Determining the variety of Ericameria cuneata var. spathulata, wide-leaved rock goldenbush, in the San Jacinto Mountains
Introduction
This page reports our study to determine the variety of our plants of Ericameria cuneata.
We initially thought we had both variety cuneata and var. spathulata in the San Jacinto Mountains. The key to distinguish the two varieties in the Jepson Manual is:
26. Leaves sessile, blades wedge-shaped, largest 3-14(18) mm, 2-9(12) mm wide ..... var. cuneata26' Leaves stalked, blades spoon-shaped, largest (9)12-25 mm, 4-16 mm wide ..... var. spathulata
The leaf stalk is also called a petiole.
The key derives from an exhaustive survey of E. cuneata plants throughout its range done by Urbatsch in 1976. He did field surveys in the early fall of 1972 and 1973, collecting five to ten plants from each of many populations. In all, he examined almost 500 herbarium specimens.
The key on the shape of the leaves masks a considerable variation in petiole lengths and leaf shapes throughout the population of both varieties. Urbatsch shows leaf silhouettes depicting variation in leaf size and shape through the ranges of these two varieties in his paper. See also that same figure with leaves of the two varieties more clearly marked.
Initially, we did not use the leaf size, and only used the presence or absence of a petiole to distinguish the varieties. This is not a trivial observation in the field, since the leaves remaining on the stem are mostly crowded together, making it difficult to see the leaf bases and their shape. Most of the time, careful study of the plant could find some viewing angles in which the leaf base could be noted. If any petioled leaves were found, we called it var. spathulata. If no petioled leaves were found, we called it var. cuneata.
From a number of field observations, it appeared to us that we had two different varieties, since a number of plants appeared to have sessile leaves that were wedge-shaped at base, and others appeared to have petioled leaves that were spoon-shaped (spatulate); see Figs. 1 and 2 below, and compare them to the leaf shapes of the two varieties from Urbatsch 1976. Furthermore, there appeared to be geographic separation of the varieties, with plants that appeared to have sessile leaves found at higher elevations on the coastal side of the San Jacinto Mountains. Plants that appeared to have petioled spoon-shaped leaves were found at lower elevations on the coastal side, as well as at all elevations along the Desert Divide on the desert side of the San Jacinto Mountains. This geographical distribution was as expected from Urbatsch 1976.
Everything seemed nice and consistent, and we were happy with our determinations of the varieties, although we did wonder if our two varieties intergraded at times.
Recently, we were working on a separate study to note which species were found at higher elevation in the San Jacinto Mountains than was given in the Jepson Manual. That study is still in progress, but we have been somewhat amazed at how often the peak elevation found in our survey is very close to the highest elevation in the Jepson Manual.
It was therefore a surprise when we found that our highest var. spathulata, at 8110 feet elevation on the Skyline Trail, was a whopping 1880 feet above the highest elevation of 6230 feet given in the Jepson Manual. Since var. cuneata was given as occuring up to 9200 feet elevation, our first thought was that we might have misdetermined the variety on the Skyline Trail. But we hadn't; our pictures showed that plant was a very clear var. spathulata from its petioled spoon-shaped leaves; see Fig. 1.
Fig. 1. Photograph of a var. spathulata at 8110 feet elevation. Note the leaves are clearly petioled and spatulate, and not wedge-shaped at the base. The red circles highlight some of the leaves with clear petioles. Click on the picture for a version showing more area.
We then checked our photographs of what we called var. cuneata to see how different it was from var. spathulata. The leaves on one plant on the Cedar Spring Trail appeared to fit var. cuneata precisely, with a wedge-shaped base with no petiole; see Fig. 2.
Fig. 2. Photograph of what appeared to be a var. cuneata at 6200 feet elevation on the Cedar Spring Trail. The red circles highlight some of the leaves with wedge-shaped bases showing no clear petioles. Click on the picture for a version showing more area.
To nail the determinations, and see how well the varieties were separated in the San Jacinto Mountains, we decided to do two special field studies in which we also sampled the largest leaves we could find on each plant, in order to use the second part of the key to distinguish the varieties. We also did a study of the wonderful dataset of 63 E. cuneata plants from the San Jacinto Mountains area at iNaturalist.
Field Studies and Results
Field Study 1 on 2 November 2020
The first field study was on 2 November 2020 where we studied plants from four areas, two of which had plants that we had determined as var. cuneata, and two of which had plants that we had determined as var. spathulata.
We sampled plants in four areas separated by 10 miles as the crow flies, at elevations from 4500 to 6130 feet. From west to east, the locations were:
- Pine Cove, Dead Man's Curve on SR243 at 5940 feet elevation.
- The east bank of Idyllwild Creek, opposite mile 0.9 of the Deer Springs Trail, at 6130 feet elevation.
- Idyllwild, on the Ernie Maxwell Trail at 6000 feet elevation.
- Quinn Flat in Garner Valley at 4500 feet elevation.
These plants were mostly in full flower, making them directly comparable to the ones in Urbatsch's study.
To our great surprise, using the largest leaves on the plants, it turned out that plants from all four areas were all clearly var. spathulata, with distinct petioles and wide, spatulate leaves.
