Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil

Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Revista Brasileira de Ornitologia, 23(4), 417-427 ARTICLE December 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil 1,3,5 1,2 4 Marcelo O. Barbosa , Renato T. Pinheiro , and Karlla V. C. Barbosa Programa de Pós-graduação em Ecologia de Ecótonos – Mestrado, Universidade Federal do Tocantins – UFT, Rua 7, Qd. 15, s/n, Jardim dos Ipês, CEP 77500-000, campus de Porto Nacional, TO, Brazil. Professor Adjunto, Universidade Federal do Tocantins – UFT and Grupo de Pesquisa em Ecologia e Conservação de Aves – ECOAVES(UFT), Av. NS 15 ALCNO 14, Estação Experimental, 109 Norte, CEP 77001-090 campus de Palmas, TO, Brazil. Instituto Natureza do Tocantins – NATURATINS, Gerência de Unidades de Conservação, Quadra 302 norte, Alameda 2, Lote 3A, CEP 77006- 336, Palmas, TO, Brazil. Museu de Zoologia da Universidade de São Paulo, Avenida Nazaré 481, Ipiranga, CEP 04263‒000, São Paulo, SP, Brazil. Corresponding author: mobarbos@yahoo.com.br Received on 17 March 2015. Accepted on 15 December 2015. ABSTRACT: The Brazilian Merganser is a threatened species and c lassified as Critically Endangered. Its world population is estimated to be fewer than 250 individuals surviving in the wild. Currently, the species is known from disjunct populations distributed in the states of Minas Gerais, Goiás, and Tocantins (in the Jalapão region). For this study, the stretches of the Preto, Novo, Soninho, and Sono Rivers were surveyed by airboat to better estimate the species distribution in the region. Brazilian Mergansers were found only on the Novo River along a stretch upstream from Da Velha waterfall. Four pairs and a few other adult individuals were identified there. A 50 km stretch of the Novo River was selected for monthly censuses (n = 13) from August 2009 to October 2010 to estimate the relative abundance and population size of the Brazilian Merganser. The A bundance Kilometric Index was 0.125 individuals/ km. The Novo River, from Da Velha waterfall until near its headwaters, is 145 km long and our survey estimated a population of 13.7 individuals. Changes caused by uncontrolled tourist activities on the Novo River, such as rafting during the species’ breeding season, riparian forest deforestation, and proposed hydropower projects, can negatively impact this small Brazilian Merganser population putting it in great risk of local extinction. KEY-WORDS: Brazilian Merganser, censuses, river width, Novo River, relative abundance. INTRODUCTION Historically, this species occurred in Paraguay, Argentina and Brazil, but is presently only found in the The Brazilian Merganser (Mer gus octosetaceus Vieillot, latter country. In Paraguay, the species occurred at the 1817) is a globally and locally endangered species headwaters of the Paraná river basin, where it is currently (MMA 2014, IUCN 2015). It is one of the most considered to be extinct in the wild. In Argentina, the species occurred within the Paraná river basin in Misiones, threatened waterfowl species of the Neotropical region but the last sightings there and within that country as a (Bartmann 1988, Collar et al. 1992), with probably less than 250 individuals surviving in the wild (Birdlife whole were made in 2002 (Giraudo & Povedano 2003). 2015). Monogamous and sedentary, it feeds on small In Brazil, the species exists as disjunct small populations, fish (Characidae), which are captured during fast dives occurring in tributaries of the upper São Francisco River, in the state of Minas Gerais (Silveira 1998, Silveira & (Silveira & Bartmann 2001). Its nests are made in tree Bartmann 2001, Lamas & Santos 2004) and right bank or rock cavities (Bruno et al. 2010, Barbosa et al. 2011) right beside water bodies (Lamas & Santos 2004), and its tributaries of the Tocantins River Basin, in the states breeding season usually ranges from June to September. It of Goiás and Tocantins (Braz et al. 2003, Bianchi et al. inhabits only shallow, clear water rivers or small streams, 2005, Barbosa & Almeida 2010, Barbosa et al. 2011). The first recor ds of the species at Jalapão State Park (Parque with stretches of rapids, and bordered by riparian Estadual do Jalapão), eastern Tocantins, were made as vegetation. Due to its low population densities and high sensitivity to environmental disturbance, the species recently as 2002 along the Novo River (Braz et al. 2003), is most threatened by the loss of suitable habitat and/ a tributary of the Sono River basin, at the border of the or reduction of water quality, complicating the species’ Jalapão State Park (Figure 1), where it was subsequently recorded (Pacheco & Olmos 2010). Between 2007 and conservational status. Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa 2010 the species was studied along the Novo River by first we have studied its occurrence and distribution on the author (Barbosa & Almeida 2010, Barbosa et al. 2011). main suitable rivers present in the region, estimating Considering the critical status of the Brazilian Merganser population sizes and evaluating the local availability of and the little knowledge about it at the Jalapão region, the species’ habitat requirements. FIGURE 1. Map of the study area showing the distribution of four pairs of the Brazilian Merganser Mergus octosetaceus identified a long the Novo River during this study as well as a pair on the Sono River (Pacheco & Olmos, 2010). The stretches of rivers surveyed are highlighted in black and grey. Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa METHODS 46°37'24.75''W) and extending toward its confluence with the Sono River. The different distances surveyed were determined by habitat suitability along each river as Study area well as access and navigability issues. The study was conducted at the Jalapão State Park, located in the eastern portion of the State of Tocantins Characterization of the rivers (10°30'34.87''S; 46°34'36.40''W). The region where the In order to characterize each section of the surveyed rivers, Jalapão State Park was created represents one of the best we measured variables of the environment and structural preserved areas of Cerrado in Brazil and it is formed by the different protected areas that together constitute a differences each 1.5-2.0 km a long the Novo (28 points in total area of roughly two million hectares. The largest 50 km), Preto (12 points in 24 km) and Soninho/Sono protected areas are Jalapão State Park (158.885 ha), Serra Rivers (14 points in 27 km). Geographic coordinates and elevation were taken using a Garmin Map60CSx GPS, Geral do Tocantins Ecological Station (645.378 ha), and this being the second measure used to verify the declivity Nascentes do Parnaíba National Park (645.378 ha). The climate in the region has two markedly distinct (relative drop in elevation) for each stretch of the rivers seasons, with a rainy season extending from October sampled. Variables such as vegetation type and clarity to April and a dry season from May to September. The of the water, were used to further characterize each river section. We photographed river banks and satellite images vegetation is Cerrado with prevalence of grasslands, which were used to aid vegetation analyses at each sampling grow on poor (oligotrophic) and sandy soils. The region is characterized also by the presence of plateaus that can point. Vegetation types were characterized on both banks reach 800 m, where the headwaters of many watercourses according to the classification of Coutinho (1978), i.e., that flow to the Tocantins river basin are located (Von campo limpo (open fields), campo úmido (wetland), campo Cerrado (cerrado field), cerrado sensu stricto, mata ciliar Behr 2004). (riparian forest), and cerradão (woods). The clearness of Local range the water was classified in three categories, accor ding to the capacity for clear vision of the river bottom, and We conducted searches for this species along the Novo, ranging from low visibility (dark water) to moderate Preto, Soninho and Sono Rivers (this latter river named visibility and high visibility (extremely clear water). after the confluence of Novo and Soninho Rivers) b y To test how the structural differences on the landscape boat. The stretches of Novo and Soninho rivers were may affect the presence of the species, we generated surveyed during two days whereas the Preto and Sono probabilistic models based on the presence/absence of the along a single day only. These rivers were selected because species at the survey points in relation to the following they present potential habitat conditions for the Brazilian variables: river width, type of flow and flow level. The Merganser such as clear water, rapids, and well-preserved width of the river was measured through analyses of high- vegetation along their banks. The choice of river sections resolution satellite images. Type of flow was classified as sampled was made based on: 1) previous visits to the “remanso” (slowly-flowing water) or “corredeira” (rapids), study area; 2) previously known observation localities for and the level of flow was obtained by eye according to the the species (Braz et al. 2003, Barbosa & Almeida 2010, intensity and declivity of each rapid, being attributed the Pacheco & Olmos 2010); and 3) satellite images. following levels: 1 – low; 2 – moderate; and 3 – high. The following stretches were surveyed: 24 km (representing 35% of the total river length) along the Preto Populational data River (n = 1), with a starting point at the municipality of Mateiros (10°34'33.96''S; 46°28'21.06''W), in To estimate the relative abundance of the Brazilian Merganser along the Novo River, we selected a section September 2009; 69 km (76%) along the Soninho with nearly 50 km, located between the bridge on the River (n = 1), with the starting point at the Chácara da Gildete (10°20'24.47''S; 46°32'53.02''W), in October TO-255 road (10°33'02.43''S; 46°45'34.73''W) and Da 2010; 20 km (9%) along the Sono River (n = 1), with Velha waterfall (10°16'17.29''S; 46°52'44.51''W), where the starting point at the confluence with t he Novo River the species had already been documented (Braz et al. 2003, Barbosa & Almeida 2010). Monthly censuses (n = (10°10'58.98''S; 46°56'21.04''W), in October 2010, 