Plants of Southern California: Heuchera species

Type localities and papers:

The number of Heuchera species in southern California has been declining with time, but not because of species extinction. Instead, taxonomists have decided that some of the previously-defined species were spurious, artificial splits of a single species. The Jepson Manual Second Edition has seven species, down from nine in the First Edition, by making two of the previous species synonymous with other species.

There are a number of taxonomic problems within Heuchera, mostly because the species all basically look alike. As a result, botanists must use subtle characters to distinguish many of the species, many of them literally just differences of a mm or so, which creates a tendency to define taxa based on spurious differences. For example, in western North America, 12 different taxa have been defined from what is now considered a single species, H. rubescens, using "almost every minor variant in morphology, often without correlation to any pattern of geographical distribution".

Of course, just because humans can't easily see or define the difference between species, it doesn't mean that the species don't exist. As always, botanists try to do the best with what we have. It will be interesting to see what molecular studies reveal about these species. Perhaps there are a multitude of cryptic species in the genus, and we'll end up with more species than we have now. Or perhaps we still have artificial splits present in our current species complement, and we'll end up with as few as four species.

Of the seven species of Heuchera in southern California, many are so similar that even vouchers of the same species are sometimes determined as another species. We can only reliably place the cismontane southern California plants into four well-defined groups, and it is problematic going further than that.

This page details the four well-defined groups, and shows thumbnail pictures of each species, as well as their geographic and elevational distribution. Click on the thumbnail pictures to get a larger version of any of them.

The species accepted in the Jepson Manual Second Edition (JM2) treatment are given in Table 1, along with their ranges and synonymy. (The elevation ranges come from vouchers, and are somewhat different than the Jepson Manual Ranges.)

Table 1. The Heuchera species of cismontane southern California

Species NameRange in soCalElevation (feet)Synonyms# Vouchers
H. abramsiiSnGb8,500 - 10,000 56
H. brevistamineaLaguna Mountains4,400 - 6,000 20
H. caespitosaTransverse Range (uncommon SnBr)3,300 - 9,000H. elegans171
H. hirsutissimaSnJt, Santa Rosa Mountains7,000 - 10,840 44
H. maximan Channel Islands250 - 700 24
H. parishiiSnBr4,400 - 11,400H. alpestris139 (150?)
H. rubescensPR (+SnBr?)4,000 - 6,400(+?)H. leptomeria29 (17?)

The uncertainty in the elevation and range for H. rubescens is due to possible confusion with H. parishii. Even though the Jepson Manual places only H. parishii at SnBr, there are 11 vouchers determined as H. rubescens from SnBr. There is even one voucher from the top of Mt. Baldy at SnGb.

Although the JM2 synonymized H. elegans with H. caespitosa, there are good arguments to keep them separate. I wrote this page before I decided to keep them separate, so even though they are combined here, be aware that I call the plants in the San Gabriel Mountains H. elegans, and the photographs below called H. caespitosa are actually of H. elegans.

The number of vouchers for each species is after the tweaking I did to some of the voucher determinations, as detailed below.

Our seven species can relatively easily be placed into four different groups, using two discriminants that are visible in most photographs:

Fig. 1 shows the difference in how exserted the stamens are for these two categories, as well as narrow vs. wide petals:

Exserted stamens / Narrow Petals
(H. rubescens shown)

Included stamens / Wide Petals
(H. caespitosa shown)

Fig. 1. In the pictures on the left, from Tom Chester, note how it is easy to see the stamens in both the frontal and side views, since they extend well past the tips of the calyx lobes. In the pictures on the right, from Michael Charters, the stamens can only be seen on the frontal view, since they do not extend past the calyx lobes. On the left, the petals are very narrow for most of their length, and widen only gradually toward their tip. The base of the petal at lower left is completely hidden by the filament (the stalk of the stamen). In contrast, the petals on the right have an obviously-widened distal portion that is much wider than its proximal (lower) portion and the filaments.

Using these two discriminants, the seven species can be divided into four groups as follows:

Two groups consist of a single species, which are therefore uniquely identified with those two discriminants, H. abramsii and H. maxima.

The other two groups contain multiple species, and it is very difficult, if not impossible, to discriminate the species within each group from normal photographs. There is even some question as to whether the species within each of those groups are distinct entities.

