GREATER RHEA Rhea americana

Although it is well distributed in Brazil (Sick 1997, Miranda 2021), the primary information on this species' breeding is from captivity (e.g., Anon. 1922, 1941, Carvalho 1938, Brito 1949, Britto 1950, França & França 1987, Mello 1987, Mendes 1989, Dani 1993, Toledo 2003), as data from the wild are still rare (e.g., Olalla 1938, Olalla & Magalhães 1956, Belton 1984, Codenotti 1997, Codenotti & Alvarez 1997, Leite & Codenotti 2005, Azevedo et al. 2006, Buzzetti & Silva 2008, Costa et al. 2014). We describe observations at two nests found in the wild, one at Alto Parnaíba, Maranhão and the other at Poconé, Mato Grosso.

Alto Parnaíba, Maranhão: this nest was found on 18 November 2002 in the middle of a large open area, with a few grasses and some low bushes. It measured: 130 cm external diameter, 60 cm internal diameter, 22 cm external height and 10 cm internal height. During the night of 19–20 November, a fire burned all of the vegetation around the nest, which remained intact because the bird had removed the immediately adjacent vegetation, thereby serving as a firebreak, preventing the eggs and a small bush beside the nest from being affected (Fig. 1). This behaviour of cleaning the environs of the nest had been described previously (Muñiz 1916, Sick 1997) and, at least based on our observations, serves as a firebreak, since the species' nesting habitat is very prone to fire. The nest contained 12 oval-shaped, slightly greenish, white eggs; mean size 121.8 ± 8.6 × 82.4 ± 4.7 mm and mass 425.4 ± 60.1 g (Table 1; n = 12). Significant variation was apparent, especially in mass, with up to 170 g between the lightest and heaviest egg (320–490 g). This difference is usual for the species and does not appear to influence hatching success (Fernández & Reboreda 2008). Ten nestlings hatched at dawn on 24 November and followed the male to the edge of the burnt area. Two rotten eggs were left in the nest.

Figure 1.

The eggs of this nest of Greater Rhea Rhea americana were not affected by a fire, which burned all of the surrounding area due to the male having removed vegetation immediately around the nest, Alto Parnaíba, Maranhão, Brazil, November 2002 (NORDESTA collection)

Poconé, Mato Grosso: we found a nest on 28 July 2011 in a pasture sometimes used by cattle. It measured 105 cm external diameter and 80 cm internal diameter, practically with no depression; only some stems of Typha sp. (Typhaceae) and other plants lined the ground where the eggs had been laid. There were 19 eggs in the nest, plus one in the area cleaned by the male around it. On 4 August, between 06.00 and 09.00 h, we observed that the male incubated constantly. Interestingly, three other adults approached the male, which remained motionless on the nest, surrounding him, but then left. The male frequently turned the eggs (Fig. 2) to warm them equally. On 8 August, the first chick hatched at 07.10 h and another eight hatched between then and 16.00 h. On 9 August, we returned to the nest at 07.00 h, but no more eggs had hatched, and the nestlings remained in the nest. At 08.08 h, the adult punctured the egg outside the nest, which was rotten and quickly attracted flies (Diptera) that were eaten by the nestlings. Throughout the day, at regular intervals, the male flapped its wings over the eggs as if wanting to help break the shells. At 10.30 h, he left to feed for ten minutes, accompanied by the chicks, then returned to the nest. At various times while on the nest, the male took beakfuls of sand and threw them over his back, often hitting the nestlings, who shook themselves down. Throughout the day, more eggs hatched; some chicks explored the surroundings as soon as they could stand (Fig. 3). By 14.20 h, all the eggs in the nest had hatched, totalling 18 chicks. At 15.00 h, they all left the nest, accompanied by the adult, making short pecks on the ground and swallowing movements with the bill turned upwards. After c.30 minutes, all of the young returned to the nest and the male brooded them.

TABLE 1

Measurements of Rhea americana eggs found in Alto Parnaíba, Maranhão, Brazil, on 18 November 2002.

Figure 2.

Male Greater Rhea Rhea americana frequently turns the eggs in the nest, Poconé, Mato Grosso, Brazil, August 2011 (NORDESTA collection)

Figure 3.

Young Greater Rheas Rhea americana are able to explore the nest's environs immediately on hatching, Poconé, Mato Grosso, Brazil, August 2011 (NORDESTA collection)

Both nests were of the ‘simple/platform’ type. Our data coincide with those already published in other literature, both from Brazil and adjacent countries, but these are the first from the north-east of the country. In this region, a few scraps of information about the rhea were noted in the 17th century, during the Dutch colonisation (Teixeira 1992), but the species is now extinct in most states there (Forbes 1881, Sick 1997, Sagot-Martin et al. 2020).

Family Tinamidae.—The only family in the Order Tinamiformes is endemic to the Neotropical region, occurring from Mexico to southern South America and numbering nine genera and between 44 (Remsen et al. 2021) and 46 species (Winkler et al. 2020b). In Brazil, five genera and 23 species occur (Pacheco et al. 2021).

SMALL-BILLED TINAMOU Crypturellus parvirostris

Found across much of Brazil south of the Amazon, it is especially well distributed in the Cerrado and Caatinga, and in formerly forested regions, expanding its range due to deforestation (Sick 1997, Melo & Godoy 2010). There are sporadic records north of the Amazon, in Amapá (Sick 1997; P. F. França,  http://www.wikiaves.com/2094143). It also occurs in south-east Peru, eastern and northern Bolivia, Paraguay and northern Argentina (Cabot et al. 2020a).

Like most Tinamidae, knowledge of breeding by C. parvirostris in the wild or captivity is scarce (Ihering 1900, Snethlage 1935, Britto 1950, Sick 1997, Marini et al. 2012). We describe observations at 20 nests found between 1980 and 2004 at Quebrangulo, Alagoas (n = 6), Arcos, Minas Gerais (n = 11) and Altamira, Bahia (n = 3).

Nests were found in almost all months, with peaks in September–October and especially February (Fig. 4). However, as shown in Fig. 4, in Bahia and Minas Gerais, nests with eggs were recorded between October and March, and in Alagoas between May and September. Breeding indicated for September in Rio de Janeiro (Britto 1950), October in Pará (Snethlage 1935) and September–November in central Brazil (Marini et al. 2012). The ovary of a female measured 3.5 × 1.5 mm in August in northern Argentina (Navas & Bó 1988). Hayes (2014) found clutches between 17 November and 21 December in Paraguay, and Maurício et al. (2013) observed a weak-flying young on 5 June in Rio Grande do Sul. A search of WikiAves (22 December 2020) found images indicative of breeding in every month except June–August. Based on this, we believe that C. parvirostris may breed almost year-round, with least activity during April–June. However, more data are needed to better understand any variation in timing across the species' distribution and the number of clutches annually. In captivity, C. parvirostris can produce up to four clutches per annum (Sick 1997).

Figure 4.

Number of active nests with eggs by month (based on date of discovery) of Small-billed Tinamou Crypturellus parvirostris in three Brazilian states.

Nests were shallow depressions in the soil, mostly under a tall grass thicket (n = 9) (Fig. 5) or near the root or trunk of a shrub. One was under a Baccharis crispa (Asteraceae). Some nests, however, were built above ground on clumps of vegetation. One in Minas Gerais was on a grass bush, 2 cm from the ground, and another, at 4 cm, on a clump of Digitaria horizontalis (Poaceae). Two nests, one each in Minas Gerais and Alagoas, were 15 cm above ground on grass. In these cases, crumpled grass was used as a base to lay the eggs. Nests were lined (n = 8), to varying extents, with dried grass leaves and/or roots (n = 5), sometimes with soft feathers. Not all nests were lined internally, and some eggs were laid directly on the ground (n = 5).

Two nests measured 14 cm internal diameter and 2–4 cm internal height and were of the ‘simple/unlined’ or ‘simple/platform’ type. The nest of this species has never been described in detail. It was considered to be a small and unlined depression (e.g., Marini et al. 2012, de la Peña 2019a), but most nests in this study had a lining (e.g., Fig. 6), as did almost all of the nests photographed on WikiAves.

Clutches were of 2–5 eggs, mean 3.8 ± 0.8 (Table 2; n = 20 nests), agreeing with most of the literature (Snethlage 1935, Britto 1950, Marini et al. 2012, Cabot et al. 2020a). Only Hayes (2014) mentioned an average clutch of 5.2 ± 1.1 eggs in Paraguay, with some nests containing up to eight eggs (n = 11 nests).

Eggs were salmon-beige to pale lilac with a lustrous appearance and oval shape. Description of colours without using a colour catalogue or spectrophotometric techniques is always subjective, as are their shapes (Gómez & Liñán-Cembrano 2017, Stoddard et al. 2017). However, egg colours described for C. parvirostris agree with our observations, being bright, light purplish-brown [‘glänzend hellviolettbraun’] (Snethlage 1935), light chocolate (Britto 1950), light violet-chocolate (Sick 1997) or chocolate-violet, two of them with tiny brown dots (Marini et al. 2012). Better knowledge of Tinamidae egg colours, and other characteristics, might enable a better understanding of the affinities between species (e.g., Bertelli 2016).

Figure 5.

Small-billed Tinamou Crypturellus parvirostris in a nest under a bush, Arcos, Minas Gerais, Brazil, October 1995 (NORDESTA collection)

TABLE 2

Measurements of Small-billed Tinamou Crypturellus parvirostris eggs found in the wild in Brazil. Brazilian states: AL = Alagoas, BA = Bahia, MG = Minas Gerais. NM = not measured.

Continued

Continued

Figure 6.

Nest of Small-billed Tinamou Crypturellus parvirostris; note grass lining, Arcos, Minas Gerais, Brazil, January 2004 (NORDESTA collection)

Mean egg size was 38.3 ± 2.1 × 28.1 ± 1.6 mm (n = 52) and mass 15.3 ± 2.9 g (n = 47) (Table 2). These measurements agree with previous data (Ihering 1900, Britto 1950, Marini et al. 2012). We draw attention to the tendency of eggs in the north to be smaller and paler than those from further south, mean 13.1 g and 37.0 × 27.5 mm in Alagoas, compared to 17.9 g and 39.9 × 29.0 mm in Minas Gerais.

