Avifaunal inventory of a Southern Amazonian transitional forest site : the São Luiz farm , Mato Grosso , Brazil

this paper describes the avifauna sampled at the são Luiz farm, in northern Mato Grosso state, a southern Brazilian Amazonian forest site. the avifauna was sampled at forested and open sites, between 29 June and 27 July 2008, using point counts, mist-nets and general observations. We recorded 194 bird species within 18 orders and 46 families. the records of this study expanded the known range limits of at least 16 bird species. despite the need for sampling in other seasons, the rarefaction curves indicate a representative sampling effort. the bird community observed at this site contains most of the species typically associated with Amazonian forests, south of the Amazon, and suggests that ‘transitional forests’ found at this site should be qualified as ‘Amazonian’ when considering their legal status. our data highlights the importance of this anthropogenically-impacted and poorly-known region of Amazonia.


INTRODUCTION
over the last thirty years, increasing populations and its development pressures have lead to extensive deforestation of Amazonian forests (nepstad et al., 1999; skole & tucker, 1993; skole et al., 1994).forest degradation became widespread in the Amazon with the boom of logging activity in the mid 1980's and in the following decades (nepstad et al., 1999;Laurance et al., 2001;Carvalho et al., 2002).About 20% of the Brazilian Amazon has been deforested during 30 years of occupation (inPE, 2010).the borders of the biome have been the most affected, including the Brazilian states of Mato Grosso, and Pará, which together comprise more than 60% of the deforested territory (inPE, 2010).Most Amazonian forests in Mato Grosso have already being changed by logging and clear-cutting, then converted to large-scale plantations and cattle ranching fields.in contrast to this extensive land exploitation, forests in Mato Grosso and neighboring areas have been poorly studied in terms of their biodiversity (i.e.Zimmer et al., 1997;Lees et al., 2008;Aleixo & Poletto, 2007;Whittaker, 2009).
Most published ornithological inventories for Mato Grosso state (and neighboring areas) include the habitats of 'cerrado' and 'pantanal' (naumburg, 1930;stone & roberts, 1934;Pinto, 1940;Pinto & Camargo, 1948;silva & oniki, 1988;Willis & oniki, 1990;dubs, 1992;silveira & d'Horta, 2002).only six bird checklists describing typical or transitional upland Amazonian forests ('terra firme') in Mato Grosso have been published to date (fry, 1970;novaes, 1976;novaes & Lima, 1991;Zimmer et al., 1997;Lees et al., 2008;Aleixo et al., 2010).the published data shows that most of the birds living in 'terra firme' forests in Mato Grosso come from the Xingu bird endemism area (Borges, 2007).some sites in this state can be home to about 350 species with the southern transitional areas poorer than northern sites (i.e.362 sp.: Lees et al., 2008;238 sp.: Aleixo et al., 2010).the region has many endemic bird species, and the relative abundance of more common species may be very different from those observed in other Amazonian areas.surveys of community composition have been recognized as being the key to improving knowledge about Amazonian biodiversity (Capobianco et al., 2001).Herein, we report the ornithological results of a bird survey at the são Luiz farm, emphasizing that these sites are some of the last patches of 'terra firme' forests located in southern Mato Grosso state, Brazil.

METHODS
this study was conducted at the são Luiz farm, located close to the Mt 110 state road in the municipality of Querência in Mato Grosso state, Brazil.the site is placed on the east bank of the tanguro river (an affluent of the Xingu river), to about 25 km from the Xingu national Park (s 12° 38' 40",W 52° 23' 5";s 12° 41' 15",W 52° 21' 38").the farm covers an area of 1,300 ha, including 650 ha of fragmented transitional Amazonian forest.only one river (the semp river) passes through the region.the cultivated areas are primarily comprised of soybean and corn, with Brachiaria sp.grass pastures for cattle in some open sites.
the avifauna was sampled at forested and open sites (s 12° 38' 44.14", W 52° 23' 4.50"; s 12° 41' 15.76", W 52° 21' 38.07"), between 29 June and 27 July 2008.We recorded birds in forested sites using mist-nets, point counts, and general observations methods, whereas those in open areas were recorded only through general observations.General observations were performed not just along the forest trails, but included main roads and plantations.
our sampling units consisted of twelve 550 m long transects, each with four points separated by 150 m from each other (i.e., 50 m, 200 m, 350 m, and 500 m).At each of these 12 transects, 28 mist-nets (12 x 2.5 m; mesh size 36 mm) were distributed in four groups of seven nets at each point.Mist-nets were opened for two days, from sunrise (6:30 h) to 13:30 h, yielding a total sampling effort of 4,700 mist-net-hours.We checked the nets hourly and closed them during periods of heavy rain.All birds captured were identified to species level, weighed, measured (standard measurements included wing, tail, bill, and total length) and, whenever possible, aged, sexed and photographed.Captured birds were banded with numbered metal bands obtained from Centro nacional de Pesquisa para Conservação de Aves silvestres (CEMAVE)instituto Chico Mendes de Conservação da Biodiversidade (iCMBio).All recaptures from the same sampling period (two days) and from the same net line were excluded from the analysis to avoid double counting.
