Flora of Deep Canyon, Santa Rosa Mountains

Jan G. Zabriskie, 1979

Updated by Tom Chester, 2006

Updated Determinations
Discarded Data
Notes On Some Taxa
Checklist of the Vascular Flora of Deep Canyon


The northeast slope of the Santa Rosa Mountains has three main drainages: Palm Canyon, Deep Canyon and Martinez Canyon, all of which drain into the Coachella Valley and then to the Salton Sea. See Overview of Flora of Santa Rosa Mountains for a location map showing these drainages, and a detailed map of the Deep Canyon drainage.

Deep Canyon is the site of the University of California's Philip L. Boyd Deep Canyon Desert Research Center, which has nearly 3,600 acres of UC-owned land and approximately 10,400 acres of public domain through long-term permits granted by the U.S. Bureau of Land Management (Ting and Jennings 1976). This is approximately half of the ~25,000 acres (~100 km2) of the entire Deep Canyon drainage.

Research in Deep Canyon began in 1956, with Boyd offering the initial land for the Center in 1958. (Boyd was an early land developer in the Palm Springs Area.) An excellent overview of the history and natural history of Deep Canyon was published in book form as Deep Canyon, a Desert Wilderness for Science, editors Irwin P. Ting and Bill Jennings, 1976, Boyd Research Center, UC Riverside, 177 pp.

Jan G. Zabriskie joined the Deep Canyon staff in 1973 as a research associate after getting his B.A. in botany, and studied the flora of Deep Canyon for the next four years. He conducted a rich set of 22 surveys over a three year period from 1974 to 1977. The surveys sampled the entire elevation range within Deep Canyon from the bottom of the Canyon at 80 feet elevation to just below the top of the mountain at 8400 feet elevation. The survey sites were spaced at 400 foot elevation intervals except for the lowest spacing of 320 feet. (He reported everything in meters, but clearly conducted all work in feet since those were the units used by topo maps.)

Zabriskie took care to select the survey sites to have the same physical aspect (site slope and direction the slope was facing), and to reflect the predominant physical and biotic aspect of each elevation. (Zabriskie's words are always given in italics on this page.) The surveys were done when the plants were growing so full canopies could be measured and perennial herbs included in the count. The survey at each site was done by stretching a measuring-tape and walking a contour line 400 m long (1312 feet). Perennial plants that intercepted the tape were counted and measured as they were encountered along the line.

Zabriskie explored many other areas in Deep Canyon as well, and found at least five taxa previously unrecorded in the Santa Rosa Mountains that were significant range extensions from elsewhere. These taxa were found in the Upper Plateau section: Chamaesyce revoluta, Tragia ramosa, Mirabilis pumila, Tridens muticus, and Marina orcuttii var. orcuttii.

In addition to collecting vouchers for each taxon, Zabriskie noted the elevational distribution for each taxon, recording the lowest and highest elevation, as well as noting whether each taxa was more or less continuously distributed with elevation or not.

He reported his work in the delightful and comprehensive book Plants of Deep Canyon and the Central Coachella Valley, California (1979, Boyd Research Center, UC Riverside, 175 pp.). His book gives insights into the plants within Deep Canyon, with detailed descriptions of the major plant communities as they change with elevation (see his map), and why they are here and distributed as they are. He also briefly mentions some species found in neighboring areas but are not present in Deep Canyon. I strongly recommend his book to any botanist wishing for a good read in botany.

I have updated his Species List to the Jepson Manual names and species circumscriptions as much as possible, and present the updated Species List separately. The rest of this page describes the changes I've made to his original list.


Zabriskie's 1979 list contains 619 taxa. Of those, 12 taxa have been deleted as separate entries because they were combined with other taxa, leaving a final total of 607 taxa. The following table details the combined taxa, some of which have interesting stories to tell thanks to Zabriskie's elevation data.

Combined Taxa

Most of the Zabriskie taxa here are now synonymous with the combined taxon name. The ones that are not are linked to an explanation below.

