Justification of Red List Category
Recent records from Brazil indicate that this species's status may be marginally better than previously thought. Nevertheless, the remaining known population is still extremely small and fragmented, and the perturbation, damming and pollution of rivers are likely to be causing continuing declines. For these reasons, it is listed as Critically Endangered. Further information on the population size (in particular whether it exceeds 250 mature individuals) and on the subpopulation structure may result in its downlisting to Endangered in the future.
The population was estimated at 250 individuals in 1992, and was thought likely to have declined since, given ongoing threats. However, there are recent suggestions that the population may exceed this figure (L.V. Lins in litt. 2012). Recent estimates from the three main areas currently known to hold the species are of 70-100 territories (140-200 mature individuals) in the Serra da Canastra area (L.V. Lins in litt. 2012, 2013), fewer than 50 individuals at Chapada dos Veadeiros (Disconzi 2012) and eight at Jalapão (IECOS Brasil 2013 per L.V. Lins in litt. 2013, Barbosa et al. 2015), but these figures require confirmation. The population is currently precautionarily maintained within the band 50-249 mature individuals.
This species's population is suspected to have declined rapidly over the last 20 years (three generations), in line with habitat loss and degradation within its range, owing to the expansion of hydroelectric power schemes, soy bean cultivation and mining operations. It appears to have been extirpated from Paraguay.
The species occurs in extremely low numbers at a few highly disjunct localities in south-central Brazil, northeast Argentina and possibly Paraguay. The species's stronghold is in and (mostly) around Serra da Canastra National Park, Minas Gerais, Brazil (Lamas 2006, Lins et al. 2013), where recent surveys yielded a rough estimate of 70-100 territories, roughly equivalent to 140-200 mature individuals (L. V. Lins unpubl. data); if confirmed this would represent a significant increase to the size of the largest known subpopulation. Tributaries of the Rio São Francisco in west Bahía were thought to hold a significant population (Pineschi and Yamashita 1999), but a 2003 survey there failed to locate any birds. It has recently been found in Patrocínio municipality, Minas Gerais (I. Lamas in litt. 2012); a record from Itacolomi State Park in the same state is though thought to refer to an accidental or escaped bird (Arvelino de Paula 2008, L. V. Lins in litt. 2012). In Goiás, there are records from Emas and Chapada dos Veadeiros national parks; the latter being surveyed in 2003-2004 with birds found just outside the park in the Rio das Pedras, and the total in the area thought to number fewer than 50 individuals (Disconzi 2012). In 1995, a small population was discovered on the Rio Tibagi, Paraná (Anjos et al. 1997), but searches in 1998 were unsuccessful (L.F. Silveira in litt. 1999). In 2002, another small population was discovered on the Rio Novo, in Jalapão State Park, Tocantins (Braz et al. 2003), and six expeditions in 2007 and 2008 surveying a c.55 km stretch of the Rio Novo located three breeding pairs (Barbosa and Almeida 2010), and four pairs along 115 km of river in 2010-2011 (IECOS Brasil 2013 per L.V. Lins in litt. 2013, Barbosa et al. 2015). It is believed extinct in Mato Grosso do Sul, Rio de Janeiro (Pacheco and Fonseca 1999), São Paulo, and Santa Catarina. In 2017, a male was photographed in Serra do Mar State Park, Salesópolis, São Paulo, but this may have been a vagrant individual (Neto et al. 2018). In Misiones, Argentina, 12 individuals were found on the Arroyo Uruzú in 2002, the first records in the country for 10 years despite extensive surveys (Benstead 1994, Hearn 1994, J. C. Chebez in litt. 1999). In Paraguay, it was last recorded in 1984 and there is little (if any) habitat left. However, local reports indicate that a few individuals may still survive (R.P. Clay in litt. 2003).