Concerning incubation, we made observations at just two nests and both clutches were almost complete when we found them. A nest found on 19 February 1991 in Minas Gerais had four eggs, with evidence of the nestlings pecking at the shells on 6 March 1991. The next day, 7 March, at dawn, the chicks were no longer in the nest, indicating an incubation period of at least 15 days. A second nest, also in Minas Gerais, had a complete clutch of five eggs on the day it was found, 29 January 2004. On 17 February, by the end of the day, two of the eggs showed peck marks made by the nestlings, but they hatched only on the next day, 18 February, at 09.00 h. The other eggs hatched in sequence, and by 11.00 h, all had hatched and departed the nest, indicating a minimum incubation of 19–20 days. Marini et al. (2012) reported a 21-day incubation period in central Brazil, whereas in captivity it is considered to be 19 days (Cabot et al. 2020a).

Figure 7.

Small-billed Tinamou Crypturellus parvirostris in warning posture beside the eggs in the nest, Arcos, Minas Gerais, Brazil, February 2004 (NORDESTA collection)

At Arcos, Minas Gerais, dogs predated a nest with three eggs; elsewhere, de la Peña (2019a) indicated Crested Caracara Caracara plancus, the lizard Salvator merianae and Crab-eating Fox Cerdocyon thous as predators of eggs and young. Nestlings are nidifugous (Campbell & Lack 1985: 394) and explore the nest environs within a few hours after hatching. As we observed at Altamira, Bahia, on 31 December 1980, nestlings hatched during the night, around 02.00 h. By dawn, at c.05.00 h, they had already left the nest and were hiding in the nearby vegetation. Nestling development is poorly known. Sick (1997: 158) mentioned that chicks of C. parvirostris are uniform brownish, whilst Marini et al. (2012) stated only that they are fluffy on hatching.

The adult leaves the nest and returns by walking discreetly; however, when surprised on the nest, it flies noisily or adopts an escape posture typical of Tinamidae, with the breast close to the ground and the tail quite erect (Fig. 7). Some authors consider this behaviour a cryptic mechanism, as described for Tataupa Tinamou C. tataupa (Smith & Morris 2018).

At Arcos, Minas Gerais, on 3 February 2004, we witnessed an aggressive adult after its chicks hatched. The bird spread its wings and assumed a threatening posture while the chicks ran to hide. Marini et al. (2012) mentioned having twice observed an adult pretending to have an injured wing to protect the eggs.

We noted the species feeding on beetles, larvae and seeds found by turning over dry leaves on the ground using its bill. We also observed it feeding on fallen flowers of Handroanthus heptaphyllus (Bignoniaceae), corn kernels after harvest, and cassava tubers exposed by rain or by an armadillo. The species appears to have a well-defined territory, returning to roost every night, usually in a slight depression at the base of a tree trunk or a termite mound, the surroundings of which are covered in their faeces.

RED-WINGED TINAMOU Rhynchotus rufescens

The largest tinamou in Brazil is common and widely distributed in open environments south of the Amazon, including also extreme south-east Peru, northern Bolivia, eastern Paraguay, northern Argentina and Uruguay (Sick 1997, Cabot et al. 2020b). In Brazil, it occurs from Rio Grande do Sul to Rio Grande do Norte and west to Maranhão and Mato Grosso, where it has benefitted from deforestation, expanding its distribution (Sick 1997).

Ecological information on the species in the wild in Brazil comes mostly from Rio Grande do Sul, designed to generate subsidies for the control of hunting, which for many years was legal in the state (e.g., Sander 1982, Menegheti 1983, Menegheti et al. 1985, Pinheiro & López 1999). For breeding, the eggs are well described (e.g., Euler 1900, Ihering 1900, Miranda-Ribeiro 1905, 1938, Berla 1946, Britto 1950, Sick 1997), but there is little information on nests (e.g., Euler 1900, Miranda-Ribeiro 1905, 1938).

Observations were made on three nests found between 1985 and 1990 at Quebrangulo, Alagoas (n = 1) and Arcos, Minas Gerais (n = 2). Nests were found in October and February in Minas Gerais and in April in Alagoas, i.e., mainly during the austral summer, in agreement with the literature (Miranda-Ribeiro 1938, Carvalho 1939, Britto 1950, Setubal & Cavalcanti 1992, Sick 1997, Maurício et al. 2013, Hayes 2014). Breeding in the north-east is possibly later, in the rainy season, as at Quebrangulo. Berla (1946) found a nest with ten eggs in the same month in Pernambuco.

Nests were on the ground, between clumps of tall grass or below shrubs, as described by prior authors (e.g., Euler 1900, Miranda-Ribeiro 1905), and were simple depressions, lined with varying quantities of grass stalks and dry leaves, with some feathers among the eggs, probably belonging to the adult itself. Mean measurements of the three nests were: external diameter 27.6 ± 2.5 cm (range 25–30 cm); internal diameter 15 ± 1 cm (14–16 cm); external height 5 ± 1 cm (4–6 cm) and internal height 2 ± 0 cm. Nest classified as ‘simple/ unlined’ or ‘simple/platform’.

Clutch size varied from five to ten eggs, with a mean of 7.3 ± 2.5 per nest (Table 3; n = 3), which is within the species' known range (Euler 1900, Ihering 1900, Miranda-Ribeiro 1905, 1938, Carvalho 1939, Berla 1946). Hayes (2014), however, mentioned a lower mean for Paraguay, of 2.4 ± 0.6 per nest, range 1–5 eggs (n = 8).

Eggs (n = 11) were uniform brown, purple-brown or greyish brown, very shiny (Fig. 8), and short oval, albeit varying greatly in precise shape, as noted by Ihering (1900). Sick (1997: 158) mentioned that white eggs may exceptionally occur. On average, eggs measured 56.3 ± 3.7 × 43.6 ± 1.7 mm, mass 55.3 ± 4.7 g (n = 11) (Table 3), largely in accord with previous data (Berla 1946, Britto 1950), although those measured by Ihering (1900) were slightly smaller.

Incubation at one nest in Minas Gerais lasted at least 21 days. It was found with nine eggs on 24 February 1988. However, on 26 February, a tenth egg had been laid, possibly indicating that incubation had started very recently. The first nestling hatched on 16 March, but on 17 March a dog predated the nest; de la Peña (2019a) mentioned only birds as predators of the eggs and nest of this species. Other authors indicated an incubation period of 19–21 days (Miranda-Ribeiro 1938, Carvalho 1939, Sick 1997).

Young hatch covered in down, dark brown and quite striped, similar to young Rhea, as mentioned by Sick (1997). Soon after hatching, they can already run and hide in the vegetation. The male follows the chicks as they leave the nest and lowers his head on perceiving danger in the distance so as to be concealed by the vegetation. On closer approach, it adopts an aggressive posture, ruffling its feathers and spreading its wings. In real danger, the adult runs away and the chicks flee in all directions. We observed a male teaching the chicks to feed by picking up an insect or berry with his bill and dropping it in front of the young, encouraging the latter to seize the food.

Figure 8.

Eggs of Red-winged Tinamou Rhynchotus rufescens, Quebrangulo, Alagoas, Brazil, April 1985 (NORDESTA collection)

TABLE 3

Measurements of Red-winged Tinamou Rhynchotus rufescens eggs found in the wild in Brazil. Brazilian states: AL = Alagoas, MG = Minas Gerais. NM = not measured.

Although hunting is currently prohibited throughout Brazil, the species continues to be a favourite quarry. One of the nests in Quebrangulo had one of its eggs broken by poachers to identify if they were fresh and could be eaten. A second nest was destroyed by rural workers engaged in clearing an area. The species' decline due to hunting and deforestation has occurred in many parts of the country (Carvalho 1939). R. rufescens almost disappeared from around Arcos in the 1960s, but by working on environmental awareness with the local population, especially farmers, the species has slowly reappeared in the region. School and village events were organised, as well as visits to landowners to explain the importance of this species, which can also attract ecotourists. By eating insects, it moreover is an ally to farmers.

SPOTTED NOTHURA Nothura maculosa

One of the commonest and best-known tinamous of open areas in Brazil. It occurs from the north-east to the mid-west and south of the country, and is absent from forested areas as well as southern Bahia, eastern Minas Gerais and Espírito Santo (Sick 1997). It also occurs in Paraguay, Uruguay, and northern and eastern Argentina, reaching quite far south in the continent (Gomes 2020).

Even more than the previous species, the natural history of N. maculosa has been extensively studied in the wild in Rio Grande do Sul (e.g., Menegheti & Marques 1981, Menegheti 1981, 1982, 1983, 1984, 1985a,b, 1988, Burger 1985, 1991, 1992, Pinheiro & López 1999). Despite this, data on nests and eggs from Brazil are uncommon in the literature, both from the wild (Ihering 1900, Sick 1997, Lima et al. 2010, Lopes et al. 2013) and captivity (Tomimori 1982). Several publications with breeding information for N. maculosa accrued elsewhere are also mentioned in the following.

Eleven nests of N. maculosa were found between 1981 and 1994 at Quebrangulo, Alagoas (n = 2), Arcos, Minas Gerais (n = 7) and Altamira, Bahia (n = 2). Active nests were found in January–April in north-east Brazil (Alagoas, Bahia) and September–January in Minas Gerais (Table 4). In both regions, these periods correspond to the early to mid-rainy seasons and match other data from Brazil (Menegheti & Marques 1981, Menegheti 1988, Setubal & Cavalcanti 1992, Lopes et al. 2013), including captivity (Tomimori 1982), and from Argentina (Pereyra 1928, Bump & Bump 1969). The species' reproductive period appears well delimited in Brazil (Burger 1991, 1992, Setubal & Cavalcanti 1992), Argentina (Arriaga et al. 1983) and Paraguay (Hayes 2014), with birds having inactive gonads during the austral winter.

Nests were sited on the ground, well hidden among tall vegetation or under a bush. One was under a shrub Uncaria tomentosa (Rubiaceae) and another among dry leaves in the middle of a sugarcane plantation, both at Arcos. One nest was found near the urban area of Quebrangulo. When building, the bird often bends the surrounding vegetation to form a shallow platform, manipulating the grass above it into a kind of dome or arch that protects and conceals the nest (Fig. 9). The nest is abundantly lined with dried grass stems and leaves, and feathers may also be present, possibly from the adult itself (Fig. 10). According to Bump & Bump (1969), of 32 nests with some material, six had only grasses, whilst the other 26 had grasses and feathers. They noted that feathers are generally few in number, c.5–10, but 34 feathers were counted in a single nest.

