The western and trans-Andean populations of Striped Woodhaunter Automolus subulatus are sometimes considered separate species. We discuss previously published data on the nesting of Striped Woodhaunter and present novel information concerning the nest, eggs, nestlings and parental care of western A. s virgatus and trans-Andean A. s. subulatus. Nest placement and architecture of the two populations are similar to each other and to other Automolus species. However, Striped Woodhaunter build shorter nest tunnels than other related species and genera. All similarities in nest design, nestbuilding behaviour and parental care presented herein support the genetic clade including Automolus, Thripadectes and Clibanornis, but do not differentiate between the subspecies of Striped Woodhaunter. More studies are required about adult attendance and nest design within this clade, taking into account more samples across the species' range.
The systematics of the non-monophyletic genus Automolus (Furnariidae: Philydorini) are still under scrutiny (Claramunt et al. 2013, Schultz et al. 2017). The most recent changes that have gained acceptance are the subsuming of Hyloctistes within Automolus (Claramunt et al. 2013) and the splits of Pará Foliage-gleaner A. paraensis from Olive-blacked Foliage-gleaner A. infuscatus (Claramunt et al. 2013, Clements et al. 2019) and Chiriquí Foliage-gleaner A. exsertus from Buff-throated Foliage-gleaner A ochrolaemus (Freeman & Montgomery 2017, Chesser et al. 2018). Like other furnariids (Irestedt et al. 2006), nesting behaviour and nest architecture of Automolus spp. may prove useful for testing DNA-based taxonomic arrangements. However, the nesting biology of Automolus species is well known only for Chiriquí Foliage-gleaner (Skutch 1952, 1969) and White-eyed Foliage-gleaner A. leucophthalmus (Euler 1900, J. C. R. Magalhães in Remsen 2003a, Marini et al. 2007, Cockle & Bodrati 2017) but poorly known or unpublished for the other seven species (Remsen 2003a).
Striped Woodhaunter A. subulatus is distributed from eastern Nicaragua south to western Ecuador and, east of the Andes, from southern Venezuela and south-east Colombia to northern Bolivia and western Amazonian Brazil (Stiles & Skutch 1995, Remsen 2003a). Some authors treat western and trans-Andean populations as separate species, Western Woodhaunter A. virgatus and Amazonian Woodhaunter A. subulatus (Ridgely & Greenfield 2001, Hilty 2003, del Hoyo & Collar 2016) based largely on vocal differences (Ridgely & Tudor 1994, Freeman & Montgomery 2017). This split is considered premature by other authorities (see Remsen 2003b) and herein we follow Clements et al. (2019). Here we discuss and clarify published data on the nesting of Striped Woodhaunter and present novel information concerning the nests, eggs, nestlings and parental care for two subspecies, A. s. virgatus and A. s. subulatus.
Historical data for A. s. subulatus and A. s. assimilis.—The first published nest description attributed to Striped Woodhaunter was presented by Sclater & Salvin (1873). They quoted the notes of E. Bartlett, which accompanied a specimen of A. subulatus collected at Chamicuros, Loreto, eastern Peru (05°30′S, 75°30′W, sensu Stephens & Traylor 1983). However, the description, of a cup nest built 2–3 m above ground among dead palm fronds and holding two spotted eggs, is clearly in error, as first noted by Zyskowski & Prum (1999). The last-named authors provided the only other published information on the nest architecture of Striped Woodhaunter, including a photograph, based on a nest collected by N. Wheelwright in western Colombia (nest 1; Table 1). This nest was described as a platform-like cup of loosely interlaced leaf petioles placed at the end of an earth tunnel, and the photograph revealed two nestlings probably less than half-grown at the time of discovery (Zyskowski & Prum 1999). K. Zyskowski (in litt. 2018) kindly provided additional details (Table 1) on this nest which, based on its locality, is attributable to A. s. assimilis.
Monitoring dates, localities, nest contents and main observations of Striped Woodhaunter Automolus subulatus nests found in western Colombia (nest 1; Zyskowski & Prum 1999, Hilty 2003; K. J. Zyskowski in litt. 2018), in central-west Costa Rica (nest 2) and in eastern Ecuador (nests 3–7).
