Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay

Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Revista Brasileira de Ornitologia, 22(2), 219-233 ARTICLE June 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople Department of Zoology, University of Cambridge, Downing Street, Cambridge, CB2 3EJ, UK. Email: rob.pople@cantab.net Current address: BirdLife International, Wellbrook Court, Girton Road, Cambridge, CB3 0NA, UK. Received on 03 September 2013. Accepted on 02 October 2013. ABSTRACT: Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay. I present the first detailed description of the breeding biology of the White-winged Nightjar (Eleothreptus candicans), based on data collected over three breeding seasons during 1998-2001 at Aguará Ñu, Canindeyú, eastern Paraguay. Male nightjars defended small territories situated on the upper slopes of ridgelines. Each territory contained one or more “display arenas” at which the male performed nuptial display flights. Aggregation indices confirmed that the primary display arenas of males were significantly clustered within the survey area. Within their territories, males apparently selected display arenas on the basis of their structural characteristics: mounds used as arenas were significantly lower and broader than random mounds. Males engaged in display activity from late August to early January. On average, males performed 0.54 ± 0.04 display flights per minute during nocturnal focal watches, but there was considerable intra-male variation in display rate. Following a burst of activity immediately after their arrival at display arenas at dusk, male display rate was best explained by ambient levels of moonlight. Males produced a previously undescribed insect-like “tik tik” call when inactive on their territories. Females made two nesting attempts per season, using a different site (360 ± 31 m apart) for each attempt. The clutch of two eggs was laid directly on the ground, adjoining a small access “clearing”, in either campo cerrado or wet grassland vegetation. The first egg was laid within two days of a full moon for all seven clutches for which laying date was confirmed. Incubation, brooding and chick provisioning duties were carried out exclusively by the female. The incubation period was 19 days and the nestling period c. 19-20 days. Overall, my data are consistent with polygyny – and even an exploded lek or “landmark” mating system – in this population of White-winged Nightjars. KEYWORDS: Caprimulgidae; Caprimulgus candicans; cerrado; male display behaviour; nesting biology INTRODUCTION chick (Capper et al. 2000, Clay et al. 2014) and male display behaviour (Clay et al. 2000). Male White-winged Owing largely to their cryptic plumage and crepuscular or Nightjars were found to perform ritualised display flights at small arenas – low anthills or termite mounds nocturnal habits, the Caprimulgidae (true nightjars) are comparatively under-studied. Their breeding biology in – located on the upper slopes of ridgelines in areas of particular is poorly known, and for many species the only open grassland. Although males were present near these published data are anecdotal or based on single nesting arenas throughout the night, display activity was highest attempts. Studies of the better-known species show them under cloudless conditions on moonlit nights. Display flights were accompanied by a dull “tk…grrrrrt” sound to be generally ground-nesting, with clutches of one or two eggs, and an essentially monogamous breeding (recording in Ranft & Cleere 1998), perhaps produced system (Cleere 1999, Holyoak 2001). The Caprimulgidae by movement of air through the outer primaries. are also one of only a few bird families in which lunar Based on the apparent clustering of male display synchrony in reproduction has been documented arenas at Aguará Ñu, Clay et al. (2000) suggested that this population of White-winged Nightjars might exhibit (Murton & Westwood 1977, Holyoak 2001). The White-winged Nightjar (Eleothreptus candicans) a lek or “landmark” mating system. Leks can be broadly is one of the rarest caprimulgids in the Americas, defined as any aggregation of males visited by females considered Endangered by BirdLife International (2013). primarily for the purpose of copulation, whereas landmark Research in Paraguay since the discovery of a population species are characterised by the use of a specialised habitat as encounter sites for mating (Höglund & Alatalo 1995). in 1995 at Aguará Ñu, Mbaracayú Forest Nature Reserve (Lowen et al. 1997), has clarified some aspects Four criteria were proposed by Bradbury (1981) to of its breeding biology, including the nest site, eggs and distinguish “classical” leks from other lek-like mating Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople systems: (1) lack of male parental care; (2) aggregation attached to certain individuals, including all adult females of males at an arena to which females come for mating; captured (see Pople 2003 for details). (3) arenas containing no resources required by females (e.g. nest sites, food) except males themselves; and (4) Male territories and display arenas free mate choice by females visiting an arena. Lekking has been suggested to occur in at least three other species of I use the term “territory” to describe the relatively small caprimulgid (see Holyoak 2001). However, no nightjars area within which a male nightjar conducted all his were included on a list of 97 lek-mating bird species breeding-season display activity (as per Clay et al. 2000). compiled by Höglund & Alatalo (1995). Each territory contained a primary “display arena” at Here I present the first comprehensive description which the male carried out most display activity, but some of the breeding biology of the White-winged Nightjar, also contained one or more secondary arenas, which were on the basis of data collected during a three-year study used more sporadically, for example when the male was in Paraguay, including male display behaviour, the disturbed from its primary arena (Clay et al. 2000). For distribution of male territories and details of nesting simplicity, the location of each territory was defined by biology based on data from multiple nesting attempts. its primary display arena. Radio-telemetry showed that c. I use these and other data to assess whether the study 95% of nocturnal breeding-season fixes for radio-tagged population exhibits a polygynous or lekking mating males fell within 100 m of their display arenas (Pople system, and consider the implications for the species’s 2003). In occasional cases when the primary display arena conservation. of a territory changed between or, more rarely, within seasons, the display arena used for longer was selected for subsequent analyses. METHODS The study site was surveyed for displaying males each season between August and December. Males were Fieldwork was conducted over three breeding seasons initially located by the “tk…grrrrrt” sound produced (September-January) between October 1998 and January during display flights, which was often audible from up to 2001 at Aguará Ñu (24°10’S, 55°17’W), a c. 5500-ha c. 200 m. Surveys were mainly conducted on dry, relatively area of open-country habitats within Mbaracayú Forest still, moonlit nights, when male display activity was at its Nature Reserve, Canindeyú department, eastern Paraguay. highest (Clay et al. 2000). Once a displaying male was Aguará Ñu is a low plateau, 