Bioacoustics of Catharus fuscater mentalis

Diagnosis Genetics: In both phylogenetic reconstructions, samples of C. arcanus sp. nov. formed a clade that was sister to C. [ f. ] mirabilis of Cerro Pirre (mean uncorrected p -distance = 0.04 ± <0.01). In the UCE tree, these taxa were reciprocally monophyletic and formed a clade that was sister to all South American taxa. ABGD and ASAP analyses both identified C. arcanus sp. nov. and C. [ f. ] mirabilis as independent genetic clusters. The estimated divergence time between C. arcanus sp. nov. and C. [ f. ] mirabilis was 1.3 Mya (95% HPD = 0.9–1.7). Morphology: Adult study skins of C. arcanus sp. nov. are paler on the dorsal surface than similarly-aged skins of C. hellmayri and C. [ f. ] mirabilis , in both sexes ( Fig. 13 ). Unlike adult females of C. [ f. ] mirabilis , which had darker crowns than mantles, there was no contrast between the pale crown and mantle of the C. arcanus sp. nov. adult female (LSUMZ 164300), which had an enlarged ovary (11 × 5 mm ), 100% pneumatized skull, and no bursa. Ventrally, study skins of C. arcanus sp. nov. and C. [ f. ] mirabilis were paler than C. hellmayri ( Fig. 13 ). Like C. [ f. ] mirabilis (and to a lesser extent C. hellmayri , see: Ridgley and Gwynne 1989), C. arcanus sp. nov. has a yellowish wash on the ventral surface in life, that renders the breast ‘olive’ and abdomen ‘pale yellow’ ( fide E. Eisenmann, on the label of AMNH 811714). Similar notes were written on the labels of LSUMZ 80201 (‘Yellowish olive instead of grey underneath in fresh specimens’) and AMNH 811713 (‘olive yellowish instead of grey underneath’). This character distinguishes these three taxa from all South American populations, but the yellow wash fades quickly after death. After a few years, museum skins of C. arcanus sp. nov. have no trace of yellow in the ventral plumage and the ‘olive’ fades to cooler than Light Drab (119C). To our knowledge, no fresh (<5 years old) material of C. arcanus sp. nov. or C. [ f. ] mirabilis exists in any collection. Voice: Catharus arcanus sp. nov. is distinguished from the South American taxa by its Type 1 punctuation call structure ( Fig. 9 ), and from C. [ f. ] mirabilis by its ‘tightly spaced’ contact call structure ( Fig. 10 ). Our dataset lacked recordings of blurred calls from C. arcanus sp. nov. , and too few recordings were available to adequately determine whether C. arcanus sp. nov. and C. [ f. ] mirabilis are divergent in song. Notwithstanding, the only triadic song contour (BCA) detected in C. arcanus sp. nov. , which accounted for 43% of songs ( N = 3/7), was not detected in C. [ f. ] mirabilis . Of the three tetradic contours detected in C. arcanus sp. nov. (CABD, CDAB, DBAC), only one (CDAB) was detected in C. [ f. ] mirabilis ( Fig. 12 ).

Diagnosis Genetics: In the UCE tree, samples from (i) western Ecuador and north-western Peru , encompassing the type localities of C. berlepschi Lawrence 1888 and C. f. caniceps Chapman 1924 , and (ii) eastern Peru from Amazonas south to the Río Apurímac ( C. b. nebulus ssp. nov.), formed reciprocally monophyletic clades. Together, these clades formed the sister-group of a clade composed of samples of C. [ f. ] opertaneus and Undescribed 2. We treat these major sister-groups as polytypic species: C. berlepschi (including C. b. berlepschi , C. b. caniceps , and C. b. nebulus ssp. nov.) and C. opertaneus (including C. [ f. ] opertaneus and Undescribed 2, see below). The subspecies C. b. berlepschi and C. b. caniceps were not genetically diagnosable, despite their plumage colour differences. A sample of C. b. berlepschi from the Yunguilla Valley, Azuay , Ecuador (ZMUC 128360) was more similar to a paratype of C. f. caniceps (AMNH 175531) than a sample of C. b. berlepschi from Imbabura , Ecuador (ZMUC 119570). In the ND2 tree, these three samples, plus one sample each from the Huancabamba and Chunchuca valleys (i.e. east of the Abra de Porculla, an east–west pass through the Western Cordillera at 2145 m ) and a sample from Loja , Ecuador , formed a clade with low within-clade divergence (mean uncorrected p -distance = 0.005 ± 0.004). Samples from across the range of the eastern Andean group ( C. b. nebulus ssp. nov.) also formed a clade in the ND2 tree with low within-clade divergence (mean uncorrected p -distance = 0.002 ± 0.001). The divergence between these sister-clades ( C. b. berlepschi + C. b. caniceps , vs. C. b. nebulus ssp. nov.) was an order of magnitude greater (mean uncorrected p -distance = 0.022 ± 0.002) than the mean divergence within either clade. ABGD analysis of ND2 data merged C. berlepschi with Undescribed 2 (uncorrected p -distance = 0.01 ± <0.01), although these were not sister-groups in the UCE tree, while treating C. b. berlepschi and C. b. nebulus ssp. nov. as independent genetic clusters. ASAP analysis merged C. b. berlepschi , C. b. nebulus ssp. nov., and Undescribed 2. Morphology: Study skins from the range of C. berlepschi in the western Andes were sexually monochromatic in the adult plumage and exhibited a north–south colour cline. At the northern end of the cline, adults had dark dorsal plumage (darker than Sepia, 119), medium-light ventral plumage ( Figs 15 , 16 ; Table 3 ), and completely black maxillae ( Table 4 ), whereas southern adults ( C. b. caniceps ), including in the Huancabamba and Chunchuca valleys, were paler on the flanks, breast, and throat ( Figs 15 , 16 ; Table 3 ) and had dichromatic irides ( Table 5 ). Specimens from Zaruma, El Oro (ANSP 83643, 83644) had somewhat intermediate characters between the endpoints of the plumage colour cline, and are here listed under the nominate form, C. b. berlepschi . Adult study skins of C. b. nebulus ssp. nov. were dichromatic in iris colour and more similar in plumage colour to C. b. caniceps than C. b. berlepschi . A juvenile specimen identified as C. b. nebulus was morphologically divergent from C. b. berlepschi and C. b. caniceps (see below), but this requires further investigation to confirm. Voice: The blurred calls of C. b. berlepschi and C. b. nebulus ssp. nov. had a ‘check-like’ sonogram shape, which distinguished them from all other taxa in the complex, including Undescribed 2, which had a downward sloping blurred call that was narrower in bandwidth ( Fig. 11 ). The subspecies C. b. berlepschi and C. b. nebulus ssp. nov. differed in punctuation call structure ( Type 2 vs. Type 3, respectively). In song, of the five triadic contours (ABC, ACB, BCA, CAB, CBA) detected in C. b. berlepschi , four (ABC, ACB, BAC, BCA) were shared with C. b. nebulus ssp. nov.; of the four tetradic contours (ACDB, BCDA, CABD, CDBA) detected in C. b. berlepschi , none was shared with C. b. nebulus ssp. nov., and only one (BDAC) was shared with Undescribed 2.