Fig. 3 shows the largest leaves found on the plants in the four separate areas. All leaves are spatulate in shape. All leaves except one or two show a clear, albeit short, petiole.
| ||
| ||
| ||
| Fig. 3. The largest leaves from plants in four different locations. Two plants were sampled at Quinn Flat. See also the original picture showing all leaves at once. | ||
The length (including petiole) and width of the leaves shown in Fig. 3 is plotted in Fig. 4, which also shows the normal ranges for the two varieties given in the key (i.e., the range doesn't include the rarer values in parentheses in the key). All the measured leaves fit perfectly in the normal range for var. spathulata, and are clearly inconsistent with any being var. cuneata.
Fig. 4. Plot of leaf length vs. width for all the leaves shown in Fig. 3. The normal range for each variety, for the largest leaf on a plant, is also shown. All the measured leaves fit perfectly in the normal range for var. spathulata, and are clearly inconsistent with any being var. cuneata.
Field Study 2 on 6 November 2020
Given the shocking result from the first day of fieldwork, we reviewed all the photographs we had of plants we had called var. cuneata. A careful examination of those photographs revealed at least some leaves with petioles for all except the plant shown in Fig. 2, from the Cedar Spring Trail.
The second field study was on 6 November 2020, devoted to the plants on the Cedar Spring Trail, to see if they were actually the var. cuneata they appeared to be. We sampled six plants from four distinct but closeby locations on the trail, at elevations from 6040 to 6350 feet, separated by just 0.15 miles as the crow flies.
We faced an immediate problem in comparing the largest leaves from these plants with the values for the two varieties in Urbatsch, since these plants were all in fruit and rapidly dropping their older, largest leaves. Most plants did not have the dropped leaves at their base, being presumably blown away by wind events. Hence it should be kept in mind that the values for the largest leaf on each plant may be smaller than the value that would have been found if the plants had been sampled while in bloom.
Looking at photographs of the plants from the first field study, the shape of the leaves appears to be unaffected by the loss of the largest leaves.
The variety of these plants were less clear. The petioles on their leaves were significantly shorter, and the leaves were narrower, than in the plants surveyed on 2 November 2020. The measured leaves do not fit either variety well, falling mostly in the region of overlap of the two varieties. Although that intermediacy might be due to the dropping of the older, largest leaves, the leaf shapes are mostly intermediate as well, not fitting either variety well.
Fig. 5 shows the largest leaves found on the six plants.
| ||
| ||
| ||
| ||
| ||
| ||
| Fig. 5. The largest leaves still present on six plants on the Cedar Spring Trail. The lowest elevation plant was at GPS #1; the highest at GPS #4. The broken leaf at top left was measured before it was broken. | ||
The length (including petiole) and width of the leaves is plotted in Fig. 6, which also shows the normal ranges for the two varieties given in the key, and shows the largest leaf from each plant in the first field survey.
Fig. 6. Plot of leaf length vs. width for the longest leaf from each plant shown in Fig. 5 (Cedar Spring Trail plants) and for the longest leaf from each plant shown in Fig. 3 (Field Study 1 plants). Values for the largest leaf on each plant for the Cedar Spring Trail plants may be smaller than the values that would have been found if the plants had been sampled while in bloom; see the text. The normal range for each variety, for the largest leaf on a plant, is also shown.
Taken by themselves, from this plot alone without comparing to the plants from Field Study 1, it would be impossible to say which variety the plants on the Cedar Spring Trail, since they fall in the intermediate zone between the two varieties. But it is highly likely that these plants are just the "smallest leaf part" of the var. spathulata in the San Jacinto Mountains. The points from these plants are very close to one point from Field Study 1, indicating they are part of the same population. Also, remember that these plants were sampled when the largest leaves at flowering time had probably fallen off. If we were able to sample these plants earlier in the year, all the points might move to larger values, placing them inside the distribution of the points from Field Study 1 and completely within the box for var. spathulata.
iNaturalist study
iNaturalist has a very rich dataset with 63 observations with pictures of our plants, nearly all of which show the leaves well enough to determine how well they fit var. spathulata. The largest contributor to iNat was Colin Barrows, with 28 observations, followed by Don Rideout with 8 observations, and "dalemeister" with 5. 17 other people contributed 1 or 2 observations each.
The vast majority of the iNat observations are clearly var. spathulata, with petioled wide leaves, that all looked the same. Five observations showed plants similar to those on the Cedar Spring Trail. They were at similar elevations and exposure, all from the PCT north of SR74 except for one on the Skyline Trail, at about the same elevation. Interestingly, two of them had both the narrow leaves of the Cedar Spring Trail plants and normal spatulate wide leaves on the same plant: one by Chris Weber and one by "larry-heronema". There must be something about this elevation and habitat that causes this leaf variation.
Don Rideout also had an observation with the longest petioles by far of any of our plants. This plant was just 0.1 mile from the Dead Man Curve's plant shown in Fig. 3, growing in a shadier habitat.
