Plants of Southern California: Stephanomeria virgata
Introduction and Preliminary Conclusions
In Southern California, Stephanomeria virgata used to be a well-understood, easily-recognized species which matched the description in A Flora of Southern California by Munz quite well. In particular, Munz stated that the number of flowers per head was quite variable, ranging from 4 to 15, so one never came across a specimen that had the "wrong" number of flowers for that id. (Note for beginners: the number of flowers per head for this species is the same as the number of "petals" seen in a single "flower", since in the Asteraceae, each head is a group of many individual smaller flowers.)
In 1972, Gottlieb realized that there were two distinct species that had both been classified as S. virgata ssp. virgata since their gross morphology is identical to the eye. He named the new species, found at Torrey Pines State Reserve in the city of San Diego, S. diegensis, and suddenly the properties of S. virgata became a bit murky. How much of the characteristics reported in Munz for S. virgata actually came from the S. diegensis population?
The main discriminant between these two species is the existence of a narrow, longitudinal groove on each of the five sides of the 2-3 mm seed, something not noticed at all by the casual observer. See a Photograph of Stephanomeria achenes, with the caption given here.
The existence or not of the groove had long been known. S. virgata ssp. pleurocarpa had been defined as having achenes lighter colored, not grooved; pappus deciduous to discriminate this subspecies from ssp. virgata (Davidson and Moxley, Flora of Southern California, 1923, p. 355). Hence Gottlieb was separating out S. diegensis from S. virgata ssp. virgata, if one accepted those two subspecies. However, Munz did not recognize those as valid subspecies.
The number of flowers per head for S. diegensis was said to average 11-13, raising the possibility that all the former S. virgata with a large number of flowers per head were actually S. diegensis.
In 1993, the Jepson Manual recognized S. diegensis, but stated that the number of flowers per head was 6-8. Furthermore, the JM also resolved S. virgata into two subspecies, one with 5-6 flowers per head, ssp. pleurocarpa, and one with 8-9 flowers per head, ssp. virgata.
This muddied the water not only for S. diegensis, but for S. virgata as well. What happened to all the plants with 10-15 flowers per head? Was it really true that plants with 7 flowers per head perforce had to be S. diegensis? How does a poor botanist assign a subspecies to a S. virgata specimen that exhibits a range of 5-9 flowers per head?
I have begun a program to sort out what is going on with these two species, along with S. exigua, in Southern California. My analysis is not yet definitive, since I have so far only analyzed in detail seven locations in Southern California. Five of those locations are in the PR: the Santa Rosa Plateau; Agua Tibia Mountain; the Santa Margarita River Trail in Fallbrook; Monserate Mountain in Fallbrook / Bonsall; Old Highway 395 in Bonsall / Escondido near Lilac Road; and the Blue Sky Ecological Reserve / Lake Poway / Mt. Woodson area in Poway. The sixth location is Torrey Pines in SCo. These locations span a distance of ~40 miles from north to south, and 0 to 30 miles from the coast. See a plot of the geographical distribution of this species and closely-related species.
My preliminary conclusions are:
- Many specimens formerly thought to be S. virgata are actually S. diegensis, and the range of properties of S. diegensis is much larger than reported in the JM.
In particular, I find no evidence of any S. virgata or S. exigua ssp. deanei at Torrey Pines despite their presence on the plant list for Torrey Pines; every specimen I have examined there is S. diegensis. See Stephanomeria diegensis for more information.
Further, S. diegensis seems to be the widespread cismontane species in Southern California. The only location for which I have found S. virgata is the farthest inland location, Agua Tibia Mountain, 30 miles from the nearest coastline. Every specimen at all the other locations I've examined is S. diegensis or S. exigua ssp. deanei. Those locations range from the coast to 16 miles inland from the nearest coastline.
It is interesting that Arthur Gibson and Barry Prigge report the same findings for the Santa Monica Mountains in the Spring 2003 newsletter of the Mildred E. Mathias Botanical Garden:many, if not all, collections of the "widespread Stephanomeria virgata" [in the Santa Monica Mountains] are S. diegensis
In particular, it is possible, though by no means proven, that most "S. virgata" specimens in all of Southern California with more than 9 flowers per head are actually S. diegensis.
- The number of flowers per head is exceedingly variable for each of these three Stephanomeria annuals in Southern California, and in general is poorly described in the floras:
- For S. exigua, Munz says "flowers mostly 5" and the JM says "flowers 5-6". I find that only 32% of the heads have 5 flowers, with 28% of the heads having 7-8 flowers. See Stephanomeria exigua ssp. deanei for more information.