Figure 9.

Spotted Nothura Nothura maculosa in nest, very well hidden by surrounding vegetation (the arrow indicates the bird's head), Arcos, Minas Gerais, Brazil, October 1992 (NORDESTA collection)

Figure 10.

Nest and eggs of Spotted Nothura Nothura maculosa, Arcos, Minas Gerais, Brazil, November 1987 (NORDESTA collection)

Two nests measured 12 and 18 cm external diameter, 6.5 and 10.0 cm internal diameter, 2 and 6 cm exterior height and 2 and 5 cm internal height, respectively. The nest is a ‘simple/ platform’ type, as noted by others (Dabbene 1923, Hudson 1928, Azpiroz 2001, Lima et al. 2010).

Clutches ranged from three to eight eggs per nest, mean 5.1 ± 1.5 (Table 4; n = 10); one nest with two eggs was excluded because laying was still incomplete when monitoring ceased. Other authors (Ihering 1900, Bump & Bump 1969, Belton 1984, Lima et al. 2010, Lopes et al. 2013) have reported similar clutch sizes, although Pereyra (1928) mentioned up to 11 eggs in a nest and Bump & Bump (1969) 12 eggs. For Paraguay, Hayes (2014) mentioned a mean of 3.3 ± 1.7 eggs per nest, range 1–5 (n = 4), but he may have included nests whose clutches were still incomplete.

TABLE 4

Measurements of Spotted Nothura Nothura maculosa eggs found in the wild in Brazil. Brazilian states: AL = Alagoas, BA = Bahia, MG = Minas Gerais. NM = not measured.

Continued

Eggs are uniform chocolate-brown or brown, very dark and shiny (Fig. 10) with a long oval, elliptical shape, and both poles rounded or even pointed. Mean size was 42.1 ± 1.7 (n = 37) × 30.8 ± 0.8 (n = 38) mm and mass 20.1 ± 2.0 g (n = 31) (Table 4), in accord with the literature (Dabbene 1923, Bump & Bump 1969, Tomimori 1982), although Ihering (1900) noted some slightly larger measurements, as did Bump & Bump (1969), who indicated that egg width can reach up to 36.1 mm. Tomimori (1982) mentioned that, in captivity, the first eggs laid by young females are smaller. Sick (1997) thought the egg of White-bellied Nothura N. boraquira less glossed than in N. maculosa.

Data on incubation were obtained at just one nest, visited every second day. It was found on 14 September 1992 with one egg. On 16 September, there were two eggs, 20 September, three eggs, 22 September, four eggs, and 24 September, five eggs, indicating that the female usually laid one egg every c.2 days, although the interval between the second and third eggs was four days. Of the five eggs, four hatched on 12 October, an incubation period of 18–19 days. These observations suggest that the species initiates incubation after the last egg is laid since all hatched on the same day, i.e., synchronously (Campbell & Lack 1985). The young left the nest on the day they hatched, with the adult. A second nest, found on 11 January 1993 in Quebrangulo, held six eggs. All hatched on 1 February, indicating at least a 21-day incubation period. In Rio Grande do Sul incubation lasts 16–18 days (Menegheti 1988). In captivity, laying occurs at intervals of 2–5 days, and artificial incubation lasts 17–19 days at 37°C (Tomimori 1982).

In November 1993, at Arcos, we observed an adult come off the nest feigning an injured wing, held slightly drooping beside the body. Once, an individual was caught in a net and, when handled, pretended to be dying with relaxed body and limp legs, but flew away quickly and noisily as soon as it was released. One adult flew from the nest when the observer was just 50 cm away. The behaviour of males at the nest tends to vary individually, ranging from those that immediately fly off to those that attack potential predators, both other birds and humans (Daguerre 1926, Tomimori 1982). Thanatosis, or feigning death, has been previously documented in the species (Pozzi 1927).

Family Anhimidae.—An exclusively South American family of paludicolous habits, Anhimidae numbers three species (Winkler et al. 2020c, Remsen et al. 2021), two of which occur in Brazil (Pacheco et al. 2021).

HORNED SCREAMER Anhima cornuta

Widespread in South America, from Colombia and Ecuador to Bolivia and Paraguay (Piland 2020). Although Sick (1997) described it as Amazonian, the species occurs in almost every state of Brazil, with historical records from many locations, including frequent records in the north-east, where it was first reported in the 17th century (Teixeira 1992).

Although breeding information including from captivity (e.g., Schomburgk 1848, Lint 1956, Gill et al. 1974, Barrow et al. 1986, Naranjo 1986) has been published from other countries, for Brazil knowledge was, until recently, anecdotal and in some cases doubtful (Euler 1900, Ihering 1900, Teixeira 1992, Pascoal et al. 2016). However, a recently published work that reviewed images from WikiAves, focused on breeding records for the Cerrado biome, has substantially improved knowledge (Tubelis 2020).

Observations were made at four nests of A. cornuta, all in the municipality of Alto Parnaíba, Maranhão, in August–December between 2003 and 2020, at the end of the dry and the start of the rainy season. For the Brazilian Cerrado, two breeding periods are indicated, a longer one during the drier period, from early June to mid October (peak August–September), and another, in the rainy period, from mid December to early February (Pascoal et al. 2016, Tubelis 2020). In Colombia, nests with eggs have been reported in August and December (Gill et al. 1974), and breeding is considered to occur between late November and early May (Naranjo 1986). Thus, significant variation in seasonality has been noted, but with breeding evidence from virtually every month of the year across the entire range. More observations and a more detailed review are needed to better understand seasonality, including the species' still poorly known moulting cycle (Haffer 1968).

The four nests were on the ground in a Buriti palm Mauritia flexuosa (Arecaceae) grove, and always near water. One was on a small island in the middle of a lagoon, at the base of three young Buriti trunks. Nests were constructed with all kinds of plant material such as grass stalks, dry grass, Cyperus sp. (Cyperaceae) and Typha sp. (Typhaceae), found mainly in the immediate surroundings, and included green leaves and other assorted vegetation (Fig. 11). Two nests measured: (A) external diameter 80 × 87 cm; internal diameter 33 × 40 cm; external height 18 cm; internal height 10 cm; (B) external diameter 79 cm; internal diameter 32 × 38 cm; external height 13 cm; internal height 8 cm. All were of the ‘simple/platform’ type. Although previous descriptions of nests are similar, we were unable to find precise measurements of nests in Brazil, with the only information derived from image analysis indicating nests with diameters of c.1–3 m (Tubelis 2020). For Colombia, Gill et al. (1974) mentioned nests with a diameter of about 1 m, probably external, and 8–10 cm internal height, whilst Naranjo (1986) mentioned a diameter of 65 cm but did not specify whether this was measured externally or internally, although we suspect the former. As far as could be ascertained, the nests studied here were not floating, unlike some described previously (e.g., Gill et al. 1974, Naranjo 1986, Todd 1996). Sick (1997) mentioned that the nest of the Anhima is ‘half floating’, whilst Tubelis (2020) concluded that most of the species’ nests are built near water on floating or non-floating vegetation, with which we concur. Pascoal et al. (2016) mentioned that the nest they found was built on macrophytes but was not floating (W. Pascoal in litt. 2021). One of the reasons we believe that nests of Anhima do not float comes from a nest near a lake, which flooded after heavy rains on 30 September 2009, with the rising water inundating the three eggs. Todd (1996) indicated that, although they float, nests can drift away.

Figure 11.

Adult Horned Screamer Anhima cornuta in nest, Alto Parnaíba, Maranhão, Brazil, September 2009 (NORDESTA collection)

Clutch size varied between three (n = 3) or four (n = 1) eggs, as reported by others (Euler 1900, Ihering 1900, Gill et al. 1974, Naranjo 1986, Pascoal et al. 2016). For the Cerrado it is two to five (Tubelis 2020) but can reach seven (Kear 2005). Coloration of the eggs was always creamy brown (Fig. 12), becoming darker with time, but other reports range from white (Euler 1900, Ihering 1900, Naranjo 1986) to brownish with yellowish spots (Naranjo 1986) or olive-brown (Sick 1997). An egg may be initially white, but changes colour with dirt brought to the nest by the adults (Sick 1997, Tubelis 2020).

TABLE 5

Measurements of Horned Screamer Anhima cornuta eggs from Alto Parnaíba, Maranhão, Brazil, 28 August 2003.

Three eggs measured on average 86.8 ± 3.2 × 59.8 ± 2.1 mm, with mass 136 ± 14.7 g (Table 5) and were oval or long oval in shape. These are possibly the first published measurements of A. cornuta eggs in the wild in Brazil. The only egg size datum for the species in Brazil is from captivity in Ceará, where eggs measured 82 × 54 mm (Dias da Rocha 1921). For Colombia, nine eggs measured 84.6 ± 3.2 × 61.0 ± 1.9 mm and weighed 149.9 ± 11 g (Naranjo 1986). In Peru, the species' eggs are sometimes consumed by humans (González 1999).

Figure 12.

View from above of a Horned Screamer Anhima cornuta nest with eggs, Alto Parnaíba, Maranhão, Brazil, November 2004 (NORDESTA collection)

Observations at two nests permit us to conjecture as to the incubation period. One nest found with two eggs on 28 August 2003, held three eggs the next day. The three chicks hatched on 9 October, or an incubation period of 41 days, from the laying of the last egg. All of the four eggs in a second nest, found on 19 November 2004, hatched on 31 December, thus indicating a minimum incubation period of 42 days. Due to the absence of sexual dimorphism, we could not ascertain if both sexes incubated, but observations in captivity, and in the wild by others, indicate that they do (Lint 1956, Naranjo 1986, Kear 2005), with the female remaining at the nest during the day and the male by night (Naranjo 1986). In captivity, artificial incubation lasts 44 days (Lint 1956), whilst in nature Naranjo (1986) estimated 40–47 days, but did not visit the nest more regularly to enable greater precision. Given that A. cornuta probably starts incubating before all the eggs are laid, as is true of Southern Screamer Chauna torquata (Stonor 1939), we hypothesise that, in the wild, incubation lasts 41–42 days.

When adults leave the nest, they usually cover the eggs, as already noted by Sick (1997). Nestlings hatch with yellowish down, resembling the young of Anatidae, and usually stay close to the adults. According to Naranjo (1986), they probably remain in the adults' territory for the first year of their lives.