The group with stamens exserted and petals narrow, which now contains two species, previously contained a third species in Munz 1974, H. alpestris, which was considered a reserved-judgment taxon in the Jepson Manual First Edition under H. parishii, and not even mentioned in the Jepson Manual Second Edition.

There is an argument that plants called H. parishii are just the form of H. rubescens in the San Bernardino Mountains (SnBr).

Another interesting possibility was mentioned in the Jepson Manual First Edition, that H. parishii might be of hybrid origin in the distant past, from H. rubescens and possibly H. hirsutissima. If so, that might account for why SnBr is missing the analog of H. caespitosa / H. hirsutissima at high elevations found at SnGb / SnJt.

If H. parishii is distinct from H. rubescens, these two species can perhaps be distinguished in pictures by their petal length relative to the stamen length. H. rubescens has long petals 3.5 to 6 mm in length, extending well past the calyx lobes, that are longer than the stamens. H. parishii has short petals just 2 to 3 mm in length, barely extending past the calyx lobes and shorter than the stamens. However, in the Flora of North America treatment, the petals of H. rubescens are said to be 2-3(-6) mm, so the usual version of the very widespread H. rubescens has petals the same length as H. parishii.

Other differences are even more subtle, slight differences in the petal widths, how equal the petals are, how symmetric the hypanthium is, and how lobed the leaves are. Although the pictures below of H. parishii (of a high-elevation plant formerly called H. alpestris) appears to show different leaf glands and lobe tip shapes, other pictures of H. parishii have leaves that look just like those of H. rubescens.

The Flora of North America treatment of H. rubescens has an interesting comment about how almost every minor variant in many characters has been defined as a different taxon, and it may well be that H. parishii is one of those.

If you want to distinguish H. parishii from H. rubescens, the easiest solution is to just use geography, and call all plants in this group that are found at SnBr H. parishii, and all plants in this group found elsewhere H. rubescens. However, I note that there are 11 vouchers at SnBr that are called H. rubescens, compared to 139 such vouchers that are called H. parishii. So ~7% of the time, plants at SnBr apparently fit the key for H. rubescens better than H. parishii. There are even four vouchers outside of SnBr determined as H. parishii.

The three species with stamens not exserted past the calyx lobes, with wide petals, are also hard to distinguish, with even more subtle characters used for discrimination. This group also used to include an additional species, H. elegans, that was combined with H. caespitosa in the Jepson Manual Second Edition. H. brevistaminea has stamens reaching only to the midpoint of the calyx lobes; the other two species have stamens reaching to the tip of the calyx lobes. Of those two species, H. caespitosa has all five of its calyx lobes well developed, whereas some of the calyx lobes of H. hirsutissima are much shorter than the others. None of these characteristics are visible in most photographs, so it is impossible to use photographs to distinguish these species based on those characteristics.

Every flora states that this is a difficult complex which needs further study, which means even the experts are not sure about the species delineations here.

What is especially troubling about using the stamen length difference to distinguish species is that the Flora of North America Heuchera treatment says that the Stamens and styles continue to elongate during flowering, and hypanthium, ovary, and fruit dimensions change over time.

Once again, the easiest solution is to use geography. H. caespitosa is found in the Transverse Ranges (uncommon at SnBr); H. hirsutissima is found at SnJt; and H. brevistaminea is found in the Laguna Mountains of San Diego County.

Pictures of all the species are given below, sorted into the four groups.

Stamens exserted, petals narrow: H. parishii, H. rubescens

H. rubescens

H. parishii

Fig. 2. Pictures on the left of H. rubescens, taken by Tom Chester on 28 May 2013 in Strawberry Creek Gorge, Idyllwild, San Jacinto Mountains. Pictures on the right of H. parishii, taken by Thomas Stoughton on 10 July 2009 at Dollar Lake, San Bernardino Mountains, and are online at Calphotos. This high-elevation plant would formerly have been called H. alpestris. These pictures appear to show that the petals are longer than the stamens for H. rubescens, and perhaps somewhat shorter than the stamens for H. parishii. This may or may not be a reliable discriminant; see the text.

Stamens exserted, petals wide: H. maxima

H. maxima

Fig. 3. Pictures of H. maxima taken by Stan Shebs in April 2007 at the Regional Parks Botanic Garden, Berkeley, California, from Wikimedia Commons.

