Breeding

Timing of breeding. Reproduction is seasonal throughout the range of Rose-throated Becards. The longevity or permanence of pair bonds remains uncertain, but some evidence suggests that these bonds may be confined to the breeding season. In Guatemala, Skutch (1969) observed that becards in November and December "seemed not to associate in pairs", at least in the mountains above 2450 m. Forcey (2002) noted that "his earliest record of paired birds" in Oaxaca, Mexico, was on 4 April during his study. During the nonbreeding season (September-March) in Costa Rica, reports of single birds in eBird checklists are frequent, but reports of two birds may or may not signify an associated male and female, even when sex is identified, as few observers who submit checklists in this region comment on the behavior of birds they see.

Information on breeding seasonality and most other aspects of reproduction is fragmentary and non-quantitative. Testes begin to enlarge at least as early as February (one specimen, Oaxaca; Shaldach et al. 1997). Broadly, nesting activity sometimes initiates in the last half of March and extends into early August. Most breeding is concentrated from May to July (e.g., Oberholser 1974, Rowley 1984, Gehlbach 1987, Binford 1989, Stiles and Skutch 1989, Russell and Monson 1998, Forcey 2002, and Corman 2005) . Extremes occur as early as mid- (Forcey 2002; Oaxaca, nest-building) to late March (Schaldach et al. 1997; Oaxaca, nest-building) and as late as early September (Russell and Monson 1998; Sonora, adult associated with young bird near nest). In Arizona, arrival of adults, when breeding was more regular in the past, was chiefly from mid- to late May, occasionally in early May. Egg-laying extended from late May to early June, and parental care was from July to early August (Corman 2005; Appendix 1 ). In Texas, breeding reports from the early 1970s were during the period late May into July (Brush 2000; Appendix 1 ). Arrival or first appearance at traditional breeding sites there has been variable (summer dates suggest breeding season movements), ranging from 28 March (2006; very early) and 24 April (1999) to 30 July (1976) (Appendix 1 ). In Oaxaca in five years, earliest intraspecific agonistic behavior was 14-23 April, first nesting was 15 April-1 June, latest date for feeding nestlings was 22 July, and last dates for breeding (feeding fledged young) was 26 June-6 July (Forcey 2002). Also in Oaxaca (Rowley 1984), active nest building was in late April, and nests with eggs were early to late May. Nest-building in July and early August likely resulted from re-nesting attempts following early nest failures (Corman 2005). Breeding birds at higher elevations in the mountains may begin nesting about a month later than in some lowland localities in Mexico (Forcey 2002). No data are available on geographic variation in becard breeding seasonality among regions.

Nest-building, nests, and nest placement. Most information on reproduction focuses on the bulky, enclosed nest. Both sexes participate in nest-building, but the female contributes the larger share of the work (Davis 1945, Skutch 1969). During one 1.5 hr period, the female made 24 deliveries of material compared to 13 made by the male (Skutch 1969). In some cases the male appeared to make no contribution to the growing nest (Phillips et al. 1964, Rowley 1984). During nest-building, the male is attentive to his mate and often followed her during forays to collect more nest material (Skutch 1969, Rowley 1984), sometimes even carrying away a mouthful of material that he had brought as the female was departing the nest on another trip (Skutch 1969). The latter behavior may be construed as mate guarding in relation to paternity protection (Kokko and Morrell 2005). The female works on the exterior and interior of the nest, while the male confines most or all of his attention to the outside (Davis 1945, Skutch 1969). Large billsful of material are collected from near the nest or farther away (Sutton and Pettingill 1942). Material often is removed from an old (sometimes, other times not [Brandt 1951]) or same-season, failed nests for use in a new nest (Skutch 1969; W. J. Sheehan, in litt.). The length of the nest-building period is somewhat variable based on limited information. Skutch (1969) reported over two weeks to complete a nest (also in Baicich and Harrison 1997). In Texas, Brush (2000) reported a first nest built by a solitary male (no female was ever seen) took 27 days (8 May to 3 June 1999), while a second nest completed by the same male was finished in 9 days. Solitary nest-building by an apparent unmated becard is not unusual at the northern edges of the range of the species in Texas and Arizona, where often (at least in recent years) only one adult shows up in the spring or early summer, localizes itself, and commences to construct a nest (Brush 2000, Corman 2005; Appendix 1  examples).