170-270 m above sea level, detected, it was observed for 10-15 minutes to identify its which forms a natural island of cerrado vegetation, primary display arena, the precise location of which was flanked by forest to the south and west, and isolated from recorded using a Magellan 2000 XL GPS receiver. It was Paraguay’s main areas of cerrado in the departments of then captured and fitted with an individually numbered Concepción and Amambay to the north-west (Jiménez metal tarsus ring, a coloured plastic ring, and a small, & Knapp 1998). uniquely shaped piece of reflective tape, affixed to the Most fieldwork was focused on an area of c. 400 ha in dorsal side of one of the non-central tail feathers. This southern Aguará Ñu, where a series of shallow valleys drain tail mark was unobtrusive when birds were at rest during southwards into the Arroyo Guyrá Kehá. The seasonally the day, but conspicuous when illuminated with a torch wet grasslands of the valley bottoms and lower slopes at night. Territories were repeatedly revisited to study are dominated by grass and sedge species, and woody male display behaviour. During these visits, the identity vegetation is almost entirely absent. The drier soils of the of the male and the location of the display arena(s) were upper slopes and ridge-tops support “campo cerrado” confirmed. vegetation, consisting primarily of grasses and herbs, but To investigate aggregative tendencies in displaying with a scattering of shrubs (e.g. Cochlospermum regium, males, nearest-neighbour analysis (Clark & Evans 1954, Caryocar brasiliense), Yata’i palms (Butia paraguayensis), Krebs 1989) was used to quantify the spatial distribution saplings and the occasional fully grown tree. Palm density of territories (see also Höglund & Stöhr 1997). The is lowest at the tips of the ridges in the south, where the distance from each primary arena to its nearest neighbour campo cerrado is relatively open. Termite mounds, 15-90 (i.e. the measure used herein to represent distances cm in height, are present throughout the campo cerrado between territories) was derived from GPS locations, and and wet grassland, and low anthills occur occasionally in the average nearest-neighbour distance (NND) within the campo cerrado. each season was calculated. Estimates of expected NND I identified breeding seasons by their principal year, were then obtained (Clark & Evans 1954), assuming a so the season from September 2000 to January 2001 is random distribution of males within the survey area. termed the “2000 season”. Adult White-winged Nightjars The latter was defined by a convex polygon (of c. 400 were captured, sexed and ringed, and lightweight (1.3- ha) encompassing the zone surveyed most thoroughly 2.5 g; <5% of adult body weight) radio-transmitters were Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople for territories (see Results). Since the small sample sizes m away. Nautical twilight (the time at which the sun is in the current study prevented the use of a boundary 12° below the horizon) was used as the point delineating strip, estimates of expected NND were corrected using dusk or dawn from true “night” (Brigham & Barclay Donnelly’s (1978) adjustment for edge-effect bias. An 1992). Until this point, lunar illumination is exceeded index of aggregation (R) was obtained for each season by by residual solar light (Austin et al. 1976). Seventy-one dividing the mean observed NND by the mean expected per cent (64 of 90) of focal watches were begun before NND. When R = 1.0 the observed spatial distribution the end of nautical twilight, and thus termed “dusk does not deviate from random, whereas values of R watches”. The remaining 26 watches were conducted between 1 and 0 indicate an increasing degree of clustering under truly nocturnal conditions and termed “night (Krebs 1989). watches”. The gross topographical characteristics of male During dusk watches, the observer arrived at the territories and an equal number of random points were observation point shortly after sunset, and the watch measured using a digital version of a 1:50000 Paraguayan commenced when the focal male arrived at the display Instituto Geográfico Militar map, imported as a layer into arena. The number and time of all displays were then a geographical information system (GIS) in ArcView 3.2 recorded, as well as details of any interactions with other (ESRI 1999). The following variables were recorded for individuals. A “display” was defined as any flight up from each site: (i) gradient, to nearest 1°; (ii) altitude, to nearest the display mound accompanied by both the “tk” and 5 m above sea level; (iii) altitude difference, from nearest “grrrrrt” sounds. Following the completion of the watch, “spot height” (peak local altitude); (iv) distance to spot the identity of the focal male was confirmed based on its height, to nearest 25 m; (v) distance to stream, to nearest 25 reflective tail mark and/or colour rings. m; (vi) aspect, allocated to one of four quadrants (north- Potential environmental correlates of male display east, south-east, south-west and north-west). rate were investigated using a multiple regression model, Qualitative observations suggested that males with the following variables: (i) time, in minutes, in did not select display arenas at random, even within relation to nautical twilight; (ii) temperature, minimum territories. To determine how mounds used as display recorded during the watch; (iii) moon phase, proportion arenas differed from available mounds, the structural of the moon face illuminated (MFI); (iv) moon height, characteristics of nine display arenas used during the estimated degrees above the horizon; (v) moonlight, 2000 season were measured and compared to those from estimated lunar illuminance (see below). Ambient a sample of random mounds. The arenas included seven temperature was measured every 15 minutes with an primary arenas and two (regularly used) secondary arenas electronic logger. Details of astronomical phenomena of seven males. For each mound I measured: (i) maximum (i.e. MFI, times of sunrise/sunset, moonrise/moonset and height (including any “towers”), to nearest 5 cm; (ii) modal twilights) were obtained from astronomical almanacs. height, i.e. height of the majority of the mound (excluding Moon phase and height significantly influence overall any “towers”), to nearest 5 cm; (iii) basal “circumference”, lunar illuminance (Austin et al. 1976), but neither factor to nearest 0.1 m; (iv) maximum basal chord, across the shows a simple linear relationship with illuminance longest axis of the mound, to nearest 0.1 m; (v) presence/ (R. Willstrop in litt.), so their combined effects were absence of “towers”, i.e. free-standing vertical structures estimated, using information in Austin et al. (1976), and projecting above the modal level of the mound (see treated as a separate measure: “moonlight”. Figure 3a). For each display arena, four random mounds Tape recordings of male display sounds and other were selected by walking 20 m from the display arena on “vocalisations” were made at the display arenas of marked each of the four cardinal points of the compass, and then males, from distances of c. 5-10 m, during the 1999 and identifying the nearest mound. On two occasions it was 2000 breeding seasons. Recordings were later digitised not possible to locate a mound, hence the total number and spectrograms of male display sounds generated in of random mounds measured was 34. Avisoft using the following settings: FFT-length = 128; frame size = 100%; window = Blackman; overlap = Male display behaviour 75%. These settings provided sufficiently fine temporal resolution (8 ms) to distinguish the separate elements To collect quantitative and qualitative data on male of the “grrrrrt” sound. The following temporal measures display behaviour, 30-minute focal watches were were obtained using the on-screen cursors (Figure 1): conducted on dry, relatively still evenings at the primary interval between “tk” and “grrrrrt” components (INT); display arenas of territory-holding males. Focal males length of “grrrrrt” component (GTLEN); number of were monitored from c. 10-20 m away with the aid of elements constituting “grrrrrt” component (GTNUM). a torch set on diffuse beam. Individuals did not appear Frequency measures were not taken, however, given the to be disturbed by the presence of an observer, and on relatively poor resolution (31 Hz) and limited frequency other occasions continued to display when I was just 5 range of display sounds. