Proposal (1060) to South American Classification Committee

Recognize Aphrastura subantarctica as a separate species from A. spinicauda

Analysis

Aphrastura spinicauda is currently considered polytypic, with a widespread nominotypic subspecies spinicauda, and two island taxa restricted to Chile: bullocki from Mocha Island and fulva from Chiloé. Rozzi et al. (2022a,b) described the Diego Ramirez population of Aphrastura spinicauda as a new taxon at the species level under the name of Subantarctic Rayadito (Aphrastura subantarctica). Their conclusions are based on morphological data of 117 individuals (three populations) and genetic data of the mitochondrial cytochrome b gene (120 samples from 20 localities) and 12 nuclear microsatellite markers from 153 individuals of five Aphrastura populations (Figure 1; Rozzi et al. 2022a).

Figure 1. Study areas for the morphological and genetic characterization of Aphrastura spinicauda and A. subantarctica. The distribution range of the nominal species is shown in light green, and the new taxonomic group from the Diego Ramírez Archipelago in light blue. (A) Sampling sites for morphology. (B) Sampling sites for mtDNA. (C) Sampling sites for microsatellite markers. The numbers correspond to the sample size. From Rozzi et al. (2022a).

The taxon was described in two articles (Rozzi et al. 2022a,b), because the first one (Rozzi et al. 2022a) did not comply with the standards of The International Code of Zoological Nomenclature (ICZN):

“Diagnosis: Morphology.— Typical Aphrastura structure with rounded wings, and an idiosyncratic tail morphology. Aphrastura’s distal third of the inner web of the rectrices is abruptly and deeply excised, giving the tips of the feathers a thornlike appearance. No other genus in the family has a similar tail structure. Aphrastura differs in these morphological characters from the phylogenetically closest related genera in the subfamily Synallaxinae present in southwestern South America: Leptasthenura and Sylviorthorhynchus. In contrast to Aphrastura, Leptasthenura’s tail is not abruptly and deeply excised at the distal portion of the inner web of the rectrices; in Sylviorthorhynchus, the rectrices are denuded of barbs. A. subantarctica differs from A. spinicauda, in having on average a larger and heavier body (~25%), a larger beak (~15%), a larger tarsus (~5%), and a shorter tail (~16%) (Fig. 2 in this proposal). The primaries and secondaries are greyish on the ventral side with whitish edges; the central rectrices are dark grey on the ventral side, but do not differ between the two species.”

Figure 2. Comparison of body weight (A), tail length (B), beak length (C), beak width (D), tarsus length (E), and wing length (F) among Aphrastura populations. Metrics (means ± SD) of birds from the northwest arm of the Beagle Channel (NW Beagle, n = 50), Navarino Island (n = 54), and the population of the proposed new species A. subantarctica on Gonzalo Island, Diego Ramírez (n = 13). Lowercase letters indicate statistically significant differences at an alpha of 0.05. From Rozzi et al. (2022a).

Morphology

Rozzi et al. (2022a) measured a total of 117 adult individuals of Aphrastura in the NW Beagle Channel (n = 50), Navarino Island (n = 54), and Diego Ramírez (n = 13). Morphological differences among populations were statistically supported for weight, tail length, tarsus length, beak length, and beak width (all p < 0.05), but not for wing length (p > 0.05). Birds from the Diego Ramírez population were significantly heavier and larger (with a longer and wider bill and longer tarsi), but they had a significantly shorter tail than birds from the other two populations. The PCA analysis shows that the Aphrastura populations of the Beagle Channel and Navarino Island overlap in body dimensions, whereas the individuals of the Diego Ramírez population form a clearly separate cluster (Rozzi et al. 2022a).

Phylogeographic patterns:

Presented in Rozzi et al. (2022a): Overall mitochondrial genetic diversity was low, revealing a short genealogy of the cytb gene. Genetic diversity was larger in populations from southern Chile between 42 and 53° S (from Chiloé to Punta Arenas), compared to that of populations in the center and the north of the distribution, and also of the populations in the extreme southern part of the distribution (Navarino and Horn islands). Noticeably, based on pairwise FST and Ф values, the Diego Ramírez population is strongly and significantly separated from all other populations. Individuals sampled from Diego Ramírez shared the exact same haplotype, which differed by one mutation from the most dominant haplotype found in A. spinicauda. The Diego Ramírez haplotype is also present on Horn Island (1 out of 3 individuals) and at low frequency in Navarino Island (1 out of 20 individuals) (Fig. 3 from Rozzi et al. 2022a).

The two articles describing the new taxon do not present a phylogenetic analysis or any explicit species delimitation analysis.

Figure 3. Mitochondrial haplotype network of Aphrastura geographic groups. Each circle in the network corresponds to a different haplotype, the size of the circles corresponds to haplotype frequencies, and the colors correspond to the different geographical groups. CNS Central, north and south, PATAG Patagonia (Chiloé to Tierra del Fuego), NAVHOR Navarino and Horn islands, DIERAM Diego Ramírez (representing the proposed new species A. subantarctica). From Rozzi et al. (2022a).