- For S. diegensis, I find a range of 5-15 flowers per head, not the 6-8 of the JM or the average of 11-13 of Gottlieb. See Stephanomeria diegensis for more information.
- For S. virgata, I find an almost uniform distribution of 5-8 flowers per head, with no evidence of any separation into two subspecies. See below for the data.
I speculate that the reason the floras are so poor in describing the flowers is due to the intrinsic variability of these species, coupled with reliance on herbarium vouchers, which results in poor statistics from a sampling of too few heads to derive the intrinsic distribution.
A number of factors conspire to limit the number of heads available from herbarium vouchers.
First, my data reported above were derived from field surveys of hundreds of heads on multiple days in multiple locations. In the field, I can take advantage of hundreds of specimens in a given area, whereas it would be unusual for a herbarium to contain more than one or two samples from a given area.
Second, I can survey all the flowers on the many branches of a 30 dm specimen, whereas at best an individual herbarium sheet will contain 10 dm of a single branch.
Third, individual Stephanomeria plants are notorious for having only a few open flowers at once, and for their habit of closing their flowers immediately after being collected. Thus the number of open flowers that can be easily surveyed in a voucher is reduced from what would have been observed in the field, unless the specimen was pressed immediately in the field.
Thus any investigation using herbarium specimens will have a small sample, probably measured only in the tens.
Furthermore, I have found that different locations can vary markedly in the average number of flowers per head. Hence it is quite important to sample a number of locations in a given area, which is not often done for vouchers.
- The existence of two separate subspecies for S. virgata seems questionable. I find no evidence to support the existence of two subspecies for S. virgata at Agua Tibia Mountain. Furthermore, it is easy to find examples online that contradict the claimed geographic locations of those two subspecies in other areas of California as well. See below for the data.
The Two Subspecies of S. virgata and Their Geographic Distribution
The Jepson Manual defines two subspecies, pleurocarpa with 5-6 flowers per head and appressed outer phyllaries, and virgata with 8-9 flowers per head and reflexed outer phyllaries. (Henceforth, I sometimes refer to this simply as the number of flowers, meaning the number of flowers per head.)
The geographic distribution of these two subspecies as stated in the JM is curious:
Area Subspecies present CA-FP except the areas below only pleurocarpa (5-6 flowers) PR both SCoRO, SCo, TR only virgata (8-9 flowers)
Hence ssp. pleurocarpa is the only subspecies found in northern and central California, except it is replaced by ssp. virgata in the Outer South Coast Ranges. Subspecies pleurocarpa is additionally found disjunctly in the PR. Subspecies virgata is supposedly the only subspecies found in Southern California, except it shares the PR with that disjunct population of ssp. pleurocarpa, and has the additional disjunct population in the Outer South Coast Ranges. Such a patchwork distribution of two subspecies is unusual.
It is easy to find online pictures of plants with 8 to 9 flowers per head from areas that are supposed to have only ssp. pleurocarpa with 5-6 flowers per head:
- 9 flowers per head in SnJV from the Kaweah River Delta Region near the cities of Farmersville, Exeter, Tulare, Visalia, and Woodlake, California
- 8 flowers per head in SnFrB from Henry Coe State Park.
In addition, Walker Pass in the Sierra Nevada, s SNH lists ssp. virgata in their plant list.
These three areas do not even adjoin regions that are supposed to contain ssp. virgata.
While not definitive, the ability to easily find examples of ssp. virgata in regions that are supposed to contain only ssp. pleurocarpa casts doubt upon at least the geographic segregation of the species.
The Number Of Flowers Per Head At Agua Tibia, PR
As part of a program to understand the characteristics of the Stephanomeria diegensis, exigua and virgata population in Southern California, I collected data on the number of flowers per head, keeping track of the range for individual plants, on the Dripping Springs Trail at Agua Tibia in the Palomar Mountain range (DS) on 17 November and 3 December 2003. I examined the phyllaries for glandularness on both trips, and collected fruit from 12 different locations along the trail for detailed examination on 3 December.
There is a large Stephanomeria population on this trail, consisting of literally hundreds of plants from mile 0.54 to at least mile 4.01. This large population in 2003 was probably due to the significant trail clearing that occurred in spring 2001 and spring 2002. Since no annuals even germinated in the 2001-2002 drought, the burst of annuals happened in 2003. This abundance is even more amazing since in 66 days of field work from September 1994 through May 1999, Banks (1999; A Vascular Flora of the Agua Tibia Mountains) reported no Stephanomeria on any part of this trail.