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople Hatching success was calculated as the percentage of eggs laid that hatched successfully. This method can bias values for hatching success upwards, as it takes no account of nests failing before they are found (Beintema INT GTLEN/GTNUM & Müskens 1987). However, my method of nest location reduced this risk, and the sample size was insufficient to use the daily exposure method (Mayfield 1961, 1975). Nests were visited every two to four days after hatching, and chicks were weighed (to nearest 0.5 g) to obtain data on their growth and development. Chicks were ringed FIGURE 1. Spectrogram of male White-winged Nightjar display at c. 10 days old, when their tarsus width was similar to sound showing “tk” and “grrrrrt” components and the three temporal variables measured for each recording (see text for definitions). adult birds. Younger chicks were individually marked with non-toxic white correction fluid on their feet. During the nestling period, nocturnal focal watches Nesting biology of 50-255 minutes were conducted at nests to characterise the division of chick brooding and provisioning duties Nest sites were located by monitoring radio-tagged by the parents. Nest sites were monitored from distances females and checking their daytime “roosts” for eggs every of c. 10-20 m, using a head-torch with diffuse beam to c. 2 days during the breeding season. The “initiation date” obtain eye-shine from adults without disturbing them. for a nesting attempt was defined as the date on which The duration (to nearest five seconds) of each visit was the first egg was laid: presence of eggs was the only factor recorded, along with the sex of the visiting adult. Radio- reliably distinguishing nest sites from roost sites. Once a tracking equipment was occasionally used to confirm the nest was initiated, its precise location was recorded using identity of tagged birds, but monitoring was primarily a GPS receiver. Gross topographical characteristics of nest visual. The data collected were used to calculate mean sites and an equal number of random sites were measured visit rates and average times spent at, or away from, as described above for male territories. To ensure that the nest site. A small number of daytime spot-checks random sites reflected those available to, but not used by, were conducted to assess parental allocation of diurnal nesting females, they were selected to fall within 400 m brooding duties. of the nest sites with which they were paired. Two radio- tagged females monitored during the 2000 breeding Statistical analysis season ranged up to 403 ± 33 m from the centre of their breeding ranges (pers. obs.). All data were tested for normality and homoscedasticity, The following variables were recorded for each site: and then analysed using standard parametric or non- (i) gradient, to nearest 1°; (ii) altitude, to nearest 5 m parametric univariate tests, as appropriate. If appropriate, above sea level; (iii) distance to nearest stream, to nearest 25 data were subsequently also analysed using multivariate m; (iv) distance to nearest male display arena, to nearest 10 techniques. When the dependent variable was continuous m; (v) aspect, allocated to one of four quadrants (north- (e.g. male display rate), multiple linear regression was east, south-east, south-west and north-west); (vi) habitat, used to investigate the effects of potential covariates. classified as one of two dominant types (campo cerrado Analyses were conducted using a forward stepwise and wet grassland). Habitat type was determined from a procedure with entry and removal probabilities set at P = simplified habitat map of the study site, added as a layer 0.05. At each step, the variable with the most significant to the GIS (see Pople 2003 for details). score statistic was entered into the model, provided that The mass of freshly laid eggs was measured to the its inclusion significantly improved model fit: significance nearest 0.5 g. Measures of length and maximum breadth was tested using partial F-tests (Hair et al. 1995). This of eggs were taken to the nearest 0.1 mm. Egg volume process was repeated until no further variables met the was estimated following Hoyt (1979), using the equation: criteria for entry or removal, at which point the model volume = 0.51 × length × (maximum breadth) . During was considered final. the egg stage, nests were visited every three to five days When the dependent variable could be allocated to confirm the sex of any adults present. These “spot- to one of two discrete categories (e.g. male territory checks” were carried out at various times of day and or random site, display mound or random mound), night, independent of ongoing radio-telemetry fixes binary logistic regression was used to identify which taken to characterise the home range and movements combination of variables best distinguished between of individuals. Although radio-tracking equipment was the two categories. Analyses were conducted using a sometimes used to confirm an adult’s identity, the bird’s forward stepwise procedure as outlined above, but with location was not known prior to any spot-check. significance tested using likelihood ratio tests in which Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople the difference in deviance (-2 × log likelihood) between arenas were significantly clustered within the survey area models with and without the variable was treated as in all three seasons (Table 2). ², with degrees of freedom (df ) equal to the number In total, 13 distinct territories were identified of parameters being added (Manly et al. 1993). When between 1998 and 2000: five of these were occupied in all comparing display mounds and random mounds, the three breeding seasons, and a further three were occupied inclusion of secondary mounds for two males raised in at least two. The ownership of certain territories the possibility of pseudoreplication. To address this, changed both between and, less frequently, within parameter estimates produced by the model were jack- seasons, probably owing to the death or emigration of knifed following Sokal & Rohlf (1995), taking the male the original territory-holding male. In nine of 11 such as the sampling unit. Jack-knifed parameter estimates changes in territory ownership, the original male was did not differ significantly from those of the original never recorded