Gingersrus Database Taxon ID 7859
Costus flammulus, possible syn.
OLD NAME: Costus scaber
NEW NAME: Costus flammulus, possible syn.
NAME CHANGE NOTES: New species CURRENTLY UNDER EVALUATION, published March 28, 2023 in Systematic Botany (2023), 48(1): pp. 145?157
FULL SCIENTIFIC NAME: Costus flammulus K.M.Kay & P.Ju?rez
STATUS : distinct form
FIELD OBSERVATIONS:(If field observations are available, you can click on the link to open in a new window.)
PHOTOS:(If photos are available, you can click on the link to open in a new window.)
This new species was published by Kathleen Kay and Pedro Ju├írez in Systematic Botany, and is currently under evaluation for inclusion in the planned revision to new world Costaceae. It was delineated as a new species based on a concert of species concepts, including phylogenetics, phenotypic traits, pollination studies, and the distribution in different geological regions and climate niches. It was determined to be endemic to the montane cloud forests of the volcanic cordilleras in northern Costa Rica. It was compared against the form of Costus pulverulentus that occurs in surrounding lowland regions which is easily distinguished from populations of Costus scaber in Costa Rica by the fibrous margins of the bracts, the stamen that far exceeds the labellum, fully exposing the thecae, and typically with corolla lobes folded outward, exactly as shown in the authors' publication. It was not compared at all against populations of Costus scaber in the form that occurs elsewhere in Costa Rica and that has bract margins entire, corolla lobes tightly closed and a floral structure exactly like their new species. In the published description there is no diagnosis that permits distinguishing the new species from the similar already-described species of Costus scaber.
According to co-author Pedro Ju├írez, the decision to compare against Costus pulverulentus was based upon its position in a molecular phylogeny that was published in 2020 in which two specimens from the Monteverde area were found "nested within C. pulverulentus". These specimens were in a clade that was in a lineage that contained two Mexican samples determined as C. pulverulentus and other samples that were from lowland Costa Rica, Panama and Colombia. It is well known that the Mexican form of Costus pulverulentus has a flower structure similar to that of Costus scaber. Furthermore, the Costa Rican specimens that were identified as Costus scaber in that phylogeny were in the same larger grouping as these two specimens and Costus pulverulentus, but were well separated from the Amazonian specimens. The boot strap values (78/80) indicate that the specimens' position here is not as strongly supported and the sampling only included two specimens of the proposed new species, both of which were listed as Kay 18-0001 collected at "Monteverde Field Station on the Ficus trail about 250meters from intersection with S Principal Trail". In the voucher at UC the species listing is "Costus aff. wilsonii". The bar code number is UC2086205.
The illustration above is excerpted from Evolution 14108. Note that the Kay collection 18-0001 that was used in the publication of C. flammulus was NOT included as a paratype for the new species.
The holotype for Costus flammulus is P. Ju├írez & C. Girvin 1620 at US, with isotypes at CR and MO. It was collected 18 August, 2022 in Monteverde. It was not found listed on the portals at US or MO. As noted above, it appears that neither the holotype nor any of the listed paratypes have been included in a molecular phylogeny, upon which this new species is largely based.
Many of the listed paratypes of this new species were previously identified by Paul Maas from herbarium specimens as Costus wilsonii, but he has reviewed available duplicates in the herbarium at Leiden and agrees that they are not Costus wilsonii. He said he had long ago determined several of the specimens as "Costus aff. wilsonii" but in reviewing them, he now thinks they are closer to Costus scaber - and not to Costus pulverulentus.
In 2013 I saw many plants at Monteverde that look similar to the photo and illustration of the new species Costus flammulus and I immediately recognized them as Costus scaber based upon the traits I have used throughout the new world tropics to identify that species. I have found despite the wide variation in bract and flower color, C. scaber can be distinguished from similar looking plants by the following traits.
The photos in the top row of the figure above were all taken from the type locality in Huanuco, Peru along the Rio Derrepente and represented by my accession number R3384. The photos in the bottom row were excerpted from the published manuscript for C. flammulus. There was no photo showing details of the leaves to illustrate the upper side main vein, and the written description only described the leaves as "adaxial surface glabrous". However, when I visited the Monteverde Cloud Forest Reserve in 2013, I examined the leaves of the plants photographed below and I did find the single line of minute hairs consistent with Costus scaber.
I have contacted the authors and explained my concerns about the new species being indistinguishable from Costus scaber. Pedro Ju├írez told me "Yes, I agree. I think it'd be impossible to distinguish any of these entities without the help of genetics and careful analysis of the populations' structure, morphological comparisons of live plants, and their ecology. Hybridization will make this task even harder. I honestly think that "Costus scaber" is either SUPER variable or a species complex. I don't think that traditional morphological distinguishability would be possible here, in the field, or in herbaria. I apologize, I know it sounds frustrating, but most likely phenotypic traits are not going to be useful to define these entities." The Inaturalist link above are observations that he has determined to be this species, presumably based upon their locality.
Kathleen Kay told me: " I have not measured a bunch of traits on herbarium sheets to tell you exactly how to tell C. scaber and C. flammulus apart on dried specimens, but I feel confident in the field after working with both for many, many years. Costus flammulus has undulate leaf margins and C. scaber does not (or very rarely but not as strong). Costus scaber has rounded to very obtuse bright red floral bracts with a very narrow and distinct bright yellow callus. Costus flammulus has angular floral bracts (almost right angles or slightly obtuse) with yellow-orange calluses that are wide and grade into the more orangish-red bracts. Costus scaber has a flower with bright red petals and bright yellow on the stamen tip and labellum tip. Costus flammulus has flowers in which all the visible parts grade from orangey-red to yellow as you move from base to tip (like a little flame, hence its name). There are probably more characters, but that has been enough to give me a good gestalt in the field. "
I have studied these character traits in my own photos of Costus scaber and in photos published on Inaturalist.org but I have not been able to distinguish the Monteverde plants from some others in Honduras, Ecuador and other regions based upon these traits. Even at the type locality in Peru, I found a broad nectar callus (not a narrow yellow line), an angular apex of the bracts, and flowers with colors graduationg from orange to yellow looking very much the same as this Costus flammulus. As for the undulating margins of the leaves, this can sometimes be helpful in distinguishing species (as in Costus pictus), but in my observations of Costus scaber I have seen much variation in this trait as well, regardless of the region where it occured.
This species has much diversity both in South America and in Central America, and the Mesoamerican plants are nearly indistiguishable from C. pulverulentus in photos that show only the inflorescence and flower. In my opinion, much further study and wider sampling of the evolutionary lineages is needed before this can be accepted as a separate species from Costus scaber. The link below contains a sampling of Costus scaber photos from Mexico to Bolivia and east to Brazil and Guayana
It is very likely that this C. flammulus form is the same form of Costus scaber that was described by Schumann as Costus nutans in Das Pflanzenreich in 1904. The type for Costus nutans was collected near Aguacate, Costa Rica, which is located in the same mountain region at the north end of Lake Arenal at about 600 meters elevation. The exact locality of the type collection is not known, but the Parque Nacional Volc├ín Tenorio is just 10 km to the north with elevations up to 1800 meters. If indeed this is a separate species from Costus scaber, it is very likely that the correct name would be Costus nutans and C. flammulus would be a synonym for that species because the earlier Schumann name would take precedence.