13) were conducted between August 2009 and October which included a previously known locality for the species (Pacheco & Olmos 2010); and a total of 130 km (85%) 2010 by boat in a controlled speed between 4 and 7 along the Novo River (Ecological Station [n = 2]; Da km/h, during five to eight hours per day. Each census Velha downstream [n = 1]; Da Velha upstream [censuses started around 08:00 h and lasted for two days. The identification and c lassification of the individuals stretches [n = 13]), with the starting point located 25 (number of individuals, sex, pairs, groups, and pairs with km downriver from its headwaters (10°51'57.22''S; Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa young) were made from direct sightings using binoculars Relative abundance of Brazilian Mergansers along (Nikon 8 x 42). Each point was georeferenced using the Novo River was estimated using the Kilometrical Garmin Map60CSx GPS. The difference in elevation on Abundance Index (KAI) (Telleria 1986), which consists each 5 km at the Novo River was measured in order to of determining the number of individuals detected verify a possible correlation between the presence of the per kilometer of survey; here we considered only adult species and the river gradient of descent. The exact point individuals detected on the census section (50 km). That of each individual/group recorded on the river was taken index was estimated for each census period in order to right after the sighting and they were counted only when verify possible seasonal variations. parallel to the boat to avoid errors. The proportion of the 50-km area that was occupied In order to determine the proportion of each river by the species was obtained using PRESENCE 3.0 (Hines section (50 km) that was effectively occupied by the 2006). That algorithm estimates the proportion of area species and to obtain an accurate estimate of the number occupied or the probability of a given site being occupied, of individuals that occupy the Novo River, we estimated through comparing the number of sites where the species the Proportion of Occupied Area (). We selected 28 was recorded in relation to the total number of surveyed equidistant points (every 1.5–2 km), defined as surveying sites (MacKenzie et al. 2002). The models generated sites (the same points used to obtain the variables through by the software were ranked according to the Akaike’s the section at the Novo River); therefore, we obtained Information Criterion (AIC: Akaike 1974). We used the the records of presence (1) and absence (0) of the species single-season analysis and the model pre-defined by the at each of the surveying sites during the censuses. The software, which considers the proportion of occupied observation of any individual located 750 m up or down area () and a detection probability (P) constant. river from each survey site was interpreted as presence at In order to verify which landscape features are that particular site. related to the presence of the species (e.g. river width, type of flow and flow level), we created seven probabilistic Statistical analyses models (Table 1) through a simple regression (with one of the variables), including a null model, and In order to identify possible differences in the mean width using the statistical package R 2.13.2, we made simple of the sampled rivers, we conducted a Kruskal-Wallis, and multiple regressions. The models were calculated non-parametric test. The differences between types of using the AIC, adjusted to small samples (Burnham & flow were evaluated using CHI-Square Test ( ). The Anderson 1998). This method is useful for comparing correlation between Brazilian Merganser presence and similar models with different complexity levels; AIC is altitude every 5 km along the sampled rivers was obtained an index that rank models according to their probability by the Pearson Linear Correlation Test, performed using of being selected given the proposed variables, i.e., the BioEstat 4.0. The significance of each test was  = 0.05 smaller the value, the more likely the model to explain (Zar 1999). the observed data. TABLE 1. Models of occurrence of the Brazilian Merganser Mergus octosetaceus along rivers of the Jalapão region, TO run by software R. R Development Core Team. 2012. Evidence Models Variables AICc R² Δi AIC AICw weight Model 0 No variables 73.423 0 4.709 0.041 10.534 Model 1 River width 68.714 0.094 0 0.441 1 Model 2Flow type (rapids) 75.045 0.005 6.33 0.018 23.695 Model 3Flow level 74.948 0.006 6.233 0.019 22.578 Model 4 River width + Flow type 70.186 0.101 1.471 0.211 2.087 Model 5 River width + Flow level 70.582 0.096 1.867 0.173 2.544 Model 6Flow type + Flow level 76.941 0.006 8.226 0.007 61.151 Model 7 River width + Flow type + Flow level 71.983 0.104 3.268 0.086 5.126 AICc; value related to the correction for small samples; ΔiAIC (delta) refers to the relative difference of a model to the lowest AIC value, being more representative those values lower than 2.; AICw (evidence weight) is the chance of a model being selected, varying between 0 and 1, on which the highest values represent a higher chance of the model best respond to the species’ presence. Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa RESULTS with rapids proportionately distributed with the calmer water sections ( = 1.33, df = 1, P = 0.24 rather than P = 0.05 or ns). In general, the rapids were moderate (62.5%) Regional distribution to low (37.5%). This river presents a moderate visibility Brazilian mergansers were recorded only in the Novo of the bottom, due to the dark color of the water mainly River at the section located above the Da Velha waterfall; near its confluence with the Novo River. The Novo River was the widest one sampled, with no records were obtained along the Preto, Soninho and the width varying between 36 and 170 m (mean = 87.2 Sono rivers. During a total sampling effort of 132 hours and 130 km along the Novo River, 125 contacts with m, SD = 32.5 m). The presence of rapids is relatively equal Brazilian mergansers were obtained along a 115 km to the presence of smooth current water (46.4% versus stretch between Da Velha waterfall and inside the Serra 53.6%, respectively), with long sections of each type and no significant difference at the proportions of each type Geral do Tocantins Ecological Station (Figure 1). ( = 0.14, df = 1, P = 0.70 rather than P = 0.05 or ns). Rivers characterization That river presents a sandy bottom and very clear water, which confers a high visibility to the river, and sandy The river sections surveyed presented similar features, banks formed mainly during the rainy season from the sand carried by the smaller tributaries whose headwaters and in general showed well preserved marginal vegetation are located in regions dominated by sandstone. with predominance of Cerrado sensu stricto, Campo and Mata Ciliar, and good water quality; however, the river sections are structurally distinct. The Soninho River is Factors related to the species occurrence relatively narrow, with a total width varying between 26 The model that best explained the species occurrence and 105 m, with a mean of 46.9 m (SD = 22.3 m); that according to the landscape features was model #1, which river flows through a steep gradient along its length, with a predominance of rapids (85.7%,  = 7.14, df = 1, P = is represented by the variable river bank/river width (Table 0.007) over smoothly-flowing waters. The flow of rapids 1). However, models #4 and #5, representing variables was predominately low (66.7%) to moderate (25%), with river width plus type of flow and river width plus level of flow, respectively, presented delta AIC values lower than a low visibility of the river bottom due to the presence of two. In fact, the width significantly varied a mong rivers humus related compounds that confer a dark color to the water. (H = 35.36, P < 0.0001), the Novo River being the widest The Preto River was the narrowest one sampled, with (mean width = 87.2 m) and the Preto River the narrower a width varying from 10 to 37 m (mean = 14.4 m, SD (mean width = 14.4 m); moreover, the target species was detected primarily at wider sections of the sampled rivers, = 7.73 m). It is a fast river, with 66.7% of the surveyed where the width varied between 50 and 100 m (Figure 2). points composed of rapids, mainly at its central portion, Rivers width FIGURE 2. Occurrence of the Brazilian Merganser Mergus octosetaceus at sampling points along the Novo, Soninho and Preto Rivers, relative to the width of rivers in the Jalapão region, TO. Present = 1 and Absence = 0. Revista Brasileira de Ornitologia, 23(4), 2015 Occurrence Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa The Brazilian Merganser was recor ded only on the river sections with presence of rapids ( = 2.18, df = 1, Novo River, with most of the contacts (87.3%) made P = 0.13 rather than P = 0.05 or ns) than in calm waters, within the lotic section, an area of highest gradient and with a predominance observed within low intensity descent and with presence of rapids (Figures. 3A and B). rapids. The contacts obtained in rapids level 1 (70.59%) At that section, there is a positive correlation (P = 0.0077) were significantly more frequent ( = 24.06, df = 1, P between the number of contacts and the difference of < 0.0001) than those obtained in remaining levels. The river gradient calculated at each five kilometers (Figure 4). results agree with the other two selected models (#4 and Even though there was no significant correlation found, #5), which associate the variable “river width” with type in general the species was more frequently recorded in and level of flow (Table 1). Census stretch (km) (A) (B) FIGURE 3. (A) Altimetry of contact points with the Brazilian Merganser Mergus octosetaceus along a stretch of Novo River (50 km), Jalapão, TO. (B) Frequency of contacts with the Brazilian Merganser Mergus octosetaceus at intervals of 5 km along a stretch of Novo River (50 km), Jalapão, TO. Declivity (m) FIGURE 4. Correlation between the number of contacts with the Brazilian Merganser Mergus octosetaceus and declivity difference calculated at every 5 km of surveyed stretch of the Novo River (50 km), Jalapão, TO (Pearson r = 0.7876, P = 0.0077). Revista Brasileira de Ornitologia, 23(4), 2015 Altitude (m) N.