Plants of Southern California: Quercus agrifolia varieties
JM key, description and illustration, pp. 658-663.
Munz description, pp. 479-480.
Fred Roberts description and key, pp. 24, 28-31 (in The Oaks of the Southern Californian Floristic Province, 1995)
Flora of North America: Quercus agrifolia
Q. agrifolia has two varieties that have been defined: var. agrifolia and var. oxyadenia. The difference is the hairiness of the leaf, especially the underside.
There is clearly some confusion about what makes var. oxyadenia in practice, as well as considerable confusion about its range. In addition, in many areas of Southern California the varieties seem to intergrade, and are not clearly separate. This makes it difficult to know for sure how to classify many specimens.
In fact, the most recent treatment on black oaks (Quercus Sect. Lobatae), by Richard J. Jensen in the Flora of North America, has made var. oxyadenia a reserved-judgment taxon! The Jepson Manual states that reserved-judgment taxa are unclearly differentiated forms, and reserves judgment about (but calls into question) whether a form should be considered taxonomically distinct. You know the varieties are confusing when a flora drops them from the main treatment!
This article reviews the distinguishing characteristics of the two claimed varieties, gives my findings on the range of properties for Q. agrifolia in five different areas of Southern California, reviews the geographic range of the two varieties as well as determinations made by others, and states the current manner in which I classify the two varieties in our plant lists.
Note that my current approach is preliminary; I may change my mind after I see a number of Q. agrifolia specimens from the Cuyamaca and Laguna Mountains, which I have not yet done.
The guidance given by the floras
- Munz says that:var. agrifolia has leaves that are glabrous or somewhat stellate-pubescent beneath especially along the veins,and distinguishesvar. oxyadenia by leaves densely stellate, especially beneath
- The JM key is:4. Lf lower surface glabrous to sparsely hairy (exc vein axils) .... var. agrifolia
4'. Lf lower surface densely tomentose ....var. oxyadenia
- Roberts' key is:13a. Underside of leaf densely hairy, often covering entire under side of leaf ...var. oxyadenia
13b. Underside of leaf with tufts of hair in the axils of the veins, otherwise without hairs on underside of leaf ....var. agrifolia
- The Flora of North America says:Plants with densely pubescent leaves, especially abaxially, have been treated as Q. agrifolia var. oxyadenia.
There is consensus that a leaf that is glabrous underneath except for the vein axils is var. agrifolia, and that a leaf that is densely hairy to tomentose underneath is var. oxyadenia.
The problem comes for leaves that are in-between. In the next section, I show some examples of such leaves.
Examples of actual leaves
The pictures below are of four leaves from four different trees from different areas that illustrate the general problem. They are placed in order of hairiness, with increasingly-hairy leaves to the right.
Fallbrook Dripping Springs mile 0.00 Santa Rosa Plateau Dripping Springs mile 0.03
Fallbrook Dripping Springs mile 0.00 Santa Rosa Plateau Dripping Springs mile 0.03
Note two caveats on the specimens:
- Ignore any color difference in the leaves. Greener leaves are simply fresher leaves, with the leftmost leaf a fresh specimen and the other specimens up to a bit more than one year old.
- The leaves are not representative specimens from each area. In general, leaves from each area span the entire range of hairiness from the leaf on the left to the leaf on the right. The leaves are simply representative of the range in hairiness for each area.
The leaf on the left is from Fallbrook in North San Diego County. Leaf #3 is from the Santa Rosa Plateau in southwest Riverside County. Leaves #2 and #4 are both from the Dripping Springs Trail, Agua Tibia Mountains, about 15 miles east of both the Santa Rosa Plateau and Fallbrook, with mileage measured from the parking area at the campground toward the trail.
The top pictures are of the entire leaves; the bottom pictures are blown-up portions of the upper left of the leaves. Although the hairs are somewhat hard to see in these pictures, it was necessary to show the entire leaf, since the discrimination in the varieties is on the entire leaf. Use the leaf on the right to tune your eye to the appearance of the stellate hairs on the leaf blade.
Descriptions of the leaves (from microscopic examination of the leaves, not from the pictures):
- The leftmost leaf is totally glabrous, without hairs anyplace on the leaf or petiole.
- Leaf #2 has a densely-hairy petiole, a hairy lower half of the main vein, dense hairs in two vein axils near the middle of the leaf, and is ~sparsely-hairy over most of the lower surface.
- Leaf #3 is fairly similar to leaf #2, but is hairier on the veins. It has a densely-hairy petiole, a densely-hairy lower half of the main vein, dense hairs in the lowermost two vein axils of the upper half of the leaf, hairs on portions of the side veins, and is ~sparsely-hairy over most of the lower surface.
- Leaf #4 has a densely-hairy petiole, densely-hairy main veins, dense hairs in all vein axils, and is densely-hairy over nearly the entire lower surface.
How would one classify these leaves?
- Leaf #1 is undisputedly var. agrifolia.
- Leaf #4 is most likely var. oxyadenia; it is nearly a perfect match for the JM illustration. The leaf is not tomentose; the hairs are dense, but they are not overlapping. But the leaf is far from sparsely hairy.
- Leaf #3 is fairly close to the leaf in the Roberts' illustration of var. oxyadenia. It might be called var. oxyadenia from that comparison, but might be called var. agrifolia using Munz or the JM.
In fact, this leaf is from the Santa Rosa Plateau, and Lathrop and Thorne (1985) probably considered this as var. oxyadenia. In their Flora of the Santa Rosa Plateau, they listed both var. agrifolia and var. oxyadenia, but dismissed the difference with the following words:Subspecific names are used where appropriate, but varieties, considered here as minor genetic variants with little if any geographic or ecological significance, are placed in parentheses or brackets when mentioned at all.
