Report on the Verbenas of West Texas
Yesterday afternoon and evening, I learned the Verbena species of West Texas. I identified 1,612 out of the 1,763 observations. Species identified included (source): V. bracteata (252), V. canescens (341), V. cloverae (2), V. halei (243), V. livermorensis (6), V. macdougalii (1), V. neomexicana var. hirtella (110), V. perennis (364), and V. plicata (26). Verbena pulchella, though identified, is not included here because it belongs to the genus Glandularia. Species not identified included: V. brasiliensis, V. bonariensis, V. hastata, V. menthifolia, V. scabra, V. urticifolia, and V. xutha. These excluded species totaled less than 100 observations.
The major goal of this exercise was to distinguish five species: V. canescens, V. cloverae, V. neomexicana var. hirtella, V. perennis, and V. plicata. Out of all the species included in Texas keys, I find these the most difficult. It became clear later in the process that V. halei was somewhat difficult to distinguish from V. perennis and V. plicata under odd circumstances where only a small part of the plant was photographed.
Challenges
Technical challenges
- Photographs lacking a view of leaves or bracts. A reply was given stating that the observation could not be identified. If the observation was labeled as a species that was common at that locality, the observation was marked as reviewed. If not, the observation was not marked as reviewed. If the observation was outside the normal range of the species, a disagreeing identification at genus level was given.
- Interpreting photos. It is very difficult to assess scale from photographs and this can only be overcome by using proportions. I wasn't able to fully understand this before IDing the last of the observations. Interpreting hair type and color is also exceptionally difficult from photographs and I (for the most part) gave up on this early in the process. I think it is possible to use hairs, but the lighting of the photo would have to be accounted for.
- Unclear language used in keys. This pertained mostly to leaves and hairs. For leaves, the description of V. plicata leaves can be challenging to understand without knowing all the other species. This was overcome by looking at and photographing one specimen of each species (particularly the hairs of the undersides of the leaves). Though hairs are almost useless when identifying iNaturalist observations, they are quite useful in preserved specimens. After understanding the differences using hairs, it is easy to pick out the how the authors of the keys tried to describe the leaves. For the challenge I had with hairs, it also helped reading the descriptions. Specifically, the key states that V. livermorensis lacks glandular hairs. However, the key is only referencing the lower stems and not the inflorescence. If you look at the inflorescences and in the description of the inflorescences, you'll notice that they are actually covered in glandular hairs (so much so that it can help distinguish this species from V. neomexicana var. hirtella).
Variability challenges
- Leaf variability in V. plicata and to a lesser extent V. cloverae. This was resolved by understanding heteroblasty and the effects of stress on leaves. Ultimately, the species of Verbena I focused on had leaves that could be divided into three categories: basal, transitional, and bracteate. Basal leaves are always more likely to appear petiolate regardless of species. Many transitional leaves in V. plicata uncomfortably stretch the definition of petiolate. Also, in V. plicata, basal leaves are rounder while transitional leaves are typically more deeply divided. One atypical instance of deeply divided basal leaves caused some difficulty.
- Bract variability in V. canescens. This was challenging as the keys used state that the bracts are longer than the calyces. In actuality, bract length varies from slightly shorter than the calyces to more than six times as long. The key here is not how long the bracts are, but how prominent they are. In V. canescens the lower bracts tend to spread and look notably separated from the calyces. In V. neomexicana var. hirtella the bracts are usually appressed to the calyces or short enough to not notice them (even the lower bracts).
- Flower variability in V. neomexicana var. hirtella. There appear to be at least two forms of this variety. They are different enough that I'm surprised they aren't recognized as different varieties or species. Further investigation may reveal this as an error on my part (i.e., misidentified V. neomexicana observations) or elevation differences among populations.
- Populational variability in V. plicata. Northern populations usually have a distinct petiole in basal-most leaves while southern populations don't. There appears to be a point around the southern High Plains where basal leaves become more like transitional leaves from northern populations (less rounded, more deeply incised, blade tissue surrounding the "petiole" becoming wider).
- A few atypical observations. These were simply set aside as unreviewed for later.