I found no S. diegensis on either trip. S. exigua ssp. deanei was found in at least six separate locations along the trail, from mile 0.57 to at least mile 4.01, with a population on the order of one hundred plants in total. S. virgata essentially lines the trail in many locations, from mile 1.04 to at least mile 3.96, although it apparently avoids the areas where S. exigua is found. The population of S. virgata numbers at least several hundred.
Both species were past their peak bloom on both observation days, with at least half of the plants totally finished flowering. The plants were also far from being completely finished, with about one hundred heads observed in bloom on each field day.
There were no problems distinguishing the two species at a glance, even though the S. exigua specimens on the trail ranged up to 17 dm, much taller than the 6 dm reported in the floras. As per the JM key to distinguish these two taxa, S. exigua ssp. deanei blooms in an open panicle, with obviously peduncled heads, and inflorescence branches typically spreading at 90° from the vertical. In contrast, S. virgata has a more robust inflorescence, some that were at least 30 dm, with mostly sessile heads, and with more robust ascending virgate branches.
In addition, I double-checked the identification of each plant by observing the phyllaries of at least one head for the presence of the stalked-glandular hairs for S. exigua ssp. deanei, using a 10x hand lens. The identifications at a glance were always confirmed with the observations using a hand lens.
The data for the S. virgata population on this trail are presented here. The data for S. exigua var. deanei are presented in a companion analysis.
Here I analyze the two characteristics that are supposed to discriminate the two subspecies of S. virgata, the number of flowers per head and the shape of the outer phyllaries, separately below. I first consider the number of flowers per head.
I have plotted, from all heads in flower for each specimen, the maximum number of flowers versus the minimum number of flowers for individual plants:
To clarify what is being plotted, consider first the point represented by the pink filled rectangle plotted in the top of the middle section of the plot. The color and shape of the point indicates that this specimen was observed on 11/17/03. The values corresponding to that point are a minimum number of flowers of 7, and a maximum number of flowers of 10. From that data point, it can be deduced that I observed at least two heads in flower on that specimen, one containing 7 flowers and one containing 10 flowers. There may have been other heads observed, but if there were, none had fewer than 7 flowers or more than 10 flowers.
Consider now the blue filled diamond near the bottom left of the plot, that is partially obscured on its right by a pink rectangle. The blue diamond indicates that this specimen was observed on 12/3/03. The values corresponding to that point are a minimum number of flowers of 5, and a maximum number of flowers of 5. There might have been only a single head in bloom on that plant, with 5 flowers, but if there were more heads, they all had 5 flowers.
Next to that blue filled diamond is a pink rectangle. It also corresponds to a minimum number of flowers of 5, and a maximum number of flowers of 5, but I have slightly shifted the position of its plotted point so that it would not obscure the blue diamond. This plot shows every specimen I observed on both days for which I kept the data separate for specimen, but the values for some data points have been slightly dithered to show each point. The value corresponding to each point is thus the closest integer to the plotted position.
The area marked with red diagonal lines in the lower right of the plot is an area where it is impossible to plot a point, which simply recognizes that it is not possible to have the minimum number of flowers exceed the maximum number of flowers for a single plant.
Also shown above are triangles that would contain all the data points on their apexes, if the two subspecies given in the Jepson Manual were accurate treatments of the entire species. That is, if there was a population with a minimum number of flowers of 5, and a maximum of 6, ssp. pleurocarpa, all the members of that population would plot at (5,5), (5,6) or (6,6). (The first number in parentheses is the minimum number of flowers, and the second number is the maximum.) Similarly, if there was a ssp. virgata population, all its members would plot at (8,8), (8,9) or (9,9).
Clearly, those two subspecies are not an accurate description of the population of this species in this area. The vast majority of the points do not fall within the circumscription for each subspecies. Furthermore, essentially the same plants were surveyed on both days, and it is easy to see the dramatic variation from day to day. On 11/17, 8 plants, out of 21, would have been classified as ssp. pleurocarpa from the number of flowers. But on 12/3, only a single plant, out of 12 plants, would have been so classified. Similarly, on 11/17, no plants would be classified as ssp. virgata; on 12/3, 2 plants would have been. Yet these are the same plants from the same area, which did not change their nature during the 16 day interval between observations.