SOUTHERN SCREAMER Chauna torquata

Occurs from south-east Peru, northern Bolivia and Rondônia south to central and eastern Argentina, Uruguay and Rio Grande do Sul (Brady 2020). In Brazil, it is especially common in the Pantanal and the wetlands of Rio Grande do Sul (Fontana et al. 1995, Sick 1997). Like the previous species, the breeding ecology of C. torquata is poorly known (Gibson 1880, Daguerre 1934, Stonor 1939, Pereyra 1940, Weller 1967, Azpiroz 2001, Fox et al. 2019), including in Brazil, from where only nest and egg measurements appear to be available (Euler 1900, Ihering 1900). We describe observations at five nests, all in the municipality of Poconé, Mato Grosso, between 2007 and 2009.

Figure 13.

Southern Screamer Chauna torquata nest, near centre of photo, amongst floating vegetation, Poconé, Mato Grosso, Brazil, June 2007 (NORDESTA collection)

Figure 14.

Nest and eggs of Southern Screamer Chauna torquata, Poconé, Mato Grosso, Brazil, July 2009 (NORDESTA collection)

TABLE 6

Measurements (in cm) of Southern Screamer Chauna torquata nests found in the wild in Brazil.

Available literature is scarce and confusing as to seasonality in this species, which probably varies geographically and depends on water levels. We found active nests in June (n = 4) and July (n = 1), i.e., the dry season in the Pantanal. In Brazil, it is said to occur July–December (Euler 1900, Ihering 1900), in Argentina in August–January (Gibson 1880, Weller 1967), and in August in captivity in England (Stonor 1939). We analysed content for the species on WikiAves (22 February 2021), finding 12 photos using the filter ‘egg’, 116 under ‘nest’ and 76 under ‘immature’. The filter ‘egg’ yielded images taken March–December (except June), ‘nest’ April–February, and ‘immature’ every month except May. Most photos retrieved via the first two filters were in July–November (92.9%). Most images, however, are from Rio Grande do Sul. When focusing the search solely on the Pantanal (Mato Grosso and Mato Grosso do Sul), photos are concentrated in the drier months, with ‘nests’ between June and August and ‘immatures’ mainly from July to September.

Nests were all on floating vegetation (Fig. 13), mainly Eichhornia crassipes (Pontederiaceae). To construct the nest, the birds use various stalks, sometimes very thick on the outside, and leaves of the same plant on the inside (Fig. 14). Viewed from above, nests are not circular, and so have a larger and smaller diameter. Average of three nests: external diameter 75.3 ± 27.3 × 83.3 ± 35.1 cm; internal diameter 30.3 ± 2.5 × 38.0 ± 5.3 cm; external height 13.7 ± 1.5 cm; and internal height 10 ±1.7 cm (Table 6). The difference between internal and external diameters demonstrates the thickness of the nest wall. These nests agree with those already described (Dunford 1877, Gibson 1880, Weller 1967, Sick 1997), although nest height is sometimes reported as being much greater, c.50 cm (Euler 1900). In captivity, possibly because of the reduced availability of material, C. torquata may use vegetation from the first nest to build the second (Stonor 1939). Weller (1967) mentioned that the bird may spend a long time ‘kneading’ the vegetation at the chosen site until the area is flat. Feathers may be present in some nests (Weller 1967). The nest is a large structure of the ‘simple/platform’ type.

In the vicinity of one nest was an adult Wattled Jacana Jacana jacana, probably a male, which was always present and was visited by the presumed female (Fig. 15). No interactions were witnessed between the two species and it is interesting to note that J. jacana appears in four photos on WikiAves next to or even in nests of C. torquata (V. Chahin,  http://www.wikiaves.com/410984; T. Machanoker,  http://www.wikiaves.com/2347819; L. C. Moura,  http://www.wikiaves.com/3086281; M. S. Hundertmarck,  http://www.wikiaves.com/3915329).

Figure 15.

Adult Southern Screamer Chauna torquata in nest with a Wattled Jacana Jacana jacana, probably a male, which was close by throughout the study period, Poconé, Mato Grosso, Brazil, July 2009 (NORDESTA collection)

TABLE 7

Measurements of Southern Screamer Chauna torquata eggs found in the wild in Mato Grosso, Brazil. NM = not measured.

Mean clutch size at five nests was 4.2 ± 0.8 eggs (Table 7), very similar to Weller's (1967) findings in Argentina, with an average of 4.7 eggs per nest (n = 18 nests). Other literature, however, indicates two or three (Sick 1997), five or six (Gibson 1880, 1920, Ihering 1900) or seven to nine eggs (Euler 1900, Gibson 1920).

Mean measurements of eight eggs were 94.2 ± 2.0 × 61.3 ± 1.1 mm and mass 177.5 ± 12.8 g (Table 7). Eggs are long, oval-shaped and white, sometimes cream-coloured, probably from accumulated dirt (Fig. 14). These observations agree with the literature (Sick 1997). However, average egg size is often smaller than we found, e.g., 86.36 × 58.42 mm (Gibson 1880, 1920), 90 × 59 mm (Euler 1900) and 87–89 × 61–62 mm in Rio Grande do Sul (Ihering 1900). Gibson (1920) mentioned that eggs are sometimes eaten by humans. In Argentina, nests are often parasitised by Black-headed Duck Heteronetta atricapilla (Cabrera et al. 2017).

We were unable to record the incubation period, but according to the literature it is 44–45 days (Sclater 1905, Weller 1967, Sick 1997), and by both sexes. Contrary to Naranjo's (1986) report for A. cornuta, the female C. torquata usually remains in the nest at night (Stonor 1939, Weller 1967).

Family Anatidae.—A cosmopolitan family that inhabits most aquatic environments, with the exception of truly pelagic waters (Winkler et al. 2020d), Anatidae are represented in South America by 47 species (Remsen et al. 2021), of which 26 occur in Brazil (Pacheco et al. 2021).

WHITE-FACED WHISTLING DUCK Dendrocygna viduata

Occurs in the Old World, in sub-Saharan Africa and on islands like Madagascar and the Comoros, and in the New World from Costa Rica to central Argentina (Sick 1997, Carboneras & Kirwan 2020a) with occasional records further north in the USA (O'Connell & McBride 2019). It probably colonised the Americas naturally from Africa, and in Brazil occurs in all states but shuns more forested regions of Amazonia (Sick 1997).

Comparatively, much more is known about the species' breeding in the Old World (Clark 1976, Jones 1978, Carboneras & Kirwan 2020a). In the Americas, information is mainly from Argentina (Casares 1935, Di Giacomo 2005, de la Peña 2005, 2013, 2019a), Bolivia (Kerr 1901), Suriname (Hellebrekers 1942) and Uruguay (Azpiroz 2001). For Brazil, we found very few published data (Britto 1950, Sick 1997, Lopes et al. 2013).

Observations were made at three nests found between 1981 and 2020 at Iguatama, Minas Gerais (n = 1), Pium, Tocantins (n = 1), and Quebrangulo, Alagoas (n = 1). Our few observations cannot define seasonality in Brazil. However, they all coincide with the rainy season in the respective regions, which pattern seems to be repeated throughout the species' distribution, including the Old World (Hellebrekers 1942, Britto 1950, Clark 1976, Belton 1984, Petrie 1998, Di Giacomo 2005, Lopes et al. 2013, de la Peña 2019a, Carboneras & Kirwan 2020a).

Figure 16.

Nest of White-faced Whistling Duck Dendrocygna viduata hidden at the base of a dense scrub, indicated by the red arrow, Quebrangulo, Alagoas, Brazil, May 2020 (NORDESTA collection)

Figure 17.

Nest and eggs of White-faced Whistling Duck Dendrocygna viduata, Quebrangulo, Alagoas, Brazil, May 2020 (NORDESTA collection)

Nests were always on the ground, in open areas with grass 30–80 cm tall and always near flooded environments (Fig. 16). The nest in Minas Gerais was c.100 m from a pond in a cattle pasture towards the top of a slope. It was a slight depression in the ground, surrounded by dense grass, and the interior was lined with dry, thinner grass and a few feathers. This nest measured 27 cm external diameter, 20 cm internal diameter, 3.5 cm exterior height and 2.0 cm internal height. The nest in Alagoas, also in a cattle pasture, measured 27 cm external diameter and 18 cm internal diameter; it was built on trodden grass and had the same internal and external height. The nest in Tocantins (not measured) was constructed under a bush and protected by dense vegetation. The material used for construction seems to have been collected in their immediate surroundings. Although the literature indicates that nests of the species are built on the ground (Britto 1950, Clark 1976, Sick 1997, de la Peña 2019a), like those we observed, there are reports of nests on tree forks (Clark 1976) or in palm hollows (Kerr 1901). In captivity, according to Britto (1950), the female oversees choosing the site and building the nest, usually under vegetation, be it grass, bushes or even trees (Britto 1950, Sick 1997, de la Peña 2019a). All nests observed by us were lined with dry grass, but there are records of green grass (Britto 1950) or even without any lining (de la Peña 2019a). In South Africa, of 16 nests, seven had feathers in the lining (Clark 1976). Few nest measurements are available, ranging from 16–25 cm external diameter and 5 cm internal height (Britto 1950, de la Peña 2019a, Carboneras & Kirwan 2020a).

TABLE 8

Measurements of White-faced Whistling Duck Dendrocygna viduata eggs found in the wild in Brazil. Brazilian states: AL = Alagoas, MG = Minas Gerais, TO = Tocantins. NM = not measured.

Our nests had 11–13 eggs (Table 8). Clutch size appears to be quite variable; the literature indicates 6–7 (Britto 1950), 7–13 (Clark 1976), 4–13 (Carboneras & Kirwan 2020a) or 16 in single cases in South Africa and Argentina (Clark 1976, de la Peña 2019a) and even up to 30 in Argentina (Casares 1935). Probably, nests with more than 14–15 eggs are the result of more than one female, sometimes including egg dumping by Fulvous Whistling Duck D. bicolor (Britto 1950, Clark 1976, Sick 1997).

