Proposal (1053) to South American Classification Committee

Treat Diglossa cyanea as two species

Effect on SACC: This would split widespread Diglossa cyanea into two species separated by the Marañon/North Peruvian Low.

Background: Diglossa cyanea (Masked Flowerpiercer) as traditionally treated consists of five subspecies from the montane northern Venezuela through the Andes to central Bolivia. Four of the five occur northern of the Marañon and are distinguished by minor, qualitative differences in plumage. The subspecies south of the Marañon (melanopis) is somewhat more distinctive in being larger, duller in plumage, and having a paler forehead and other minor plumage differences. To the best of my knowledge, no one has ever proposed that any of these taxa be treated in any way other than subspecies.

New information: Martínez-Gómez et al. (2023) analyzed vocalizations (N=88), mtDNA (N=122), and wing length (N=364) of all taxa from throughout the range.

The songs showed strong differences between the northern taxa and melanopis in spectral frequency and structure, with melanopis having up to 5 whistles at the end of the song not detected in samples from any of the northern taxa. Here are two sonograms from the paper:

Here is a graphical depiction of mean standardized differences in song parameters for melanopis vs. northern populations:

The northern populations showed strong differences in mtDNA sequence divergence (6.7%) from melanopis; in fact, the mtDNA gene tree showed paraphyly with respect to Diglossa caerulescens, which was dismissed as a likely gene tree vs. species tree problem. That degree of difference (6.7%) in mtDNA is greater than that between most currently recognized species of Diglossa (see Mauck & Burns 2009)

Discussion. Here is what we wrote in the paper:

“Taxonomic Implications

“Based on our analysis of vocal divergence combined with in- sights on their behavioral isolation consequences from play- back experiments (see Freeman et al. 2022), we recommend that the 2 primary lineages be treated as separate species under the Biological Species Concept (Mayr 1942), specific- ally D. melanopis and D. cyanea (sensu stricto). The depth of genetic divergence in ND2 (6.7%) is even higher than differences between most other currently recognized species in Diglossa (Mauck and Burns 2009). These 2 species have been evolving in isolation for at least 2.7 million years and exhibit song differences that indicate behavioral isolation (Uy et al. 2018).”

There’s not much more to add to that. Obviously, I recommend a YES on this, with the attitude that I think we shifted burden-of-proof to the single-species treatment. One could vote NO because of insufficient number of true playback experiments or lack of comparative analysis with other Diglossa taxa that are considered species or subspecies. Some will be impressed with the genetic distance, but I personally don’t think that adds anything one way or another when looking at allopatric, sedentary tropical taxa. Also, samples near the northern boundary of melanopis distribution are missing (in contrast to samples from near the southern extreme of nominate group distribution), so one could use that as a reason to vote NO..

Note on English names: We recommended “Warbling Masked-Flowerpiercer” and “Whistling Masked-Flowerpiercer”. The rationale was to emphasize that vocal differences were the way to distinguish them while continuing the connection to the parental name “Masked Flowerpiercer”. My inclination is to do a separate proposal on the names, but if you are happy with these and don’t want to go through the work of a separate proposal, please say so.

Lit Cit:

MARTÍNEZ-GÓMEZ, S. C., C. E. LARA, J. V. REMSEN, JR., AND R. T. BRUMFIELD. 2023. Unmasking hidden genetic, vocal, and size variation in the Masked Flowerpiercer along the Andes supports two species separated by Northern Peruvian Low. Ornithology 140: 1–14.

Van Remsen, May 2025

Voting Chart: https://www.museum.lsu.edu/~Remsen/SACCPropChart1044+.htm

Comments from Stiles: “This split seems quite clear-cut, based upon definite differences in plumages and vocalizations , the depth of the split in time (2.5 mya, as deep as the splits of several other species in the genus, and the well-known geographic barrier separating them, so YES to the recognition of two species. With regard to E-names, the suggested vocalization-based ones of Van are certainly accurate, but to use them one would have to hear them sing! I might suggest two plumage-based ones: Bright-blue vs. Dusky (Masked) Flower-piercers.”

Comments from Lane: “YES. The study and results seem pretty straightforward. Not sure I think “whistling” quite describes the song of the melanopis group as well as “screeching” but that’s a conversation to have at a later time.”

Comments from Areta: “YES to splitting a monotypic D. melanopis from a polytypic D. cyanea. The vocalizations, wing length, and genetics all support the split across the dry Marañón Valley. The plumage distinctions are subtle, but that is why they had different names in the first place. It seems that obscura (whose diagnosability was questioned by Martínez-Gómez et al. 2023) and tovarensis (with its somewhat different song, which I´ve recently recorded) do not differ to an extent that we should worry about them being split in the future. So, I am happy to consider obscura and tovarensis as subspecies of D. cyanea.”

Comments from Naka: “YES, I fully agree in splitting the two allopatric populations across the Marañon into D. melanopis and D. cyanea. This study is as good as it gets, with morphological, vocal and deep genetic variation associated with a very known geographical barrier. Vocal differences are stunning, and I do like the English names proposed by Silvia and co-authors.”

Comments from Robbins: “YES for recognizing two species. The distinct songs alone support species recognition. The limited genetic data and subtle plumage differences are consistent with the genetic break at the Marañon/North Peruvian Low.”

Comments from Stiles: “YES.”