Combined Taxon NameZabriskie TaxaMin elevation (feet)Max elevation (feet)# Elevations*
Gnaphalium canescens ssp. thermaleGnaphalium thermale740074001
Gnaphalium wrightii440063009
Machaeranthera canescens var. ziegleriMachaeranthera canescens var. ziegleri680076008
Machaeranthera canescens var. canescens400040001
Cryptantha micranthaCryptantha micrantha ssp. micrantha032009
Cryptantha micrantha ssp. lepida340075008
Sambucus mexicanaSambucus mexicana480048001
Sambucus caerulea760084002
Lotus strigosusLotus strigosus30045008
Lotus tomentellus052009
Lupinus concinnusLupinus concinnus ssp. orcuttii60052009
Lupinus pallidus20052009
Quercus cornelius-mulleriQuercus turbinella ssp. californica270064009
Quercus dumosa390065009
Krameria erectaKrameria parvifolia var. imparata70053009
Krameria parvifolia var. glandulosa390053009
Eriogonum inflatumEriogonum inflatum10036009
Eriogonum inflatum var. deflatum10036009
Claytonia exigua ssp. exiguaClaytonia spathulata var. exigua570057001
Claytonia spathulata var. tenuifolia160016001
Mimulus guttatusMimulus guttatus100076009
Mimulus nasutus100076009
Nolina parryiNolina parryi390064009
Nolina wolfi250052009
* This column gives the number of separate elevations if the elevation range is not continuous. A continuous range is indicated by the value 9; a continuous range plus a disjunct single elevation or range is indicated by the value 8.

The taxa above fall into three different categories of the relationship of the elevation ranges of the Zabriskie taxa:

  1. Zabriskie taxa with substantially the same elevation range. Five of the 11 taxa have identical or nearly identical elevation ranges: Lotus strigosus, Lupinus concinnus, Krameria erecta, Eriogonum inflatum, and Mimulus guttatus, and another, Quercus cornelius-mulleri, has the range of one taxon contained within the range of the other.

    Four of the five are not even given as judgment-reserved taxa in the Jepson Manual, and hence the Jepson Manual authors considered these to be clearly-artificial splitting of these taxa. The fifth, Eriogonum inflatum, still has the two Zabriskie taxa listed as separate varieties in the Jepson Manual, but these varieties have been shown to be environmentally-deduced by the amount of available moisture. See further comments.

    The similar elevation range found by Zabriskie is strong supporting evidence that these taxa were artificially split, and the formerly-separate taxa were just extremes found in the species itself.

  2. Taxa that grade from one Zabriskie taxon at lower elevations to another Zabriskie taxon at higher elevations. Two of the 11 taxa fall into this category: Cryptantha micrantha and Nolina parryi, with Quercus cornelius-mulleri being gradational at lower elevations.

    All three of these taxa also do not have any judgment-reserved taxa in the Jepson Manual. The variation with elevation is thus probably an environmental effect.

    For example, Zabriskie says for Quercus cornelius-mulleri:

    The scrub oak (Quercus turbinella ssp. californica) is the most abundant oak of the upper plateau. Another scrub oak, Q. dumosa, is nearly identical to Q. turbinella and occurs with it on the upper plateau and on higher surrounding slopes. Q. dumosa, the green-leaved species, is more common coastward and Q. turbinella, the gray-leaved species, is more common on the desert side of the mountains. Where their ranges overlap, variability in the diagnostic characters makes differentiation between the hybridizing species somewhat arbitrary. The two species are often considered together in a description of the Deep Canyon scrub oaks.

    The difficulty in telling these "species" apart, as well as their "hybridization", are both now removed, since these plants are all now considered Quercus cornelius-mulleri, which was recognized in 1981, after Zabriskie published his flora. A key characteristic of this species is the dense minute overlapping white stellate hairs on the bottom of the leaf surface, which is a characteristic of both of the above Zabriskie taxa.

    Q. turbinella has a lower blade surface with glandular, yellowish hairs in addition to planar stellate hairs, and is found only from the e DMtns (New York Mtns) to Colorado and Texas. Prior to the definition of Q. cornelius-mulleri, plants of that species with gray upper leaf surfaces were assigned to Q. turbinella.

    Q. dumosa was a catch-all term for nearly all scrub oaks in southern California. It was clear from Zabriskie's description, as well as my observations of these same plants on the Cactus Spring and Sawmill Trails, that Zabriskie was referring to the green-upper-leaved form of Q. cornelius-mulleri, and not to one of the three other scrub oak species found in the Santa Rosa Mountains.

    It is recognized that the upper leaf surface for Q. cornelius-mulleri ranges from shiny green to dull gray-green. This color change is entirely due to the density of hairs on the upper leaf surface. If the hairs are scraped partially off the gray-green leaf, it becomes identical in color and shininess to the green leaf. Green-leaved forms have hairs covering less than ~25% of the blade surface; gray-leaved forms have hairs covering more than ~50% of the blade surface. Intermediate-colored leaves have hairs covering between ~25% and ~50% of the blade surface.

    A simple explanation for the variation in the upper leaf surface is again availability of water. Coastward, and at higher elevations, more water is available, and there is less need for the plants to have hairs to restrict the amount of sunlight falling on the upper leaf surface.