in addition to mist-netting, Luiz Mestre carried out 96 point counts along the same 12 transects and points also sampled with mist-nets.Any given point was sampled twice on different days, but not on the same days as mist-netting.Point counts were conducted between 6:30 h and 9:00 h and lasted ten minutes each.Most birds recorded during point counts were registered with a Marantz PMd 671 digital recorder coupled to a sennheiser MKH 60 directional microphone.Whenever possible, identifications of birds recorded during point counts were confirmed visually with the aid of binoculars.Unknown vocalizations were subsequently checked against known calls and songs and, if necessary, confirmed by consulting other experienced ornithologists (i.e., A. Aleixo and s. dantas).for each record obtained at point counts, the distance from the observer and the height at first detection were noted.We excluded from the richness and abundance estimates all birds flying over or recorded outside a 50 m radius from each point count.We also did not include birds that came from directions of other sampling points (primarily wide-ranging canopy groups such as Psittacidae, Emberizidae, and thraupidae) to avoid double counting.We complemented mist-netting and point count data with general observations (total of ca. 100 hours), to obtain the final checklist.
We used point counts and mist-net data to assemble rarefaction species curves.these curves were calculated with incidence data from point counts, mist-nets, and both methods together using Estimates v.7 software (Colwell, 2004).to maintain independence between point counts (and to avoid any potential double counting) all detections > 100 m from the observer and all registrations of birds flying the point counts were excluded from the analysis.Estimates of the 'true' species richness in each habitat were also calculated using Estimates v.7, using the mean of the four commonly employed abundance-based estimators (ACE, CHAo1, JACK1 and BootstrAP).Avian guild classification was based on terborgh et al. (1990), and species sensitivity and number of habitats used by each species were taken from stotz et al. (1996).We followed the CBro (2009) list for nomenclature and taxon ordinance.

RESULTS AND DISCUSSION
We registered 194 bird species from 18 orders and 46 families, occurring at the são Luiz farm, between 29 June and 27 July 2008.the most representative families were tyrannidae, thraupidae, and thamnophilidae with 28, 16 and 15 species respectively.Mist-nets captured 455 individuals of 74 species belonging to 22 families, and point counts recorded 1,880 individuals of 144 species belonging to 34 families.
the rarefaction curves from mist-nets, point counts, and the two methods combined did not reach an asymptote.However, the combined point count and mist-netting data suggested the accumulation curves were beginning to reach an asymptote (figure 1). the point counts estimate of the mean 'true' richness was 159.6 (± 9.6) species, using ACE, CHAo1, JACK1 and BootstrAP means.Mist-nets estimated richness was 110.5 (± 19) using the mean of the above estimators (table 1).Estimated species richness from combining both sampling methods was 213.7 (± 28.2) species, which is still short of the potential regional richness, which can reach over 350 species (362 sp.: Lees et al., 2008, 238 sp.: Aleixo et al., 2010).some of the differences between our richness estimates and those of other studies are related to the sampling effort, which was smaller in our study and restricted to just a single season.species richness would doubtlessly increase with an increasing sampling effort including additional seasons.However, the region is clearly a transition between Amazonia and 'cerrado', and can contain poorer communities compared to other sites farther north.
despite the fact that all sampling effort was conducted in forested sites and the bulk of species were Amazonian birds, we did encounter some typical 'cerrado' (open areas) bird species, such as Rhea americana, Brotogeris chiriri, Picumnus albosquamatus, Casiornis rufus, and Xenopsaris albinucha (table 2). the Brazilian Amazonian endemic species Automolus paraensis was captured (and photographed) twice in recently and post-burned sites (it was also observed by Aleixo et al., 2010).
We highlight some bird species recorded here, which are at their range limits and/or had their ranges expanded.We found that Pteroglossus bitorquatus, despite its large range (Brazilian Amazonia, and Bolivian Amazon border), had not previously been recorded on the east side of the tanguro/Xingu river (Erize et al., 2006;infonatura, 2007;BirdLife, 2010).We captured and photographed one individual of the sub-species P. b. sturmii (which has an all dark mandible, see Grosset, 2011) in a secondary postburned site.We also recorded individuals during point counts in undisturbed and disturbed sites, and they were found in smaller numbers than the congener P. aracari.this last cited species was mostly found in undisturbed sites and can be considered to be near its southern range limit (Erize et al., 2006).P. aracari was also recorded 30 km south by Aleixo et al. (2010).We also registered Picumnus albosquamatus in its northern range limit (infonatura, 2007).A female of this species was recaptured once (and photographed) in a recently burned site.