This plot highlights the variability of these plants, caused by small number statistics for the number of heads on each plant, as well as the intrinsic variability among plants.
The variability in the population is directly shown in the histogram of the number of heads observed with a given number of flowers, with each day treated separately:
On 12/3, the histogram is almost bimodal, with peaks at 5 and 7-8 flowers. You can see it might be tempting to recognize two "subspecies" from the 12/3 data, one with 5-6 flowers, and one with 7-8 flowers. Yet on 11/17, the histogram is very smooth, with almost a uniform number having 6, 7 or 8 flowers. I suspect it is this variability that has led to the definitions of the two subspecies.
The two days are combined in the next histogram, which shows that the number of flowers per head ranges fairly uniformly from 5 to 8, with one outlier at 10:
A total of 108 heads were surveyed for the above plot. Propagated one sigma sqrt(N) error bars are shown for the percentages.
To within the error bars, the histogram is consistent with ~25% of the heads having 5 flowers, another ~25% having 6 flowers, another ~25% having 7 flowers, and the final ~25% having 8 flowers. Note that a single head was observed to have 10 flowers. More data would probably show some heads with 9 flowers.
The Shape Of The Outer Phyllaries
Information on the shape of the outer phyllaries, as well as the number of flowers, was gathered on 11/17/03 at the Dripping Springs Trail.
I measured 42 heads from 20 S. virgata plants. The number of flowers ranged from 5 to 10.
The two subspecies of S. virgata are supposed to separate on whether the outer phyllaries are appressed (ssp. pleurocarpa) or reflexed (ssp. virgata). For convenience, in the following, I will just refer to the outer phyllaries as the phyllaries.
All the phyllaries I observed were either completely appressed, or their tips were spreading at an angle of 0 to 90° from the appressed position. The bases of all phyllaries I observed were always appressed. I classified each head as having phyllaries whose tips were appressed, an angle of 0°; ascending, an angle of ~45°; or spreading, an angle of 90°. I assigned intermediate classifications such as appressed / ascending when some of the phyllaries were appressed, and some were ascending.
In particular, none of the observed phyllaries had any portion that were reflexed, which means "abruptly bent or curved downward or backward" according to the JM glossary.
Note that this definition means abruptly (bent or curved) (downward or backward); if it had meant abruptly bent, or curved downward or backward, it would have had that comma in the description. (Every other glossary I consulted simply says abruptly bent downward or similar phrasing.) So a phyllary with a spreading tip does not qualify as a reflexed phyllary, even if the bend in the phyllary is abrupt.
Hence even without any further analysis of the data the existence of two subspecies, one of which has appressed phyllaries, and one of which had reflexed phyllaries, is highly suspect. Nearly all the phyllaries observed had tips that were ascending to spreading.
In order to plot the nature of the phyllaries, I have assigned a numerical value as follows:
Assigned Value Nature of Phyllary Tips Angle From
Appressed Position (°)
1.00 appressed 0 1.25 appressed / ascending 0 - 45 1.50 ascending 45 1.75 ascending / spreading 45-90 2.00 spreading 90 2.50 reflexed at 45° from reflexed / appressed 135 3.00 reflexed-appressed 180
Hence a reflexed phyllary would have an angle of 90 to 180 degrees from appressed, and a numeric value of 2.00 to 3.00 here.
The following plot shows all the individual heads I measured:
The pink line at a value of 1.00, for number of flowers 5 to 6, represents ssp. pleurocarpha, with its appressed phyllaries. The yellow rectangle, at values of 2.00 to 3.00 and number of flowers 8 to 9, represents ssp. virgata. The data points have been dithered around their actual values in order to display all the points with a given value. For example, there were 4 heads with 7 flowers that had spreading phyllaries. If I plotted them all at (7,2.0) only one point would show in the plot. Hence I plotted them at values differing by 0.1 of each coordinate, to display each point separately.
There is a grand total of a single point that fits ssp. pleurocarpha, and one has to stretch the definition of reflexed (to include spreading) to obtain a grand total of 3 points that fit ssp. virgata.
Essentially 100% of the observed heads do not fit either subspecies, producing strong evidence that there is only a single species at this location in the PR.
There is a very weak trend in the data, that at most is barely statistically significant, for the phyllaries to be slightly more spreading as the number of flowers increases. Perhaps in the n CA-FP this trend becomes more pronounced and formed the basis for the definition of the subspecies. But at least here in the PR, no such differentiation into two subspecies was observed.
Copyright © 2003-2004 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: 13 April 2004