Eggs we found were dirty white (Fig. 17) with mean size 48.3 ± 1.0 × 37.2 ± 0.5 mm, and mass 34.9 ± 1.3 g (n = 12) (Table 8), agreeing with known variation in the species (Casares 1935, Hellebrekers 1942, Britto 1950, Clark 1976, de la Peña 2019a). Only Di Giacomo (2005) mentioned much longer eggs, reaching 57.5 mm.

Incubation was followed at the nest in Minas Gerais, which contained 11 eggs on 16 December 1981 and 12 eggs the next day, which was the complete clutch. The 12 eggs were still being incubated on 15 January 1982, but all hatched next day. By 11.00 h, the nest was already abandoned by the chicks indicating a 30-day incubation period from the laying of the last egg. Incubation in the literature is reported as 26–30 days (Britto 1950, Sick 1997, Kear 2005, Carboneras & Kirwan 2020a). Eggs were also laid daily in the nest in Alagoas. It contained four eggs when found on 4 May 2020, six on 6 May, eight on 8 May, 12 on 12 May, and the complete clutch of 13 eggs on 15 May. In South Africa, Clark (1976) also reported that eggs are laid at 24-hour intervals.

The Alagoas nest was probably predated by dogs after the farmer released cattle into the area accompanied by dogs the day before the nest was lost.

BLACK-BELLIED WHISTLING-DUCK Dendrocygna autumnalis

Unlike other species of Dendrocygna in Brazil, D. autumnalis occurs only in the New World, from the southern USA to northern Argentina and in all Brazilian states (Sick 1997, James & Thompson 2020).

Information on the species' breeding is abundant; however, the vast majority of this knowledge pertains to D. a. fulgens, which occurs further north, mainly from Texas, USA (e.g., Meanley & Meanley 1958a,b, Bolen 1962, 1967, Bolen et al. 1964, Bolen & Cain 1968, Cain & Arnold 1974, Delnicki & Bolen 1975, 1976, 1977, Delnicki et al. 1976, Heins-Loy 1986, Schneider et al. 1993, Edmonds & Stolley 2008, Stolley et al. 2008, James et al. 2012). For South America, where the nominate occurs, information becomes scarce (Casares 1935, Hellebrekers 1942, Haverschmidt & Mees 1994, Di Giacomo 2005, de la Peña 2013), including Brazil, where publications are very few (Snethlage 1935, Sick 1997, Camacho & Pimentel 2012).

Observations were made at eight nests, all found during 2005–10 in the municipality of Poconé, Mato Grosso. Records were concentrated in the rainy season in the Pantanal, with the earliest nests in September and the latest in March (Table 9). The peak, 50% of records, was in December, one of the wettest months in the region (Cardoso & Marcuzzo 2010). This matches other observations in the country (Snethlage 1935, Camacho & Pimentel 2012). There must, therefore, be some variation in seasonality across the species' range, depending on local rainfall. In the USA, where it is migratory, the species arrives to breed between March and September (Schneider et al. 1993), with a peak in May–July (James & Thompson 2020). In Argentina, the few records also coincide with the rainy season (Di Giacomo 2005, de la Peña 2019a), whilst in Suriname the species starts breeding at the end of the rainy season (Haverschmidt & Mees 1994).

Except one nest, in a box specially built to provide for breeding Psittacidae (see Table 10, nest 7), the other breeding records were in palm cavities, probably Acrocomia aculeata (Arecaceae). These were broken and already dead palms, which had lost their crowns and had rotten stems, with vertical cavities of different depths. Internally, there was no lining to the egg chamber, only small wood fragments from the palm itself, almost like dust (Fig. 18). The adult entered either through the top of the stump or via a lateral crack. One of the palms had been used for nesting by Turquoise-fronted Parrot Amazona aestiva, which hatched chicks 15 months earlier. A nest in Rio de Janeiro state was also in an Acrocomia palm (Camacho & Pimentel 2012). Sick (1997) indicated that the species nests in tree hollows, palm fronds or even on the ground. Although apparently less common, there are several records of nests outside cavities (Bolen et al. 1964, Haverschmidt & Mees 1994, Markum & Baldassarre 1989a); these are usually less successful (Edmonds & Stolley 2008). The species commonly uses nest boxes in some regions (Bolen 1967, McCamant & Bolen 1979, Feekes et al. 1992, Croft et al. 2020).

Figure 18.

Adult Black-bellied Whistling Duck Dendrocygna autumnalis in nest, Poconé, Mato Grosso, Brazil, November 2005 (NORDESTA collection)

TABLE 9

Measurements of Black-bellied Whistling Duck Dendrocygna autumnalis eggs found in the wild in Mato Grosso, Brazil. NM = not measured.

TABLE 10

Measurements of Black-bellied Whistling Duck Dendrocygna autumnalis nests found at Poconé, Mato Grosso, Brazil. NM = not measured.

Nest measurements varied greatly, especially as they were sited in natural hollows; for example, the entrance to the cavities, high in palms, averaged 5.6 ± 3.0 m above ground but varied from 2–10 m (n = 7). The ‘corridor’ between the entrance and egg chamber measured 116.4 ± 75.5 cm but varied from 41–210 cm (n = 7). All measurements are shown in Table 10. We found just one report with nest measurements from Brazil, mostly inside the same range of variation (Camacho & Pimentel 2012).

The number of eggs varied from 16 to 30 per nest, with a mean 23.3 ± 4.8 (Table 9). This is very high and certainly not the product of single females. In fact, it is difficult to quantify the number of eggs laid by each female because intraspecific parasitism occurs in up to 100% of nests in some regions (James & Thompson 2020). Single-nest means of between 21.4 and 29.9 eggs have been documented in North America (McCamant & Bolen 1979, James & Thompson 2020), and our observations are similar. Delnicki et al. (1976) described an exceptional case of a nest box that held 17 eggs, but 48 hours later had 50; an increase of 33 eggs in just two days, the product of at least 17 different females. The same nest five days later held 90 eggs and reached a max. 101 eggs, of which 38 hatched. At one nest we observed, 13 eggs were laid in just six days. Thus, parasitic behaviour is common in D. autumnalis. Our nest with 30 eggs was also in a box, indicating how boxes may be essential for the species. Possibly, parasitic behaviour is regulated by the availability of suitable cavities, as this is usually a scarce and much-competed for resource (e.g., Kappes 1997, Brightsmith 2005, Botero-Delgadillo et al. 2015) and could even justify interspecific parasitism, with D. autumnalis documented egg dumping in nests of other cavity-nesting Anatidae, e.g., Wood Duck Aix sponsa (Bolen & Cain 1968) and Muscovy Duck Cairina moschata (Sick 1997). The only counterpoint, a record of parasitism in a ground nest, involved Laughing Gull Leucophaeus atricilla (Ballard 2001). For more information, see Delnicki & Bolen (1975), who discuss cavity availability and breeding in Texas.

Figure 19.

Newly hatched chicks of Black-bellied Whistling Duck Dendrocygna autumnalis, Poconé, Mato Grosso, Brazil, December 2009 (NORDESTA collection)

Size of 23 eggs (Table 9) was on average 51.5 ± 1.5 × 37.0 ± 1.2 mm, mass 33.2 ± 3.8 g. Eggs were all white and oval or, less often, short oval in shape. Coloration and measurements agree with previous data (James & Thompson 2020). However, mean mass tends to be somewhat higher elsewhere, both in Argentina with 38.1 g (Di Giacomo 2005) and in Texas with 41 g (James & Thompson 2020).

Incubation in three different nests lasted: 29 days, 26–30 days, and 32 days. As more than one female laid in each nest, it was difficult to know when laying finished and incubation started. The literature indicates a c.28-day incubation period (Bolen et al. 1964, James & Thompson 2020). The chicks (Fig. 19) jumped from the nest when only one day old.

At the nest with 30 eggs in the nest box, 25 chicks hatched and left the cavity, 24 in the morning and one in the afternoon. In the nest with 23 eggs, 18 hatched and jumped out. At the nest with 25 eggs, 19 hatched and departed, and finally, at a nest with 16 eggs, 13 hatched and fledged. Two nests suffered complete or partial egg predation and were abandoned, and a third was abandoned with all 26 eggs. In Costa Rica, hatching success was 13–66%, although lower values are more frequent (McCoy et al. 1992).

Figure 20.

A palm with a natural hollow used for breeding by Muscovy Duck Cairina moschata, the arrow indicates the entrance (see also inset), Poconé, Mato Grosso, Brazil, December 2007 (NORDESTA collection)

MUSCOVY DUCK Cairina moschata

The only bird species domesticated by pre-Columbian peoples, it was brought to the Old World as early as the first half of the 16th century (Sick 1997, Gamboa 2019). In the wild, C. moschata occurs from Mexico to northern Argentina and throughout Brazil (Donkin 1989, Sick 1997, Eitniear et al. 2020). According to archaeological studies, some South American populations may be derived from domestic forms translocated by native peoples (Stahl 2005, Stahl et al. 2006). There is often some difficulty in identifying wild individuals from domestic ones, both in live birds and by skeletal characteristics (Nacinovic 1982, Stahl 2005).

Information about the species' breeding in the wild in Brazil is rare (Ihering 1900, Sick 1997). According to Kear (2005), there is no information on breeding success for wild C. moschata, and it is one of the New World Anatidae whose ecology is most poorly known (Markum & Baldassarre 1989b).

Observations were made at four nests found between 2007 and 2021 at Poconé, Mato Grosso (n = 2) and Arcos, Minas Gerais (n = 2). At Poconé, one nest was found on 19 December 2007 and the other on 15 November 2011, whereas at Arcos one was found on 29 October 2021 and one on 17 November 2021, or the rainy season in both regions, in accord with the pattern elsewhere (Wetmore 1965, Di Giacomo 2005, de la Peña 2019a, Eitniear et al. 2020).