º of contacts Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa Populational data Species’ distribution at the Novo River Relative abundance of the Brazilian Merganser along the Brazilian Mergansers were present in all censuses at Novo River was 0.125 individuals/km. Based on that the Novo River, and a total of four distinct pairs were value, we estimate a total of 6.23 individuals for the 50 identified, with t hree of them (Pairs 1, 2 and 3) being km section of the Novo River selected for the censuses. previously recorded (Barbosa & Almeida 2010). A total Considering the length of 145 km between Da Velha of 14 ducklings were recorded for those three pairs (see waterfall and Riacho Verde (near its headwaters) – the only Barbosa & Almeida 2010 for 2007–2009 data). However, section where the species were confirmed – we estimate a it was observed that there appeared to be a reduction of population of 18 adult individuals inhabiting the Novo the survival of the ducklings of pairs 2 and 3, which river, which has a total length of roughly 160 km (Table 2). occupy an area in which local people use to practice Estimation of Occupied Area Proportion of the rafting (Table 4). The fourth pair was recor ded at Serra Novo River was  = 0.7563, i.e., 75.6% of the surveyed Geral do Tocantins Ecological Station (Pair 4, Figure 1). section was occupied (Table3). Based on these data and During the survey throughout the Novo River, located considering that the estimate for the 145-km section with inside the Serra Geral do Tocantins Ecological Station, the species occurrence is equivalent to that obtained at during the reproductive seasons of 2009 and 2010, no the sections where the censuses were made, we estimate ducklings were recorded accompanying that pair. The a population of 13.7 individuals for the Novo River. characteristics used to define a pair were the presence of a Subsequent surveys conducted in September and October close relationship between individuals and the presence of 2009 and July and August 2010 in a 115-km section of a worn nuchal crest on the female, regarded as evidence of the Novo River (including the sections of census and recent mating. Single individuals and/or groups of up to at the Serra Geral do Tocantins Ecological Station), the three individuals presenting adult plumage were observed observed population size is similar to the estimated, with on the Novo River as well, sometimes occupying the same 12 individuals recorded in 2009 and 14 in 2010. river section as the pairs. TABLE 2. Relative abundance and population estimates of the Brazilian Merganser, Mergus octosetaceus, along the Novo River, Jalapão, TO. N.° of adults observed Census KAI (50 km)PopEst (145 km)PopEst* per census Aug 2009 7 0,14 20,3 15,4 Aug 2009 9 0,18 26,1 19,8 Sep 2009 4 0,08 11,6 8,8 Nov 2009 7 0,14 20,3 15,4 Dec 2009 4 0,08 11,6 8,8 Feb 2010 9 0,18 26,1 19,8 Mar 2010 5 0,1 14,5 11 Apr 2010 5 0,1 14,5 11 May 2010 8 0,16 23,2 17,6 Jun 2010 7 0,14 20,3 15,4 Jul 2010 5 0,1 14,5 11 Sep 2010 6 0,12 17,4 13,2 Oct 2010 5 0,1 14,5 11 Mean (±SE) 6,23 (±0,48) 0,125 (±0,01) 18,07 (±1,40) 13,7 (±1,06) (KAI) Kilometric Abundance Index for the census stretch - 50 km; (PopEst) Populacional Estimate for a stretch of 145 km; () Proportion of Area Occupied (PAO) obtained at the census stretch ( = 0,7593). TABLE 3. Modeling of the proportion of area occupied by the Brazilian Merganser Mergus octosetaceus on a stretch (50 km) of the Novo River, Jalapão, TO, obtained with the software PRESENCE. Model AIC AIC wgt P (± SE)  (± SE)  Observed Constant P 300,97 1.0 0.1954 (0.0266) 0.7593 (0.0931) 0.7142 AIC - Akaike’s Information Criterion; AIC wgt – model weight; P - probability of detection;  – Proportion of area occupied ( Observed, i.e, ration between the total sampled sites by the number of sites with presence of species ). Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa Revista Brasileira de Ornitologia, 23(4), 2015 TABLE 4. Number of ducklings recorded of three pairs of the Brazilian Merganser Mergus octosetaceus on a stretch of Novo River during 2009 and 2010. N.º of Coordinates (Datum Year Pairs Date Locality/stretch Details/observation ducklings WGS 84) Pair not detected on the stretch. It was occupying the stretch of river (APA Jalapão) upstream of bridge over TO-255 Pair 1 ? ? road, out of the census stretch. 02 August 2009 Novo River, 3 km downstream of Corta 10°19'S, 46°52'W (*) Pair observed on September 23 without Pair 2 1 * 20 August 2009 Perna stream (700 m away from Pair 3). 10°21'S, 46°52'W ducklings. 02 August 2009 10°19'S, 46°52'W Novo River, 2.3 km downstream of Corta (*) On November 25, the pair was observed Pair 3 2 * 20 August 2009 10°19'S, 46°52'W Perna stream. without ducklings. 23 August 2009 10°16'S, 46°52'W Novo River/APA Jalapão, upstream of Pair observed during survey into the Serra Pair 1 5 12 August 2010 10°36'S, 46°45'W Korubo camping. Geral do TO (SGTES). Ducklings with about a week of age; (*) 27 July 2010 10°28'S, 46°47'W On 16 September 2010 the pair was found Novo River, 3 km downstream of Praia dos Pair 2 6 * 16 September 2010 10°24'S, 46°52’W with 5 ducklings (with about two months of Crentes. 15 October 2010 10°24'S, 46°53'W age) and on 15 October 2010 4 ducklings already capable of flight were observed. 28 July 2010 10°19'S, 46°52'W Nest found on 10 June 2010 with a female Pair 3 x 16 September 2010 Novo River, stretch near the nest. 10°19'S, 46°52'W inside it. Probable nest desertion due to the 15 October 2010 10°18'S, 46°52'W presence of the boat. (*) Observed decrease in the number of surviving ducklings during subsequent surveys. 2010 2009 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa DISCUSSION well. However, when variables such as type of flow, flow level and river width are associated (models #4 and #5), it becomes evident that the merganser selects a set of The Jalapão region includes an important portion of the water courses that converge within the Sono river basin, landscape features as their habitat. which flows into the Tocantins River. Considering the A few studies highlight that the Brazilian Merganser exceptional features of the streams in this region, the occurs exclusively on upland rivers, inhabiting the upper portions of rivers where the presence of large, low density of human population (Mamede et al. 2002) aquatic predators (ictiofauna) would be a limiting and the vast portion of protected, well preserved riparian areas, we expected that the region could shelter a larger factor (Partridge 1956, Willis & Oniki 2003). On the population of Brazilian Merganser than what we found, upper Novo River, the Da Velha waterfall works as a as observed at Serra da Canastra, in the State of Minas natural barrier to large predatory fishes, preventing the occurrence of species such as Jaú (Zungaro zungaro) and Gerais, where approximately 80 individuals are known to Caranha (Pyaractus brachipomus), among others that occur (Lamas 2006). Most of the water courses in the Jalapão region represent small to narrow streams, forming occur at Soninho, Sono and the lower portions of the the typical veredas and with marginal vegetation and low Novo River. Thus, the absence of those predators along water volume, thus not providing the most favorable Novo River, upstream from Da Velha waterfall, may play some role in the observed presence of the species on that conditions for the species occurrence. Therefore, we portion of the river. believed the larger streams would represent areas of greater potential occurrence of the species in the study Pacheco & Olmos (2010) mention the presence of a site. However, we observed this greater presence only on pair of Brazilian Merganser on a sandy bank at the Sono the Novo River. River, roughly 5 km downstream from the confluence of Novo and Soninho Rivers. The pair was spotted during The Novo River presents some distinctive features a flyb y, as part of a rapid ecological assessment of the that distinguish it from the others surveyed (Preto, Soninho and Sono), especially concerning its width. It is region. In the present study, we surveyed 69 km along considerably wider than the others, which we believe is the Soninho River and 20 km along the Sono River, an important condition for the presence of the Brazilian including the spot where Pacheco & Olmos (2010) obtained the aforementioned record, but without Merganser in the Jalapão region. Nevertheless, in other observing any Brazilian Merganser there. Considering sites where the species is known to occur (e.g. Serra da Canastra and Chapada dos Veadeiros) it reportedly occurs that the species was detected in all censuses along the in narrower streams than the Novo River. It is important Novo River, and that it has been reported along the Sono to highlight that the river width itself is not an essential River as well, we expected that the species would be recorded there too. feature for the presence of the species throughout its range. Beginning in 2011, new fruitless searches were In general, records were obtained in portions of the River Novo that varied between 50 and 100 m in width (Figure made along the Soninho River (Valtécio Carvalho pers. 2), an interval distinct from the mean width of Preto comm.). Moreover, during a study for the implementation and Soninho Rivers. Wider sections allow birds to have a of a hydroelectric plant at Perdida River, an important tributary of the Sono River, intensive searches for the wide visual field, which would facilitate protection from species were also unsuccessful. Although our sampling natural predators and of humans disturbance (humans are commonly present at Novo River, mostly due to the effort has been small (n=1), these data and information practice of rafting and of camping in river banks). seem to indicate the absence of the species at Sono and Clearness of the water is also an important factor Soninho Rivers is real and may mean they are extinct there at present. that distinguishes the Novo River from the others present at the region. The low depth and presence of sand and sandstone sediment favor seeing and capturing prey by Conservation at the Jalapão region the mergansers, which could also be a important factor for presence of the species along this