again; in two instances the male moved to model, suggesting that pseudoreplication was not a major a territory vacated by another “missing” male. problem. Male territories were located almost exclusively on Categorical data were analysed using the G-test of the upper slopes of ridgelines running perpendicular to independence. A form of the G-test for goodness of fit, the main Arroyo Guyrá Kehá valley (running north-west adjusted using the Williams correction for the two-cell to south-east across the lower half of the maps in Figure 2). case (Sokal & Rohlf 1995), was used to compare the Preliminary analysis of gross topographical characteristics observed sex ratio of adult captures with that expected using univariate tests suggested that territories differed on the basis of a 1:1 sex ratio. Statistical significance was significantly from random points for only one variable set at P < 0.05 unless otherwise specified, and means are (Table 3). A binary logistic regression model, explaining given ± one standard error (SE). Since the repeated use of c. 27% of the deviance in site use, also revealed that a statistical test increases the probability of committing a Type I error, the sequential Bonferroni technique was employed to adjust the initial a level when making many simultaneous comparisons (Rice 1989). RESULTS Forty-nine White-winged Nightjars (34 adult and 15 young birds; Table 1) were captured and ringed between 1998 and 2001. The sex ratio of captured adults was biased towards males in all three field seasons, and the overall ratio of 3.25:1 male:female differed significantly from parity. The four females captured in 1998 were all caught late in the breeding season, and none of them was confirmed to (re-)nest that season; two individuals captured in late December had already begun post- nuptial moult. Hence, all data on nesting biology were obtained from females caught during the 1999 and 2000 breeding seasons. Male territories Breeding-season territories were identified for 10 males in 1998 and eight males in each of 1999 and 2000. One territory, discovered late in the 2000 breeding season, was excluded from the nearest-neighbour analysis as it was located outside the normal survey area (Figure 2). The remaining territory-holding males occurred at a mean density of one male per 50 ± 5 ha (n = 3 seasons) within the area surveyed. The closest male display arenas FIGURE 2. Location of male White-winged Nightjar territories were just 70 m apart (in 1998), but the average nearest- (closed circles) during: (a) 1998; (b) 1999; and (c) 2000 breeding neighbour distance across all three seasons was 254 ± 8 seasons. Pale brown lines are contours, dashed lines are watercourses. m. Aggregation indices confirmed that primary display The polygon represents the c. 400-ha survey area. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople “distance to stream” was the only significant predictor deviance in mound use, showed “modal height” to be (likelihood ratio test: ² = 9.89, df = 1, P < 0.005): on the only significant predictor (likelihood ratio test: ² = average, male territories were further from streams than 11.01, df = 1, P < 0.001): on average, display mounds random points. The model correctly reclassified 69% of were lower than random mounds. The model correctly sites overall, with an equal degree of classification success reclassified 91% of random mounds, but only 22% of for territories and random sites. Aspect was not entered mounds used as display arenas, resulting in an overall into the model, but univariate tests suggested that male correct reclassification rate of 77%. territories were more likely to have a westerly than easterly aspect compared to random points (11 of 13 versus 5 of Male display behaviour 13; G-test: G = 5.80, df = 1, P < 0.05). At a finer scale, preliminary analysis of the structural In both 1999 and 2000, sporadic display activity was characteristics of mounds used as display arenas in 2000 noted on certain nights leading up to the August full suggested that they differed significantly from random moon, but male display activity did not commence in mounds for a number of variables (Table 4). However, a earnest until September. Activity levels remained high binary logistic regression model, explaining 25% of the through to December, but subsided soon after: the latest date on which display activity was noted was 6 January. Although an attenuated version of the mechanical “grrrrrt” sound was occasionally heard at the start of foraging sallies, the full “tk… grrrrrt” display sound was only ever noted during male display flights at arenas. Of 132 display flights by five males in 2000, the majority (63%) were initiated from a nearby perch, but 23% immediately followed the return of the male from a foraging sally, and the remaining 14% involved males already sitting on their display arena. In these latter cases, the male was observed to jump up and down on the display arena to produce the “tk” (probably by contact with the mound), before flying up with the “grrrrrt” sound as per usual. Display activity was not solely restricted to males in “definitive” plumage. A young male moulting into definitive plumage, but still retaining five or six (brown) juvenile outer primaries, was observed to display on two nights in early November 1998 (see below). Recordings of 8-19 display sounds were obtained for eight marked males during the 1999 and 2000 breeding seasons. Spectrograms showed that display sounds had a relatively fixed structure, with the “tk” and “grrrrrt” components separated by 500 to 700 ms, but with broadly similar frequency ranges (c. 100-400 Hz; Figure 1). The display sounds produced by the pre-definitive male in 1998 were of a noticeably higher frequency, but a recording of its display was not obtained for direct comparison. Two of the three temporal measures (GTLEN and GTNUM) were highly correlated (Spearman rank correlation: r = 0.96; n = 103; P < 0.001). Only GTNUM was considered further, as it was judged to be less susceptible to measurement error. Males showed significantly more inter- than intra-individual variation for both INT (Kruskal-Wallis test: H = 57.82, df = 7, P < 0.001) and GTNUM (H = 60.40, df = 7, P < 0.001), suggesting consistent inter-individual differences in these temporal measures. A total of 90 focal watches was conducted at male display arenas over the course of three breeding seasons FIGURE 3. Typical male display arena with “towers” (a), eggs (b) and chicks aged 14 days old (c) of White-winged Nightjar at Aguará Ñu. (1998: n = 10; 1999: n = 20; 2000: n = 60). No significant Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople TABLE 1. Sex and age composition of White-winged Nightjars captured. Season Males Females Young birds Total Sex ratio (males/female) G P 1998 20 4 3 27 5.00 11.41 < 0.001 1999 14 3 5 22 4.67 7.50 < 0.01 2000 9 3 9 21 3.00 3.01 ns Overall 26 8 15 49 3.25 9.89 < 0.005 (a) Column totals do not equal “Overall” total due to recapture of certain individuals between seasons. TABLE 2. Mean ± SE nearest-neighbour distances (NND) and degree of aggregation for breeding-season territories of male White-winged Nightjars. Season No. male Observed Expected Aggregation index Significance territories NND (m) NND (m) (R) 1998 10 253 ± 97 372 ± 69 0.68 z = -1.71, P = 0.044 1999 8 241 ± 35 425 ± 89 0.57 z = -2.07, P = 0.019 2000 7 268 ± 34 460 ± 103 0.58 z = -1.87, P = 0.031 (a) Assuming random distribution of males