The bush dog Speothos venaticus occurs from Panama to southern Brazil and is naturally rare or difficult to observe in nature, making it one of the least known Neotropical canids (Ihering, Reference Ihering1911; Zuercher et al., Reference Zuercher, Swarner, Silveira, Carrillo, Hoffmann and Sillero-Zubiri2004; DeMatteo & Loiselle, Reference DeMatteo and Loiselle2008; Oliveira, Reference Oliveira2009). It is categorized on the global IUCN Red List as Near Threatened (DeMatteo et al., Reference DeMatteo, Leite-Pitman and Michalski2011) and as Vulnerable in Brazil, where the main threats are habitat loss and fragmentation, prey depletion from poaching, and transmission of disease from domestic dog (Oliveira & Dalponte, Reference Oliveira, Dalponte, Machado, Drummond and Paglia2008). In the threatened Atlantic Forest (Myers et al., Reference Myers, Mittermeier, Mittermeier, Fonseca and Kent2000) occurrence data for this species are extremely rare (Beisiegel & Zuercher, Reference Beisiegel and Zuercher2005; Beisiegel, Reference Beisiegel2009) and in southern Brazil are limited to a few records from > 15 years ago in the states of Santa Catarina (Ihering, Reference Ihering1911; Cherem et al., Reference Cherem, Simões-Lopes, Althoff and Graipel2004) and Paraná (Margarido & Braga, Reference Margarido, Braga, Mikich, Bérnils and Pizzi2004). Here, we document the occurrence of bush dogs in protected areas in a continuous remnant of coastal Atlantic Forest in Paraná using sightings and camera trapping, and reports from park guards.

As part of a project to examine the richness and distribution of medium- and large-sized mammals we have collected data on the occurrence of mammal species in four protected areas (in the category Reserva Particular do Patrimônio Natural) in the Área de Proteção Ambiental de Guaraqueçaba (Environmental Protection Area of Guaraqueçaba). This region lies in one of the largest continuous remnants of the Atlantic Forest in Brazil, connecting preserved forest areas across a coastal mountain range in the states of Paraná and São Paulo (Fig. 1). The four reserves are private areas comprising a total of 21,050 ha of forest on plains and hills (0–800 m altitude), monitored by park guards on a monthly basis. They are connected by a landscape of continuous forest where there are also small rural settlements close to the reserves. From March to September 2009 we placed camera traps on trails at 40 sites (500 m apart) in the 8,700 ha Reserva Natural Rio Cachoeira. We repeated this from April to October 2011, when we also surveyed the 3,300 ha Reserva Natural Morro da Mina, 6,650 ha Reserva Natural Serra do Itaqui and 2,400 ha Reserva Natural Salto Morato. In this second period we surveyed 46 sites (1.5–2 km apart), with camera traps installed both on and off trails (2,354 and 684 trap days, respectively; Table 1). Off-trail camera traps were installed on paths made by animals. All camera traps were installed 20–30 cm above the ground. Consecutive photos of the same species were considered independent records when the interval between them was at least 12 hours. Additionally, we gathered reports of observations of the bush dog from the park guards. Confirmation of species was by photographic identification and descriptions of morphological characteristics.

Fig. 1 The study area in a continuous remnant of coastal Atlantic Forest in Paraná state, southern Brazil (indicated by the rectangle on the regional map), with the location of the four reserves where surveys were conducted in 2009 and 2011. The black triangle and black circles indicate the locations of sighting and camera-trap records, respectively, of the bush dog Speothos venaticus.

Table 1 Sample effort and records of bush dogs Speothos venaticus in four reserves in the coastal Atlantic Forest, southern Brazil (Fig. 1).

* Reserves with previous sightings of bush dogs by park guards

The surveys of 2009 and 2011 resulted in a total of one sighting and three photographic records of the bush dog, in three of the reserves (Fig. 1, Table 1). We saw three adult individuals in Reserva Natural Rio Cachoeira at 09.45 on 14 September 2009. They were moving on the opposite margin of the Turvo River, at c. 5–8 m from us. That record was opportunistic, whilst we were walking the trails to remove our camera traps. In the same area park guards reported the sighting of a single individual in 2006 (J. Veiga & V. Veiga, pers. comm.).