river. The Novo Among the main factors that can potentially threaten the local population of the Brazilian Merganser at Jalapão River differs from t he other rivers in the region also in region are the establishment of hydroelectric plants and the higher percentage of rapids along its course, due to having more sections with higher gradient descent. Due the practice of rafting. The modification of the flow is to the fact that the lotic parts of the river correspond to considered the most serious and permanent threat to the sites where the species was detected, it is important the ecological sustainability of rivers and their banks (Sparks 1995, Lundqvist 1998, Ward et al. 1999). The to highlight that the presence of rapids itself is not an impoundment of a river represents the interruption essential feature for the presence of the species, as all the other rivers surveyed presented rapids on their courses as of an open-end transport system to a closed and Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa accumulation system (Junk & Mello 1990). According ACKNOWLEDGMENTS to Bunn & Arthington (2002), the modification of the water flow changes the habitat and influences the species’ We are grateful to Instituto Natureza do Tocantins – distribution and abundance, as well as the diversity and NATURATINS, to Programa de Apoio à Pós-Graduação composition of the aquatic communities. Regarding the of the Secretaria da Ciência e Tecnologia do Estado do rivers in the Jalapão region, changes in water flow b y Tocantins, to Neotropical Grassland Conservancy and impoundment caused by future hydroelectric ventures to The Association of Field Ornithologists for financial planned for the Novo and Sono rivers, similar to eight support. We are also grateful to Instituto Chico Mendes plants constructed on the Palmeiras River (at the outskirts de Conservação da Biodiversidade – ICMBio for licenses of Serra Geral do Tocantins Ecological Station) and two supplied (SISBIO 19649-1 and 2), to Serra Geral do under construction on the Perdida River (tributary of the Tocantins Ecological Station and to Jalapão State Park Sono River), could affect and t hreaten the survival of the for logistics support, to Valtécio Fernandes Carvalho for already small population of Brazilian Mergansers that friendship, partnership in the works and dexterity with inhabits the region. the boat, to Thiago Vernaschi V. Costa for first E nglish An important threat to the species’ survival within version. We appreciate the improvements in English the Novo River is the practice of rafting by tourists where usage made by Phil Whitford through the Association of the species was previously known to reproduce (Barbosa Field Ornithologists' program of editorial assistance and & Almeida 2010, Barbosa et al. 2011). The peak of that by A. Aleixo. sporting activity coincides with the species’ reproductive season and the disturbance caused by such activity may REFERENCES affect the species’ reproductive dynamics and interfere with mating, nest construction, incubation, and duckling Akaike, H. 1974. A new look at the statistical model identification. survival. We have witnessed that during approach by IEEE Transactions on Automatic Control, 19: 719-723. boats, ducklings disperse and are temporarily abandoned Barbosa, M. O. & Almeida, M. L. 2010. Novas observações e dados by the parents. In some cases the disturbance keeps the reprodutivos do pato-mergulhão Mergus octosetaceus na região do Jalapão, Tocantins, Brasil. Cotinga, 32: 109-113. ducklings away from the parents for a prolonged time, Barbosa, M. O; Pinheiro, R. T & Carvalho, V. F. 2011. Descrição with the risk of getting lost and/or being more exposed de ninhos do pato-mergulhão (Mergus octosetaceus) em cavidade to predation. We also noticed an extensive degree of arbórea na região do Jalapão, Tocantins, Brasil. Cotinga 33: 71-75. movements along the river of adults and ducklings Bartmann, W. 1988. New observations on the Brazilian Merganser. following disturbance by rafting that may contribute Wildfowl, 39: 7-14. Bianchi, C. A.; Brant, S.; Brandão, R. A. & Brito, B. F. 2005. New to a lower efficiency of ducklings’ foraging and lower records of Brazilian Merganser Mergus octosetaceus in the rio das reproductive success and recruitment. Pedras, Chapada dos Veadeiros, Brasil. Cotinga, 24: 72-74. The fact that the Brazilian Merganser presently Birdlife International. 2015. Species factsheet: Mergus octosetaceus. occurs only on the Novo River upstream from Da http//:www.birdlife.org (access on 10 December 2015). Braz, V. S.; Abreu, T. L. S.; Lopes, L. E.; Leite, L. O.; França, F. Velha waterfall in the Jalapão region, reinforces the G. R.; Vasconcelos, M. M. & Balbino, S. F. 2003. Brasilian importance of that river section for the survival of the Merganser Mergus octosetaceus discovered in Jalapão State Park, species at its northernmost range limit. Concerning Tocantins, Brazil. Cotinga, 20: 68-71. the pairs we recorded, only one can be considered to Bruno, S. F.; Andrade, R. D.; Lins, L. V.; Bessa, R. & Rigueira, be protected, since it occurred inside the Serra Geral S. E. 2010. Breeding behavior of Brazilian Merganser Mergus octosetaceus, with a tree-cavity nest in Serra da Canastra National do Tocantins Ecological Station. The other three pairs Park, Minas Gerais, Brazil. Cotinga, 32: 84-89. inhabit stretches at Área de Proteção Ambiental do Bunn, S. E. & Arthington, A. H. 2002. Basic Principles and Jalapão (reserve of sustainable use), or at the borders of Ecological Consequences of Altered Flow Regimes for Aquatic Jalapão State Park, which are not entirely protected. For Biodiversity. Environmental Management, 30: 492-507. Burnham, K. P. & Anderson, D. R. 1998. Model selection and the effective protection of such pairs it seems important inference: a practical information-theoretic approach. 2ª Ed., to enlarge the area of the Jalapão State Park, or create Springer Verlag, New York. new protected areas that encompass the left bank of Collar, N. J.; Gonzaga, L. A. P.; Krabbe, N.; Madroño Nieto, A.; the Novo River, as suggested by Barbosa and Almeida Naranjo, L. G.; Parker, T. A. & Wege, D. C. 1992. Threatened (2010). It is also important that an effective control birds of the Americas: the ICBP/IUCN red data book. Cambridge, U.K. International Council for Bird Preservation. of the practice of rafting is established to avoid high Coutinho, L. M. 1978. O conceito de cerrado. Revista Brasileira de disturbance especially during the reproductive season, Botânica, 1: 17-23. and that education initiatives focused on informing Giraudo, A. R. & Povedano, H. 2003. The Atlantic Forest of South tourists and local communities about the importance of America: biodiversity status, threats, and outlook, p. 181-193. In: conservation of the Brazilian Merganser in the Jalapão Galindo-Leal, C. and Câmara, I. G. (eds). Threats of extinction to flagship species in the Interior Atlantic Forest. Washington, DC: region be put in place. Island Press. Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa Hines, J. E. 2006. PRESENCE. Software to estimate patch occupancy Pacheco, J. F. & Olmos, F. 2010. As Aves do Tocantins, Brasil - 2: and related parameters. USGS-PWRC. http://www.mbr-pwrc. Jalapão. Revista Brasileira de Ornitologia, 18(1): 1-18. usgs.gov/software/presence.html (access on 10 September 2010). Partridge, W. F. 1956. Notes on the Brazilian Merganser in Argentina. IUCN 2015. IUCN Red List of Threatened Species. http://www. Auk 73: 473-488. iucnredlist.org (access on 16 November 2015). R Development Core Team. 2012. R: a language and environment Junk, W. J. & Nunes de Mello, J. A. S. 1990. Impactos ecológicos for statistical computing v. 2.13.2. R Foundation for Statistical das represas hidrelétricas na bacia amazônica brasileira. Estud. Av., Computing, Vienna, Austria. http://www.R-project.org (access 4(8): 126-143. on 04 April 2013). Lamas, I. R. 2006. Census of Brazilian Merganser Mergus octosetaceus Silveira, L. F. 1998. The bir ds of Serra da Canastra National Park and in the region of Serra da Canastra National Park, Brazil, with adjacent areas, Minas Gerais, Brasil. Cotinga, 10: 55-63. discussion of the threats and conservation. Bird Conserv. Intern., Silveira, L. F. & Bartmann, W . 2001. Natural history and conservation 16: 145-154. of the Brazilian Merganser Mergus octosetaceus at Serra da Canastra Lamas, I. R. & Santos, J. P. 2004. A Brazilian Merganser Mergus National Park, Minas Gerais, Brazil. Bird Conserv. Intern., 11: octosetaceus nest in a rock crevice, with reproductive notes. Cotinga 287-300. ,22: 38-41. Sparks, R. E. 1995. Need for ecosystem management of large rivers Lundqvist, J. 1998. Avert looming hydrocide. Ambio 27: 428-433. and floodplains. BioScience, 45: 168-182. Mackenzie, D. I.; J. D. Nichols; G. B. Lachman; S. Droege, J. A. Telleria, J. L. 1986. Manual para el censo de vertebrados terrestres. Ed. Royle & C. A. Langtimm. 2002. Estimating site occupancy rates Raíces, Madrid, 278p. when detection probabilities are less than one. Ecolog y, 83(8): Von Behr, M. 2004. Jalapão: sertão das águas. São José dos Campos, 2248-2255. SP. Somos Editora. Mamede, F.; Garcia, P . Q. & Sousa, W. C. 2002. Análise da viabilidade Ward, J. V.; Tockner, K. & Schiemer, F. 1999. Biodiversity of sócio-econômico-ambiental da transposição de águas da bacia do floodplain ecosystems: Ecotones and connectivity. Regulated rio Tocantins para o rio São Francisco região do Jalapão, TO. In: Rivers: Research and Management, 15: 125-139. Caderno de Política Ambiental, 1. Brasília. Conservation Strategy Willis, E. O. & Oniki, Y. 2003. Aves do Estado de São Paulo. Ed. Fund/Conservation International of Brazil/Instituto Internacional Divisa. Rio Claro, SP, 400p. de Educação do Brasil. Zar, J. H. 1999. Biostatistical Analysis. Prentice Hall, Upper Saddle MMA 2014. Lista Nacional Oficial de Espécies da Fauna Ameaçadas River, 663p. de Extinção. http://www.icmbio.gov.br (access on 16 November 2015). Associate Editor: Luís Fábio Silveira Revista Brasileira de Ornitologia, 23(4), 2015 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ornithology Research Springer Journals

Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil

Download PDF
11 pages

Loading next page...