Nest and egg of A. s. virgatus.—We examined a nest of A. s. virgatus collected by J. E. Sánchez & E. M. Carman at Finca Rafiki Safari Lounge, Costa Rica (nest 2; Table 1), and deposited at the Museo Nacional de Costa Rica, San José (MNCR-ONH772). Sánchez et al. (2004) provided a habitat description for the locality. Nest 2 was collected from a burrow excavated in a dirt bank adjacent to a forest trail (cavity with tunnel, sensu Simon & Pacheco 2005), with an expanded inner chamber at the end of a tunnel. The nest itself was a shallow, platform-like cup composed entirely of loosely interwoven leaf rachises (Fig. 1A). We detected both leaflet scars and extra-floral nectaries on most of the rachises, suggesting that they were probably from a plant in the family Fabaceae. No additional details concerning the burrow are provided on the specimen label, but we were able to measure the nest platform (Table 2; on Fig. 2, see measurements 10–15). A single, unmarked white egg accompanied the nest (MZUCR-H205; Table 1; Fig. 1B) but the specimen label provides no details regarding clutch size or egg development. The size of the hole opened in the egg, however, suggested that it may have contained a well-developed embryo when collected.
Nests of A. s. subulatus.—HFG studied five nests of A. s. subulatus found between 2004 and 2013 at four localities in eastern Ecuador (Table 1): near La Selva Jungle Lodge (nest 3; Fig. 3A), Cabañas Yankuam (nest 4), Boanamo (nest 5; Fig. 3B–C) and Gareno Lodge (nests 6–7; Fig. 3D). Habitat was similar at all four localities, all representing mosaics typical of relatively undisturbed western Amazonian forest (see Greeney 2017, Greeney et al. 2018 for detailed descriptions). HFG visited the nests periodically to ascertain their status and contents. He checked nest contents either directly or using a small lighted mirror, and made direct observations of adult behaviours. When possible to document nest attendance by adults, he filmed nest activity at nests 3, 6 and 7 (Table 1) by placing a video camera on a tripod 1.5 m tall, 3 m from the nest entrance. Due to its position the video camera could not film activity within the inner chamber. Behaviour of the adults appeared to be unaffected by the presence of the camera.
All Ecuadorian nests were sited in earth burrows (cavity with tunnel, sensu Simon & Pacheco 2005) as described for assimilis and virgatus with entrances at a mean height of 124 cm (range = 60–230 cm; SD = 63.5 cm; Table 2) above ground (Fig. 2, measure 1). Nest 3 was in the root mass of an overturned Cecropia tree (Urticaceae), nest 4 was in a 1.5 m-tall bank with a 60 cm overhang along a road-cut, nest 5 was in streamside bank below an overhang of dirt and roots, and the other two nests were in the large (c.3 m tall) root masses of trees felled by wind action. The burrows' entrances led to tunnels that varied in slope from downward at a c.30° angle (nest 3) to sloping slightly upward, and opened into enlarged chambers containing the nest (nests 6 and 7). Nest cups of A. s. subulatus were platform-like structures of loosely arranged, stiff, unbranched leaf rachises that were barely sufficiently cohesive to remain intact when removed from the burrow. In the case of nest 5 (Fig. 3B), all rachises appeared to be from the same species of plant, but the taxonomic affinities of the nest materials were not examined closely at the other nests.
Measurements of A. s. subulatus burrows (Table 2; Fig. 2, measurements 2–9) were: entrance height = 7.1 cm (6.0–8.5 cm; SD = 1.1 cm; n = 4) and entrance max. diameter = 9.1 cm (8–11 cm; SD = 1.3 cm; n = 4); min. tunnel height = 5 cm (n = 1); tunnel width = 8.5 cm (7–10 cm; SD = 2.1 cm; n = 2); tunnel length (from entrance lip to start of inner chamber) = 26.3 cm (20–31 cm; SD = 5.2 cm; n = 4). Inner chamber height = 11 cm (10–12 cm; SD = 1.4 cm; n = 2); inner chamber max. diameter = 18 cm (13–23 cm; SD = 7.1 cm; n = 2); inner chamber min. diameter = 14 cm (12–16 cm; SD = 2.8 cm; n = 2). The inner chamber max. and min. diameter were measured perpendicular to each other on the horizontal plane (Fig. 2). HFG measured only the platform of nest 5 (Fig. 2).
