notes Found growing on limestone substrate a few hundred feet into a large, fairly steep-walled, amphitheater-like ''canyonette'' off the south side of lower Titus Canyon. Using the Jepson eFlora Key to Brassicaceae, the first couplet requires deciding between ''Hairs simple or 0'' vs. ''At least some hairs branched''...which is not an easy call (at least for me:-). The plants looked quite glabrous to the naked eye...with only scattered, fairly isolated patches of relatively sparse and quite minute hairs (see the 'high-resolution' version of my 2nd image in this series). Those hairs mostly appear to be simple (though somewhat 'curly'), but when such hairs are branched at the very base it can be very hard to clearly discern.
However, either way one goes...whether with ''Hairs simple or 0'' (leading to 'Group 2'); or ''At least some hairs branched'' (leading to 'Group 3')...the end result of working through the key is genus Sibara, which was erected by E. L. Greene in 1896, and is an anagram of Arabis (just spelled in reverse :-).
As for species here, when I found these plants in 2009, I got to Sibara rosulata using the key in JEPS93. In the current Jepson eFlora treatment, that species is no longer recognized...having been synonymized under S. deserti. (But to get to S. deserti in the current Jepson eFlora key...one has to go with 'plant hairy', which is not how I would describe the plants here!)
The species here was first described in 1908 as Thelypodium deserti by the illustrious Marcus E. Jones, and later treated under the name Arabis deserti by Abrams in 1944 (see Fig. 2098 here). Soon thereafter, in 1947, Reed Rollins split the taxon into two species, to which he gave names Sibara deserti and Sibara rosulata...see Rollins' descriptions & discussion on pp. 140-143 here. The plants I photographed here agree much better with Rollins' S. rosulata than his S. deserti, i.e. the cauline leaves are mostly entire and narrowly linear; the fruits are ascending (rather than loosely reflexed); the styles become elongate in fruit and somewhat expanded distally; and the the leaves and fruits here are only sparsely hairy to glabrous, with hairs simple to merely forked (vs. highly-branched and more abundant in Rollins' circumscription of S. deserti). Also, Rollins' type specimen for S. rosulata was collected in Titus Canyon, and (thus not surprisingly) the calcareous, canyon wall habitat where I found these plants fits well among the habitats described for S. rosulata in Rollins' 1993 opus on North American 'Cruciferae'.
On pg. 142 of his 2010 synopsis of Sibara, Ishan Al-Shehbaz synonymized S. rosulata under S. deserti (though without any remarks explaining why in that paper). In his remarks under the FNA description of S. deserti, Al-Shehbaz gives a reasoned basis for his synonymy (even if the current Jepson eFlora key doesn't work well for the plants here).
An image of a sheet holotypes for S. rosulata, collected in 1947 by Roxanna Ferris and identified by Reed Rollins, appears at this JSTOR link (though, somewhat annoyingly, the image can't be viewed at a decent size without special access privileges).