 
/lp/springer-journals/population-estimate-of-mergus-octosetaceus-in-the-jalap-o-region-2snqVF9ks0

References (37)

Publisher
Springer Journals
Copyright
Copyright © Sociedade Brasileira de Ornitologia 2015
eISSN
2178-7875
DOI
10.1007/bf03544318
Publisher site
See Article on Publisher Site

Abstract

Revista Brasileira de Ornitologia, 23(4), 417-427 ARTICLE December 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil 1,3,5 1,2 4 Marcelo O. Barbosa , Renato T. Pinheiro , and Karlla V. C. Barbosa Programa de Pós-graduação em Ecologia de Ecótonos – Mestrado, Universidade Federal do Tocantins – UFT, Rua 7, Qd. 15, s/n, Jardim dos Ipês, CEP 77500-000, campus de Porto Nacional, TO, Brazil. Professor Adjunto, Universidade Federal do Tocantins – UFT and Grupo de Pesquisa em Ecologia e Conservação de Aves – ECOAVES(UFT), Av. NS 15 ALCNO 14, Estação Experimental, 109 Norte, CEP 77001-090 campus de Palmas, TO, Brazil. Instituto Natureza do Tocantins – NATURATINS, Gerência de Unidades de Conservação, Quadra 302 norte, Alameda 2, Lote 3A, CEP 77006- 336, Palmas, TO, Brazil. Museu de Zoologia da Universidade de São Paulo, Avenida Nazaré 481, Ipiranga, CEP 04263‒000, São Paulo, SP, Brazil. Corresponding author: mobarbos@yahoo.com.br Received on 17 March 2015. Accepted on 15 December 2015. ABSTRACT: The Brazilian Merganser is a threatened species and c lassified as Critically Endangered. Its world population is estimated to be fewer than 250 individuals surviving in the wild. Currently, the species is known from disjunct populations distributed in the states of Minas Gerais, Goiás, and Tocantins (in the Jalapão region). For this study, the stretches of the Preto, Novo, Soninho, and Sono Rivers were surveyed by airboat to better estimate the species distribution in the region. Brazilian Mergansers were found only on the Novo River along a stretch upstream from Da Velha waterfall. Four pairs and a few other adult individuals were identified there. A 50 km stretch of the Novo River was selected for monthly censuses (n = 13) from August 2009 to October 2010 to estimate the relative abundance and population size of the Brazilian Merganser. The A bundance Kilometric Index was 0.125 individuals/ km. The Novo River, from Da Velha waterfall until near its headwaters, is 145 km long and our survey estimated a population of 13.7 individuals. Changes caused by uncontrolled tourist activities on the Novo River, such as rafting during the species’ breeding season, riparian forest deforestation, and proposed hydropower projects, can negatively impact this small Brazilian Merganser population putting it in great risk of local extinction. KEY-WORDS: Brazilian Merganser, censuses, river width, Novo River, relative abundance. INTRODUCTION Historically, this species occurred in Paraguay, Argentina and Brazil, but is presently only found in the The Brazilian Merganser (Mer gus octosetaceus Vieillot, latter country. In Paraguay, the species occurred at the 1817) is a globally and locally endangered species headwaters of the Paraná river basin, where it is currently (MMA 2014, IUCN 2015). It is one of the most considered to be extinct in the wild. In Argentina, the species occurred within the Paraná river basin in Misiones, threatened waterfowl species of the Neotropical region but the last sightings there and within that country as a (Bartmann 1988, Collar et al. 1992), with probably less than 250 individuals surviving in the wild (Birdlife whole were made in 2002 (Giraudo & Povedano 2003). 2015). Monogamous and sedentary, it feeds on small In Brazil, the species exists as disjunct small populations, fish (Characidae), which are captured during fast dives occurring in tributaries of the upper São Francisco River, in the state of Minas Gerais (Silveira 1998, Silveira & (Silveira & Bartmann 2001). Its nests are made in tree Bartmann 2001, Lamas & Santos 2004) and right bank or rock cavities (Bruno et al. 2010, Barbosa et al. 2011) right beside water bodies (Lamas & Santos 2004), and its tributaries of the Tocantins River Basin, in the states breeding season usually ranges from June to September. It of Goiás and Tocantins (Braz et al. 2003, Bianchi et al. inhabits only shallow, clear water rivers or small streams, 2005, Barbosa & Almeida 2010, Barbosa et al. 2011). The first recor ds of the species at Jalapão State Park (Parque with stretches of rapids, and bordered by riparian Estadual do Jalapão), eastern Tocantins, were made as vegetation. Due to its low population densities and high sensitivity to environmental disturbance, the species recently as 2002 along the Novo River (Braz et al. 2003), is most threatened by the loss of suitable habitat and/ a tributary of the Sono River basin, at the border of the or reduction of water quality, complicating the species’ Jalapão State Park (Figure 1), where it was subsequently recorded (Pacheco & Olmos 2010). Between 2007 and conservational status. Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa 2010 the species was studied along the Novo River by first we have studied its occurrence and distribution on the author (Barbosa & Almeida 2010, Barbosa et al. 2011). main suitable rivers present in the region, estimating Considering the critical status of the Brazilian Merganser population sizes and evaluating the local availability of and the little knowledge about it at the Jalapão region, the species’ habitat requirements. FIGURE 1. Map of the study area showing the distribution of four pairs of the Brazilian Merganser Mergus octosetaceus identified a long the Novo River during this study as well as a pair on the Sono River (Pacheco & Olmos, 2010). The stretches of rivers surveyed are highlighted in black and grey. Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa METHODS 46°37'24.75''W) and extending toward its confluence with the Sono River. The different distances surveyed were determined by habitat suitability along each river as Study area well as access and navigability issues. The study was conducted at the Jalapão State Park, located in the eastern portion of the State of Tocantins Characterization of the rivers (10°30'34.87''S; 46°34'36.40''W). The region where the In order to characterize each section of the surveyed rivers, Jalapão State Park was created represents one of the best we measured variables of the environment and structural preserved areas of Cerrado in Brazil and it is formed by the different protected areas that together constitute a differences each 1.5-2.0 km a long the Novo (28 points in total area of roughly two million hectares. The largest 50 km), Preto (12 points in 24 km) and Soninho/Sono protected areas are Jalapão State Park (158.885 ha), Serra Rivers (14 points in 27 km). Geographic coordinates and elevation were taken using a Garmin Map60CSx GPS, Geral do Tocantins Ecological Station (645.378 ha), and this being the second measure used to verify the declivity Nascentes do Parnaíba National Park (645.378 ha). The climate in the region has two markedly distinct (relative drop in elevation) for each stretch of the rivers seasons, with a rainy season extending from October sampled. Variables such as vegetation type and clarity to April and a dry season from May to September. The of the water, were used to further characterize each river section. We photographed river banks and satellite images vegetation is Cerrado with prevalence of grasslands, which were used to aid vegetation analyses at each sampling grow on poor (oligotrophic) and sandy soils. The region is characterized also by the presence of plateaus that can point. Vegetation types were characterized on both banks reach 800 m, where the headwaters of many watercourses according to the classification of Coutinho (1978), i.e., that flow to the Tocantins river basin are located (Von campo limpo (open fields), campo úmido (wetland), campo Cerrado (cerrado field), cerrado sensu stricto, mata ciliar Behr 2004). (riparian forest), and cerradão (woods). The clearness of Local range the water was classified in three categories, accor ding to the capacity for clear vision of the river bottom, and We conducted searches for this species along the Novo, ranging from low visibility (dark water) to moderate Preto, Soninho and Sono Rivers (this latter river named visibility and high visibility (extremely clear water). after the confluence of Novo and Soninho Rivers) b y To test how the structural differences on the landscape boat. The stretches of Novo and Soninho rivers were may affect the presence of the species, we generated surveyed during two days whereas the Preto and Sono probabilistic models based on the presence/absence of the along a single day only. These rivers were selected because species at the survey points in relation to the following they present potential habitat conditions for the Brazilian variables: river width, type of flow and flow level. The Merganser such as clear water, rapids, and well-preserved width of the river was measured through analyses of high- vegetation along their banks. The choice of river sections resolution satellite images. Type of flow was classified as sampled was made based on: 1) previous visits to the “remanso” (slowly-flowing water) or “corredeira” (rapids), study area; 2) previously known observation localities for and the level of flow was obtained by eye according to the the species (Braz et al. 2003, Barbosa & Almeida 2010, intensity and declivity of each rapid, being attributed the Pacheco & Olmos 2010); and 3) satellite images. following levels: 1 – low; 2 – moderate; and 3 – high. The following stretches were surveyed: 24 km (representing 35% of the total river length) along the Preto Populational data River (n = 1), with a starting point at the municipality of Mateiros (10°34'33.96''S; 46°28'21.06''W), in To estimate the relative abundance of the Brazilian Merganser along the Novo River, we selected a section September 2009; 69 km (76%) along the Soninho with nearly 50 km, located between the bridge on the River (n = 1), with the starting point at the Chácara da Gildete (10°20'24.47''S; 46°32'53.02''W), in October TO-255 road (10°33'02.43''S; 46°45'34.73''W) and Da 2010; 20 km (9%) along the Sono River (n = 1), with Velha waterfall (10°16'17.29''S; 46°52'44.51''W), where the starting point at the confluence with t he Novo River the species had already been documented (Braz et