During field surveys of dpto. Beni in 2003–05 (Tobias & Seddon 2007), I visited a range of habitats close to the main highway from Riberalta to Guayaramerin, an area put on the ornithological map by the late Sjoerd Meyer, who rediscovered Masked Antpitta Hylopezus auricularis in swampy forest by the río Beni on the outskirts of Riberalta in 1994 (Meyer 1998). Meyer also surveyed the road to the Brazil border from Riberalta, finding some scarce grassland species, including Rufous-sided Pygmy Tyrant Euscarthmus rufomarginatus, in campo at San Lorenzo de Pampa. Subsequent visitors to the area focused on these localities—in easy reach of Riberalta—but tended not to venture further towards Guayaramerin where different habitats appear.

Approaching Guayaramerin on this road, I noticed a small area of dense stunted forest, similar to habitats known in Brazil as campina. I spent two mornings in April 2005, and another three mornings in October 2005, mist-netting and making sound-recordings at the most accessible patch of habitat, c.16 km from Guayaramerin at 10°55.027′S, 65°26.805′W (referred to as Guayaramerin site B in Tobias & Seddon 2007). During these visits, I made the first documented records for Bolivia of Green-tailed Goldenthroat Polytmus theresiae, Pale-bellied Mourner Rhytipterna immunda and Ruby-topaz Hummingbird Chrysolampis mosquitus, as well as the first records of Black Manakin Xenopipo atronitens and Red-shouldered Tanager Tachyphonus phoenicius for dpto. Beni (Tobias & Seddon 2007).

During my second visit, on 25 April 2005, I trapped an adult Turdus sp. with a black bill in shrubby vegetation with dwarf trees (canopy height 1–5 m) interspersed with rough grassland. This was the lowest-stature vegetation at the site, and supported a relatively small number of bird species, including Green-tailed Goldenthroat, Red-shouldered Tanager and White-fringed Antwren Formicivora grisea rufiventris. In the same habitat, I had recently trapped 2–3 Creamy-bellied Thrushes Turdus amaurochalinus, migrants or non-breeding visitors from a breeding range further south, easily identified by their pale yellowish bills. The bird with the black bill was slightly unfamiliar in overall pattern (Fig. 1), so I considered all possible options. It was clearly not a Hauxwell's Thrush T. hauxwelli, which has a browner bill and more rufous plumage, and is generally restricted to the lower storey of tall humid forest. Bill colour also eliminated Unicoloured Thrush T. haplochrous, a mysterious Bolivian endemic that I had searched for unsuccessfully on several occasions in dpto. Beni. The individual appeared most similar to Black-billed Thrush T. ignobilis, which is common nearby in forest edge, overgrown clearings and gardens. Nonetheless, the stunted vegetation where it was mist-netted seemed anomalous, lacking the usual array of species found alongside Black-billed Thrush. I was also struck by the plumage tones and head pattern, leading me to wonder whether it might be a different subspecies than I was used to.

Figure 1.

Adult Campina Thrush Turdus arthuri, near Guayaramerin, dpto. Beni, Bolivia, October 2005 (Joseph A. Tobias)

I was unable to determine the subspecies based on available literature, so the images remained labelled simply as T. ignobilis for 15 years. In 2020, I finally had time during the Covid-19 pandemic to transfer my records to eBird, and started to work on a checklist for ‘site B’ from notebooks and images ( https://ebird.org/checklist/S66602837). When I uploaded the images of T. ignobilis as media files, I was again surprised at the bird’s unfamiliar appearance. By this time, Black-billed Thrush had undergone recent taxonomic revision with some populations split as Campina Thrush T. arthuri (Cerqueira et al. 2016). Given that the habitat I surveyed in 2005 was structurally very similar to types of Brazilian campina, it seemed likely that the thrush I trapped was T. arthuri and comparison of plumage features confirms this. The key features that separate this individual from T. ignobilis are the darker face with dusky or slaty feathering around the eyes, well-defined streaking or flecking evenly spaced on the pale throat, and the cold grey head, wings and underparts (Fig. 1). In comparison, the local race of Black-billed Thrush T. ignobilis debilis has slightly warmer ‘mousey-brown’ plumage, a less dusky face (Figs. 2–3), and the throat is either largely unstreaked in the centre or a more uniform smudgy brown, similar to the rest of the underparts (Figs. 2–4).

Although cryptic and unremarkable in plumage, T. arthuri has a very distinctive jay-like call. I did not notice this in the field and made no recordings that can be ascribed to the taxon. Intriguingly, T. arthuri appears to sing only rarely, if ever, and there are as yet no confirmed recordings for this species of the complex songs known in all closely related taxa.

Figure 2.

Black-billed Thrush Turdus ignobilis debilis, Los Amigos Research Station (CICRA), Madre de Dios, Peru; the lower throat of this subspecies is often white and unstreaked (Joseph A. Tobias)

Figure 3.

Black-billed Thrush Turdus ignobilis debilis, Amazonia Lodge, Madre de Dios, Peru (Joseph A. Tobias)

Figure 4.

Black-billed Thrush Turdus ignobilis, Anorí, Antioquia, Colombia; this individual illustrates the characteristics of subspecies ignobilis and goodfellowi, which have darker throats, lacking white, creating a much more uniform impression to the underparts (Joseph A. Tobias)

Biogeographical considerations.—Previous records of T. arthuri are from southern Venezuela, Guyana, Suriname, and Amazonian Brazil south-west as far as Rondônia (Cerqueira et al. 2016). My record from near Guayaramerin extends the range another 250 km to the south-west, into Bolivia. The presence of T. arthuri in Bolivia makes biogeographic sense because campina specialist birds are generally distributed widely albeit very patchily over the Brazilian Shield, which here extends narrowly over the upper rio Madeira into Bolivia (Moskovits et al. 2003). Several species, including Citron-bellied Attila Attila citriniventris, Rhytipterna immunda, Tachyphonus phoenicius and Xenopipo atronitens, have similar patchy distributions across weathered formations of the Brazilian Shield, with a few records into easternmost Bolivia.

Further field work may reveal that the range of T. arthuri extends further south into eastern dpto. Santa Cruz given that other stunted forest species (e.g., Tachyphonus phoenicius, Xenopipo atronitens) occur in Noel Kempff Mercado National Park (Killeen & Schulenberg 1998). However, T. ignobilis was not recorded during intensive field work—including mist-netting—in and around this protected area (Killeen & Schulenberg 1998), suggesting that T. arthuri must be rare and local if it does indeed occur there. The species may also occur in dpto. Pando, where patches of similar habitat are widely scattered and poorly surveyed.

The stunted forests of dpto. Beni are not referred to locally as ‘campina’ and they appear to grow on different substrates to most Brazilian campinas. The ground at ‘site B’ was sandy in a few patches, but mainly covered in loose iron-rich volcanogenic deposits, suggesting a different geological history (Tobias & Seddon 2007). Despite this difference, the vegetation structure was reminiscent of campina growing on poorly drained, weathered clays or ancient rock outcrops more widely in South America. This retrospective record from Guayaramerin confirms that T. arthuri is widespread in a variety of different ‘stunted forest’ types or shrubby thickets across the Guiana and Brazilian Shields, being replaced on younger, more fertile, sedimentary soils in much of western Amazonia by T. ignobilis debilis.