within 406 ha survey area (see Methods). (b) P-values from one-tailed z-tests (Campbell 1996). TABLE 3. Mean ± SE values of topographical variables for male White-winged Nightjar breeding-season territories and an equal number of random points. Variable Male territories Random points Significance (n = 13) (n = 13) Gradient (°) 2.2 ± 0.6 4.5 ± 0.7 W = 220, P = 0.023 Altitude (m) 232 ± 4 214 ± 6 t = -2.51, P = 0.019 Altitude difference (m) 11 ± 3 29 ± 7 W = 220, P = 0.023 Distance to spot height (m) 406 ± 77 683 ± 110 t = 2.06, P = 0.050 Distance to stream (m) 567 ± 74 310 ± 46 W = 122, P = 0.007 Aspect – number of sites facing: NE (0-90°) 0 (0%) 0 (0%) SE (90-180°) 2 (15%) 8 (62%) SW (180-270°) 6 (46%) 2 (15%) NW (270-360°) 5 (38%) 3 (23%) (a) From unpaired t-tests or Mann-Whitney U-tests. Values underlined indicate table-wide significance at an adjusted initial a level of 0.01 (Rice 1989). TABLE 4. Mean ± SE values of the structural characteristics of male White-winged Nightjar display arenas used during 2000 breeding season compared to a sample of random mounds. Variable Display arenas Random mounds Significance (n = 9) (n = 34) Maximum height (m) 0.42 ± 0.10 0.61 ± 0.04 W = 831, P = 0.014 Modal height (m) 0.24 ± 0.03 0.51 ± 0.05 W = 842, P = 0.006 Basal “circumference” (m) 6.0 ± 0.8 3.7 ± 0.5 W = 665, P = 0.013 Maximum basal chord (m) 2.1 ± 0.3 1.1 ± 0.1 W = 654, P = 0.005 “Tower(s)”: present 3 (33%) 8 (24%) G = 0.32, ns absent 6 (67%) 26 (76%) (a) From Mann-Whitney U-tests or G-test. Values underlined indicate table-wide significance at an adjusted initial a level of 0.01 (Rice 1989). Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople differences in display rate were apparent between seasons was subsequently recorded on several occasions during (Kruskal-Wallis test: H = 1.18, df = 2, P > 0.05), so data display watches (see Figure 5 for spectrogram), and were pooled for subsequent analyses. On average, males was only observed from males perched in or near their performed 0.54 ± 0.04 displays per minute during focal display arena. The thin, high-pitched nature of the call watches, although considerable variation in display rate made it difficult to detect under certain conditions, but was apparent (range: 0.00-1.37 displays/min; n = 90). it was apparently produced at rates of 1.24 ± 0.20 “tik However, when comparing the six males for which five tiks” per minute (range: 0.00-3.33 calls/min; data from or more focal watches were conducted, no evidence was 26 display watches). There was no significant correlation found for consistent inter-male differences in display rate between the rate of “tik tik” production and male display (H = 5.95, df = 5, P > 0.05). rate during display watches (Pearson correlation: r = 0.31, On average, males arrived at their display arenas 26.6 n = 26, P > 0.05). ± 0.6 minutes after sunset and 27.9 ± 0.7 minutes before Aggressive interactions between males were the end of nautical twilight. On arrival, males showed a observed during 14 focal watches (30% of watches for burst of high display activity. The mean display rate in the which details were noted) and on various occasions first half of dusk watches was more than twice that in the during ad hoc observations at display arenas. Interactions second half (0.75 ± 0.06 versus 0.35 ± 0.04 displays/min; typically comprised rapid chases, sometimes involving Wilcoxon signed ranks test: T = 1742, n = 64, P < 0.001), up to three males, and occasionally also physical contact. whereas there was no significant difference between the A soft, liquid “gurgling” call produced during chases was two halves of night watches (0.50 ± 0.07 versus 0.54 ± noted on a number of occasions. Aggressive interactions 0.06 displays/min; paired t-test: t = -1.01, P > 0.05). were observed most frequently at one particular display When considering night watches alone, “moonlight” arena (11 of 14 watches), and this bias was significant was the only environmental variable that explained a when allowing for the distribution of watches among significant amount of the variation in display rate (overall display arenas (G-test: G = 7.64, df = 1, P < 0.01). There fit of model: R = 0.166, F = 4.77, P < 0.05). Display was a non-significant trend for higher rates of display 1,24 rate was positively related with this estimate of overall activity during watches that included male–male chases lunar illuminance, reflecting a tendency for males to compared to during those that did not (0.74 ± 0.10 display more when the moon was fuller and higher in versus 0.53 ± 0.06 displays/min; unpaired t-test: t = the sky (Figure 4). However, even when conditions were -1.76, P = 0.085). not suitable for display activity, males were generally still It was not possible to record female visits to present – either sallying or resting – on their territories. male display arenas reliably, owing to difficulties in There was no significant relationship between display rate distinguishing unmarked females from juveniles or from and sallying rate during focal watches in which the latter females of other similar-sized species (e.g. Little Nightjar was measured (Spearman rank correlation: r = -0.26; n = Setopagis parvulus). However, during one dusk watch in 36; P > 0.05). October 1998, five nights prior to full moon, a female During one display watch in November 1998, the visit and presumed copulation were documented. Over focal male produced a previously undocumented, insect- the course of c. 20 minutes, the female landed on the like “tik tik” call whilst perched in the vicinity of its display display arena at least twice, and three presumed copulation arena. This vocalisation (or a single “tik” equivalent) attempts by the focal male were witnessed. Prior to one 1,4 1,2 1,0 0,8 0,6 0,4 0,2 0,0 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 'Moonlight' FIGURE 5. Spectrogram of previously undescribed, insect-like “tik tik” call given by male White-winged Nightjars when inactive FIGURE 4. Male White-winged Nightjar display rate in relation to in vicinity of display arenas. Note background insect noise at c. 3.5 “moonlight”, an estimate of lunar illuminance. Data from “night” and 5.0 kHz. Recordings made by M. C. Velázquez at Aguará Ñu in watches only (Pearson correlation: r = 0.41, n = 26, P < 0.05). December 2000. Revista Brasileira de Ornitologia, 22(2), 2014 Male display rate (displays/min) Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople attempt, the male appeared to display to the female whilst Nesting biology they were both perched on the arena, by spreading its wings and cocking and splaying its tail. When the female I documented nine breeding attempts by four different left the display arena after copulation, the male followed females : five in 1999 and four in 2000 (Table 5). A fifth and flew parallel to her, with wings raised, in a slow female captured in early December 2000 was heavily buoyant flight, conspicuously different to the rapid flight gravid, but not subsequently relocated. Eggs were laid typical of male–male encounters. On another occasion, directly on the ground, in campo cerrado (six nests) or wet in December 2000, a male was observed to conduct a grassland (three nests) vegetation, within a small “tunnel” similar “escort” flight, with raised wings, for a presumed between tussocks of grass or herbaceous plants and facing female flying through his territory. onto a small area of bare earth (c. 25-50 cm in diameter), FIGURE 6. Distribution of White-winged Nightjar nest sites in relation to male display arenas during: a) 1999 and b) 2000 breeding seasons. Nest sites displayed as filled circles; multiple nesting attempts distinguished by “A”, “B”, etc. Display arenas displayed as open circles; labels refer to the territory-holding male. Note the territorial fidelity in successive seasons shown by males M001, M052, M055, and M057. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople the latter apparently serving as a “runway” for the adult asymmetry in growth rate was observed in two broods: bird (as at roost sites; pers. obs.). There was no evidence in one (circular symbols in Figure 7), the larger chick to suggest that the gross topographical characteristics of (26.3 g) weighed more than twice its sibling (10.9 g) by nest sites differed significantly from those of random sites day 10 (see also Figure 3c). The latter had disappeared available to nesting females (Table 6). by day 15. Initiation date was confirmed for seven clutches, Nocturnal focal watches conducted at six nests (of and consistently fell within two days of a full moon four females) showed that chick brooding and provisioning (1.3 ± 0.3 days; range: 0-2). The earliest initiation date was undertaken solely by the female. In almost 25 hours observed was 25 September (in 1999), although two of monitoring (mean watch length: 113.6 ± 17.3 min, n chicks discovered on 13 October 2000 were estimated to = 13), no male visits were observed, whereas 67 female be approximately 10 days old, and had probably hatched visits were recorded, at a mean rate of 3.04 ± 0.80 visits/ from eggs laid around the full moon of 14 September. hour (range: 1.20-4.94; n = 4 females). Females spent Females made two nesting attempts per season, using a on average 7.7 ± 1.3 minutes (range: 0.1-46.0 min; n = different site for each attempt (mean distance between 58) at the nest during each visit, and 11.8 ± 1.3 minutes nest sites: 360 ± 31 m, range: 296-461 m, n = 4; Figure (range: 1.5-59.9 min; n = 55) away from the nest between 6). The female (F067) captured in late November 1998, visits. It was not possible to confirm whether chicks whilst brooding a single chick, was probably making a were provisioned on every visit. Time spent at the nest second breeding attempt (Table 5). during nocturnal focal watches was negatively correlated Clutch size was two for all seven clutches discovered with minimum ambient temperature (Spearman rank at the egg stage. It was not possible to check each nest correlation: r = -0.81, n = 13, P < 0.005). In 100% (10 daily, but the second egg was generally laid within of 10) of daytime checks during the chick period, the 24 hours of the first, although in one case there was a female was present at the nest site, either brooding or delay of at least 30 hours. Eggs were pale cream-beige, immediately adjacent to the chicks. with variable dark brown and mauve-grey speckling and Females displayed a range of anti-predator blotching (Figure 3b). Differences in egg patterning were behaviours whilst incubating or brooding. When initially observed within most clutches, with one egg being more approached, they relied on their cryptic plumage, densely and uniformly covered in fine markings, and crouching motionless with eyes closed to slits. If the the other showing fewer, larger markings, often with a intruder approached to within 0.5-2.0 m, they would subapical ring of grey blotches at the obtuse pole. Egg shuffle forward into the adjacent clearing in preparation morphometrics are summarised in Table 7. Both eggs for take-off. On one occasion, a female incubating a of a clutch weighed on day 1 and day 18 of incubation single, recently laid egg performed a presumed threat decreased in mass by 0.7 g (11% and 12%) during the display: drooping the wings slightly, puffing up the body intervening period. Incubation began with the laying of the first egg, and was carried out exclusively by the female: no male 45 was ever found at the nest site. Thirty-four spot-checks carried out at seven nests during the incubation period found the female in attendance during 100% (20 of 20) of daytime and 57% (8 of 14) of nocturnal checks. During nine daytime checks, the female was present but sitting slightly behind one (n = 3) or both (n = 6) of the eggs. There was no significant difference in ambient air temperature on these occasions compared to 11 occasions when the female was incubating both eggs (unpaired t-test: 23.4 ± 1.8°C versus 23.7 ± 1.3°C; t = -0.13; P > 0.05). Although I could not confirm the hatching date for every clutch, the modal incubation period was 19 days, and chicks probably hatched within 24 hours of 0 5 10 15 each other. Age of chick (days) Hatching success was 86% (12 of 14 eggs) for the seven nests for which initial clutch size was known. At FIGURE 7. Growth rate of eight White-winged Nightjar chicks hatching, chicks had open eyes and were covered in a between hatching and 15 days of age. Each brood represented by different shape; open and closed symbols represent two chicks within pale buff-coloured down, with greyish legs and pale- same brood. Four chicks weighed on day 0 had masses of c. 5 g (square tipped greyish bill. The average mass of four chicks (from and diamond symbols). two clutches) on day of hatching was 5.2 ± 0.1 g. Clear Revista Brasileira de Ornitologia, 22(2), 2014 Mass of chick (grams) Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople feathers, raising and slightly splaying the tail, whilst pattern. After about 10 days, the developing flight feathers rocking the body and tail from side-to-side. When began to project from their feather sheaths. Although I eventually flushed (off eggs or chicks), females typically could not confirm the exact age of fledging, extrapolation performed an injury-feigning distraction display, flying of observed growth rates suggests that chicks attain mean 5-10 m away from the nest with a low, laboured flight, adult mass (c. 50 g) at c. 19 days. The sole surviving chick before landing clumsily in the vegetation. If approached of one clutch present on day 18 could not be relocated again, the female would fly off normally, although on on day 20. most occasions she was subsequently found to have Owing to uncertainty over the fate of most broods, remained near the nest site. I could not estimate fledging success or overall nesting At one week old, chicks were already beginning to success. I recorded two instances of predation (once of lose their natal down and attain the juvenile plumage eggs, once of chicks; Table 5), but could not identify TABLE 5. Summary of White-winged Nightjar nesting attempts during 1999 and 2000 breeding seasons. First nesting attempt Second nesting attempt Season Female Date first Incubation Hatching Chicks last seen Date first Incubation Hatching Chicks last seen egg laid period date at nest site egg laid period date at nest site a b † 1999 F063 26 September – – – 25 October 19 days 13 November 1 December c † d 1999 F066 25 September 19 days 14 October 25 October 21 November 19 days 10 December 14 December e † 1999 F067 – – – – – – – 10 December f g † † 2000 F010 – – – 18 October 12 November 19 days 1 December 14 December h † 2000 F066 11 October 17 days 28 October – 13 November 19 days 2 December 17 December (a) Eggs taken by unknown predator within one week of laying. (b) Smaller of two chicks disappeared during afternoon of 28 November; larger chick no longer