al. 2003, Barbosa & Almeida 2010). Monthly censuses (n = (10°10'58.98''S; 46°56'21.04''W), in October 2010, 13) were conducted between August 2009 and October which included a previously known locality for the species (Pacheco & Olmos 2010); and a total of 130 km (85%) 2010 by boat in a controlled speed between 4 and 7 along the Novo River (Ecological Station [n = 2]; Da km/h, during five to eight hours per day. Each census Velha downstream [n = 1]; Da Velha upstream [censuses started around 08:00 h and lasted for two days. The identification and c lassification of the individuals stretches [n = 13]), with the starting point located 25 (number of individuals, sex, pairs, groups, and pairs with km downriver from its headwaters (10°51'57.22''S; Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa young) were made from direct sightings using binoculars Relative abundance of Brazilian Mergansers along (Nikon 8 x 42). Each point was georeferenced using the Novo River was estimated using the Kilometrical Garmin Map60CSx GPS. The difference in elevation on Abundance Index (KAI) (Telleria 1986), which consists each 5 km at the Novo River was measured in order to of determining the number of individuals detected verify a possible correlation between the presence of the per kilometer of survey; here we considered only adult species and the river gradient of descent. The exact point individuals detected on the census section (50 km). That of each individual/group recorded on the river was taken index was estimated for each census period in order to right after the sighting and they were counted only when verify possible seasonal variations. parallel to the boat to avoid errors. The proportion of the 50-km area that was occupied In order to determine the proportion of each river by the species was obtained using PRESENCE 3.0 (Hines section (50 km) that was effectively occupied by the 2006). That algorithm estimates the proportion of area species and to obtain an accurate estimate of the number occupied or the probability of a given site being occupied, of individuals that occupy the Novo River, we estimated through comparing the number of sites where the species the Proportion of Occupied Area (). We selected 28 was recorded in relation to the total number of surveyed equidistant points (every 1.5–2 km), defined as surveying sites (MacKenzie et al. 2002). The models generated sites (the same points used to obtain the variables through by the software were ranked according to the Akaike’s the section at the Novo River); therefore, we obtained Information Criterion (AIC: Akaike 1974). We used the the records of presence (1) and absence (0) of the species single-season analysis and the model pre-defined by the at each of the surveying sites during the censuses. The software, which considers the proportion of occupied observation of any individual located 750 m up or down area () and a detection probability (P) constant. river from each survey site was interpreted as presence at In order to verify which landscape features are that particular site. related to the presence of the species (e.g. river width, type of flow and flow level), we created seven probabilistic Statistical analyses models (Table 1) through a simple regression (with one of the variables), including a null model, and In order to identify possible differences in the mean width using the statistical package R 2.13.2, we made simple of the sampled rivers, we conducted a Kruskal-Wallis, and multiple regressions. The models were calculated non-parametric test. The differences between types of using the AIC, adjusted to small samples (Burnham & flow were evaluated using CHI-Square Test ( ). The Anderson 1998). This method is useful for comparing correlation between Brazilian Merganser presence and similar models with different complexity levels; AIC is altitude every 5 km along the sampled rivers was obtained an index that rank models according to their probability by the Pearson Linear Correlation Test, performed using of being selected given the proposed variables, i.e., the BioEstat 4.0. The significance of each test was  = 0.05 smaller the value, the more likely the model to explain (Zar 1999). the observed data. TABLE 1. Models of occurrence of the Brazilian Merganser Mergus octosetaceus along rivers of the Jalapão region, TO run by software R. R Development Core Team. 2012. Evidence Models Variables AICc R² Δi AIC AICw weight Model 0 No variables 73.423 0 4.709 0.041 10.534 Model 1 River width 68.714 0.094 0 0.441 1 Model 2Flow type (rapids) 75.045 0.005 6.33 0.018 23.695 Model 3Flow level 74.948 0.006 6.233 0.019 22.578 Model 4 River width + Flow type 70.186 0.101 1.471 0.211 2.087 Model 5 River width + Flow level 70.582 0.096 1.867 0.173 2.544 Model 6Flow type + Flow level 76.941 0.006 8.226 0.007 61.151 Model 7 River width + Flow type + Flow level 71.983 0.104 3.268 0.086 5.126 AICc; value related to the correction for small samples; ΔiAIC (delta) refers to the relative difference of a model to the lowest AIC value, being more representative those values lower than 2.; AICw (evidence weight) is the chance of a model being selected, varying between 0 and 1, on which the highest values represent a higher chance of the model best respond to the species’ presence. Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa RESULTS with rapids proportionately distributed with the calmer water sections ( = 1.33, df = 1, P = 0.24 rather than P = 0.05 or ns). In general, the rapids were moderate (62.5%) Regional distribution to low (37.5%). This river presents a moderate visibility Brazilian mergansers were recorded only in the Novo of the bottom, due to the dark color of the water mainly River at the section located above the Da Velha waterfall; near its confluence with the Novo River. The Novo River was the widest one sampled, with no records were obtained along the Preto, Soninho and the width varying between 36 and 170 m (mean = 87.2 Sono rivers. During a total sampling effort of 132 hours and 130 km along the Novo River, 125 contacts with m, SD = 32.5 m). The presence of rapids is relatively equal Brazilian mergansers were obtained along a 115 km to the presence of smooth current water (46.4% versus stretch between Da Velha waterfall and inside the Serra 53.6%, respectively), with long sections of each type and no significant difference at the proportions of each type Geral do Tocantins Ecological Station (Figure 1). ( = 0.14, df = 1, P = 0.70 rather than P = 0.05 or ns). Rivers characterization That river presents a sandy bottom and very clear water, which confers a high visibility to the river, and sandy The river sections surveyed presented similar features, banks formed mainly during the rainy season from the sand carried by the smaller tributaries whose headwaters and in general showed well preserved marginal vegetation are located in regions dominated by sandstone. with predominance of Cerrado sensu stricto, Campo and Mata Ciliar, and good water quality; however, the river sections are structurally distinct. The Soninho River is Factors related to the species occurrence relatively narrow, with a total width varying between 26 The model that best explained the species occurrence and 105 m, with a mean of 46.9 m (SD = 22.3 m); that according to the landscape features was model #1, which river flows through a steep gradient along its length, with a predominance of rapids (85.7%,  = 7.14, df = 1, P = is represented by the variable river bank/river width (Table 0.007) over smoothly-flowing waters. The flow of rapids 1). However, models #4 and #5, representing variables was predominately low (66.7%) to moderate (25%), with river width plus type of flow and river width plus level of flow, respectively, presented delta AIC values lower than a low visibility of the river bottom due to the presence of two. In fact, the width significantly varied a mong rivers humus related compounds that confer a dark color to the water. (H = 35.36, P < 0.0001), the Novo River being the widest The Preto River was the narrowest one sampled, with (mean width = 87.2 m) and the Preto River the narrower a width varying from 10 to 37 m (mean = 14.4 m, SD (mean width = 14.4 m); moreover, the target species was detected primarily at wider sections of the sampled rivers, = 7.73 m). It is a fast river, with 66.7% of the surveyed where the width varied between 50 and 100 m (Figure 2). points composed of rapids, mainly at its central portion, Rivers width FIGURE 2. Occurrence of the Brazilian Merganser Mergus octosetaceus at sampling points along the Novo, Soninho and Preto Rivers, relative to the width of rivers in the Jalapão region, TO. Present = 1 and Absence = 0. Revista Brasileira de Ornitologia, 23(4), 2015 Occurrence Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa The Brazilian Merganser was recor ded only on the river sections with presence of rapids ( = 2.18, df = 1, Novo River, with most of the contacts (87.3%) made P = 0.13 rather than P = 0.05 or ns) than in calm waters, within the lotic section, an area of highest gradient and with a predominance observed within low intensity descent and with presence of rapids (Figures. 3A and B). rapids. The contacts obtained in rapids level 1 (70.59%) At that section, there is a positive correlation (P = 0.0077) were significantly more frequent ( = 24.06, df = 1, P between the number of contacts and the difference of < 0.0001) than those obtained in remaining levels. The river gradient calculated at each five kilometers (Figure 4). results agree with the other two selected models (#4 and Even though there was no significant correlation found, #5), which associate the variable “river width” with type in general the species was more frequently recorded in and level of flow (Table 1). Census stretch (km) (A) (B) FIGURE 3. (A) Altimetry of contact points with the Brazilian Merganser Mergus octosetaceus along a stretch of Novo River (50 km), Jalapão, TO. (B) Frequency of contacts with the Brazilian Merganser Mergus octosetaceus at intervals of 5 km along a stretch of Novo River (50 km), Jalapão, TO. Declivity (m) FIGURE 4. Correlation between the number of contacts with the Brazilian Merganser Mergus octosetaceus and declivity difference calculated at every 5 km of surveyed stretch of the Novo River (50 km), Jalapão, TO (Pearson r = 0.7876, P = 0.0077). Revista Brasileira de Ornitologia, 23(4), 2015 Altitude (m) N.