Eggs and nestlings of A. s. subulatus.—The complete clutch at three of the five A. s. subulatus nests comprised two immaculate white eggs, although some were slightly stained pale brown, probably from the surrounding earth of the inner chamber (Fig. 3A–B).When the adults were not at the nests, HFG measured and photographed the eggs (nests 3, 5 and 7; Table 1; Fig. 3A–B) and one newly hatched nestling (nest 7; Table 1; Fig. 3D). Mean measurements of six eggs were 24.3 mm (22.8–25.4 mm, SD = 1.0 mm; Table 1) × 17.8 mm (17.1–18.5 mm, SD = 0.6 cm; Table 1). The masses of three eggs with advanced embryonic development were 4.1 g (3.8–4.3 g; SD = 0.3 g; Table 1). An undeveloped and slightly damaged egg weighed during the latter half of incubation had a mass of 3.5 g (nest 5). On HFG's final visit to nest 7 (Table 1), at 16.30 h, it contained a single nestling that probably hatched on the morning of the same day based on its physical appearance and mass. The second egg was lightly pipped, suggesting that the eggs' hatching would occur c.24 hours apart. The nestling weighed 4.7 g. It had long, densely plumose, grey natal down on its capital, spinal dorsal, spinal pelvic, alar, ventral sternal, femoral and crural regions (sensu Proctor & Lynch 1993). The skin was pinkish, including the tarsi and toes, with the cloaca and surrounding skin noticeably more whitish. Its nails were dusky white, as was the bill, except the dusky-grey tip. The prominent egg tooth, tomia and inflated rictal flanges were bright white and the mouth lining was pale pink, similar to the skin colour (Fig. 3D).
Measurements (cm) of all Striped Woodhaunter Automolus subulatus nests that we found, in central-west Costa Rica (nest 2) and eastern Ecuador (nests 3–7). Ent. = entrance, max. = maximum, diam. = diameter, I. cham. = Inner chamber, min. = minimum, Ext. = external, Int. = internal. Numbers in parentheses correspond to the measurement's numbers in Fig. 2.
Behaviour of A. s. subulatus adults.—At two nests (4 and 6) HFG observed nest construction over the course of 3–4 days. Nest 4 was visited three times during burrow excavation and, when first found, was c.10 cm deep. Three days later the tunnel was c.1.5 cm deeper, and eight days after discovery it was c.15 cm deep. A single adult was flushed from the nest on both the first (09.30 h) and final (14.30 h) visits. On both occasions the adult emerged with soil on its bill, flew directly into dense vegetation 3–5 m from the nest, and vocalised continually for the 4–5 minutes that HFG remained at the nest. The vocalisation, presumably an alarm call, was nearly identical to that recorded at a nearby locality by B. M. Whitney, given by an adult in response to playback of the same vocalisation ( www.xenocanto.org/86344). HFG was unable to determine if both sexes participated in excavation. The burrow of nest 6, when discovered, contained an empty but apparently fully formed nest. HFG visited the nest six times over the course of five days, between 06.15 h and 17.00 h, without observing an adult. Three days after discovery, during one hour of video observation (08.30–09.30 h), HFG recorded a single adult visit. The adult, of unknown sex, arrived with a single leaf petiole in its bill and remained in the burrow for c.3 minutes before flying away.
When first encountered, nest 3 contained a single egg showing no signs of development and a second with a tiny (>1 mm) embryo. Based on the experience of HFG with the embryonic development of numerous tropical suboscine passerines, we estimate that the clutch was completed 2–4 days prior and suspect that the undeveloped egg was inviable. HFG recorded adult incubation rhythms (on/off-bouts), between 06.00 h and 18.00 h (sunrise to sunset) on the five consecutive days following discovery of nest 3. As he was able to record only entrances and exits at the nest burrow, he inferred that the eggs were covered during the entire period an adult was inside. Both adults incubated the eggs, as evidenced by the observation of adults replacing each other at the nest, but HFG could not distinguish the sexes. During the entire observation period, adults spent 63.5% of daylight hours warming the eggs. Daily percentage attendance for the five days was 64.5, 43.4, 37.2, 79.2 and 80.0%, respectively. On HFG's final visit to nest 7 (Table 1), direct observations of adults at the nest revealed that both brooded the nestling and delivered single, very small (1–3 mm) prey items.