We obtained in this study a significant increase in the range of Myrmotherula sclateri, which was also observed by Aleixo et al. (2010) 30 km to the south.We recorded figure 1. rarefaction curves obtained with the Estimates software based on mist-nets (Mn), point counts (PC), and both methods combined (Mn+PC).Estimates are solid lines and 95% confidence intervals are dotted lines.the species on three occasions, only in undisturbed sites.
Myrmotherula hauxwelli was recorded at the southern limit of its range (ridgely & tudor, 1994; infonatura, 2007), and was detected in undisturbed and old-growth secondary forests.Herpsilochmus rufimarginatus had not previously been registered east of the Xingu river (ridgely & tudor, 1994; infonatura, 2007), but had been registered at the tanguro farm, 30 km to the south (Aleixo et al., 2010).the species was commonly recorded in secondary and undisturbed forests at the study sites.the species Myrmornis torquata was recorded once in a post-burned site and was also recorded at the tanguro farm, 30 km to the south (Aleixo et al., 2010); these records expand the known species range by about 1,000 km to the south.Willisornis poecilinotus was commonly captured at the study sites and had its range expanded eastward by about 700 km, as also observed by Aleixo et al.(2010) at the tanguro farm, 30 km to the south (ridgely & tudor, 1994; infonatura, 2007).Hemitriccus griseipectus was recorded only once in an undisturbed forest site, expanding its range about 700 km eastward (infonatura, 2007).Another important record was Hemitriccus minimus which has patchy distribution, centered mainly in western Amazonia.We recorded five and captured three individuals of H. minimus (photographed), mostly in secondary habitats, about 700 km east to its published distribution.H. minimus was also recorded by Aleixo et al. (2010) in the tanguro farm.Poecilotriccus fumifrons was captured once (and photographed) in secondary forest, and registered here about 300 km south of the previously published limit (ridgely & tudor, 1994; infonatura, 2007).Iodopleura isabellae was recorded twice in secondary forests, and, along with those records for tanguro farm (Aleixo et al., 2010), expand the known species range by about 600 km.Tangara chilensis was recorded seven times in secondary and undisturbed sites, more than 700 km east of its published range (ridgely & tudor, 1989; infonatura, 2007).Tangara gyrola was found 1,000 km east its published range (ridgely & tudor, 1989; infonatura, 2007), being recorded five times in disturbed and undisturbed sites.Euphonia rufiventris was registered four times in disturbed and undisturbed forests and, along with those records for tanguro farm (Aleixo et al., 2010), expand its published range 700 km eastward (ridgely & tudor, 1989; infonatura, 2007).Xenopipo atronitens is a white sand forest specialist, and we captured this species only once in a disturbed site (photographed).We also captured two males and two females of Turdus subalaris in undisturbed forests.this record is an important northern range extension for this austral migrant (also recorded by Aleixo et al., 2010 at tanguro farm).We also emphasize the importance of this site for large and medium-sized granivorous species that are highly sensitive to human disturbances, such as Tinamus tao, Tinamus major, Crypturellus strigulosus, Pionites leucogaster, and Pyrrhura perlata.
our results indicate that the avifaunal community of primary forests at são Luiz farm (and surrounding region) is typically Amazonian.More than 75% of the bird species recorded have previously been observed in regions of northern Mato Grosso characterized by typical Amazonian 'terra firme' forests (infonatura, 2007;Lees et al., 2008).Although Brazilian environmental laws ('Código florestal'; Brasil, 1965) classify the legal Amazon as forests north of 13 o s, our results, obtained some kilometers to the south of this (s 12° 38' 44.14", W 52° 23' 4.50"; s 12° 41' 15.76", W 52° 21' 38.07") indicate that the local avifauna is representative of an Amazonian bird community.the environmental act 'Código florestal' (Brasil, 1965) determines the percentage table 2. Bird species list of são Luiz farm in northern Mato Grosso, southern Amazonia.nomenclature, taxonomy, Portuguese and English names, and status (resident -r and endemic -E) follow the Brazilian Check-list Committee (CBro, 2009).Abbreviations for methods of bird surveys are as follows: point counts (PC), mist-nets (Mn), and general observations (Go).Habitats where species were registered: 1-Primary forests; 2-second growth sites including post burned; 3-Plantations and open habitats; and 4 -Air (flying of private properties that can be converted into farmland, and non-Amazonian sites have more flexibility and greater areas for exploitation.thus, classification of these areas as no-Amazonian by environmental laws in Brazil can lead to overexploitation, rapidly fragmenting and degrading the local forest cover.Herein, our data provided a contribution to the knowledge of the avifauna of this comparatively poorly-studied transitional area of the Amazon, underscoring the need to conserve it with the same policies directed at preserving the main Amazonian biome. University, and Museu Paraense Emílio Goeldi funded by the national Aeronautics and space Administration (grant nnX07Af16G).Birders Exchange and idea Wild donated important materials for this research