Cyornis sumatrensis (Sharpe, 1879)

• C. sumatrensis . Indochinese jungle-flycatcher:

○ C. sumatrensis sumatrensis (Sharpe, 1879) – S Myanmar, S Thailand, Peninsular Malaysia and NE Sumatra.

○ C. sumatrensis indochina Chasen & Kloss, 1928 – SE Myanmar, Thailand (except C & S), Cambodia, S Laos and C Vietnam.

○ C. sumatrensis lamprus Oberholser, 1917 – Anamba Island (off E Peninsular Malaysia).

Although the Sundaic C. olivaceus is a widespread species and relatively common in Sumatra and Java, limited studies have been carried out on this jungle-flycatcher. Until recently, it was mistakenly treated as a Rhinomyias due to its seemingly different plumage in comparison to most other Cyornis species. Our bioacoustic analyses detected limited geographical variation between the Bornean and Javan populations and between the Sumatran andThai populations ( Table 3). However, strong vocal differentiation is absent among the four populations ( Fig. 1F View Figure 1 ), thus supporting the taxonomic treatment of a single species. Similarly, the greater Sundanese C. caerulatus is very poorly known in life, especially the Sumatran C. c. albiventer. Although our bioacoustic analyses did not detect strong vocal distinction between the Sumatra and Sabah populations, we encountered two different song types of C. c. caerulatus : one of which is a consistent four-element phrase, while the other consists of two to three elements in a descending phrase (see Fig. 1G View Figure 1 ).

CONSERVATION IMPLICATIONS

Overall, we propose taxonomic revisions to six out of the nine Cyornis species complexes in the present study, resulting in the elevation of ten taxa to species status. Our recommendation for a taxonomic elevation of C. kalaoensis generates an additional small-island endemic, but we do not believe that it would currently classify as threatened given that it is common in secondary habitat ( Eaton & Rheindt, 2017). However, limiting the distribution range of C. banyumas to Java calls for an urgent need to reconsider the conservation status of this endemic species. The Javan C. banyumas population was found to be declining as a result of the bird trade ( Eaton et al., 2015). The species is now very rarely encountered in the wild, and a network of Javan birders and professional bird guides only recorded the taxon a handful of times in 2018 (pers. obs.), thus the species may warrant elevation in conservation status to Critically Endangered according to criterion A2 ( IUCN, 2018).

Our study demonstrates that a relatively small panel of vocal parameters can be employed as a useful tool in cryptic species detection in Cyornis flycatchers, which are characterized by great levels of hidden species diversity ( Eaton et al., 2016; del Hoyo et al., 2018) and that their application was in agreement with genetic data (present data; Zhang et al., 2016; Garg et al., 2018). Using recordings collected by both professional and recreational ornithologists, and vetted by us for correct species identification (see Methods), we were able to detect vocal divergences, or a lack thereof, in various species complexes, thereby providing important taxonomic and conservation insights by flagging cryptic conservation units that may otherwise slip under the radar and face extinction threats. Despite great progress in the ongoing taxonomic revolution that incorporates bioacoustic and genomic methodologies, our study suggest knowledge gaps remain even in bird taxonomy, the best-studied animal group on Earth. Given that vocal traits are of critical importance in the maintenance of species integrity in songbirds, and can be even more appropriate for species delimitation than plumage in some groups ( Rheindt et al., 2008), we propose a routine application of simple vocal parameter sets – adjusted to taxonomic grouping – to scan for additional cryptic species-level diversity across songbirds. However, we emphasize that the application of such simple standard parameter sets, whether bioacoustic or morphological (e.g. Tobias et al., 2010), can only ever serve to flag units of interest and produce provisional new treatments, but not to override well-established treatments based on integrative genomic, bioacoustic and morphological data.