At Poconé, one nest was in a palm (Arecaceae) cavity, with a lateral/superior opening 8 m above ground and a ‘corridor’ to the nest 60 cm long (Fig. 20). The opening measured 35 × 25 cm, and the internal diameter of the egg chamber was 30 cm. This nest was densely lined with a large amount of down below and around the eggs. At the base was a layer of wood dust from the palm itself. The second Poconé nest was in a wooden nest box, on a Handroanthus albus (Bignoniaceae), 12 m above ground (Fig. 21). The tree was adjacent to a cattle pen near an occupied house; the nest box was erected in February 2011 and occupied in November of the same year. The box measured 110 cm external height and 55 × 55 cm external diameter, with an opening 40 cm high by 25 cm wide, 65 cm above the base of the box. Like the other nest above, it was lined with much down. At Arcos, both nests were in the same high sandy cliffs formed by erosion. The first was inside a cavity 6 m above ground, the entrance opening measured 50 × 32 cm, the base of the cavity was 40 cm, and the nest had external diameter 38 cm, internal diameter 24 × 35 cm and external height 55 cm. The second nest was about 50 m away, 15 m above ground, the entrance opening was 40 × 40 cm, the base of the cavity was 45 cm, and the nest’s external diameter was 35 cm, internal diameter 25 × 30 cm and external height 40 cm. As at Poconé, both were lined with down probably from the female’s body. Other observations indicate that it is common for C. moschata to use feathers to line the nest (e.g., Casares 1933, Sick 1997, Di Giacomo 2005) in addition to plant materials (de la Peña 2019a).

Figure 21.

Nest box used by breeding Muscovy Duck Cairina moschata, with nestling leaping from the nest, 12 m above ground, Poconé, Mato Grosso, Brazil, November 2011 (NORDESTA collection)

Use of natural cavities in palms and trees is common (Lloyd 1897, Ihering 1900, Casares 1933, Wetmore 1965, Sick 1997, Di Giacomo 2005), including the crevice of a cave wall (Eitniear et al. 1998). Nest boxes are used in some regions (Woodyard & Bolen 1984, Markum & Baldassarre 1989b), including the Pantanal (Guedes 2004). But non-cavity nests have been reported in reedbeds (Wetmore 1965), abandoned nests of hawks or Jabiru Jabiru mycteria (Sick 1997).

Clutches of 14 and 15 eggs were recorded at Poconé and 15 and 16 eggs at Arcos. Eggs are oval to long in shape, and mean size was 63 ± 2.7 × 45.7 ± 0.8 mm (n = 46) and mass 73.7 ± 3.8 g (Table 11; n = 45). Clutches elsewhere ranged from 8–21 eggs (Lloyd 1987, Ihering 1900, Casares 1933, Wetmore 1965, Markum & Baldassarre 1989b, Eitniear et al. 2020), which must be viewed with care, considering that a single nest may contain the eggs of more than one female (Wetmore 1965, Woodyard & Bolen 1984, Markum & Baldassarre 1989b). Markum & Baldassarre (1989b), who were able to follow such events by monitoring nest boxes, found a mean 13.6 ± 3.7 eggs per nest (n = 13), range 9–21 eggs, but in 31% of nests more than one female had laid. When only nests with more than one female were analysed, the authors found an average of 17.7 ± 3.2 eggs per nest (n = 4), range 15–21 eggs, whilst nests visited by just one female had 9–15 eggs, mean 12.6 ± 1.9 (n = 9). Based on this, we are unsure if more than one female laid eggs in the same nest at our sites. In general, our size and mass data are within the norms described in the literature (Ihering 1900, Casares 1933, Hellebrekers 1942, Britto 1950, Markum & Baldassarre 1989b, Di Giacomo 2005). On laying, the eggs were white but acquired a light brownish-yellow hue, in accord with descriptions of yellowish or greenish-white eggs (Ihering 1900, Casares 1933, Hellebrekers 1942, Wetmore 1965). Often, they can be stained by the wood they are in contact with (Di Giacomo 2005).

TABLE 11

Measurements of Muscovy Duck Cairina moschata eggs found in the wild in Brazil. Brazilian states: MG = Minas Gerais, MT = Mato Grosso. NM = not measured.

Continued

In the nest box at Poconé, laying was completed with the 15th egg on 15 November 2011, and the chicks hatched on 19 December, or after 34 days of incubation. In Argentina, the incubation period is 31 or 32 days, by the female alone (Di Giacomo 2005), whilst in Mexico it is 30 or 31 days (Markum & Baldassarre 1989b). Periods of up to 35 days are known (Sick 1997, Eitniear et al. 2020) but have been considered to pertain to captive birds (Markum & Baldassarre 1989b), although Harun et al. (1998) disagreed.

The nest in the palm cavity, found on 19 December, initially held 14 eggs, but on 7 January there were 13 eggs and the shell of one egg was on the ground outside the cavity; the female was still incubating. On 10 January, several shells were on the ground and the nine eggs still in the nest were cold and abandoned. Of the 15 eggs in the nest box, six did not hatch. The nine chicks were attracted by two adults vocalising at the base of the tree, behaviour initiated at c.09.00 h, and the first chick appeared at the opening and jumped at 09.30 h. After the first, the others followed; once all nine were on the ground, they followed the adults away at 10.00 h. The ground where the chicks landed was hard and stony, with few grasses.

The nest at Arcos found on 29 October 2021, 6 m above ground, held 15 eggs, but on 17 November, there were only nine and no evidence of others in the environs. On 24 November, the chicks began to hatch early morning, and between 13.00 and 13.30 h six leapt to the ground into tall wet grass, where the female covered them from rain. At 16.00 h, the three remaining chicks jumped. The female walked in the direction of a 3 km-distant river. It was raining, and the ground was slippery, aiding the progression of the family towards the river. The second nest at Arcos, found on 17 November, 15 m above ground, had 16 eggs being incubated by a female. On 25 November 2021, hatching commenced and a few hours later the chicks began to leap to the ground where the female was calling to them. It was not possible to count how many chicks fledged, but they were observed walking in the direction of the river along a marshy ditch. Both nests at Arcos are subject to competition for use and, in previous years, were occupied by White-eyed Parakeet Psittacara leucophthalmus, Collared Forest Falcon Micrastur semitorquatus, American Kestrel Falco sparverius, Laughing Falcon Herpetotheres cachinnans, Great Horned Owl Bubo virginianus and Toco Toucan Ramphastos toco.

BRAZILIAN TEAL Amazonetta brasiliensis

Exclusively South American, with records in all Brazilian states (Sick 1997, Carboneras & Kirwan 2020b), it inhabits varied aquatic environments, both natural and anthropogenic, but is less common in forested zones (Sick 1997). The sexes are easily distinguished, especially by the colour of the bill, which is reddish in males and dark in females (Carboneras & Kirwan 2020b).

Knowledge of breeding is still minimal, both elsewhere (Wilson 1977, Madriz & Pacheco 1978, Azpiroz 2001, Cejas 2003, Di Giacomo 2005, de la Peña 2019a) and in Brazil (Ihering 1900, Snethlage 1935, Coimbra-Filho 1964, Belton 1984, Nascimento & Antas 1990, Sick 1997, Lopes et al. 2013), especially compared to other common wildfowl.

Observations were made at a nest found on 23 July 2021 at Quebrangulo, Alagoas. It was c.100 m from a small weir, in the middle of a pasture, near a dirt road (Fig. 22), a situation similar to that described by Belton (1984) in Rio Grande do Sul. At the time, the grass was relatively tall (50–80 cm) but after the eggs hatched the owner grazed cattle in the area.

Figure 22.

Aerial image of lagoon, near nest of Brazilian Teal Amazonetta brasiliensis, with detail of female (left) and male and their seven chicks, Quebrangulo, Alagoas, Brazil, July 2021 (NORDESTA collection)

The nest was on the ground, with grass arranged in a circular shape, some feathers and had an external diameter of 25 cm, internal diameter 20 cm, external height 6 cm and internal height 3 cm. It was a ‘simple/platform’ type and similar to other .0descriptions for the species (Belton 1984, Di Giacomo 2005).

There were seven eggs, mean size 48 ± 0.3 × 36.3 ± 1.1 mm and mass 27.5 ± 0.6 g (Table 12). Their original coloration should have been white or beige, but they were very dirty, and the background colour was intangible. Clutch size varies widely in this species, from 3–14 eggs (Belton 1984, Cejas 2003, Di Giacomo 2005, de la Peña 2019a). Size is within previously reported parameters (Di Giacomo 2005).

TABLE 12

Measurements of Brazilian Teal Amazonetta brasiliensis eggs found at Quebrangulo, Alagoas, Brazil, 23 July 2021.

On 25 July, only two days after the nest was found, seven chicks were observed at the pond accompanied by a pair of adults (Fig. 22). After 28 July, they were not seen again, probably having moved to another area.

WHITE-CHEEKED PINTAIL Anas bahamensis

Widespread in South America but absent from the central part of the continent, especially Amazonia and the Cerrado; also found in the Caribbean and Galápagos (Carboneras & Kirwan 2020c). In Brazil, it occurs from Rio Grande do Sul to the Pantanal and the northeast, being commonest near the coast (Sick 1997). It inhabits a wide range of aquatic environments, from salty and brackish waters, e.g., mangroves, to inland lagoons, natural or artificial (Sick 1997).

Information on the species' breeding outside Brazil is not rare (Hellebrekers 1942, Marchant 1960, Rumboll 1967, McKinney & Briggers 1983, Meier et al. 1989, Buden 1992, Sorenson 1992, 1994, Sorenson et al. 1992, González-Bruzual & Marín-Espinoza 2013, Davis et al. 2017, de la Peña 2019a). In Brazil, however, we found only sparse data (Ihering 1900, Lima 2006, Nacinovic 2018).

Observations were made at a nest found on 3 August 2021 in the municipality of Pão de Açúcar, Alagoas. It was in a marshy area on the left bank of the São Francisco River, where some small ponds are used for fish farming, and was c.1 m from one of these lagoons. A water pump was in operation c.10 m away. The nest was on the ground amid marginal vegetation, comprising grass and herbaceous plants, at the base of a small bush (Fig. 23). The nest had an external diameter of 22 cm, internal diameter 12 cm, external height 10 cm and internal height 8 cm. It was a ‘simple/platform’ type, constructed of the surrounding grass and lined with many fine, dark feathers.

TABLE 13

Measurements of White-cheeked Pintail Anas bahamensis eggs found at Pão de Açúcar, Alagoas, Brazil, 3 August 2021.

Figure 23

Nest site of White-cheeked Pintail Anas bahamensis, Pão de Açúcar, Alagoas, Brazil, August 2021 (NORDESTA collection)

Figure 24

Nest and eggs of White-cheeked Pintail Anas bahamensis, Pão de Açúcar, Alagoas, Brazil, August 2021 (NORDESTA collection)

The clutch was seven eggs, white and long, oval-shaped (Fig. 24), mean size 50.3 ± 0.1 × 35.4 ± 0.1 mm and mass 29.9 ± 0.7 g (Table 13). Clutch and egg size are within previous parameters reported (Lima 2006, Carboneras & Kirwan 2020c), although de la Peña (2019a) mentioned larger and heavier eggs.