Conclusions
Taken as a whole, the plants in the San Jacinto Mountains fit var. spathulata best. The vast majority of our plants have leaves that are clearly petioled, with a spatulate shape that matches the shape only of var. spathulata. However, a small number of plants from the Cedar Spring Trail, the nearby portion of the PCT, and the Skyline Trail, at about the same elevation in each area, have leaves that appear to be intermediate between the two varieties, but which are probably just the "smallest leaf part" of our var. spathulata plants. We have no specimens that clearly fit only var. cuneata.
The presence here of some specimens intermediate between the two varieties is similar to what Urbatsch found for plants in the Tehachapi and Greenhorn Mountains in Kern County, which he attributed to a habitat where "their usually distinct habitats gradually intergrade". He writes "var. cuneata occurs at higher elevations, [and] is adapted to cooler, more mesic habitats", whereas var. spathulata is "adapted to xeric areas". The habitat in the San Jacinto Mountains is indeed intermediate between the more mesic habitat of the Sierra Nevada where var. cuneata is best expressed, and the more xeric habitat in the mountains to the east such as Joshua Tree National Park.
Having some specimens with intermediate leaf sizes and shapes does not imply that there are two taxa here. The leaf properties of both varieties of E. cuneata vary significantly across the range of each variety, probably driven by evolution of the plants in each area creating local forms. Also, since those "intermediate" specimens here are surrounded by the other good specimens of var. spathulata, it is extremely unlikely they are in actuality a different taxon, breeding only with themselves. Instead, they are almost surely just one end of the distribution of leaf characteristics of the population as a whole here. We therefore call all of our plants var. spathulata.
This assignment agrees with Hall's Compositae of Southern California 1907, which places only var. spathulata in Southern California.
There is conflicting information in Urbatsch 1976 as to whether he places var. cuneata in Riverside County. In his taxonomic treatment of var. cuneata in words, he doesn't place it south of the San Gabriel Mountains in Los Angeles County. However, in his geographic distribution plot of vouchers of the varieties, he places two locations in western Riverside County, one that plots just north of Temecula, and one that plots in the San Jacinto Mountains. The Temecula location must be a badly georeferenced voucher, since there are no known E. cuneata plants near there. It is possible the other voucher is similarly poorly georeferenced, or simply a specimen on the borderline between the two taxa that he chose to call var. cuneata.
There are only four vouchers of var. cuneata from Riverside County online at the Consortium of California Herbaria on 7 November 2020. Two of them are by Hall from 1901, and should have been determined as var. spathulata after Hall's monograph in 1907. Hall called these just to the species name in his 1902 Flora of the Pine Belt of San Jacinto Mountain, and often plants determined only to the species name get assigned to the nominate species later by people who assumed such plants would otherwise have a non-nominate subspecies name.
The other two vouchers are older vouchers from known locations of var. spathulata, one "near Tahquitz Peak", and one from "Pine Cove" that has a duplicate voucher determined as var. spathulata. There are 13 vouchers determined as var. spathulata from the San Jacinto Mountains area.
One reader asked an excellent question about how reliable the variety separation can be just using leaf characteristics, since leaves in many species vary tremendously from environmental causes (sun leaves vs. shade leaves; wet areas vs. dry areas; orientation of the habitat, such as west-facing vs. east-facing, etc.) We are certainly not experts in the variation of this species across its range. Here is what Urbatsch wrote about that:
On the basis of habitat, leaf shape, and to some extent leaf size, two taxa can be delimited among the aforementioned populations. One occurs at higher elevations, is adapted to cooler, more mesic habitats, and has smaller, cuneate, sessile leaves; the other has larger, spatulate, petiolate leaves and is adapted to xeric areas.
These taxa are fairly distinct throughout most of their ranges, but they intergrade in areas of sympatry.
Distribution in the San Jacinto Mountains
The locations of E. cuneata plants in the San Jacinto Mountains is shown in Fig. 4. Only points with accurate locations are shown; 57 locations from iNaturalist; 36 locations from our GPS points; 29 locations from vouchers with fairly-specific localities; and 19 points from Calflora GPS points, mostly from Keir Morse.
Fig. 4. Geographic distribution of all known accurate locations of E. cuneata in the San Jacinto Mountains and nearby.
E. cuneata is a widespread species, seemingly found anywhere that has its habitat of rocky outcrops, not only here in the San Jacinto Mountains but throughout its range from the northern Sierra Nevada and Coast Ranges to central Baja California and southeastern Arizona.
The elevation of these locations is plotted against longitude in Fig. 5 at approximately the same longitude scale, making it fairly easy to associate points in the elevation plot with points in the geographic map. See also larger version of Fig. 5 with groups of points labeled by location.
The elevation was derived from the bulk data query of the National Map from the USGS.
Fig. 5. Elevation vs. longitude plot of all known accurate locations of E. cuneata in the San Jacinto Mountains and nearby. See also larger version of this plot with groups of points labeled by location
The elevation distribution is essentially continuous from 4000 to 9000 feet, with a small number of points between 1470 and 4000 feet elevation.
There is no obvious indication of two taxa in either the geographic map or the elevation plot. Gaps are usually clearly due to unsurveyed areas, or areas with unsuitable habitat.