratericola

And the Jepson eFlora couplet (I’ve expanded the geographic abbreviations in the original):

41. Leaves straight or ± sickle-shaped; flowers 20–30; Klamath Ranges, North Coast Ranges, Sierra Nevada Foothills, southern High Sierra Nevada, Tehachapi Mountains, Inner South Coast Ranges, Western Transverse Ranges, San Jacinto Mountains ….. Allium cratericola
41′ Leaves sickle-shaped; flowers generally 5–10; Klamath Ranges, High Cascade Range, High Sierra Nevada, Great Basin ….. Allium parvum

The situation is complicated because Allium parvum is conspicuously heterotypic. I don’t know how many species are included in Allium parvum in these treatments, but the number is at least three. Let’s tackle geography first, because that’s easier. The “Allium parvum” in the Klamath Ranges and High Cascade Range is probably some subset of the Allium falcifolium / Allium siskiyouense complex, likely the white-flowered plants that I am provisionally including under Allium siskiyouense. The High Sierra Nevada Plants are… I don’t know what they are but they look like this. I have a hunch that they’re affiliated with Allium burlewii. If there isn’t a published name for them, there should be. I might look into that one more later.

That leaves the Great Basin plants. There’s a set of morphologically homogeneous observations in a polygon approximately bounded by—starting at the south and moving clockwise—Topaz Lake, Truckee, Susanville, Summit Lake, and Fernley. Kellogg’s protologue for Allium parvum mentions plants from Mount Davidson and Washoe. Kellogg’s specimens, unfortunately, were probably lost in the 1906 San Francisco fire, and I haven’t yet found a neotypification of Allium parvum. “Washoe” presumably refers to Old Washoe City, Washoe Valley, or an adjacent site. I believe “Mount Davidson” is the Mount Davidson in Storey County, Nevada. These sites are within the polygon described above, and all the observations around Washoe Valley and Mount Davidson that might be called “Allium parvum” seem to belong to this homogeneous grouping of Great Basin plants. So, the Great Basin plants are presumably Allium parvum.

We can simplify the geography a bit: Great Basin—and probably some distance into the lower elevations of the north end of the Sierra Nevada—is Allium parvum, anywhere else is Allium cratericola. Judging by the numbers of iNaturalist observations, Allium parvum appears to be reasonably common around Reno, and presumably in similar but less populated habitats along the east base of the Sierra Nevada, but rarely observed in California. The High Sierra Nevada and Klamath / Cascade plants are also infrequently observed in California, though more often than Allium parvum. You can find iNaturalist observations of Allium cratericola here, and I’ve aggregated a set of observations of Allium parvum sensu stricto here.

For the pragmatic task of differentiating the two, we have a clear & easy geographic distinction and we could just stop there. But that wouldn’t be any fun, would it?

Imagine a linear, flat leaf. It is arcuate if it is curved in a plane more or less perpendicular to the wide axis of the leaf’s cross-section. It is falcate if it is curved in a plane more or less parallel to the wide axis of the leaf’s cross-section. The leaves of both Allium cratericola and Allium parvum are arcuate and not, or only weakly and inconspicuously, falcate. The leaves of Allium parvum sometimes look falcate from a distance but, on closer examination, are arcuate and twisted—the distal portion of the leaf is curved in a plane parallel to the ground and perpendicular to the wide axis of the leaf’s cross-section. Only a few Allium parvum appear to actually have falcate leaves. The “falcate or not” distinction in the key will work approximately none of the time.

There does appear to be a difference in the average number of flowers per inflorescence. It doesn’t look consistent enough to differentiate the two species reliably, but this part of the key will work the majority of the time.

There appears to be a consistent difference between the two in leaf width: wide and often truly flat in Allium cratericola (sometimes variable among leaves of a single plant, but at least one of the leaves wide), narrow and always channeled (U-shaped in cross-section) in Allium parvum. One of the limitations of using iNaturalist is that I can’t measure the leaves, so I can only say “wide” and “narrow”, but there is a pretty obvious and apparently consistent difference between the two. The difference disappears if we adopt the heterogeneous circumscription of Allium parvum, though—the leaves of both Allium siskiyouense and the mystery entity in the High Sierra Nevada are wide. Those two also have falcate leaves, as it happens, making the “falcate or not” actively misleading rather than merely uninformative—Allium parvum sensu stricto falls on the “cratericola” side of the line.

Anther color is informative some of the time. Allium parvum has purplish anthers, while Allium cratericola has yellow or purplish anthers. The variation in Allium cratericola may warrant some further exploration. The observations of this species in iNaturalist occur in five population clusters: in the northern Coast Ranges; around the west base of the northern Sierra Nevada (centered on Table Mountain); around the west base of the central Sierra Nevada (mostly east of Stockton); around the Tehachapi Mountains; in the San Jacinto Mountains. The northern two, in the northern Coast Ranges and around Table Mountain, have purplish anthers. (There are occasional plants with the flowers entirely unpigmented, but apparently none with yellow anthers and the typical pinkish color & dark midribs of the tepals.) The remaining three, east of Stockton, around the Tehachapi Mountains, and in the San Jacinto Mountains, have yellow anthers. There seem to be some other subtle differences between these two groups, as well, though nothing I can easily put my finger on at the moment. It might be worth recognizing two taxa, at least at the varietal level. Although anther color alone doesn’t seem like an especially significant character, the differentiation is cleaner than the boundaries between many species of Allium in California. The geographic isolation between these population clusters also implies some degree of reproductive isolation, although we should suspect that all five are pretty well isolated from each other under current conditions. I also note, without further comment, that the FNA description for Allium cratericola simply says “anthers yellow”.

I omit a couplet for this entry. One remaining difficulty is that, while I think I have a reasonable idea what the form of Allium parvum in northwestern Nevada and adjacent California looks like, and this form seems to be the relevant one in the context of distinguishing Allium cratericola and Allium parvum, I definitely do not understand Allium parvum as a whole. In any case… just don’t use the “falcate or not” character, OK? Use geography if you merely want to put names on plants. Use a combination of leaf width, number of flowers, and anther color if you want to do it the hard way.

One last, minor detail: I think the inclusion of the southern High Sierra Nevada for Allium cratericola is simply an error. Neither the Consortium of California Herbaria nor SEINet have herbarium specimens of Allium cratericola, and I can find nothing on iNaturalist that would plausibly be included in the species. Populations in the western Transverse Ranges and south Coast Ranges, on the other hand, are missing from iNaturalist but documented by herbarium specimens.