The nest description below is based mainly on details provided in Bent (1942), Brandt (1951), Davis (1945), Phillips (1949), and Skutch (1969). Nests of this species are typically pendant and domed, and nearly always placed at the slender end of a drooping branchlet in tall deciduous trees (see G. M. Sutton’s illustration of one such nest in Brandt 1951). The egg chamber is enclosed above the bottom of the structure, which is usually attached to a single support, or sometimes to two adjacent supports. The single opening into the chamber through a short entry passage is placed on the side of the nest chiefly near the bottom, or sometimes closer to the middle, and often is obscured by nest material. The plant binding that secures the nest to a branch (usually) or vine (Smithe 1966) may include bark strips (Baicich and Harrison 1997), but I am unable to confirm this in the primary literature, which remains silent on the matter. Nests are often globular or rounded in general outline, but others were described as pear-like (Levy 1958), oval (Smithe 1966), cylindrical, "bushel-basket" shaped (Phillips et al. 1964), or bell shaped (Stiles and Skutch 1989). The composition of the nest is variable but typically contains long, pliable material including inner bark strips, dry grass leaves, stems of herbs, and sometimes even supplies of spider webbing ("cobweb", Skutch 1969) that may help to bind dry material. Other items mentioned in different nests were strands of hemp rope, dead leaves, feathers, small branches, rootlets, "insect webs", spider cocoons with tufts of silk, thistle down, pine needles (sometimes in fascicles), plant down, and sheep’s wool (Eaton and Edwards 1947, Phillips 1949, Brandt 1951, Levy 1958, Skutch 1969, and Corman 2005). One nest was described as comprised of "small sticks" (Griscom 1932). Another nest was composed of masses of Spanish moss (Tillandsia usneoides), with strands hanging below the nest and a mass protruding to the side (Sutton and Pettingill 1942). One singular, unorthodox nest was built on a horizontal branch of a small tree about 1.2 m above the ground (Eaton and Edwards 1947). It was non-pendant, as it was supported underneath by the branch (to which it was fastened) and on one side by the tree bole. As is typical for the species, the nest was a mass of dead leaves and grasses that enclosed a nest chamber. Sometimes nests are adorned with fresh, green moss or sycamore seed balls on the outside (Chapman in Bent 1942, Corman 2005). The well-padded lining of the egg cradle inside the chamber likewise varied with local availability and suitability of finer material, including plant down (e.g., silk cotton fibers of Ceiba aesculifolia; Rowley 1984), bark fiber, and feathers (Phillips 1949, Skutch 1969).

Measurements of pendent, elongated nests in the published sample ranged from 25-74 cm long (top to bottom) and 20-38 cm wide (F. M. Chapman in Bent 1942, Zimmerman and Harry 1951, Oberholser 1974, Brandt 1951). One globular nest was estimated to be 38-46 cm wide (Brandt 1951). An oval nest entrance was about 5 cm wide and 7.5 cm high (Brandt 1951), while a rounded one was 6.5 cm in diameter. The nest cavity in one nest was described as shallow and about 13 cm in diameter and 18 cm high (Brandt 1951). Nest walls were reported to be about 4-6.5 cm thick (Baicich and Harrison 1957).

Initially, the nest is "irregular and formless" (Skutch 1969). The material in the nest superstructure is layered and intermeshed, not tightly woven, giving the affair a messy appearance with sticks or other material protruding to the sides, and often streamers hanging below (Bent 1942, Phillips 1949, Brandt 1951, Skutch 1969). The female apparently creates the modestly-sized nest chamber, perhaps secondarily by forcing nest fibers aside after the bulk of the nest is well formed (Skutch 1969). The female often uses time off the eggs during incubation to add material inside and outside the nest during this breeding stage (Skutch 1969, Rowley 1984). Most observers mention the mobility of the nest, given its pendent nature and siting, as influenced by even light breezes. Stronger winds can whip the nest back and forth (Sutton and Burleigh 1940), yet wind-driven nest motion seems to be a normal trade-off for the apparent inaccessibility of the nest.

As noted, nests are chiefly placed at the supple, outer end of a lateral branch. Under the weight of the bulky nest, the slender supporting branch droops downwards. Often supporting branches are over open spaces (e.g., in a woodland, over a road corridor [Bent 1942], in a tree isolated in a plowed field [Zimmerman and Harry 1951]) or above water in a riparian situation. Tree species used depend on the region and plant community in which becards live. In Arizona, for example, cottonwoods (Populus fremontii) and sycamores (Platanus wrightii), which are dominant in that region along streams in arid regions in becard habitat, often are used (Brandt 1951, Corman 2005). Along the lower Rio Grande of Texas in riparian subtropical evergreen woodland, a local elm (Ulmus crassifolia), ash (Fraxinus berlandieriana) and live oak (Quercus virginiana) are frequently the trees of choice among other species (Gehlbach 1987, Brush 2000). One nest was suspended from a small vine in a supporting palm (Smithe 1966). Nest heights of pendant nests vary from 2.5 to 19 m. Mean height across the species range is 9.6 ± 5.32 (SD) m (n = 19), based on estimates from various sources cited above.