Future challenges
- Overuse of V. canescens. By far the most common misidentifications where as V. canescens. It wouldn't surprise me if this number was in the hundreds. Meanwhile, misidentifications of V. canescens as another species were less than ten. This was particularly prevalent where V. canescens over the range of V. plicata. This probably reflects either a bias in the computer vision or a strong bias of eastern botanists assuming that the western plants are the species they know. There were also many misidentified as V. halei but this is to be expected as it is a widespread species. I have included a couple of comments to help with explaining the differences:
- V. plicata misidentified as V. canescens: In V. plicata, I tend to think of the lower leaves as looking like a pan where the "winged petiole" is the handle and the main part of the blade is the main part of the pan. If the leaves look like they have a handle or a "winged petiole", it's not V. canescens. The reason I describe it this way is that "winged petiole" is somewhat ambiguous if your not familiar with the other species in the genus.
- V. plicata misidentified as V. halei: V. halei has narrower upper leaves (or leaflets) and much smaller, inconspicuous bracts on the spikes. The inflorescences are usually much taller and highly branched.
Conclusion
At some point, I'll probably write up a more digestible guide based on this little study. For the time being, I'll provide a short key to help distinguish the five I focused on plus V. halei. I especially recommend reviewing the map. It's easy to toggle off certain species to make comparisons.
Maps
West Texas distribution map
Observations I've IDed including outside of West Texas
Key
'1. Leaves linear, hairs appressed
'2. Inflorescences usually not branched, lower leaves simple or when just a few teeth, plants short, mountains of the Trans-Pecos, especially the Guadalupe Mountains......V. perennis
'2. Inflorescences branched, lower leaves compound, plants tall, widespread......V. halei
'1. Leaves not linear
'3. Mature inflorescence branched, hairs appressed, bracts inconspicuous......V. halei
'3. Mature inflorescence simple (often three coming from nodes with transitional leaves), hairs spreading, bracts conspicuous or inconspicuous
'4. Proximal 1/4 - 1/2 of leaves narrowed to a petiole-like structure, apical half rounded or deeply divided, especially noticeable in basal leaves
'5. Bracts prominent, usually spreading, most longer than the calyces; flowers small, usually purple; plants greener due to less dense covering of hairs......V. plicata
'5. Bracts not prominent, usually appressed, shorter than the calyces; flowers large, usually lavender; plants grayer due to denser covering of hairs.......V. cloverae
'4. Proximal 1/3rd of leaves usually the same width as the narrowest part of the middle of the leaf, especially in transitional leaves
'6. Bracts prominent, usually spreading, plants short, Trans-Pecos and a large part of central Texas......V. canescens
'6. Bracts not prominent, usually appressed, plants taller, confined to the Big Bend region of the Trans-Pecos......V. neomexicana var. hirtella
Sources
Correll and Johnston, 1970. Manual of the Vascular Plants of Texas
Powell and Worthington, 2018. Flowering Plants of the Trans-Pecos and Adjacent Areas
Diggs et al., 1999. Shinners and Mahler's Illustrated Flora of North Central Texas
Nesom, 2010. Revision of Verbena Ser. Tricesimae (Verbenaceae)
Nesom, 2010. Infrageneric Classification of Verbena (Verbenaceae)
Peralta and Múlgura, 2011. El Género Glandularia (Verbenaceae) En Argentina (for notes on Glandularia pulchella)
Referenced observations
Long-bracted V. canescens
This one is particularly instructive: https://www.inaturalist.org/observations/117245060
https://www.inaturalist.org/observations/117787574
https://www.inaturalist.org/observations/116893001
https://www.inaturalist.org/observations/116365975
https://www.inaturalist.org/observations/73097437Short-bracted V. canescens
https://www.inaturalist.org/observations/74677366Broad "petiolate" V. plicata
https://www.inaturalist.org/observations/153857768Southern V. plicata
https://www.inaturalist.org/observations/153857768Comparison of small and large flowered forms of V. neomexicana var. hirtella
https://www.inaturalist.org/observations/206686181Misidentified specimens on Symbiota
Andrews County: https://swbiodiversity.org/seinet/collections/individual/index.php?occid=24928718&clid=0
No lower leaves present, only upper transitional leaves, but some clearly pseudopetiolate.