    Had the species Quercus cornelius-mulleri been recognized in 1979, Zabriskie would most likely have had only one taxon here in his flora. He was forced into shoehorning the plants into two species, since the far desert taxon Quercus turbinella has upper leaf surfaces that are dull gray-green due to hairs. The specimens with green shiny upper leaf surfaces had to go elsewhere!

  3. A Zabriskie taxon with a single disjunct occurrence of another Zabriskie taxon. Three of the 11 taxa fall into this category: Gnaphalium canescens ssp. thermale, Machaeranthera canescens var. ziegleri, and Claytonia exigua ssp. exigua. The Zabriskie Claytonia taxa are now synonymous; the other two taxa are discussed below.
    • Gnaphalium wrightii is separated from G. thermale in the Munz key on how decurrent the leaves are. In Munz, G. thermale has non-decurrent leaves, whereas G. wrightii has decurrent leaves. Munz stated it was uncertain whether G. wrightii is in California.

      The Jepson Manual does not give G. wrightii as being in California, and stated that G. thermale = G. canescens ssp. thermale has decurrent upper leaves and non-decurrent lower leaves. I suspect the degree of decurrence varies from plant to plant, and may vary with elevation. This was most likely the basis for Zabriskie thinking there were two taxa present here.

      The USDA Plants Database and the Flora of North America both treat G. wrightii as a synonym for G. canescens ssp. canescens, which, oddly, in the Jepson Manual is said to have leaves gen not decurrent. (Note that both of these references have moved this taxon to the genus Pseudognaphalium.) Presumably, plants formerly called G. wrightii have the exception to this trait.

      The Jepson Manual also provides a much cleaner key to separate out G. thermale = G. canescens ssp. thermale from G. canescens ssp. canescens based on the length of the heads. I measured the heads of plants on the Cactus Spring Trail, just below the elevation of 4400 feet given by Zabriskie for "G. wrightii", as 3-5 mm, exactly that of G. thermale, and not the 5-6 mm of G. canescens ssp. canescens.

      Thus I treat Zabriskie's G. wrightii at 4400-6300 feet as the same G. thermale he found at 7400 feet.

    • Machaeranthera canescens var. ziegleri is endemic to the Santa Rosa Mountains, and the population of plants here is strikingly different from var. canescens in the field. However, as is true for many subspecies and varieties, there is sometimes overlap in some characteristics with other subspecies and varieties found elsewhere.

      Var. ziegleri is a large plant, with stems from 1-12 dm, with large heads of ~10-20 mm, and is a long-lived perennial to a shrub. It grows in dry forests, as well as in open areas such as roadsides and in desert chaparral, at 4000 - 8100 feet. Note that the ranges for all the characteristics here are minimum ranges, since rare taxa are almost uniformly not well described in the floras for their true ranges due to their small number of vouchers. Var. ziegleri lines SR74 from Santa Rosa Summit at 5000 feet elevation east for a considerable distance down to at least roughly 4000 feet.

      In contrast, var. canescens is a small plant, with stems from 1-5 dm, with smaller heads of 6-14 mm, and never has woody stems. It grows in open areas in montane habitats, from 6500 to 10,000 feet, and wouldn't be caught dead growing along the desert-y roadsides at lower elevation where var. ziegleri grows.

      It surprised me greatly when I initially surveyed the high San Jacinto Mountain trails, since I expected to see var. canescens there from my previous experience in the San Bernardino Mountains and I found no M. canescens plants at all. It then surprised me even more when I saw my first var. ziegleri at the base of the pine forest at Santa Rosa Summit; I wasn't even sure it was a Machaeranthera canescens since it was so large!

      Zabriskie probably came across plants in a single location at 4000 feet that looked similar to var. canescens and keyed to that variety. Such plants in Deep Canyon might well be somewhat depauperate plants, since this is the lowest elevation at which either of these taxa are known, and this location is the farthest into the desert portion of the Santa Rosa Mountains for var. ziegleri. It is also possible he came across first-year plants of var. ziegleri, which would look a lot like var. canescens

      Such plants have also been found before amidst the higher-elevation var. ziegleri population. There are 16 vouchers of M. canescens from along the Santa Rosa Mountain Road to Toro Peak; 15 of them are determined as var. ziegleri and a single one is determined as var. canescens.

      Clearly, it is much more reasonable to treat such plants as being on the depauperate or young end of the var. ziegleri population than to treat them as a different variety.

      Munz, perhaps the greatest student of the southern California flora, botanized the San Jacinto and Santa Rosa Mountains heavily. Munz was also a clear splitter; he defined a number of taxa from the Santa Rosa Mountains that were not even considered judgment-reserved taxa in the Jepson Manual. Munz found only var. ziegleri here, and has var. canescens only in the San Jacinto Mountains northward. If there had been any var. canescens in the Santa Rosa Mountains, Munz would have noted it.

