New distribution records of Anairetes alpinus ( Carriker , 1933 ) ( Aves , Tyrannidae ) , a highly specialized and endangered bird of Polylepis forests in Junín , Peru

The Ash-breasted Tit-tyrant, Anairetes alpinus (Carriker, 1933) (Aves, Tyrannidae), is an endangered and highly specialized bird that inhabits the Polylepis forests (Rosaceae) of Peru and Bolivia. Here, we report 2 new localities in the department of Junín, central Peru, where it was detected in forests dominated by Polylepis canoi, P. rodolfo-vasquezii, and a third undescribed Polylepis species. We conclude that a residential population is the more likely possibility although the possibility of dispersal individuals should also be considered.


Introduction
The Ash-breasted Tit-tyrant, Anairetes alpinus (Carriker, 1933) (Aves, Tyrannidae), is a highly specialized bird species that inhabits the Polylepis forests (Rosaceae) of Peru and Bolivia (Fjeldså 1993, Fjeldså 2002, Lloyd 2008a, b, Birdlife International 2018).Given its small geographic range of occurrence and small population size, currently estimated at 150-700 individuals, this species has been categorized as Endangered in the IUCN Red List of threatened species (Birdlife International 2018).Its restricted elevational range (3,700-4,600 m) (Schulenberg et al. 2007) makes this species particularly vulnerable to various human activities that occur at these elevations, such as heavy grazing, uncontrolled fire, and local changes in land use (Fjeldså andKessler 2004, Gómez 2009).These activities have been postulated as the main drivers of the ongoing habitat loss and degradation (Renison et al. 2002, Purcell and Brelsford 2004, Renison et al. 2010, Sylvester et al. 2017), simultaneously preventing the natural regeneration of Polylepis (Gómez 2009, Morales 2017, Morales et al. 2018).
In Peru and Bolivia, roughly 3 main areas where this species occurs have been identified (Fig. 1).The northernmost population occurs along the high Andes of northern Peru, mainly in the mountain chains of the Cordilleras Blanca, Huayhuash, La Viuda, Central, and Chonta (Sevillano-Ríos et al. in press).There, A. a. alpinus inhabits small, high-elevation Polylepis patches of apparently high-quality habitat (Sevillano-Ríos and Rodewald 2017), where several resident groups of up to approximately 10 individuals can easily be observed (Sevillano-Ríos 2016, CSSR per.obs.).In southern Peru, in the departments of Abancay and Cuzco, A. a. bolivianus occurs along the Cordilleras Vilcabamba and Vilcanota (Lloyd 2008 a, b, Lloyd andMarsden 2008).Here, some of the principal populations are estimated to be about 20-30 individuals (Engblom et al. 2002), although the densities seem to be low (ca 4.9 individuals/km 2 ) (Lloyd 2008a).Finally, the southernmost populations occur mainly in the Cordilleras Real and Apolobamba, Bolivia, where A. a. bolivianus also occurs; it was estimated to be roughly 150 individuals after visiting 5 different localities (Gómez 2009).
These 3 main areas maintain the core populations of the A. alpinus, where different conservation strategies have been proposed to save this and many other threatened birds of the Polylepis forests (BirdLife International 2018).However, areas in between these core populations are poorly studied.Our discovery of new localities is important, not only for a better understanding of the distributional patterns of this species but to understand its ecology and to locate new areas for its conservation.We report 2 new localities in the department of Junín, central Peru.

Methods
To assess the diversity of high Andean birds in central Peru, fieldwork was conducted between May 2016 and October 2017 at 23 locations (17 in the department of Junín, 3 in Huancavelica, and 3 in Lima).At all these sites, typical high-Andean ecosystems (3,450-4,550 m above sea level) such as Polylepis forest, mixed shrublands, and puna pastures were surveyed.Each site was visited for 3 or 4 consecutive days twice over 2 years.Point counts of 10 minutes duration were made, separated by a minimum distance of 150 m.These took place in the mornings (06:00-12:00) and afternoons (14:00-18:00) and were complemented by mist-netting efforts (12 m; n = 2) placed in open areas within the forests to passively capture passing birds.

Results
New records.Peru, Junín, province of Concepción, district of Andamarca, Alhuanya (11°41ʹ30ʺ S, 074°54ʹ01ʺ W, 4100 m elev.), observ.by H.R.Quispe-Melgar, 13 The specimen of A. alpinus captured in Alhuanya was measured (Table 1) and photographed (Figs 2, 3 The second observation was in a 5 ha patch dominated by Polylepis canoi (Figs 4,5).On this occasion, 2 individuals were seen together, moving from branch to branch through the Polylepis canopy.The humidity seemed to be higher in this forest compared to Alhuanya, although the vegetation appeared similar, and we also observed additional species such as Escallonia, Hesperomeles, and Clethra, with few signs of human activity.
Although intensive searching was conducted in both places, including the use of playbacks during the following days, we failed to record the bird again.The 2 localities were approximately 28 km apart.Anairetes alpinus is distinguished from its congeners by having a plain ashy color pattern, 2 pronounced white wing bars, absence of stripes on the chest, a white patch on the belly, and 3 outer rectrices with white coloration (Carriker 1933).However, the identification of the subspecies A. a. alpinus and A. a. bolivianus is more complicated, given that an in-hand comparison of specimens seems to be required to observe the degree of black coloration on the crown and the white central belly, among other distinguishing characteristics (Carriker 1933, 1935, Fitzpatrick and Sharpe 2017).