Eggs. The ground color of the eggs is somewhat variable, even within a clutch, ranging from pale whitish, dull gray, creamy, light buffy, pinkish cinnamon or light olive-cinnamon, to pale lilac. Brandt (1951) noted that fresh, intact eggs had a blush of "pale pink to salmon pink" that faded after the contents were removed and the shells dried following preparation. Egg shells are marked by spots, blotches, and scrawls, most heavily at the larger end, in shades of brown, reddish-brown, lilac, or gray (e.g., Bent 1942, Brandt 1951, Rowley 1984; see illustration in Baicich and Harrison (1997: plate 53). Shell luster has been described as glossy (Baicich and Harrison 1997), slightly glossy (Bent 1942), and dull (Brandt 1951), so this character may be somewhat variable among clutches.

Egg shape varies from long-ovate or sub-elliptical to ovate. Mean measurements (n = 22 eggs) are 23.2 x 16.9 mm (length x width) (Bent 1942). Extremes of length in this sample were 20.3-26.0 mm and extremes of width were 14.9-18.1 mm. In another sample of 16 eggs (from three nests), mean length and width was 23.6 x 17.1 mm. The corresponding extremes in this case were 22.0-25.1 (length) and 16.2-17.9 mm (width) (Rowley 1984). In one nest containing six eggs, mean dimensions were 22.8 x 17.7 mm (Brandt 1951). Brandt (1951) emphasized his impression, based on experience in preparing eggs for collections, that the shells of Rose-throated Becards were relatively strong given the size of the bird. He regarded this characteristic, and the heavy padding of the nest lining, as possible adaptations to the vulnerability of pendant nests to winds in exposed locations in tall trees.

Clutch size and incubation. Clutch size varies from 3-6 eggs. Oberholser (1974) gave the range from 2-4 eggs, but he is the only source to report a two-egg clutch. Baicich and Harrison reported the range as 4-6, which may be the core of the clutch size range. Rare two egg clutches may be incomplete or result from egg loss, while three egg clutches are a more likely lower limit than four eggs, given still inadequate data on the subject. No information is available on the modal clutch size in a sample. In a compilation of clutch sizes from published sources (n = 8 nests; Bent 1942, Phillips 1949, Rowley 1962, 1984), mean clutch size is 5.0 ± 1.20 (SD). This limited sample is almost certainly biased towards larger clutch sizes; the modal clutch was six eggs.

Only the female is reported to incubate eggs (Skutch 1969). During the day, attentive periods at one nest (n = 49) averaged 11.6 min (range 3-38 min), while recesses (n = 50) averaged 9.0 min (range 2-19 min) (Skutch 1962, 1969). During 17 hr of watching a nest during the day in Guatemala, under chilly, wet conditions, the female incubated only 56.3 percent of the time. One timed overnight session lasted 13.07 hr between waning light in the evening when the female last entered the nest to first departure the following morning. The male frequently called to the female on the nest, and she often seemed to respond with calls of her own as she sat on the eggs (Skutch 1969).

Based on information from specimen (skin) collectors in Oaxaca, Mexico, Parkes (1990) remarked that "Both [sexes] are said to have had incubation patches". This contradiction needs to be resolved, and probably arises from the limited life history information available on this species. Most of what we know on this subject comes from Skutch’s (1969) observations on two failed nests and one that was successful in the highlands of Guatemala. The incubation period is unknown.

Nestlings. The female broods recently hatched nestlings, but both parents cooperate in feeding young and providing other parental care (Skutch 1969, Davis 1945, Oberholser 1974, Forcey 2002). Nocturnal brooding ceased in the Guatemalan nest when the young were about 13 days old, while apparent daytime brooding ended six days earlier, when seven brooding sessions ranged from two to nine minutes. Initially, small, inconspicuous food items were brought to young nestlings (barely visible in the closed beak), but later, as the nestlings became older, larger items were more readily visible. Among these were green [probably lepidopteran] larvae and small, adult moths and butterflies with wings still attached. In one two-hour morning watch, when the young were estimated to be about a week old, Skutch (1969) observed the female bring food to the nest 9 times and the male 11 times. But, overall when the female no longer brooded the young, her rate of food delivery was somewhat greater than that of the male. In a total of six hours of watching over six days, the female made 68 food deliveries and the male contributed 58 deliveries. When the nestlings were about ten days old, Skutch commented he could hear faint calls from the nest, but that soon "they [uttered] typical calls of the adults, with all their varied inflections, but in a weaker voice….". The young remained in the nest for 18 or 19 days before the nest was destroyed by human vandalism. Skutch speculated that the young may have been able to fly at that age (contra Baicich and Harrison’s [1997] more assertive summary of Skutch’s observations). So the nestling period in this species, although apparently long and perhaps about 19 days, remains uncertain.