Population genetic structure:

Presented in Rozzi et al. (2022a): For the PCA analysis of the five sampled populations, retained the first 43 dimensions, which explained 85% of the total genetic variation. A plot with the first three components (~ 26% of variation explained) showed that Navarino Island and the southern continental populations (Bariloche and Tierra del Fuego) formed a homogeneous group (Fig. 4). The northernmost continental population (Manquehue, central Chile) was slightly separated, but relatively well mixed with the southern continental group (Fig. 4). In contrast, the Diego Ramírez population appeared well isolated from all other sampled localities, regardless of the combination of principal components being examined (Fig. 4). According to the AIC values obtained from the ‘snapclust’ method, the optimal number of genetic clusters within the group of sampled individuals was three. This estimate is congruent with the number of clusters emerging from the PCA analysis: (1) a central cluster, comprising the Manquehue population in the center of the species’ distributional range; (2) a southern cluster, composed of Bariloche, Tierra del Fuego Island, and Navarino Island populations; and (3) Diego Ramírez (Rozzi et al. 2022a).

Figure 4. Genetic variation among five populations of Aphrastura, based on 153 individuals genotyped at 12 polymorphic microsatellite loci. (A) Scatterplot from the first three principal components that explained 25.5% of the genetic variance. 95% CI ellipses are shown. (B) Individual assignment to genetic clusters for the five sampled populations. Bars represent individual posterior membership probabilities to each of the three genetic clusters found using the ‘snapclust’ method. MA Manquehue, BA Bariloche, TF Tierra del Fuego, NI Navarino Island, DR Diego Ramírez (representing the proposed new species A. subantarctica). From Rozzi et al. (2022a).

Vocalizations

There is no detailed comparison of the vocalizations of Aphrastura subantarctica concerning the rest of the populations/subspecies, only a general mention that indicates (see Rozzi et al. 2022a,b):

"Across the entire range of Aphrastura spinicauda, individuals respond to intruders near their nest sites with alarm calls. In the continental populations, mobbing calls have minimum and maximum frequencies of 2.82 and 13.01 kHz, respectively, with at least six notes per second in central and southern Chile (Ippi et al. 2011). However, based on a previous record in 2001, it seems that minimum and maximum frequencies of mobbing calls of Aphrastura subantarctica are lower, with the same number of notes (see Imberti 2011). While these preliminary observations have to be confirmed by future studies, the low call frequencies of Aphrastura on Diego Ramírez could be related to the high ambient noise, as well as their larger body size (Mikula et al. 2021)."

A work by Ippi et al. (2011) reported that vocalizations among the five populations showed some variation in the repetitive trill. In contrast, no differences were found between alarm calls and loud trills. Variation in repetitive trills among populations and forest types suggests that sound transmission may explain vocal differences in suboscines. Acoustic differences help distinguishing subspecies bullocki from spinicauda and fulva, but not the latter two subspecies from each other. Ippi et al. (2011) suggests that the geographical differentiation in vocalizations observed among Thorn-tailed Rayadito populations could result from different ecological pressures.

Vocalization examples and comparison across the range of Aphrastura spinicauda:

Aphrastura spinicauda, MN Cerro Ñielol, Temuco, Chile, 11-8-2012 Heraldo Norambuena Ramirez

Aphrastura spinicauda, PN Los Glaciares, Argentina, 6-1-2020 Gabriel Leite

Aphrastura spinicauda, PN Radal Siete Tazas, Maule, Chile 2-11-2024 Vicente Pantoja Maggi

Aphrastura spinicauda, Rio Negro, Chaitén, Chile 20-9-2021 Freddy Sepulveda

Aphrastura spinicauda, Islas Diego Ramirez, Chile 6-1-2001 Santiago Imberti

Aphrastura spinicauda, Islas Diego Ramirez, Chile 6-1-2001 Santiago Imberti

Aphrastura spinicauda, Islas Diego Ramirez, Chile 6-1-2001 Santiago Imberti

Conclusion

Rozzi et al.'s (2022a,b) approach to describing a new taxon was based on population genetics. Differences in the cytb marker only show a single mutational step between widely separated geographic areas. As expected with macrosatellite markers, divergence is more substantial using this source of information. However, the cluster analysis not only shows a cluster for Diego Ramírez, but also a cluster for the northernmost part of Maquehue. This result could suggest an identification of genetic structuring rather than a speciation event. Due to the winds found on subantarctic islands, short tails are apparently favored (there are no trees to climb and the tail acts as a sail in the wind) and short wings in relation to mass/size (which is why birds on these islands tend to become flightless).

The habitat used by Aphrastura in Diego Ramirez does not differ from that found in Cape Horn and the Mitre Peninsula, Tierra del Fuego (Argentina) where rayaditos are locally common at great distances from forest. Here, rayaditos are locally abundant in areas with tussock grass and low scrub, where they rarely fly and move close to the ground in dense vegetation avoiding the omnipresent windy conditions. Rozzi et al. make no attempt to compare any aspect (morphological, behavioural, plumage or vocal) with these closest populations of Aphrastura that occur in the same habitat which we believe should have been a fundamental starting point of their study. Furthermore, a comparison of the closest forest populations and those occurring in stunted Nothofagus antarctica would be desirable.