Stamens not exserted past calyx lobes, petals narrow: H. abramsii

H. abramsii

Fig. 4. Pictures of H. abramsii. The most unique characteristic of this species is the near absence of the long hairs in the inflorescence, with just a few long hairs on and near the calyx lobes. Top two pictures are crops from a picture taken by "stickpen" on 23 April 2009 from a labeled plant at Rancho Santa Ana Botanical Garden. Bottom picture is from Denver Botanic Gardens, used with their permission.

Stamens not exserted past calyx lobes, petals wide: H. brevistaminea; H. caespitosa; H. hirsutissima

H. brevistaminea

H. caespitosa

H. hirsutissima

Fig. 5. These three species are only distinguished by fairly minute differences, which are difficult, if not impossible, to show in photographs. H. brevistaminea is distinguished by having somewhat shorter stamens. In the middle picture on the right, you can see that the stamens reach to the tip of the calyx lobes for H. hirsutissima, which is also supposed to be the case for H. caespitosa. The stamens for H. brevistaminea only reach to the middle of the calyx lobes.

The other two species are distinguished by the calyx lobes. H. caespitosa has all five of its calyx lobes well developed, whereas some of the calyx lobes of H. hirsutissima are much shorter than the others. It is difficult to show that difference in pictures in the field since it is hard to show all the calyx lobes with the same perspective at the same distance from the camera, and without obscuration from the petals and hairs. The above pictures do not show that difference. I will try to get a photograph showing that difference in the future.

Pictures of H. brevistaminea are from Patrick Alexander, taken on 1 June 2009 from Stephonson Peak in the Laguna Mountains, at SEINet. Pictures of H. caespitosa were taken by Michael Charters in May 2013 from the PCT from Three Points to Cloudburst Summit in the San Gabriel Mountains. Pictures of H. hirsutissima were taken by Michael Charters at San Jacinto Mountain and Toro Peak in (2011?).

The geographic distribution of our seven species in shown in Fig. 6:

Fig. 6. The geographic distribution of each species in cismontane southern California. Filled black circles are voucher locations obtained from the Consortium of California Herbaria on 3 June 2013. I have outlined the geographic distribution for each species, using mostly the Jepson Manual Second Edition ranges, as well as my knowledge of where the species are found at SnJt and in the PR, ignoring some of the voucher determinations when they are suspect. In particular, H. rubescens is found at the lower elevations of SnJt, even though no vouchers are plotted there. The labels are in black text except font colors corresponding to the outline color are used where it would be ambiguous if black text were used. I have no idea whether the concentration of H. rubescens vouchers shown amidst the H. parishii vouchers at SnBr are correctly determined or not, or even if these two species are distinct entities.

A plot of elevation vs. longitude in shown in Fig. 7, and of latitude vs. elevation in Fig. 8. Since some vouchers are misdetermined; some vouchers are badly georeferenced; and some vouchers have poor elevations, not every point in these plots is trustworthy. For these plots, I've changed the determination of ten vouchers to align them with the Jepson Manual ranges or my knowledge of the species at a particular location (three vouchers of H. abramsii to H. brevistaminea; one voucher of H. abramsii to H. caespitosa; two vouchers of H. brevistaminea to H. rubescens; three vouchers of H. rubescens to H. hirsutissima, and one voucher of H. parishii to H. hirsutissima).

The points of H. hirsutissima below 7000 feet elevation may actually be H. rubescens, and the determination of those vouchers should be rechecked.

Fig. 7. Elevation vs. longitude for each species in cismontane southern California, as determined from vouchers, with only a little review yet of whether the voucher is correctly determined, and no review of whether each voucher has reliable georeferencing. The longitude is on the x-axis so that west corresponds to the left in this plot. SnJt and SnBr have similar longitudes, and hence their data points overlap in this plot.

Fig. 8. Latitude vs. elevation for each species in cismontane southern California, as determined from vouchers, with no review yet of whether the voucher is correctly determined, or has reliable georeferencing. I've put the elevation on the x-axis in this plot so that north corresponds to up in this plot.

Text by Tom Chester; Pictures by Tom Chester, Michael Charters, Denver Botanical Garden, Thomas Stoughton, "Stickpen", Stan Shebs.

This page has been significantly improved by discussion with Jane Strong.

Go to:

Copyright © 2013-2018 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
Last update: 6 June 2013 (note on H. elegans added on 22 May 2018)