Richness , composition and detectability of Psittacidae ( Aves ) in three palm swamps of the Cerrado sensu lato in central Brazil Riqueza , composición y detectabilidad de los Psittacidae ( Aves ) en tres pantanos del Cerrado sensu lato en el Brasil central

The objectives of this study were: (1) to investigate the richness and composition of the Psittacidae in palm swamps; (2) to determine if the sizes of fl ocks vary seasonally; (3) to compare detectability rates according to species. Psittacid communities were observed from November 2006 to November 2007 in semimonthly visits to the area using fi xed-point observation. A total of 1616 individuals from eight species were registered in 199 contacts. The palm swamps were highly similar (H > 0.62), and their diversity was above 1.56. The Yellow-chevroned Parakeet was considered resident in the three palm swamps (Fr > 60 %). The White-eyed Parakeet and Red-bellied Macaw were the most abundant species in the three areas. Most of the species (n = 5) were more abundant in one of the seasons. The most abundant species in the rainy season were the White-eyed Parakeet and Yellow-chevroned Parakeet (42.16 ± 30.77 individuals), whereas in the dry season the most prevalent was the Red-bellied Macaw (68.50 ± 39.35 individuals). The highest detectability rate was for the Yellow-chevroned Parakeet, which demonstrates the inter-species differences and the need to adapt the methods of research, unclear for this family.

Even though research on psittacids has been increasing over the last decade (Masello & Quillfeldt 2002), including studies on palm swamps (González 2003, Brightsmith & Bravo 2006), little is known about the ecology of this group (e.g., sizes of the fl ocks, species abundance, and detectability).Thus, in order to contribute information about psittacids in palm swamps, the objectives of this study were: (1) to determine the richness and composition of psittacids in palm swamps; (2) to evaluate whether the sizes of the fl ocks vary between the rainy and the dry seasons; (3) to compare detectability rates according to species.
The weather of the region is Aw according to the Köppen climate classification system, featuring hot, humid summers and cold, dry winters (Rosa et al. 1991).During the study period, the median monthly precipitation was 22.5 mm during the dry season and 254.36 mm during the rainy season (Source: Climatology Laboratory of Universidade Federal de Uberlândia).

Data collection
Before the beginning of data collection, a thirty-six hour observational period was carried out between June and November of 2006 to determine the peak activity times.These preliminary observations occurred between 6 AM and 6 PM.Data were collected between November of 2006 and November of 2007, during both the rainy and dry seasons.A total of 24 semi-monthly visits were made to each site.Each visit included a one-hour observation session that alternated either between 6:00-9:00 AM or 4:00-6:00 PM.This study used the fi xed-point method to estimate the size of the psittacid populations.This technique has been recommended for studies performed in areas of tall, dense vegetation or in rough terrain (Reynolds et al. 1980), as well as for studies involving the entire community, despite being a less selective method (Anderson et al. 1979, Develey 2003).
All species fl ying into or leaving the palm swamp were registered.The only number considered was the highest number of individuals seen simultaneously, in order to avoid duplicate counting of individuals.All species sighted during the one-hour period were used to determine richness.The birds were identifi ed visually using binoculars (7 x 35) and field guides (Sigrist 2006).The nomenclature of the CBRO 2009 -Brazilian Ornithological Records Committee was adopted for this study.

Data analysis
The communities were compared using the Jaccard similarity coeffi cient (Cj).The Shannon-Wiener index and Pielou's evenness index were used to verify species homogeneity (richness and abundance).
In all three palm swamps, the relative abundance of each species was calculated by the total number of visits in which it was observed (1 to 24).In order to verify any interseasonal variation in population, the abundance of each species was calculated according to season by the number of visits ( 12) independent of sightings.
The detectability test (D) (Pomeroy & Dranzoa 1997) was applied to each species.This test gives a score from 1 to 6 every 10 minutes during a one-hour interval when the species is sighted; the species recorded in the fi rst minute of that hour will have higher scores.

Richness and composition
CP and NV were the most similar palm swamps (Cj CP; NV = 0.85); the least similar were CP and CM (Cj CP; CM = 0.62).The three palm swamps had similar values on the similarity and evenness indices (1.53 ≤ H' ≤ 1.70 and 0.80 ≤ J ≤ 0.95) (Table 1).The only resident species were the White-eyed Parakeet in three palm swamps (Fr > 60 %), Red-bellied Macaw in CP and CM (Fr > 80 %), and White-eyed Parakeet in CM (Fr = 66.6 %) (Table 1).

Population size and variation
The White-eyed Parakeet and Red-bellied Macaw were the most abundant in the three areas (Table 1), whereas the least abundant were the Blue-fronted Parrot in CP and NV, and Peach-fronted Parakeet in CM (Table 1).

Detectability rate
The species with the highest D were the Yellowchevroned Parakeet, which was generally detected within 10 min (D = 5.04) and Whiteeyed Parakeet and Red-bellied Macaw, which were mainly detected between 11-20 min (D = 4.22 and 4.12, respectively) after the beginning of the visit (Table 4).