º of contacts Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa Populational data Species’ distribution at the Novo River Relative abundance of the Brazilian Merganser along the Brazilian Mergansers were present in all censuses at Novo River was 0.125 individuals/km. Based on that the Novo River, and a total of four distinct pairs were value, we estimate a total of 6.23 individuals for the 50 identified, with t hree of them (Pairs 1, 2 and 3) being km section of the Novo River selected for the censuses. previously recorded (Barbosa & Almeida 2010). A total Considering the length of 145 km between Da Velha of 14 ducklings were recorded for those three pairs (see waterfall and Riacho Verde (near its headwaters) – the only Barbosa & Almeida 2010 for 2007–2009 data). However, section where the species were confirmed – we estimate a it was observed that there appeared to be a reduction of population of 18 adult individuals inhabiting the Novo the survival of the ducklings of pairs 2 and 3, which river, which has a total length of roughly 160 km (Table 2). occupy an area in which local people use to practice Estimation of Occupied Area Proportion of the rafting (Table 4). The fourth pair was recor ded at Serra Novo River was  = 0.7563, i.e., 75.6% of the surveyed Geral do Tocantins Ecological Station (Pair 4, Figure 1). section was occupied (Table3). Based on these data and During the survey throughout the Novo River, located considering that the estimate for the 145-km section with inside the Serra Geral do Tocantins Ecological Station, the species occurrence is equivalent to that obtained at during the reproductive seasons of 2009 and 2010, no the sections where the censuses were made, we estimate ducklings were recorded accompanying that pair. The a population of 13.7 individuals for the Novo River. characteristics used to define a pair were the presence of a Subsequent surveys conducted in September and October close relationship between individuals and the presence of 2009 and July and August 2010 in a 115-km section of a worn nuchal crest on the female, regarded as evidence of the Novo River (including the sections of census and recent mating. Single individuals and/or groups of up to at the Serra Geral do Tocantins Ecological Station), the three individuals presenting adult plumage were observed observed population size is similar to the estimated, with on the Novo River as well, sometimes occupying the same 12 individuals recorded in 2009 and 14 in 2010. river section as the pairs. TABLE 2. Relative abundance and population estimates of the Brazilian Merganser, Mergus octosetaceus, along the Novo River, Jalapão, TO. N.° of adults observed Census KAI (50 km)PopEst (145 km)PopEst* per census Aug 2009 7 0,14 20,3 15,4 Aug 2009 9 0,18 26,1 19,8 Sep 2009 4 0,08 11,6 8,8 Nov 2009 7 0,14 20,3 15,4 Dec 2009 4 0,08 11,6 8,8 Feb 2010 9 0,18 26,1 19,8 Mar 2010 5 0,1 14,5 11 Apr 2010 5 0,1 14,5 11 May 2010 8 0,16 23,2 17,6 Jun 2010 7 0,14 20,3 15,4 Jul 2010 5 0,1 14,5 11 Sep 2010 6 0,12 17,4 13,2 Oct 2010 5 0,1 14,5 11 Mean (±SE) 6,23 (±0,48) 0,125 (±0,01) 18,07 (±1,40) 13,7 (±1,06) (KAI) Kilometric Abundance Index for the census stretch - 50 km; (PopEst) Populacional Estimate for a stretch of 145 km; () Proportion of Area Occupied (PAO) obtained at the census stretch ( = 0,7593). TABLE 3. Modeling of the proportion of area occupied by the Brazilian Merganser Mergus octosetaceus on a stretch (50 km) of the Novo River, Jalapão, TO, obtained with the software PRESENCE. Model AIC AIC wgt P (± SE)  (± SE)  Observed Constant P 300,97 1.0 0.1954 (0.0266) 0.7593 (0.0931) 0.7142 AIC - Akaike’s Information Criterion; AIC wgt – model weight; P - probability of detection;  – Proportion of area occupied ( Observed, i.e, ration between the total sampled sites by the number of sites with presence of species ). Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa Revista Brasileira de Ornitologia, 23(4), 2015 TABLE 4. Number of ducklings recorded of three pairs of the Brazilian Merganser Mergus octosetaceus on a stretch of Novo River during 2009 and 2010. N.º of Coordinates (Datum Year Pairs Date Locality/stretch Details/observation ducklings WGS 84) Pair not detected on the stretch. It was occupying the stretch of river (APA Jalapão) upstream of bridge over TO-255 Pair 1 ? ? road, out of the census stretch. 02 August 2009 Novo River, 3 km downstream of Corta 10°19'S, 46°52'W (*) Pair observed on September 23 without Pair 2 1 * 20 August 2009 Perna stream (700 m away from Pair 3). 10°21'S, 46°52'W ducklings. 02 August 2009 10°19'S, 46°52'W Novo River, 2.3 km downstream of Corta (*) On November 25, the pair was observed Pair 3 2 * 20 August 2009 10°19'S, 46°52'W Perna stream. without ducklings. 23 August 2009 10°16'S, 46°52'W Novo River/APA Jalapão, upstream of Pair observed during survey into the Serra Pair 1 5 12 August 2010 10°36'S, 46°45'W Korubo camping. Geral do TO (SGTES). Ducklings with about a week of age; (*) 27 July 2010 10°28'S, 46°47'W On 16 September 2010 the pair was found Novo River, 3 km downstream of Praia dos Pair 2 6 * 16 September 2010 10°24'S, 46°52’W with 5 ducklings (with about two months of Crentes. 15 October 2010 10°24'S, 46°53'W age) and on 15 October 2010 4 ducklings already capable of flight were observed. 28 July 2010 10°19'S, 46°52'W Nest found on 10 June 2010 with a female Pair 3 x 16 September 2010 Novo River, stretch near the nest. 10°19'S, 46°52'W inside it. Probable nest desertion due to the 15 October 2010 10°18'S, 46°52'W presence of the boat. (*) Observed decrease in the number of surviving ducklings during subsequent surveys. 2010 2009 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa DISCUSSION well. However, when variables such as type of flow, flow level and river width are associated (models #4 and #5), it becomes evident that the merganser selects a set of The Jalapão region includes an important portion of the water courses that converge within the Sono river basin, landscape features as their habitat. which flows into the Tocantins River. Considering the A few studies highlight that the Brazilian Merganser exceptional features of the streams in this region, the occurs exclusively on upland rivers, inhabiting the upper portions of rivers where the presence of large, low density of human population (Mamede et al. 2002) aquatic predators (ictiofauna) would be a limiting and the vast portion of protected, well preserved riparian areas, we expected that the region could shelter a larger factor (Partridge 1956, Willis & Oniki 2003). On the population of Brazilian Merganser than what we found, upper Novo River, the Da Velha waterfall works as a as observed at Serra da Canastra, in the State of Minas natural barrier to large predatory fishes, preventing the occurrence of species such as Jaú (Zungaro zungaro) and Gerais, where approximately 80 individuals are known to Caranha (Pyaractus brachipomus), among others that occur (Lamas 2006). Most of the water courses in the Jalapão region represent small to narrow streams, forming occur at Soninho, Sono and the lower portions of the the typical veredas and with marginal vegetation and low Novo River. Thus, the absence of those predators along water volume, thus not providing the most favorable Novo River, upstream from Da Velha waterfall, may play some role in the observed presence of the species on that conditions for the species occurrence. Therefore, we portion of the river. believed the larger streams would represent areas of greater potential occurrence of the species in the study Pacheco & Olmos (2010) mention the presence of a site. However, we observed this greater presence only on pair of Brazilian Merganser on a sandy bank at the Sono the Novo River. River, roughly 5 km downstream from the confluence of Novo and Soninho Rivers. The pair was spotted during The Novo River presents some distinctive features a flyb y, as part of a rapid ecological assessment of the that distinguish it from the others surveyed (Preto, Soninho and Sono), especially concerning its width. It is region. In the present study, we surveyed 69 km along considerably wider than the others, which we believe is the Soninho River and 20 km along the Sono River, an important condition for the presence of the Brazilian including the spot where Pacheco & Olmos (2010) obtained the aforementioned record, but without Merganser in the Jalapão region. Nevertheless, in other observing any Brazilian Merganser there. Considering sites where the species is known to occur (e.g. Serra da Canastra and Chapada dos Veadeiros) it reportedly occurs that the species was detected in all censuses along the in narrower streams than the Novo River. It is important Novo River, and that it has been reported along the Sono to highlight that the river width itself is not an essential River as well, we expected that the species would be recorded there too. feature for the presence of the species throughout its range. Beginning in 2011, new fruitless searches were In general, records were obtained in portions of the River Novo that varied between 50 and 100 m in width (Figure made along the Soninho River (Valtécio Carvalho pers. 2), an interval distinct from the mean width of Preto comm.). Moreover, during a study for the implementation and Soninho Rivers. Wider sections allow birds to have a of a hydroelectric plant at Perdida River, an important tributary of the Sono River, intensive searches for the wide visual field, which would facilitate protection from species were also unsuccessful. Although our sampling natural predators and of humans disturbance (humans are commonly present at Novo River, mostly due to the effort has been small (n=1), these data and information practice of rafting and of camping in river banks). seem to indicate the absence of the species at Sono and Clearness of the water is also an important factor Soninho Rivers is real and may mean they are extinct there at present. that distinguishes the Novo River from the others present at the region. The low depth and presence of sand and sandstone sediment favor seeing and capturing prey by Conservation at the Jalapão region the mergansers, which could also be a important factor for presence of the species along this river. The Novo Among the main factors that can potentially threaten