Discussion
Prior to our new records from Junín, the presence of this species in this region was unknown.These records, therefore, are important for understanding the distributional pattern of this endangered species.These records might represent non-resident individuals that were dispersing between known main populations or they might serve as evidence of a resident population in Junín that was previously overlooked.
The first hypothesis is supported by the fact that in places where this species is resident, such as Ancash, Cusco, and La Paz, it tends to be locally common and easily detected (Lloyd 2008a, Gómez 2009, Sevillano-Ríos and Rodewald 2017).For example, in the Cordillera Blanca, A. alpinus occupies the same Polylepis forest patches year after year, where a few groups of normally 4 or 5 individuals are often observed foraging only in Polylepis trees.These observations suggest a high degree of specialization and territoriality, which explain why they easily respond to playback through vocalizing from the top of the trees, fluttering, and short display flights (Servat 2006, CSSR unpublished data).In Junín, however, though we used playback and searched intensively throughout the forest, we failed to detect the species more than once.Additionally, no evidence (e.g.brood or individuals carrying nest materials) suggests a breeding population in these areas, which opens the possibility that these records are non-resident individuals dispersing between main populations (but from where?).
If we observe the spatial distribution of the new records with respect to the known main populations, we see that they are located between the A. a. alpinus (northwest) and A. a. bolivianus (southeast) resident populations (Figure 1).The closest of these main populations are approximately 168 km west in Carampoma, department of Lima (11°37ʹ34ʺ S, 076°26ʹ09ʺ W) and about 290 km northwest in Aquia, department of Ancash (10°02ʹ09ʺ S, 077°08ʹ38ʺ W); both belong to the subspecies A. a. alpinus.On the other hand, the closest populations of A. a. bolivianus are at approximately 265 km southeast in the Cordillera Vilcabamba, department of Cuzco (13°21ʹ03ʺ S, 073°07ʹ12ʺ W).In general, the regions of Lima, Ancash, and Cuzco have a high concentration of Polylepis forests where the populations of A. alpinus may be functioning as "sources" given that, at least, they look like stable populations.The areas in between, though not devoid of Polylepis patches and other vegetation communities that include Gynoxys, may only be suitable for the dispersal of A. alpinus but unsuitable for a stable population.Certainly, A. alpinus does forage on other plant species like Gynoxys, Baccharis, and Buddleia, among others (Servat 2006, Gomez 2009), and it appears that the degree of specialization to Polylepis decreases with latitude (Servat 2006, CSSR unpublished data).Although these records might be non-residential individuals dispersing between main populations; the large distance in between, their small population sizes, and the high specialization to Polylepis makes this claim dubious.
On the other hand, the possibility that the new records are from a small, previously unreported population is supported in that we detected A. alpinus only at 2 of the 23 sites.This agrees with this species' inherent pattern of rarity observed in other areas such as Ancash, Cusco, and La Paz (Lloyd 2008a, Gómez 2009, Sevillano-Ríos and Rodewald 2017) and might suggest that it was overlooked because of its rarity.Indeed, studies in Ancash, where this species is known to occur, failed to detect them even after intensive survey effort (Sevillano-Ríos et al. 2011, Servat 2006).In Ancash, A. alpinus mainly occurs in very high elevation Polylepis forests (> 4,200 m) while those surveys were placed on the extensive Polylepis forests that occur at lower elevations (Sevillano-Ríos and Rodewald 2017).Unpublished observations suggest that interspecific competition between A. alpinus and Anairetes nigrocristatus Taczanowski, 1884 (Black-crested Tittyrant) could be responsible for this pattern in Ancash.In Junín, these 2 species also occur (Schulenberg et al. 2007, Sullivan et al. 2009, CSSR unp. Obs.) and it is reasonable to think that these records correspond to a resident population of A. alpinus that might inhabit higher elevations that we had not surveyed.Moreover, while the record of 1 individual could be used to support the dispersal hypothesis, it seems unlikely that 3 individuals were dispersing through a whole new area almost at the same time.This is especially unlikely considering their high degree of specialization and territoriality.
In the absence of genetic material, dispersal data, or an adequate plumage comparison to subspecies , it is impossible to know if these individuals are dispersing or residents.Still, we believe that a resident population is more likely, although the possibility of a dispersal corridor should also be considered.If these records correspond to an undescribed resident population, only further exploration or records could confirm this.Nevertheless, if we consider the possibility that Junín it is part of a biological corridor, the importance of these areas in between would be as sink habitats.That could function as corridors that connect and prevent populations from inbreeding depression, which is important if the objective is the conservation of endangered species that live in highly fragmented habitats such as Polylepis forests (Bennett 1999, Christie andKnowles 2015).Another area that we should consider for further studies (because it could serve as a corridor between Cuzco and La Paz), is the one formed by the Cordillera Carabaya, right next to Cordillera Apolobamba, along the eastern Andean slope between Peru and Bolivia.This suggestion is based on the record of one isolated individual of A. alpinus in Sandia (Maldonado pers.comm.).Finally, studies that integrate dispersion, genetic, and population demography data at a regional scale would help us to understand more about the ecology of this little-known species of threatened birds and improve current conservation efforts.

Figure 1 .
Figure 1.Records of the presence of Anairetes alpinus.Triangles indicate the presence of the A. alpinus alpinus while the circles indicate A. alpinus bolivianus.Records were taken from the eBird database (eBird 2017), GBIF (GBIF Secretariat 2017) and Greeney (2013).The white (Alhuanya) and black (Tasta) asterisks indicate the new records in department of Junín, central Peru.
).The bird was identified as an adult of undistinguished sex, with no signals of molting in the wings although other parts of the body were not assessed.On the wings, lice were detected.The bird was captured in a mist net located in an open area inside a 1.5 ha mixed patch dominated by Polylepis rodolfo-vasquezii and another undescribed species of Polylepis.The nearby area was characterized by open and shrubby vegetation composed by Buddleja, Gynoxys, Miconia, and some single Polylepis stands surrounded by puna pastures.Signs of human activities where evident all over the area.