On the afternoon of 5 August, two days after the nest was found, six eggs hatched. Based on its duller bill, probably the female returned to the nest and began to vocalise from the lagoon's shore and was soon accompanied by the newly hatched ducklings.

Family Cracidae.—Occurs exclusively in the Americas, from the southern USA to northern Argentina, predominantly in forest environments, but sometimes adapts to modified landscapes (Sick 1997, Delacour & Amadon 2004). Eleven genera and 54 species (Winkler et al. 2020e), mostly in the South American tropics, of which 26 occur in Brazil (Pacheco et al. 2021). Among the most threatened bird families in the world (Pereira & Brooks 2006).

RUSTY-MARGINED GUAN Penelope superciliaris

Occurs south of the Amazon and east of the Madeira Rivers across most of Brazil but is uncommon in the southernmost states of Santa Catarina and Rio Grande do Sul. Also occurs in eastern Bolivia, Paraguay and north-east Argentina (del Hoyo & Kirwan 2020a). Many races and intrapopulation plumage variations are known but need detailed study (Neumann 1933, Nacinovic 2012, Evangelista-Vargas et al. 2017).

Figure 25.

Adult Rusty-margined Guan Penelope superciliaris in nest built on a tangle of vines, Arcos, Minas Gerais, Brazil, November 2017 (NORDESTA collection)

Figure 26.

Nest and eggs of Rusty-margined Guan Penelope superciliaris; note cluster of vines around nest, Arcos, Minas Gerais, Brazil, October 2017 (NORDESTA collection)

Figure 27.

Adult Rusty-margined Guan Penelope superciliaris in nest, with agglomerated materials supported in a fork, Arcos, Minas Gerais, Brazil, November 2017 (NORDESTA collection)

Figure 28.

Dead nestling Rusty-margined Guan Penelope superciliaris in nest, probably killed by ants, Arcos, Minas Gerais, Brazil, November 2017 (NORDESTA collection)

Although little is known about the species' breeding, especially in the wild, a number of publications are available (Euler 1900, Bertoni 1901, Taibel 1953, Sick 1997, Bodrati & Castillo 2008, Toledo-Lima et al. 2013, de la Peña 2019a). Observations were made at seven nests found between 1991 and 2017 at Alto Parnaíba, Maranhão (n = 2), Arcos, Minas Gerais (n = 3) and Poconé, Mato Grosso (n = 2).

TABLE 14

Measurements of Rusty-margined Guan Penelope superciliaris eggs found in the wild in Brazil. Brazilian states: MA = Maranhão, MG = Minas Gerais, MT = Mato Grosso. NM = not measured.

Although our observations covered well-separated localities, all nests were found in October–November (Table 14). In Paraguay, nesting records are available for September and late October (Bertoni 1901, Bodrati & Castillo 2008), in Argentina for November (de la Peña 2019a) and in Brazil the species nests in February in Rio Grande do Norte (Toledo-Lima et al. 2013) and probably October–January in Paraná (Mikich 1996).

Nests were sited 2–5 m above ground (Table 15) and were supported by tangles of vines (Figs. 25–26). In two cases where vines were scarce, the nest was in the fork of a tree or shrub, ensuring greater support (Fig. 27). Nests were platforms, sometimes deeper, with accumulating sticks on the outsides, and all were lined with dry and fresh leaves. They were in forest fragments, one near a pigsty on a farm and another in a tree in the middle of a pasture, which was somewhat intermediate in form between ‘simple/platform’ and ‘low cup/base’. Nest measurements are presented in Table 15. They matched descriptions in the literature (Euler 1900, Bertoni 1901, Bodrati & Castillo 2008, Toledo-Lima et al. 2013, de la Peña 2019a). However, Sick (1997) mentioned a nest in Espírito Santo built on a rock inside forest and Toledo-Lima et al. (2013) a nest close to the ground on terrestrial bromeliads.

Except one nest that had two eggs, all others contained three eggs (Table 14). Eggs were completely white (Fig. 27), two with beige-coloured spots, and shape was long oval. Mean size was 63.7 ± 3.1 × 44.9 ± 1.7 mm and mass 64.3 ± 10.9 g (Table 14; n = 14). Clutch and egg sizes match previous reports (Bertoni 1901, Bodrati & Castillo 2008, Toledo-Lima et al. 2013, de la Peña 2019a). The eggs found in Minas Gerais were larger and heavier, whilst eggs from Mato Grosso were smallest and lightest (Table 14). These variations could be mere artefacts of our small sample or reflect population differences, perhaps even linked to morphological variation. Most of our observations are from the range of P. s. jacupemba, whereas southern Maranhão is within that of P. s. superciliaris (del Hoyo & Kirwan 2020a), but further studies are required in order to make subspecific comparisons. However, eggs of very distinct sizes can be found in the same nest, as noted by Toledo-Lima et al. (2013) for Rio Grande do Norte, which were 55.5–64.3 × 38–42.1 mm. We consider that Euler's (1900) report of clutches numbering up to four eggs requires confirmation.

TABLE 15

Measurements of Rusty-margined Guan Penelope superciliaris nests found in the wild in Brazil. Brazilian states: MA = Maranhão, MG = Minas Gerais, MT = Mato Grosso. NM = not measured.

The incubation period, from the last egg being laid to the first hatching, could be calculated for only one nest. The bird finished laying two eggs on 28 October 2017, and both hatched on 26 November, for a total of 29 days. Sick (1997) mentioned a period of 28 days, Toledo-Lima et al. (2013) 29–30 days and de la Peña (2019a) and Delacour & Amadon (2004) only 26 days. The latter, shorter period, seems to be more common in captivity (del Hoyo & Kirwan 2020a).

Of the seven nests, one could not be followed, so its fate is unknown. Of the other nests, the eggs were predated at four, and two hatched. Of the latter, three eggs in one nest all hatched and the young fledged. At the second, with two eggs, both hatched but one of the nestlings was killed in the nest, probably by ants (Fig. 28), whilst the other fledged. If we discount the three eggs at the seventh nest, of the 17 remaining eggs, just five (29.4%) hatched, from which one nestling was predated, thus total reproductive success was 23.5%. This is probably the first datum of this type for the species and among the first of its kind for the genus Penelope.

CHACO CHACHALACA Ortalis canicollis

Inhabits the Chaco region of Bolivia, Paraguay, northern Argentina and, in Brazil, the states of Mato Grosso and Mato Grosso do Sul, on the Pantanal floodplain (Sick 1997, del Hoyo & Kirwan 2020b).

Very little is known about breeding (Pereyra 1939, Di Giacomo 2005, Schaaf et al. 2014, de la Peña 2019a), especially in Brazil (Sick 1997, Previatto et al. 2017), where eggs have been only briefly described from two clutches each of three eggs in captivity (Britto 1950). Observations were made at a single nest found on 12 December 2007 in the municipality of Poconé, Mato Grosso.

The nest was on a heavy fork of a Curatella americana (Dilleniaceae) amid a few vines, and constructed of thick, dry twigs and lined internally with dry leaves. It was 4 m above ground, with external diameter 20 × 24 cm, internal diameter 14 cm × 16 cm, external height 10 cm and internal height 6 cm. Like the previous species, the form is intermediate between ‘simple/platform’ and ‘low cup/base’. For Argentina, Pereyra (1939) indicated that O. canicollis often builds its nest atop the large nest of Firewood-gatherer Anumbius annumbi, but such behaviour has not been reported by others (see de la Peña 2019a).

TABLE 16

Measurements of Chaco Chachalaca Ortalis canicollis eggs found at Poconé, Mato Grosso, Brazil, 12 December 2007.

The nest contained three white, oval (n = 2) or long oval (n = 1) eggs, with a mean size of 61.1 ± 1.5 × 45.6 ± 0.2 mm, and mass 60.1 ± 0.2 g (Table 16). Five days later, one egg had disappeared, and by 20 December, the entire nest was destroyed and the eggs missing, presumably predated. Egg coloration has been described as yellowish (Britto 1950), and clutch and egg sizes agree with previous literature (Britto 1950, Delacour & Amadon 2004, Di Giacomo 2005, Schaaf et al. 2014, de la Peña 2016).

Figure 29.

Adult East Brazilian Chachalaca Ortalis araucuan in nest, Quebrangulo, Alagoas, Brazil, January 2020 (NORDESTA collection)

EAST BRAZILIAN CHACHALACA Ortalis araucuan

Occurs in eastern Brazil from Espírito Santo and Minas Gerais to Rio Grande do Norte (Teixeira 1992, Sick 1997, del Hoyo et al. 2020).

Very little is known of the species' natural history and breeding (Euler 1900, Velho 1932, Dantas & Silva 2003, Toledo-Lima et al. 2013). Some care is needed when analysing data from the literature for genus Ortalis, as taxonomy can be confusing. For O. araucuan, some publications list it as O. guttata, of which O. araucuan was frequently considered a subspecies (Pinto 1978).

TABLE 17

Measurements of East Brazilian Chachalaca Ortalis araucuan eggs found at Quebrangulo, Alagoas, Brazil, 21 January 2020.

Observations were made at a nest found on 21 January 2020 in the municipality of Quebrangulo, Alagoas, within the Pedra Talhada Biological Reserve. The nest, in a tangle of vines atop a banana tree at the edge of a marsh (Fig. 29), was lined with dry green leaves. The vines were used as a support, forming a platform. There were three white eggs, stained green from contact with decomposing green leaves in the lining. Eggs were long, oval-shaped and averaged 55.2 ± 0.0 × 35 ± 0.6 mm; mass 38.9 ± 1.4 g (n = 3) (Table 17). Two eggs hatched on the same day, but the third was not fertilised.

In Rio Grande do Norte, a nest was found on 3 February and nestlings seen on 8 May (Toledo-Lima et al. 2013), coinciding with the nesting period in Alagoas. A nest in Pernambuco, c.1 m above ground with two eggs, was photographed in December 2021 (J. A. V. Filho,  https://macaulaylibrary.org/asset/394363991). The eggs found by Toledo-Lima et al. (2013) indicate significant variation in weight and size in the same clutch, in accord with our observations, although the eggs we measured did not vary in length (Table 17).