It inhabits shallow, fast-flowing rivers, requiring rapids and clear waters. It occurs especially in the upper tributaries of watersheds, but ranges into small rivers with patches of gallery forest surrounded by 'cerrado' (tropical savanna) or within Atlantic Forest. It is non-migratory and does not abandon the stretch of river where it establishes its territory (Lamas 2006). Pairs have used 8-14 km stretches of river (Bartmann 1996, L. F. Silveira in litt. 1999), nesting in tree-cavities and rock-crevices (C. Yamashita in litt. 2000, Lamas and Santos 2004, Bruno et al. 2010, Barbosa 2018) as well as riverbank holes (Carboneras et al. 2016). Breeding activity has been recorded between June and August (Lamas and Santos 2004, Bruno et al. 2006, 2010), but timing may vary geographically. Incubation may last c.33 days (Bruno et al. 2010). Young birds have been observed between August and November (Lamas and Santos 2004, Bruno et al. 2006). The diet comprises fish, small eels, insect larvae, dobson flies (Corydalis spp.) and snails. In Serra da Canastra it eats mainly lambari Astyanax fasciatus. Territory size is believed to be related to the number of rapids, edgewaters, water speed, fish abundance and conservation of riparian vegetation (Lamas 2006). Its dispersal ability is unknown, but one young male banded in September 2010 was found breeding on another river 20 km away in June 2011 (Ribeiro et al. 2011).
Perturbation and pollution of rivers results largely from deforestation, agricultural expansion and, in the Serra da Canastra area, diamond-mining (Bartmann 1994, 1996; M. Diniz in litt. 2013). Previously, the species was thought to rely on gallery forest which, although protected by law in Brazil, has been cleared illegally throughout much of the species's range. However, evidence suggests it will occur on unforested, undisturbed stretches of river through 'cerrado'. All recent records of the species in the Serra da Canastra region refer to unprotected sites north of the National Park. These are sites under increasing pressure from mining, development of hydropower infrastructure and agriculture (L. V. Lins in litt. 2013). Mining has ceased in the immediate area of its known range, but there is no additional habitat for dispersing birds (L. F. Silveira in litt. 1999), and it is thought that diamond mining will resume at Serra da Canastra in the near future (L. V. Lins in litt. 2012). Expanding agriculture and the construction of hydroelectric dams are considered the principal threats to the species (Braz et al. 2003). Dam-building has already caused severe declines across much of its range, and is increasing in scale (L. V. Lins in litt. 2012). Tourist activities result in river perturbation and have been recorded within known territories and inside national parks (Ibama 2006).
Conservation Actions Underway
The species is legally protected in all three range states. It occurs in three Brazilian national parks, two state parks and one private protected area (Braz et al. 2003), although there are no recent published records from Emas National Park. A species action plan has been published which outlines in detail its current status, ecology, threats and proposed conservation actions. It is considered nationally Critically Endangered in Brazil (Instituto Chico Mendes de Conservação da Biodiversidade 2018). In Argentina, sections of the Arroyo Uruzú are protected within the Uruguaí Provincial Park (P. Benstead verbally 2004).
Regular monitoring of the population is carried out in Serra da Canastra National Park, Brazil. Recent research has increased knowledge of the species's movements and ecology (Braz et al. 2003, WWT 2008, Lins et al. 2013), sexual maturation and dispersal ability (Ribeiro et al. 2011), and reproductive ecology (Ribeiro et al. 2018). Further research is underway to learn more about the species's behaviour, ecology, genetics and conservation actions (Instituto Terra Brasilis n.d., Maia et al. 2017).
Nest boxes have been installed within the Serra da Canastra National Park (L.V. Lins in litt. 2012). A captive breeding programme was initiated in 2011 at the Poços de Caldas Breeding Center in Minas Gerais. In 2017, it was announced that four young had hatched at a zoo in Itatiba, Sao Paulo State (Agencia Brasil 2017). An environmental education project began in 2004 and has been implemented at São Roque de Minas and Bonita. The Pato Aqui Água Acolá Project has been engaging with students in the Serra da Canastra region to improve understanding of environmental problems related to the Cerrado biome (Instituto Terra Brasilis 2014a). The project has set up the Conscious Decision Campaign to engage with those responsible for environmental licensing, construction and tourism activities that could have an impact on water quality within the species's range (Instituto Terra Brasilis 2015). The project is also working to promote sustainable land-use by farmers across the São Francisco river basin, with the aim of reducing pollution and silting of rivers, and is promoting ecotourism as a sustainable source of income (Instituto Terra Brasilis n.d.).
Conservation Actions Proposed