      Caveats: I haven't seen either of the vouchers that purport to be var. canescens; I'll try to review them in the future. Also, my only experience with the Santa Rosa Mountains Flora are visits in October and November of 2006, so I don't have familiarity with other areas that contain M. canescens.

Updated Determinations

The following Zabriskie taxa were considered misidentifications, and were updated to the following determinations:

Zabriskie TaxonUpdated TaxonComments
Arctostaphylos patula ssp. platyphyllaArctostaphylos parryana ssp. deserticumSee discussion below
Gnaphalium wrightiiG. canescens ssp. thermaleSee discussion above
Machaeranthera canescens var. canescensMachaeranthera canescens var. ziegleriSee discussion above
Quercus engelmanniiQuercus acutidensSpecimen does not have Engelmann hairs nor Engelmann acorns, although the leaves have the shape of Engelmann leaves.

Arctostaphylos parryana ssp. deserticum. This is a new taxon defined by Keeley et al in 1997. Zabriskie's plants clearly correspond to this taxa, as he writes:

A. patula ssp. platyphylla is not generally reported to have a burl, but the clones of manzanita on the montane slopes have large basal burls as well as rooting branchlets. These populations may be hybrids of A. glandulosa ssp. adamsii and A. patula ssp. platyphylla. The ecological consequences of the combined ability to crown-sprout and root by branches presents an attractive subject worthy of investigation.

It is mysterious to me why Zabriskie took these plants to be A. patula and not A. parryana since Arctostaphylos parryana ssp. deserticum is supposed to have non-glandular branchlets and inflorescences. I'll check his voucher sometime.

The following Zabriskie taxa were given only to the species level, and were updated to the following subspecies level, using online vouchers determined to this level and / or the distribution of the subspecies if there is no other possible subspecies:

Zabriskie TaxonAssigned Taxon
Acamptopappus sphaerocephalusAcamptopappus sphaerocephalus var. sphaerocephalus
Arenaria macradeniaArenaria macradenia var. macradenia
Chaenactis carphocliniaChaenactis carphoclinia var. carphoclinia
Ericameria cuneataEricameria cuneata var. spathulata
Eriophyllum ambiguumEriophyllum ambiguum var. paleaceum
Erysimum capitatumErysimum capitatum ssp. capitatum
Hymenoclea salsolaHymenoclea salsola var. salsola
Nemacladus glanduliferusNemacladus glanduliferus var. glanduliferus
Opuntia basilarisOpuntia basilaris var. basilaris
Stephanomeria exiguaStephanomeria exigua ssp. deanei
Stephanomeria paucifloraStephanomeria pauciflora var. pauciflora

Discarded Data

Only one Zabriskie observation was discarded - an observation that Helianthus niveus ssp. canescens was present at 8000 feet. If a specimen of this taxa was growing at 8000 feet, it most likely was a waif. This desert floor taxon is given as gen < 300 m in the Jepson Manual, and grows in open, sandy places, far different from the top of the Santa Rosa Mountains.

Notes On Some Taxa

The names in this section are the determinations given in the updated checklist.

Agropyron sp. I left this name the way Zabriskie gave it, since it is not clear how to update it. Most Agropyron species in Munz are now Elytrigia or Elymus.

Opuntia phaeacantha. Zabriskie gives Opuntia phaeacantha var. discata as the abundant prickly-pear cactus at middle elevations. If one believes the synonymy given in the Jepson Manual, this taxon would update to O. engelmannii var. engelmannii.

However, the six online vouchers for this taxon are all determined as O. phaeacantha. Hence I have chosen to go with that determination for now.

The determinations of similar prickly-pears in southern California are often incorrect; many of the Opuntia phaeacantha var. discata have actually turned out to be O. vaseyi. Interestingly, Lyman Benson, in his book The Native Cacti of California, places five O. vaseyi specimens in this location, and only two O. phaeacantha specimens here. I plan on checking the determination when these plants bloom in Spring 2007.

Quercus acutidens. The wrong elevation was plotted by Zabriskie. These plants are at an elevation of 3500 feet in Horsethief Creek, at the intersection with the Cactus Spring Trail. Zabriskie plotted this at 1700 feet elevation.

Checklist of the Vascular Flora of Deep Canyon

The checklist is given on a separate page:

Checklist of the Vascular Flora of Deep Canyon

Go to:

Copyright © 2006 by Tom Chester.
Permission is freely granted to reproduce any or all of this page as long as credit is given to me at this source:
Comments and feedback: Tom Chester
Updated 30 November 2006.