Variation in the H. drummondii Complex

Label of holotype of H. serpyllifolia at NY — determined as H. drummondii by Epling & Stewart.

Hybridization & Variation in the H. drummondii Complex

by Bob Harms email-here

In their treatment of H. drummondi Epling & Stewart 1939 note that with the exception of central Texas:

it is remarkably constant in the size and configuration of the flower parts, as well as the general aspect of the plant (p36)

For Texas, however, they note variability among herbarium collections, attributing that to ontogeny and “genetic variation”, the latter said to terminate at one extreme as H. reverchonii, which they recognize as a distinct species. But they found that distinguishing H. drummondii and H. reverchonii was often difficult:

... it is a practical impossibility to segregate these two species even approximately where they come together. H. Reverchoni is restricted to a few counties in central Texas. (p. 36)

Epling & Stewart consider H. serpyllifolia an intermediate form, and it is listed with synonyms under H. drummondii (p. 35) — apparently a hybrid — although this decision seems heavily qualified. H. serpyllifoia is said to be “H. Drummondii in flower characteristics but with the habit more like that of H. Reverchoni;” and further (boldface by Harms):

For purposes of convenience one might determine the Texas plants as follows: the name H. Reverchonii might be applied to the plants of central Texas: … Travis, Hays … Counties. H. serpyllifolia might be used to designate the intermediate forms in the following counties: … Hays, … Travis …. (p. 37)

In his 1968 dissertation Irving (147–148) gives the following assessment of the problem:

Four generalized factors can be envisioned as being responsible for this complex pattern of variation:

(1) intravarietal genetic variation,

(2) hybridization and/or introgressive hybridization,

(3) phenotypic plasticity, and

(4) phenological variability.

For the H. drummondii complex, the most important factor seems to be genetic variation, presumably uncomplicated to any large extent by extant intravarietal or specific hybridization. Superficially, hybridization or even introgression would seem to play a dominant role in the variation patterns observed. Indeed, this has already been suggested as a principal source of variation by Epling & Stewart (1939).
For example, the var serpyllifolia is an excellent candidate for an F1 hybrid for it stands between var drummondii and var reverchonii in practically all of the characters which can be used in a morphological appraisal of hybridization. These include

habit and height;

leaf shape and size;

calyx length,

teeth disposition, and pubescence;

corolla size and shape.

Further, this variety is frequently found growing together with both of the other varieties on the Edwards Plateau.

Irving examined the above 4 factors and concluded “evidence for the hybrid nature of var serpyllifolia was emphatically negative”. This evidence included:

(1) analysis of a synthesized F1 generation between var reverchonii and var drummondii — which did not match var. serpyllifolia;
(2) a survey of hybrid combinations in a natural population known to include hybrids had only 1% hybrids, and of these most were var. drummondii × var. reverchonii;

Irving later published results of hybridization experiments within the H. drummondii complex, providing even more compelling evidence in support of var. serpyllifolia as a valid taxon (1) based on monoterpene patterns in F1 & F2 hybrids [R. Irving & R.P. Adams 1973: Genetics and biosynthetic relationships of monoterpenes] and seed fertility [R. Irving et al. 1979: Artificial Hybridization in Hedeoma (Labiatae)].

Hedeoma reverchonii var. serpyllifolia was treated as a distinct species (H. serpyllifolia) until Irving (1970) accorded it varietal status on the basis of morphology. The high level of fertility between H. r. var. reverchonii and H. r. var. serpyllifolia would appear to support such a treatment. The fertility data also indicate that despite a close morphological similarity, H. drummondii and H. r. var. serpyllifolia are more distantly related than H. drummondii and H. r. var. reverchonii.

In considering the variation found among herbarium collections, one source of confusion seems to stem from incorrect determinations. To some extent these have resulted from lack of information as to the scent of a plant's leaves, because earlier (prior to Irving's work) the value of this property was not known and even later collectors may have lacked a sense of smell. One example from personal experience: I was present when a collection was being identified as H. drummondii, so I crushed a leaf to test the scent. It was clearly camphor, thus H. r. var. serpyllifolia. This scent was indicated on the collection label, but even so the plant was catalogued as H. drummondii. I must note, however, that I consider determinations by Irving that I have examined to be correct.

Plant Resources Center Home Page — Flora of Texas — Hedeoma