Richness and composition
Psittacid species such as macaws and parrots are among the principal birds that use the resources of the palm swamps.Although these areas suf fer from fragmentation, they still sustain large populations (Bonadie & Bacon 2000).During this study, eight species were found using these resources.In Peru, between  seven to fi fteen psittacid species were found nesting in Mauritia palms (González 2003, Brightsmith 2005).Other studies carried out in palm swamps have demonstrated that palms become key resources for avian frugivores, including many psittacids, helping to maintain them during periods of resource scarcity (Bonadie & Bacon 2000, Brightsmith & Bravo 2006).
Of the eight species registered during this census, the least frequent, Scaly-headed Parrot and Blue-winged Parrotlet, despite conspicuous vocalization and size (12 to 28 cm) (Sick 1997, Sigrist 2006), were probably least obser ved because the Scaly-headed Parrot prefers forests and pine groves (Galetti 1993) and the Bluewinged Parrotlet is typically found in forest borders (Sick 1997, Juniper & Parr 1998).The sporadic sightings of these species, therefore, could have been a result of environmental alterations to their original habitats and/ or oppor tunism.It has been obser ved that the White-eyed Parakeet, Peach-fronted Parakeet and Yellow-chevroned Parakeet are flexible regarding habitat, frequenting even anthropogenic areas (Galetti & Pizo 2002b).Thus, their seeking refuge in palm swamps could be seen as an advantageous strategy.In addition to palms such as Mauritia, other plant species such as Miconia chamissois might offer resources during the dry season (Maruyama et al. 2007), maintaining the local avifauna during unfavorable conditions.Detectability rate (± SD) of each species in the three palm swamps.0 = no species record.
Tasa media de la detectabilidad (± DE) de cada especie en los tres pantanos.0 = sin registro de especies.The elevated frequency and relative abundance of the Red-bellied Macaw in the three areas is a result of its dependency on the buriti, primarily during the reproductive period when it utilizes the hollow trunks of dead palms for nesting and feeds on the fruit.As a specialist frugivore it remains near buriti, occasionally foraging in cultivated areas (Juniper & Parr 1998).

Population size and variation
A social gathering pattern is typical of psittacids that relocate seasonally in search of food (Sick 1997, Juniper & Parr 1998).However, as a result of the reproductive cycle and seasonal fl uctuation in abundance of food resources, the fl ock sizes of these species can vary throughout the year (Pizo et al. 1995, Alonso 2001, Galleti & Pizo 2002a).This probably applies to the Redbellied Macaw in which the average number of individuals varied signifi cantly between rainy and dry seasons.
Alternatively, these seasonal fluctuations could be related to a less conspicuous vocalization principally during the mating season.This is when pairs form and the species reduces its abundance (Cannon 1984, Chapman et al. 1989, Rodriguez-Estrella et al. 1992), which makes sightings diffi cult.A low number of Redbellied Macaw individuals were also observed by Roth (1984), who attributed this reduction to the mating season when the species explores a larger area, resulting in a reduced density of individuals, many of whom may be in search of nesting sites.
The majority of plants has one or more periods of relative scarcity during the year (Van Schaik et al. 1993), forcing frugivorous birds either to alter their diet (Galetti 1993) or migrate in search of new feeding grounds (Wheelwright 1983, Loiselle & Blake 1991, Kinnaird et al. 1996).Mauritia fl exuosa yields fruit throughout the year, which could account for the large quantity of individuals, mostly species such as the Red-bellied Macaw (PA Silva, pers. comm. 2009) and Red-shouldered Macaw.The results obtained in this study, along with those of other authors (Bonadie & Bacon 2000, González 2003, Brightsmith 2005, Brightsmith & Bravo 2006), demonstrate the importance of the palm swamps in maintaining the regional populations of psittacids.
Detectability rate E m l e n ( 1 9 7 1 ) d e m o n s t r a t e d t h a t t h e detectability coef ficient of each species is dependent on its conspicuousness, and represents the propor tion of an area's population that is actually detectable by an obser ver.Sur veys and census studies have standardized the maximum time per point at 20 min (Sutherland et al. 2005).In this study, however, the time necessar y to sight two species was longer.The Blue-fronted Parrot and Peach-fronted Parakeet were detected around 40 minutes after the beginning of the observation session.For this reason, in order to register the presence of some species of lessconspicuous psittacids such as the Blue-fronted Parrot and Peach-fronted Parakeet, a longer observation session is advisable.

Fig. 3 :
Fig. 3: Average numbers of individuals in the rainy and dry seasons, with a standard error.Media general de individuos de acuerdo con las estaciones lluviosa y seca, con un error estándar.

Fig. 4 :
Fig. 4: Sizes of fl ocks in the dry and rainy seasons, with a standard deviation.Tamaño de los grupos en las estaciones seca y lluviosa, con una desviación estándar.