Although the frequency in vocalizations (compared to a single set of recordings) seems to be slightly lower, this negligible difference can be attributed to other factors rather than be considered a factor to separate this population into a different species.

The structure recovered by the analysis of microsatellites (which portray very recent phylogenetic information and are thus of limited utility when assessing species limits) might be explained by and endogamic population in Diego Ramírez, perhaps in part due to the difficulty of crossing from the continent to the island.

The morphological differences might be adaptive, as the cold and wind-barren Diego Ramírez islands would exert pressure favoring a larger size and a reduced wing length in relation to weight/size, while the shorter tail could be related to the lack of trees and the advantage of not acting as a "ship sail". We note that the diagnosis does not include any plumage or vocal feature, but rests exclusively in morphometric differences. Although we do not dispute that these differences exist, a broader comparison to understand how the geographic variation is structured in Aphrastura spinicauda would have provided a much better yardstick than limiting the comparison to the southernmost populations. The lack of trees in Diego Ramírez provides a drastic ecological contrast to the forest places in which A. spinicauda lives in most of its distribution, however, populations in the Hornean Islands and the Mitre Peninsula use tussock (Poa flabellata grasslands) and shrub lands, showing that Aphrastura is quite plastic. The subspecies fulva from Chiloé, with its striking plumage differences from the continental forms, indicates that Aphrastura populations can exhibit marked phenotypic differences without a need for stringent geographical barriers. Finally, vocalizations should be properly studied, even if alarm calls are shown to be lower pitched in subantarctica than in the other taxa, this could just be an incidental byproduct of size, and not in itself evidence for species status of the Diego Ramírez population. In sum, we think that the data provided by Rozzi et al. (2022a, b) is more consistent with subantarctica being a subspecies of spinicauda.

Recommendation

We recommend a NO vote. The reduced to null levels of genetic differentiation in a single mtDNA marker (cyt b) coupled to a lack of vocal differences in calls (and unknown, but apparently trivial levels of vocal differentiation in song; pers. obs.) indicate to us that subantarctica can be a good, mildly differentiated subspecies of A. spinicauda but not a separate species from it.

References

Imberti, S. (2001). Internet Bird Collection: horn-tailed Rayadito (Aphrastura spinicauda). https://macaulaylibrary.org/asset/204019791

Ippi, S., Vasquez, R. A., van Dongen, W. F. & Lazzoni, I. (2011). Geographical variation in the vocalizations of the suboscine Thorn-tailed Rayadito Aphrastura spinicauda. Ibis 153, 789–805.

Mikula, P. et al. (2021). A global analysis of song frequency in passerines provides no support for the acoustic adaptation hypothesis but suggests a role for sexual selection. Ecol. Lett. 24, 477–486.

Rozzi, R.; Quilodrán, C.S., Botero-Delgadillo, E., Napolitano, C., Torres-Mura, J.C., Barroso, O., Crego, R.D., Bravo, C., Ippi, S., Quirici, V., Mackenzie, R., Suazo, C.G., Rivero-de-Aguilar, J., Goffinet, B., Kempenaers, B., Poulin, E., Vásquez, R.A. (2022a). «The Subantarctic Rayadito (Aphrastura subantarctica), a new bird species on the southernmost islands of the Americas». Scientific Reports. 12(1): 13957. ISSN 2045-2322. doi:10.1038/s41598-022-17985-4.

Rozzi, R., Quilodrán, C.S., Botero-Delgadillo, E., Crego, R.D., Napolitano, C., Barroso, O., Torres-Mura, J.C., Vásquez, R.A. (2022b). «El Rayadito subantártico: disponibilidad del binomio Aphrastura subantarctica (Passeriformes, Furnariidae)». Boletín Museo Nacional De Historia Natural 71(2): 9-15. ISSN 2045-2322. doi:10.54830/bmnhn.v71.n2.2022.222

Heraldo V. Norambuena, Juan I. Areta, Santiago Imberti and Mark Pearman

August 2025

Vote tracking chart: https://www.museum.lsu.edu/~Remsen/SACCPropChart1044+.htm

Comments from Remsen: “NO. This research team is producing some great data on microevolution among islands in this species. Fascinating results that reveal important factors relative to the early stages of differentiation. Thus, my “no” vote only refers to the effort to apply taxonomy to this level of differentiation. Taxonomy is a totally artificial construct devised by humans to provide labels for degrees of differentiation because we have a tough time dealing with continuous variation. Using that scheme, these populations do not fit our definition of species, and I would really have to work hard to accord them subspecies (diagnosable unit) rank.”

Comments from Areta: “NO. As described in the proposal, the evidence is insufficient to consider subantarctica as a different species, and I would be willing (in a positive spirit) to grant it at most subspecific status.”