the local population of the Brazilian Merganser at Jalapão River differs from t he other rivers in the region also in region are the establishment of hydroelectric plants and the higher percentage of rapids along its course, due to having more sections with higher gradient descent. Due the practice of rafting. The modification of the flow is to the fact that the lotic parts of the river correspond to considered the most serious and permanent threat to the sites where the species was detected, it is important the ecological sustainability of rivers and their banks (Sparks 1995, Lundqvist 1998, Ward et al. 1999). The to highlight that the presence of rapids itself is not an impoundment of a river represents the interruption essential feature for the presence of the species, as all the other rivers surveyed presented rapids on their courses as of an open-end transport system to a closed and Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa accumulation system (Junk & Mello 1990). According ACKNOWLEDGMENTS to Bunn & Arthington (2002), the modification of the water flow changes the habitat and influences the species’ We are grateful to Instituto Natureza do Tocantins – distribution and abundance, as well as the diversity and NATURATINS, to Programa de Apoio à Pós-Graduação composition of the aquatic communities. Regarding the of the Secretaria da Ciência e Tecnologia do Estado do rivers in the Jalapão region, changes in water flow b y Tocantins, to Neotropical Grassland Conservancy and impoundment caused by future hydroelectric ventures to The Association of Field Ornithologists for financial planned for the Novo and Sono rivers, similar to eight support. We are also grateful to Instituto Chico Mendes plants constructed on the Palmeiras River (at the outskirts de Conservação da Biodiversidade – ICMBio for licenses of Serra Geral do Tocantins Ecological Station) and two supplied (SISBIO 19649-1 and 2), to Serra Geral do under construction on the Perdida River (tributary of the Tocantins Ecological Station and to Jalapão State Park Sono River), could affect and t hreaten the survival of the for logistics support, to Valtécio Fernandes Carvalho for already small population of Brazilian Mergansers that friendship, partnership in the works and dexterity with inhabits the region. the boat, to Thiago Vernaschi V. Costa for first E nglish An important threat to the species’ survival within version. We appreciate the improvements in English the Novo River is the practice of rafting by tourists where usage made by Phil Whitford through the Association of the species was previously known to reproduce (Barbosa Field Ornithologists' program of editorial assistance and & Almeida 2010, Barbosa et al. 2011). The peak of that by A. Aleixo. sporting activity coincides with the species’ reproductive season and the disturbance caused by such activity may REFERENCES affect the species’ reproductive dynamics and interfere with mating, nest construction, incubation, and duckling Akaike, H. 1974. A new look at the statistical model identification. survival. We have witnessed that during approach by IEEE Transactions on Automatic Control, 19: 719-723. boats, ducklings disperse and are temporarily abandoned Barbosa, M. O. & Almeida, M. L. 2010. Novas observações e dados by the parents. In some cases the disturbance keeps the reprodutivos do pato-mergulhão Mergus octosetaceus na região do Jalapão, Tocantins, Brasil. Cotinga, 32: 109-113. ducklings away from the parents for a prolonged time, Barbosa, M. O; Pinheiro, R. T & Carvalho, V. F. 2011. Descrição with the risk of getting lost and/or being more exposed de ninhos do pato-mergulhão (Mergus octosetaceus) em cavidade to predation. We also noticed an extensive degree of arbórea na região do Jalapão, Tocantins, Brasil. Cotinga 33: 71-75. movements along the river of adults and ducklings Bartmann, W. 1988. New observations on the Brazilian Merganser. following disturbance by rafting that may contribute Wildfowl, 39: 7-14. Bianchi, C. A.; Brant, S.; Brandão, R. A. & Brito, B. F. 2005. New to a lower efficiency of ducklings’ foraging and lower records of Brazilian Merganser Mergus octosetaceus in the rio das reproductive success and recruitment. Pedras, Chapada dos Veadeiros, Brasil. Cotinga, 24: 72-74. The fact that the Brazilian Merganser presently Birdlife International. 2015. Species factsheet: Mergus octosetaceus. occurs only on the Novo River upstream from Da http//:www.birdlife.org (access on 10 December 2015). Braz, V. S.; Abreu, T. L. S.; Lopes, L. E.; Leite, L. O.; França, F. Velha waterfall in the Jalapão region, reinforces the G. R.; Vasconcelos, M. M. & Balbino, S. F. 2003. Brasilian importance of that river section for the survival of the Merganser Mergus octosetaceus discovered in Jalapão State Park, species at its northernmost range limit. Concerning Tocantins, Brazil. Cotinga, 20: 68-71. the pairs we recorded, only one can be considered to Bruno, S. F.; Andrade, R. D.; Lins, L. V.; Bessa, R. & Rigueira, be protected, since it occurred inside the Serra Geral S. E. 2010. Breeding behavior of Brazilian Merganser Mergus octosetaceus, with a tree-cavity nest in Serra da Canastra National do Tocantins Ecological Station. The other three pairs Park, Minas Gerais, Brazil. Cotinga, 32: 84-89. inhabit stretches at Área de Proteção Ambiental do Bunn, S. E. & Arthington, A. H. 2002. Basic Principles and Jalapão (reserve of sustainable use), or at the borders of Ecological Consequences of Altered Flow Regimes for Aquatic Jalapão State Park, which are not entirely protected. For Biodiversity. Environmental Management, 30: 492-507. Burnham, K. P. & Anderson, D. R. 1998. Model selection and the effective protection of such pairs it seems important inference: a practical information-theoretic approach. 2ª Ed., to enlarge the area of the Jalapão State Park, or create Springer Verlag, New York. new protected areas that encompass the left bank of Collar, N. J.; Gonzaga, L. A. P.; Krabbe, N.; Madroño Nieto, A.; the Novo River, as suggested by Barbosa and Almeida Naranjo, L. G.; Parker, T. A. & Wege, D. C. 1992. Threatened (2010). It is also important that an effective control birds of the Americas: the ICBP/IUCN red data book. Cambridge, U.K. International Council for Bird Preservation. of the practice of rafting is established to avoid high Coutinho, L. M. 1978. O conceito de cerrado. Revista Brasileira de disturbance especially during the reproductive season, Botânica, 1: 17-23. and that education initiatives focused on informing Giraudo, A. R. & Povedano, H. 2003. The Atlantic Forest of South tourists and local communities about the importance of America: biodiversity status, threats, and outlook, p. 181-193. In: conservation of the Brazilian Merganser in the Jalapão Galindo-Leal, C. and Câmara, I. G. (eds). Threats of extinction to flagship species in the Interior Atlantic Forest. Washington, DC: region be put in place. Island Press. Revista Brasileira de Ornitologia, 23(4), 2015 Population estimate of Mergus octosetaceus in the Jalapão region, Tocantins, Brazil Marcelo O. Barbosa, Renato T. Pinheiro, and Karlla V. C. Barbosa Hines, J. E. 2006. PRESENCE. Software to estimate patch occupancy Pacheco, J. F. & Olmos, F. 2010. As Aves do Tocantins, Brasil - 2: and related parameters. USGS-PWRC. http://www.mbr-pwrc. Jalapão. Revista Brasileira de Ornitologia, 18(1): 1-18. usgs.gov/software/presence.html (access on 10 September 2010). Partridge, W. F. 1956. Notes on the Brazilian Merganser in Argentina. IUCN 2015. IUCN Red List of Threatened Species. http://www. Auk 73: 473-488. iucnredlist.org (access on 16 November 2015). R Development Core Team. 2012. R: a language and environment Junk, W. J. & Nunes de Mello, J. A. S. 1990. Impactos ecológicos for statistical computing v. 2.13.2. R Foundation for Statistical das represas hidrelétricas na bacia amazônica brasileira. Estud. Av., Computing, Vienna, Austria. http://www.R-project.org (access 4(8): 126-143. on 04 April 2013). Lamas, I. R. 2006. Census of Brazilian Merganser Mergus octosetaceus Silveira, L. F. 1998. The bir ds of Serra da Canastra National Park and in the region of Serra da Canastra National Park, Brazil, with adjacent areas, Minas Gerais, Brasil. Cotinga, 10: 55-63. discussion of the threats and conservation. Bird Conserv. Intern., Silveira, L. F. & Bartmann, W . 2001. Natural history and conservation 16: 145-154. of the Brazilian Merganser Mergus octosetaceus at Serra da Canastra Lamas, I. R. & Santos, J. P. 2004. A Brazilian Merganser Mergus National Park, Minas Gerais, Brazil. Bird Conserv. Intern., 11: octosetaceus nest in a rock crevice, with reproductive notes. Cotinga 287-300. ,22: 38-41. Sparks, R. E. 1995. Need for ecosystem management of large rivers Lundqvist, J. 1998. Avert looming hydrocide. Ambio 27: 428-433. and floodplains. BioScience, 45: 168-182. Mackenzie, D. I.; J. D. Nichols; G. B. Lachman; S. Droege, J. A. Telleria, J. L. 1986. Manual para el censo de vertebrados terrestres. Ed. Royle & C. A. Langtimm. 2002. Estimating site occupancy rates Raíces, Madrid, 278p. when detection probabilities are less than one. Ecolog y, 83(8): Von Behr, M. 2004. Jalapão: sertão das águas. São José dos Campos, 2248-2255. SP. Somos Editora. Mamede, F.; Garcia, P . Q. & Sousa, W. C. 2002. Análise da viabilidade Ward, J. V.; Tockner, K. & Schiemer, F. 1999. Biodiversity of sócio-econômico-ambiental da transposição de águas da bacia do floodplain ecosystems: Ecotones and connectivity. Regulated rio Tocantins para o rio São Francisco região do Jalapão, TO. In: Rivers: Research and Management, 15: 125-139. Caderno de Política Ambiental, 1. Brasília. Conservation Strategy Willis, E. O. & Oniki, Y. 2003. Aves do Estado de São Paulo. Ed. Fund/Conservation International of Brazil/Instituto Internacional Divisa. Rio Claro, SP, 400p. de Educação do Brasil. Zar, J. H. 1999. Biostatistical Analysis. Prentice Hall, Upper Saddle MMA 2014. Lista Nacional Oficial de Espécies da Fauna Ameaçadas River, 663p. de Extinção. http://www.icmbio.gov.br (access on 16 November 2015). Associate Editor: Luís Fábio Silveira Revista Brasileira de Ornitologia, 23(4), 2015

Journal

Ornithology Research – Springer Journals

Published: Dec 1, 2015

Keywords: Brazilian Merganser; censuses; river width; Novo River; relative abundance

There are no references for this article.