Comments We placed the name ‘ C. f. caniceps ’ into the synonymy of C. berlepschi because, in both reconstructions, genetic samples from western Ecuador and north-western Peru formed a clade that encompassed the type localities of both taxa, and the older name has priority. Based on the shallow ND2 divergence between samples from the western Andes, on one hand, and the eastern side of the Abra de Porculla in the Huancabamba (LSUMZ 170103) and upper Chunchuca valleys (LSUMZ 170104) on the other, we hypothesize that C. berlepschi crossed the Abra de Porculla, from the west, sometime after the LGM. Despite its monophyly in both genetic datasets, we treat the taxon from the eastern Andes as a subspecies of C. berlepschi , instead of its sister-species, because additional sampling is needed to assess the degree of vocal and morphological divergence. We note that C. b. nebulus ssp. nov. is apparently divergent from C. b. berlepschi in juvenile plumage and punctuation call structure, but this may be an artefact of low sample size. Study skins of C. b. nebulus ssp. nov. also exhibit subtle sexual dichromatism, unlike the sexually monochromatic (but clinally variable) C. b. berlepschi + C. b. caniceps clade. Males of C. b. nebulus ssp. nov. (e.g. FMNH 474413, LSUMZ 129006) were darker and less brown on the dorsal surface than adult females, including on the crown, and some adult males of C. b. nebulus ssp. nov. approached the brownish plumage of the adult female (e.g. LSUMZ 106496, with enlarged testes: 9 × 15 mm ). This suggests that C. b. nebulus ssp. nov. may exhibit polychromatic plumage ‘types’ like C. [ f. ] mentalis (see below). Notably, the strongest colour differences between C. b. nebulus ssp. nov. and the C. b. berlepschi + C. b. caniceps clade occur where their ranges approach each other in northern Peru , which may be indicative of character displacement. The plumage of a juvenile from central Peru (LSUMZ 174204), identified as C. b. nebulus ssp. nov., was boldly marked relative to a juvenile C. b. berlepschi (ANSP 181216). LSUMZ 174204 has tawny streaks on the mantle, tawny spots on the breast, a whitish throat and abdomen, and a pale mandible ( Fig. 19 ). At first glance, this specimen resembles the juvenile of C. maculatus , but there are multiple reasons to doubt that identification. CM 85107, a genuine C. maculatus juvenile from Cochabamba , Bolivia , has a dark mandible (vs. pale with dusky tip in LSUMZ 174204), a darker and more heavily marked belly (vs. white), and a brown ‘collar’ across the upper breast (vs. no collar). LSU 174204 was also collected at 2300 m , at the same site as an adult female C. b. nebulus ssp. nov. (LSUMZ 174203), and about 500 m upslope of the known upper range limit of C. maculatus in central Peru (Schulenberg et al. 2007). As noted by Ridgely and Tudor (1989: 110), C. maculatus and C. [ fuscater ] populations in South America ‘seem rarely if ever actually to occur together, rather, replacing each other in an as yet not understood way’. Freeman et al . (2022) played recordings of C. b. nebulus ssp. nov. in ‘territories’ of C. b. berlepschi and recorded the response of the ‘territorial bird’ (i.e. how closely the playback target approached the speaker, and whether and how much it vocalized). In each of eight different experiments, one recording was played in the vicinity of a singing individual, and three different stimuli were used overall (XC 175873 for two experiments, XC 30494 for three experiments, XC 144268 for three experiments). We reiterate that interpreting the results of these experiments (and ‘sympatric’ experiments using audio playback of C. b. berlepschi recordings in the range of C. b. berlepschi ) is not straightforward, because the function of song (and, therefore, any response to it) is not clearly defined in species that exhibit cooperative parental care and/or facultative territoriality (e.g. Goetz et al. 2003 , Halley 2014 , Greeney et al. 2015 , Halley et al. 2016 ).