The three photographic records were obtained between 07.00 and 11.00 at two sites on trails, parallel to and c. 50 m from a perennial river. In Reserva Natural Serra do Itaqui an adult individual was recorded on 2 consecutive days on 18–19 August 2011. In Reserva Natural Morro da Mina two adults were recorded (Plate 1) on one occasion on 6 October 2011. In all photographs the bush dogs were walking along trails. Despite our photographic records the park guards in Reserva Natural Serra do Itaqui reported no evidence of the species there. In Reserva Natural Morro da Mina two park guards reported a sighting of two bush dogs in 2004 (A. Gonçalves & A. Pinheiro, pers. comm.). In Reserva Natural Salto Morato we did not obtain confirmation by camera trapping that the bush dog is present, although a guard observed the species there in 2000 (P. Moraes-Filho, pers. comm.).

Plate 1 Camera-trap photograph of bush dogs Speothos venaticus in Reserva Natural Morro da Mina, Paraná state, southern Brazil (Fig. 1).

Lima et al. (Reference Lima, Pinto and Dalponte2009) and Beisiegel (Reference Beisiegel2009) recorded bush dogs crossing trails or roads but our photographic records indicate they also walk along trails. Our limited data support suggestions that the bush dog is diurnal (Beisiegel & Zuercher, Reference Beisiegel and Zuercher2005), moves or forages near water courses (Langguth, Reference Langguth and Fox1975; Zuercher et al., Reference Zuercher, Swarner, Silveira, Carrillo, Hoffmann and Sillero-Zubiri2004) and is rarely observed (Ihering, Reference Ihering1911; Barnett et al., Reference Barnett, Shapley and Engstrom2001; DeMatteo, Reference DeMatteo2008; Beisiegel, Reference Beisiegel2009; Michalski, Reference Michalski2010). Even using camera traps, which have been widely used to detect and study elusive species of mammals (O'Connell et al., Reference O'Connell, Nichols and Karanth2011), the probability of detecting bush dogs seems to be extremely low. Studies reporting photographic records of the bush dog had a large sampling effort and a low capture success: Michalski (Reference Michalski2010) 6,721 trap days, two records; Beisiegel (Reference Beisiegel2009) 4,818 trap days, one record; our study 4,112 trap days, three records. With such limited data, including untested and potentially incorrect assumptions (e.g. constant detection probability across space and time, positive linear relationship between photographic capture rates and true abundance; O'Connell et al., Reference O'Connell, Nichols and Karanth2011) it is difficult to make comparisons of occupancy or relative abundance between areas. Regardless of the method or the region studied, the difficulty of locating and studying the bush dog is presumed to be associated with a low population density, a large home range, and the secretive habits of the species (Zuercher et al., Reference Zuercher, Swarner, Silveira, Carrillo, Hoffmann and Sillero-Zubiri2004; Beisiegel, Reference Beisiegel2009; Lima et al., Reference Lima, Pinto and Dalponte2009).

Our records and the previous sightings by park guards indicate that the bush dog occurs in the four reserves surveyed, and presumably throughout the forest within which these areas lie. This suggests that the coastal Atlantic Forest of Paraná is an important region for the conservation of this species, is this forest lies within one of the largest remnants of this biome and includes a number of protected areas and remote places that are difficult to access. However, there are potential threats (prey depletion from poaching and transmission of diseases from domestic dogs; DeMatteo, Reference DeMatteo2008; Oliveira, Reference Oliveira2009) to the bush dog in the region, with evidence of hunting pressure and of the occurrence of domestic dogs within and outside the protected areas (Andriguetto-Filho et al., Reference Andriguetto-Filho, Krüger and Lange1998; pers. obs.). Our results reinforce the importance of these four reserves, which bush dogs may be using as shelter for feeding and reproduction. Given the apparent rarity of this species we recommend that long-term efforts are taken to guarantee its conservation in this region, in particular the reduction of poaching and control of domestic animals. These recommendations have been passed to the managers of the reserves, who work with the environmental police to reduce threats to wildlife.