This leaf is typical of the most-hairy leaf found on any trees at the Santa Rosa Plateau, and hence is probably typical of what they would call var. oxyadenia.
- Leaf #2 would be classified as var. agrifolia using Munz or the JM, but its classification is unclear using Roberts. Perhaps it would be called an intermediate between the two varieties using the criteria of Roberts.
This is clearly a confusing situation. Apparently, at each of these locations, both varieties are present, along with intermediates. Hence it is of interest to review the claimed ranges of these varieties.
The Ranges Of These Varieties
The range of var. agrifolia:
Both Munz and the JM give the range of var. agrifolia to be the range of the species (NCoRO, CW, SW), which includes the range of var. oxyadenia, but with var. agrifolia confined to elevations below 900 m (3000 feet).
The range plotted by Roberts is this same area except that it excludes var. agrifolia from the Cuyamaca and Laguna Mountains. This is essentially the same distribution as in Munz and the JM, since his contour roughly follows the 3000 foot contour.
The range of var. oxyadenia:
The range for var. oxyadenia, as shown in Roberts, is confined to the Cuyamaca and Laguna Mountains in San Diego County above approximately 3000 feet. Munz says at 2000 - 4600 feet; interior cismontane Riverside and San Diego Counties. The JM simply says PR; 600 - 1500 m (2000 - 5000 feet).
Beauchamp gives an additional location for var. oxyadenia of Doane Valley at Palomar Mountain.
Banks states that var. oxyadenia is abundant in oak woodlands and scattered along perennial drainages, 160-1090 m (500-3600 feet) in his Flora of the Agua Tibia Mountains (Palomar Range), with var. agrifolia being common in oak woodlands, 240-715 m (800-2350 feet), with no clear geographic or elevational separation between the two varieties except for var. oxyadenia being the only taxon at higher (and lower!) elevations.
As mentioned above, Lathrop and Thorne (1985) listed both var. agrifolia and var. oxyadenia in their Flora of the Santa Rosa Plateau, but dismissed the difference between the varieties as being insignificant.
At the UC Riverside herbarium, there are vouchers for var. oxyadenia in:
- the San Gabriel Mountains, along SR39, canyons north of Azusa along Morris Reservoir (Reifner, August 2000)
- the Verdugo Mountains, La Tuna Canyon (Reifner, August 2000)
From my work, in addition to the localities mentioned above that seem to have at least some apparent var. oxyadenia specimens, I have seen a similar range of hairiness in leaves from the Santa Margarita Ecological Reserve, 5 miles southeast of the Santa Rosa Plateau; and from the Mt. Wilson Trail, Sierra Madre, in the San Gabriel Mountains. These are the only areas I have checked so far; I have yet to find an area in Southern California that does not have this range of variation.
Needless to say, many of these observations are wildly at variance with those reported in the floras, both for geographic range and range in elevation.
The Classification Problem
In each of the areas mentioned above, I commonly see the entire range of the illustrated variation in hairiness on nearby trees! If these are indeed separate varieties, then there seems to be a huge area in which both varieties are found, and in fact often with many intermediate specimens.
Clearly, many other botanists have also found var. oxyadenia in the midst of var. agrifolia. Also, since floras and vouchers don't often mention specimens that are troubling to classify, it is quite possible that many other botanists have faced the same difficulty in deciding how to classify some specimens.
So what's a poor botanist to do, when confronted by plants like these?
Here are some possible choices:
- Take the approach of Lathrop and Thorne, and dismiss the significance of the two varieties. This is one small step beyond what Nixon has done in the Flora of North America.
- Accept the existence of the two varieties as end points of a continuous gradation, and recognize that there are a large number of intermediate specimens that are not clearly one variety or the other. These intermediate specimens may or may not be confined to a fairly large area from the southern edges of the San Gabriel Mountains to the foothills of the Cuyamaca and Laguna Mountains.
- Accept the existence of the two varieties, expand the definition of one variety to encompass the intermediates, and shoehorn the intermediates into that variety.
- Accept the existence of the two varieties, and accept a fairly large variation in the properties of each variety. Classify a population in a given area based on the midpoint of the variation in each area.
After struggling back and forth about how to classify some of these specimens, it finally occurred to me that the situation with these two varieties is almost exactly analogous to the two main varieties of Eriogonum fasciculatum, var. foliolosum and var. polifolium. Just as for Q. agrifolia var. oxyadenia, E. fasciculatum var. polifolium is the only variety found at higher altitudes. Just as for Q. agrifolia var. agrifolia, E. fasciculatum var. foliolosum is the only variety found in most of the CW, and the low-elevation coastal regions of the SW. And the two varieties of E. fasciculatum completely intergrade in many areas of Southern California, exactly as described here for Q. agrifolia.
Thus I am inclined to take choice #2, which treats the two varieties of Q. agrifolia in the same manner as I treat the two varieties of E. fasciculatum. Operationally, that means I will list a given variety whenever a specimen gets close to that taxon; list both varieties as being present in an area if there are examples close enough to both taxa; and recognize that many specimens will not be perfect examples of the end point.
This approach has the benefit of informing botanists that a given area has either examples of only one taxon, or that a given area has examples that at least approach both taxa.
However, if I find examples close to var. agrifolia are widespread in the Cuyamaca and Laguna Mountains, I will probably change my position and agree with Lathrop and Thorne in dismissing the significance of the two varieties.
Copyright © 2005 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: 31 January 2005