Apparently common in north-east Brazil in the 1940s (Berla 1946) and 1990s (Teixeira 1992). Around Quebrangulo, observations by AS suggest that populations have increased in recent years (since 2000).

Family Podicipedidae.—Of global distribution, only absent in Antarctica, polar regions of the Northern Hemisphere and some heavily forested or desert areas (Winkler et al. 2020f). Their morphology is so well adapted to swimming that they can barely stand upright on land (Sick 1997). Worldwide, six genera and 22 species are recognised (Winkler et al. 2020f), nine of them in South America (Remsen et al. 2021) and five in Brazil (Pacheco et al. 2021). Some are of very restricted distribution and flightless. They possess some of the most interesting reproductive behaviours among birds, with complex mating rituals (Winkler et al. 2020f).

LEAST GREBE Tachybaptus dominicus

The smallest American grebe. It occurs from the southern USA and West Indies south to central Argentina, including all of Brazil (Sick 1997, Storer 2020).

Although one of the least known American grebes in terms of breeding (Storer 2020), there are data from throughout most of the species' range, for example, the USA (Bancroft 1930, McMurry & Monson 1947, Hasse & Hasse 2009, Patrikeev 2009, Pistone 2015), Mexico (Castillo-Guerreiro et al. 2002, Ortega-Álvarez 2013, Pérez-Valadez 2017), El Salvador (Miller 1932, Dickey & Rossem 1938), Honduras (Hayes 2018), Panama (Wetmore 1965), Ecuador (Marchant 1960), Venezuela (Gonzalez & Espinoza 2015) and Argentina (Lorenzón et al. 2015, Vitale & Ferrari 2017, de la Peña 2019b). In Brazil, although there is some published information (e.g., Euler 1900, Ihering 1900, Santos 1938, Pinto 1941, Mitchell 1957, Belton 1984, Sick 1997, Lopes et al. 2013), very little is known, with just one description of egg measurements (Euler 1900) and another of a single nest (Belton 1984).

Figure 30.

Least Grebe Tachybaptus dominicus in nest, Quebrangulo, Alagoas, Brazil, March 1995 (NORDESTA collection)

Observations were made at six nests found between 1995 and 2001 at Quebrangulo, Alagoas. Nests with eggs were found in March, and the last newly hatched chicks in late August, coinciding with the principal rainy season. As a species dependent on waterbodies, especially lakes, this period should guarantee the best conditions for breeding. Similarly, the literature indicates breeding in the rainy months elsewhere, e.g., November in Rio de Janeiro (Euler 1900) and Minas Gerais (Lopes et al. 2013), and March and October in Rio Grande do Sul (Belton 1984). Where the climate is favourable, however, T. dominicus can breed year-round (Storer 2020), as well as in places with propitious artificial environments, such as ponds and tanks that can locally favour the species (Ortega-Álvarez 2013).

Nests were floating but fixed (Fig. 30). Four were in different-sized lagoons with aquatic shoreline vegetation, and the nests were mainly close to this. Two other nests were in shallow lagoons, completely covered by floating vegetation, similar to eutrophic pools. One nest had external diameter 41 cm, internal diameter 11 cm, external height 19 cm and internal height 3 cm. It was constructed from several aquatic plant species. These observations are consistent with those already described (Bancroft 1930, Belton 1984, Storer 2020). There is just one report of a non-floating nest of T. dominicus, in Honduras, where the bird nested on a concrete base (Hayes 2018). An exact categorisation of this nest under the Simon & Pacheco (2005) scheme is not possible, the closest being the ‘simple/platform’ type, but this is primarily a floating nest.

Mean clutch size was 4.2 ± 0.8 (n = 6) pale beige eggs. Shape ranged from oval (n = 6), long oval (n = 7) to long pointed oval (n = 3). Mean size was 35.6 ± 1.4 × 24.2 ± 0.7 mm, mass 10.5 ± 0.7 g (n = 16) (Table 18). These measurements accord with the few available data (Ihering 1900, Marchant 1960, Storer 2020). In coloration, the species' eggs vary greatly, simply due to the bird's behaviour of covering the eggs when it leaves the nest. As the plant material is usually wet, the eggs become stained with brownish hues (Storer 2020) and can become increasingly dark, to the point of being wood-coloured (Santos 1938).

TABLE 18

Measurements of Least Grebe Tachybaptus dominicus eggs found at Quebrangulo, Alagoas, Brazil. NM = not measured.

At a nest with four eggs, found on 19 May 2000 between 10.50 and 12.20 h, the alarm call of a nearby Wattled Jacana Jacana jacana agitated the adult in the nest. Once, when the J. jacana vocalised, the grebe quickly covered the eggs and hid in the aquatic vegetation, returning to the nest ten minutes later. On another occasion, when the grebe spotted a person in the vicinity, it left the nest without covering the eggs. It dived into the vegetation but returned to the eggs five minutes later. Three adults gathered around this nest at one point, and one climbed up to incubate the eggs. Covering the eggs is well known in the species (Storer 2020) but when the adult leaves quickly due to perceived danger, it does not always do so, as also observed in Central America (Miller 1932). With the nest covered, it may stay away for many hours; no adult, for example, returned to a nest with covered eggs during three hours (between 06.00 and 09.00 h).

The egg-laying interval at one nest was c.24 hours. This nest contained one egg on 6 July 2000 and four eggs on 10 July. The hatching process, however, was somewhat distinct in two nests: the first contained five eggs on 22 May 2000, and by day 24 all five had hatched; at the second, which had five eggs on 19 June, one chick hatched per day, whilst two eggs failed. This suggests that incubation begins with the laying of the first egg. Nestlings are nidifugous; one jumped into the water only 15 minutes after hatching. Bancroft (1930) commented that less than an hour after hatching, the nestling is already able to dive and swim.

In one nest with two chicks, one and two days old, and one egg, between 14.10 and 15.30 h both adults fed the chicks. One adult left with a nestling on its back on several occasions, while the other remained in the nest with the other nestling and incubated the remaining egg. Twice, realising they were being observed, the adults quickly covered the egg and even the nestlings, using the nest material, dived and returned shortly afterwards. Nestlings were fed mainly in the nest, just once in the water. At another time, between 07.40 and 09.45 h, an adult was observed diving 86 times, 39 of which were followed by the delivery of food to a single 15-day-old chick accompanying it. The chick chirped each time it received food. At 08.19 h, the chick climbed on the adult's back and stayed for 15 minutes, during which time the adult did not dive. During the rest of the period, there was a total of one dive every 64 seconds and a food delivery every 2.5 minutes.

Only one nest was predated at the egg stage, 19 days after the first egg was laid; the predator was not identified.

PIED-BILLED GREBE Podilymbus podiceps

Occurs from northern North America, where it is migratory, through Central America and the West Indies, to Patagonia, but is absent from desert areas and most of the Orinoco and Amazon floodplains (Muller & Storer 2020). In Brazil, it occurs mainly in the east, outside Amazonia (Sick 1997).

Figure 31.

Nest of Pied-billed Grebe Podilymbus podiceps; hatching is asynchronous, so the first and fourth chicks hatched c.60 hours apart, Quebrangulo, Alagoas, Brazil, August 2001 (NORDESTA collection)

Figure 32.

The adult Pied-billed Grebe Podilymbus podiceps covered its eggs before leaving the nest, using the nest's vegetation, Quebrangulo, Alagoas, Brazil, August 2001 (NORDESTA collection)

Breeding ecology better known than the previous species (e.g., Rockwell 1910, Allen 1914, Peck 1919, Bancroft 1920, 1930, Bullock 1923, Deusing 1939, Miller 1942, Glover 1953, Marchant 1960, Chabreck 1963, McAllister & Storer 1963, Wetmore 1965, Borrero 1971, McCowan 1973, Kirby 1976, Fugle & Rothstein 1977, Sealy 1978, Davis et al. 1984, Forbes & Ankney 1987, 1988a,b, Arnold 1989, Palmer-Ball 1991, Muller 1995, Azpiroz 2001, Benítez et al. 2004, Macana-García 2010, de la Peña 2019b, Routhier et al. 2020). For Brazil, however, information is still anecdotal (Ihering 1900, Belton 1984, Sick 1997) and sometimes error-prone (Miranda-Ribeiro 1927, Santos 1938).

TABLE 19

Measurements of Pied-billed Grebe Podilymbus podiceps eggs found at Quebrangulo, Alagoas, Brazil, 7 August 2001.

A nest found on 7 August 2001 in the municipality of Quebrangulo, Alagoas, was in a large farm weir, with much aquatic vegetation concentrated near the bank. It was a heap of aquatic vegetation near the bank, supported by the immersed branches (Fig. 31). The nest had an external diameter of 43 cm, internal diameter 12 cm, external height 12 cm and internal height 4 cm, in accord with Muller & Storer (2020). Use of mud in the nest, suggested by Eurico Santos (1938), is unlikely. Like the previous species, categorisation of the nest under the Simon & Pacheco (2005) scheme is difficult, being closest to the ‘simple/ platform’ type but floating.

The female laid four eggs, with mean size 47.5 ± 0.4 × 31.1 ± 0.5 mm and mass 21.3 ± 1.1 g (Table 19). Egg measurements, although larger, are close to those of Ihering (1900), and to Fugle & Rothstein (1977) and Muller & Storer (2020) for other countries. Egg mass tends to vary individually, as the first eggs to be laid are usually lighter (Forbes & Ankney 1988a), and egg 2 in Table 19 may have been the first laid.

Hatching was asynchronous, with one chick hatching in the morning and another in the afternoon of 20 August. The third hatched on the morning of the 21st, and the last only on the 22nd, in the afternoon. Asynchronous hatching was studied by Forbes & Ankney (1987).

The chicks returned to the nest for up to two days after hatching, either being carried on the adults' backs or swimming alongside them. Both adults helped care for the offspring; while one incubated the eggs or brooded the chicks in the nest, the other, in the water, carried one of the chicks on its back. The latter adult would climb onto the nest with the nestling still aboard, which then slid down its back and snuggled between the adult's feet while the other adult continued brooding. One of the adults left the nest with the three chicks, while the other incubated the final egg. When leaving the nest, this adult covered the eggs using the surrounding vegetation (Fig. 32). The adults dived frequently and brought food to the nestlings, primarily small fishes.