The nest placement and architecture of the three Striped Woodhaunter subspecies reported here are similar to those reported for other Automolus (A. leucophthalmus: Euler 1900, J. C. R. Magalhães in Remsen 2003a, Marini et al. 2007, Cockle & Bodrati 2017; A. ochrolaemus: Van Tyne 1926; A. paraensis: Snethlage 1935, Pinto 1953; A. exsertus: Skutch 1952, 1969). In particular, the exclusive use of leaf rachises in nest construction appears to be ubiquitous in Automolus, but their nests are otherwise similar in form and placement to the nests of related genera (Thripadectes, Clibanornis and Buff-fronted Foliage-gleaner Philydor rufum: Derryberry et al. 2011) being platforms of loosely woven material placed at the end of upward-angled earth burrows (Skutch 1969, Kiff et al. 1989, Strewe 2001, Remsen 2003a, Maillard et al. 2006, Faria et al. 2008, Botero-Delgadillo & Guayara 2009, Zyskowski & Greeney 2010, Miller et al. 2012, Smith & Londoño 2013, Cockle & Bodrati 2017). The nest descriptions reported here demonstrate that general nest placement and design support the strong relationship within genera of the Automolus-Thripadectes-Clibanornis clade (see Cockle & Bodrati 2017), and do not appear to vary between currently recognised subspecies or populations of Striped Woodhaunter (Remsen 2003a, Clements et al. 2019).
Perhaps of significance, we found that tunnel length of Striped Woodhaunter burrows is generally shorter (26.3 cm; 20–31 cm; SD = 5.2 cm) than reported for related genera (81.4 cm; 38–200 cm; SD = 47.5 cm; Van Tyne 1926, Remsen 2003a, Marini et al. 2007, Faria et al. 2008, Botero-Delgadillo & Guayara 2009, Zyskowski & Greeney 2010, Miller et al. 2012, Cockle & Bodrati 2017). We know from other burrow nesters that habitat and nest design (e. g., entrance size and orientation, and tunnel length) may be important for the regulation of appropriate nest microclimates (Ellis 1982, Haggerty 1995, Ke & Lu 2009). The single nest of A. s subulatus that we observed with a downward-sloping entrance tunnel (nest 3), may reflect regional variation in architecture based on local microclimate or, alternatively, may have been an error by the adults or one forced by roots, rocks, or other obstructions within the substrate.
At present, we are unable to confirm that both sexes of Striped Woodhaunter participate in burrow excavation and nest construction, as is known for some species in the Automolus-Thripadectes-Clibanornis clade (see Cockle & Bodrati 2017). We can confirm, however, that both parents participate in incubation and chick provisioning. This behaviour is shared among most furnariids including all members of the Automolus-Thripadectes-Clibanornis clade studied to date (Remsen 2003a, Cockle & Bodrati 2017) but differs vs. other relatives in the Philydorini (sensu Derryberry et al. 2011), e.g., Ochre-breasted Foliage-gleaner Anabacerthia lichtensteini and Sharp-billed Treehunter Heliobletus contaminatus, which have uniparental care (Cockle & Bodrati 2017). Although based on relatively small sample sizes, it appears that parental attendance during incubation may be higher in Striped Woodhaunter (63.5%) than has been reported for Chiriquí Foliage-gleaner (58.0%: Skutch 1952). A preliminary interpretation of these data might be that the shorter entrance tunnels of Striped Woodhaunter burrows, which may promote more rapid loss of heat within the nest (Ke & Lu 2009), may promote improved attendance. We suggest that further information on adult attendance and nest design within this group may uncover interesting correlations. Furthermore, the seemingly rare occurrence of downward-inclined burrows in Striped Woodhaunter merits further investigation.
We thank M. Corrales-Ugalde, J. Azofeifa-Solano, G. M Kirwan and K. Cockle for their comments that improved the manuscript; and A. Pérez ‘Cope’ for the illustration of the Costa Rican nest. For permitting access to their collections, we also thank the Departamento de Historia Natural, Museo Nacional de Costa Rica, and Museo de Zoología, Universidad de Costa Rica, especially C. Pineda, G. Alvarado, S. Bolaños and M. Springer. LS acknowledges the support provided by Vicerrectoría de Investigación, Universidad de Costa Rica, under project B9123 ‘Biología reproductiva de las aves tropicales: plasticidad fenotípica e historia natural'. HFG is grateful to Rudolphe A. Gelis, Eliot T. Miller and José Simbaña for their help with field work, to John V. Moore, Matt Kaplan, and the Population Biology Foundation for supporting his field work, and Pollywog Productions and Ryan Killackey for enabling his trip to Boanamo as part of the documentary feature film Yasuni Man.