A New Treatment

Given the notable discrepancies between the two primary treatments (Guy Nesom and John Semple) for Heterotheca, the genus is a prime candidate for a novel taxonomic treatment driven by quantitative data rather than re-hashing species based on subjective morphological interpretations. The most substantial issue with previous work is the lack of molecular phylogenetic data to inform how species, populations, and collections are related to one another. Several species have been proposed which bear narrow distributions or restrictive niches, indicating that the continued existence of certain species or varieties may be under threat from a number of anthropogenic changes.

The approaches between treatments largely differ in the traditional splitting vs. lumping approach although changes between the treatments are complex in several instances. The former may identify sensitive species which warrant conservation concern more rapidly but which often lack robust data to support their distinction from similar populations. The lumping approach, conversely, requires substantial evidence of one form or another to identify taxa. A species complex may exist, but individual units (whether they be separate species, varieties, subspecies, disparate populations, or whatnot) may not be recognized without adequate proof. While the late 1900’s were marked by more conservative approaches to species recognition, some have chosen more drastic splitting of taxa in lieu of molecular data, morphometrics, greenhouse experiments, or even peer-review. In many cases, the cart has come before the horse and drastically muddied the waters so-to-speak taxonomically for complicated groups such as Heterotheca.

Semple’s Treatment

John Semple provided the first comprehensive look at the genus, collaborating with numerous students and fellow biologists to generate hundreds of chromosome counts and extensive morphometric data. While much of the latter has not been formally published, preliminary analyses and comparisons drove a number of important decisions relating to how different populations would be treated and how species complexes were delimited. This, along with exhaustive (and at times resolve-testing) measurements of difficult-to-quantify features (including hair/gland densities), led to the production of the most thorough treatment to date. Over ten-thousand herbarium specimens were annotated over the decades by the principal investigator, and dozens of wide-ranging trips to generate hundreds of collections were made to become intimately familiar with the many species and populations of the perennial species in particular. From these, countless line drawings for at least all of section Phyllotheca (which Nesom split into Phyllotheca and Chrysanthe) were generated. Further, distribution maps were hand-drawn along with a genus-wide dichotomous key (including in The Flora of North America in 2006) – challenging feats given the considerable variation across species and varieties whose ranges often overlap. The range maps for select varieties within species complexes overlap greatly (especially for the H. fulcrata s.l. and H. villosa s.l. complexes), indicating that numerous closely-related taxa may grow sympatrically. The varieties proposed were noted to be incompletely differentiated, raising questions as to whether they should be formally recognized as discrete taxa. Given the limitations of the treatment, however, this treatment is by far the better supported of the two especially provided the technological limitations when it was published in the late 1980’s to 2000’s.

Nesom’s Treatment

Guy Nesom self-published three treatments corresponding to sections Chrysanthe (most of the non-Californian species), Phyllotheca, and a combined paper of Ammodia and Heterotheca. Section Chrysanthe was split into several units and all published in Phytoneuron, a journal that Nesom alone edits and authors with little to no peer review when he is the sole author. The collective treatment utilized no morphometric analyses (separately noting that morphometrics are “unreliable and misleading in Heterotheca“), disregarded significant differences found in previous morphometric work (particularly for H. oregona s.l.), misrepresented available chromosome count data, and lacked the extensive review of specimens which Semple spent years accomplishing. Different species (varieties were rarely used) instead were made based on subjective interpretations of morphological features which differed greatly from previous work. Many of these features are accompanied by qualifiers excusing deviations or irregularities in the species proposed, especially for populations in the northern United States and Canada. Notes were made referencing gene flow among taxa, but no real data were provided for these claims to stand on. Photographs were utilized rather than illustrations throughout, both of herbarium specimens and live material. Maps were generated at the county level generally, but keys rarely included more than a few taxa. While some proposed species appear to represent coherent groups of populations, the treatment by and large has only confused matters (especially on iNaturalist) and should not be given the same weight as more robust data-informed and peer-reviewed science. Nesom himself noted in a 2007 paper on the H. villosa complex that “it has not seemed necessary to re-assemble and re-cite the massive number of specimens (10,300) Dr. Semple had on hand for his monographic study of sect. Phyllotheca” – a statement I wholeheartedly agree with but which was evidently later reversed by Nesom.