present by 3 December. (c) Nest not checked for five days; no sign of chicks on 31 October. (d) Both chicks still present – no further fieldwork in 1999. (e) Single chick discovered on 23 November; estimated by mass and stage of feather development as approximately 10 days old. (f ) Two chicks discovered on 13 October; estimated by mass and stage of feather development as approximately 10 days old. (g) Female and chicks found to have moved to new site by 20 October; not seen subsequently. (h) Chicks disappeared, presumably taken by a predator, within one week of hatching. (†) Denotes chick(s) believed to have fledged successfully. TABLE 6. Mean ± SE values of topographical variables for White-winged Nightjar nest sites and an equal number of random sites. Nest sites Random sites Variable Significance (n = 9) (n = 9) Gradient (°) 5.6 ± 1.2 4.8 ± 0.9 t = -0.52, ns Altitude (m) 218 ± 5 217 ± 5 t = -0.07, ns Distance to nearest stream (m) 306 ± 61 319 ± 54 t = 0.17, ns Distance to nearest male display arena (m) 227 ± 49 208 ± 56 t = -0.25, ns Aspect – number of sites facing: NE (0-90°) 1 (11%) 0 (0%) SE (90-180°) 1 (11%) 1 (11%) SW (180-270°) 5 (56%) 6 (67%) NW (270-360°) 2 (22%) 2 (22%) Habitat – number of sites in: Campo cerrado 6 (67%) 8 (89%) Wet grassland 3 (33%) 1 (11%) (a) From unpaired t-tests. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople the predator on either occasion. One female and her nightjar species (e.g. Jackson 1985, Mills 1986, Vilella chicks were found to have moved to a new clearing, c. 1995, Perrins & Crick 1996, Pichorim 2002), and authors 10 m from the original nest site, when the chicks were typically emphasise the advantages of moonlight for about 10 days old. Assumptions about the fate of other foraging during the first two weeks of the nestling period. broods, based solely on the disappearance of chicks from The modal incubation period of 19 days resulted in chicks the original nest site, therefore seem inappropriate. Of hatching between the new moon and first quarter, and 11 chicks ringed at the nest site, I did not recapture any hence during a period of increasing moonlight. However, (post-fledging) before the end of my fieldwork in June with male investment in reproduction apparently limited 2001. However, one chick ringed in October 2000 was to genes, it might also be argued that females use nights subsequently recaptured as a yearling male in December with high moonlight to assess the quality of males before 2001 (J. Mazar Barnett in. litt.). securing copulations (although these two considerations are not incompatible). Male White-winged Nightjars displayed for a TABLE 7. Summary of morphometrics of 14 White-winged Nightjar eggs from seven clutches. relatively prolonged period between late August and early January, broadly coinciding with the vocal activity of other nightjar species breeding at the study site (e.g. Measure Mean ± SE Range Common Pauraque Nyctidromus albicollis, Little Nightjar, Spot-tailed Nightjar Hydropsalis maculicauda and Rufous Mass (g) 6.7 ± 0.1 5.8-7.5 Nightjar Antrostomus rufus). The considerable variation in display rate within this period was likely due more to Length (mm) 27.7 ± 0.3 26.7-29.9 environmental conditions than to consistent differences Maximum breadth (mm) 21.0 ± 0.1 19.8-21.7 between males. The usually rather short-lived high levels of activity observed when males first arrived at display Volume (cm ) 6.2 ± 0.1 5.4-7.1 arenas at dusk were perhaps analogous to the dawn peak of song activity noted for many diurnal birds. Following (a) Mass provided only for eggs weighed within two days of this initial burst, male display rate was best predicted by laying (n = 10). ambient levels of moonlight. Increases in vocal activity under conditions of increased moonlight have been noted DISCUSSION for various nightjar species (e.g. Cooper 1981, Mills 1986, Wilson & Watts 2006), so it is not surprising that During the breeding season, male White-winged Nightjars a similar relationship should exist for a species with such a strong visual component to its display. The “tik tik” call defended small territories on the upper slopes of low ridgelines in the southern sector of Aguará Ñu. The patterns produced by males while in the vicinity of their display of territory use and ownership I observed supported the arenas seems likely also to play a role in mate attraction suggestion that male territories are traditional (Clay or territory defence. et al. 2000). Territories were significantly aggregated My results support previous suggestions that the “tk” and “grrrrrt” components of male display sounds are within the survey area, with the primary display arenas of neighbouring males separated on average by distances both mechanical, rather than vocal, in origin (Clay et al. of c. 250 m. Males played no part in the parental care 2000). Careful observation indicated that the “tk” sound of eggs or chicks, and females were only rarely observed was not produced by clapping of the wings below the near display arenas, with anecdotal evidence suggesting body, as postulated by Clay et al. (2000), but probably by contact with the display mound. However, there was little that they mainly visited around full moon when assessing males and seeking copulations. Nest sites were on average doubt that the “grrrrrt” sound was produced by the wings 227 m from the nearest display arena, with no evidence during the male’s near-vertical ascent from the display to suggest that they fell within the boundaries of male mound. Although not as highly modified as those of the territories. Chick paternity analyses using the amplified Sickle-winged Nightjar Eleothreptus anomalus (which is known to produce mechanical wing sounds; Straneck fragment length polymorphism (AFLP) technique (Vos et al. 1995) suggested that the nearest territory-holding & Viñas 1994), the rigid, curved outer primaries of male was not always the father of a brood, and that one definitive male White-winged Nightjars probably play male sired the chicks of at least two different females a key role in generating the “grrrrrt” (Clay et al. 2000, during 1999 (K. Dasmahapatra in litt.). Pople 2003). The production of mechanical sounds is relatively rare in birds (Bostwick 2000), but its evolution The timing of White-winged Nightjar nesting attempts showed remarkable synchrony with the full is often associated with acrobatic, polygynous courtship moon. A relationship between nest initiation date and displays (Prum 1998), as occur in the White-winged moon phase has been demonstrated for a number of Nightjar system. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople Overall, my results indicate that the population management of its habitats. The sex-ratio bias of adult of White-winged Nightjars at Aguará Ñu employs a captures is likely to have been an artefact of the greater polygynous mating system. Although the breeding conspicuousness of males (particularly during the biology of most nightjar species remains poorly breeding season), rather than a true reflection of a skewed known, polygyny is apparently relatively uncommon population sex ratio. However the finding that the study in the Caprimulgidae, and has only been demonstrated population probably exhibits a polygynous, possibly even convincingly for two other species: the Standard-winged lekking, mating system does have potential consequences Nightjar Macrodipteryx longipennis and Pennant-winged for reproductive skew and effective population size Nightjar M. vexillarius (Fry et al. 1988, Holyoak 2001, (Sutherland 1998). Furthermore, if male display arenas are Jackson 2004). Male emancipation from parental care located at traditional sites, these areas could be of special is an important prerequisite for any form of polygyny importance to the population, with limited possibilities (Höglund & Alatalo 1995), but is unexpected in altricial for relocation if they are destroyed (Clay et al. 2000). bird species if fledging success is limited by the amount As with most species of grassland bird at the study site of food brought to the nestlings (Oring 1982). Snow & (pers. obs.), the White-winged Nightjar nests between Snow (1979) proposed three ecological attributes that September and December, on the ground amongst campo favour the evolution of male emancipation in altricial cerrado or wet grassland vegetation. Any wildfires during bird species: frugivory; the ability to regurgitate food; and the nesting period would have a major impact on breeding roofed nests (in rainy season breeders). Nightjars are well success, and extensive burns in the latter half of August known to regurgitate food for their chicks (e.g. Cramp could also disrupt nesting if vegetation did not regrow 1985, Sick 1993, Cleere 1999), and this ability probably sufficiently before the onset of breeding activity. Wildfires facilitates maximal provisioning efficiency per trip away are a regular occurrence at Aguará Ñu, particularly during from the nest. In addition, for species living in savanna the austral spring (e.g. August and September), so my habitats, such as the White-winged and Standard-winged findings add weight to recommendations elsewhere for Nightjars, the mass hatches of winged termites and ants more active fire management within the few protected that occur during summer rains (e.g. Jackson 2000, areas where this globally threatened species is known to Pinheiro et al. 2002) could be analogous to the periodic persist (e.g. Rodrigues et al. 1999, Capper et al. 2000, super-abundance of food associated with frugivory. Pople 2003). Although the population of White-winged Nightjars at Aguará Ñu appeared to fulfil all four of the criteria stipulated by Bradbury (1981) for lek-mating species, the ACKNOWLEDGEMENTS exact nature of the mating system remains uncertain. The inter-male distances I recorded far exceed those typical This work was carried out as part of a Ph.D. funded by the Biotechnology and Biological Sciences Research of so-called “classical” lekking species, but were within the range observed for other “exploded” or “dispersed” Council and supervised by Michael Brooke at the avian leks (e.g. Théry 1992, Alvarez Alonso 2000, Gray Department of Zoology, University of Cambridge, UK. et al. 2009). Moreover, the question of whether males Permission to work within the Mbaracayú Forest Nature were aggregated for sexual reasons, or solely as a result of Reserve was kindly granted by the Comité de Asuntos Científicos of the Fundación Moisés Bertoni (FMB). the patchy availability of suitable display habitat, remains unanswered. If the latter is true, and males were spaced Thanks are also due to several other FMB staff (current regularly within the available habitat (e.g. Pruett-Jones & and former), particularly Alberto Yanosky, René Palacios, Pruett-Jones 1982), the study population might best be Claudia Mercolli, Tito Fernández, Myriam Velázquez (for described as exhibiting a “landmark” mating system. This help with fieldwork and sound-recordings) and Rafael González (for assistance with male display watches). term has traditionally been reserved for insect species in which males aggregate on hilltops or in forest clearings Additional help with fieldwork was also kindly provided (e.g. Alcock 1981), but is applicable to any taxon using a by Alejandro Bodrati, Osvaldo Carrillo, Hugo del Castillo, specialised habitat to provide encounter sites for mating Rob Clay, Nigel Cleere, Alberto Esquivel, Estela Esquivel, (Höglund & Alatalo 1995). Although the evolutionary Lies Hiers, Mario Ledesma, Arne Lesterhuis, Alfredo Stroessner, and Sergio Villanueva. In the UK, particular causes of aggregation in landmark species differ from those for classical lekking species, females of the former thanks are due to Kanchon Dasmahapatra (for chick do also visit male aggregations primarily for the purpose paternity analyses) and Roderick Willstrop (for advice of mating, and hence the phenomenon still fits the broad on astronomical phenomena). Mike Brooke, Ian Burfield definition of lekking proposed by Höglund & Alatalo and Andy Radford provided valuable feedback on an early version of the manuscript. Rob Clay and Nigel Collar (1995). My results have various implications for the provided constructive input and encouragement during conservation of the White-winged Nightjar and the preparation of the final manuscript, which was further Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople Fry, C. H.; Keith, S. & Urban, E. K. 1988. The birds of Africa. Vol. improved by the review comments of Mark Brigham III. London: Academic Press. and Des Jackson. Finally, special mention must be made Gray, T. N. E.; Hong, C.; Collar, N. J. & Dolman, P. M. 2009. of the late Juan Mazar Barnett, whose discovery of the Sex-specific habitat use by a lekking bustard: conservation population of White-winged Nightjar at Aguará Ñu in implications for the Critically Endangered Bengal Florican 1995 set in motion the series of events that ultimately led (Houbaropsis bengalensis) in an intensifying agroecosystem. The Auk, 126: 112-122. to my study. Hair, J. F.; Anderson, R. E.; Tatham, R. L. & Black, W. C. 1995. Multivariate data analyses with readings. Englewood Cliffs: Prentice-Hall. REFERENCES Höglund, J. & Alatalo, R. V. 1995. Leks. Princeton: Princeton University Press. Höglund, J. & Stöhr, S. 1997. A non-lekking population of Black Alcock, J. 1981. Lek territoriality in the tarantula hawk wasp Grouse Tetrao tetrix. Journal of Avian Biology, 28: 184-187. Hemipepsis ustulata (Hymenoptera: Pompilidae). Behavioral Holyoak, D. T. 2001. Nightjars and their allies: the Caprimulgiformes. Ecology and Sociobiology, 8: 309-317. Oxford: Oxford University Press. Alvarez Alonso, J. 2000. The breeding system of the Orange-crowned Hoyt, D. F. 1979. Practical methods of estimating volume and fresh Manakin. Condor, 102: 181-186. weight of bird eggs. Auk, 96: 73-77. Austin, R. H.; Phillips, B. F. & Webb, D. J. 1976. A method for Jackson, H. D. 1985. Aspects of the breeding biology of the calculating moonlight illuminance at the earth’s surface. Journal of Fierynecked Nightjar. Ostrich, 56: 263-276. Applied Ecology, 13: 741-748. Jackson, H. D. 2000. Food of the nightjars in Zimbabwe. Ostrich, Beintema, A. J. & Müskens, G. J. D. M. 1987. Nesting success of 71: 404-407. birds breeding in Dutch agricultural grasslands. Journal of Applied Jackson, H. D. 2004. Courtship displays of the Pennant-winged Ecology, 24: 743-758. Nightjar, Semeiophorus vexillarius. Ostrich, 75: 1-4. BirdLife International. 2013. Species factsheet: Eleothreptus candicans. Jiménez, B. & Knapp, S. 1998. Palms of the Reserva Natural del http://www.birdlife.org/datazone/speciesfactsheet.php?id=2390 Bosque Mbaracayú (Mbaracayú Natural Forest Reserve). Principes, (access on 3 August 2013). 42: 65-79. Bostwick, K. S. 2000. Display behaviors, mechanical sounds, Krebs, C. J. 1989. Ecological methodology. New York: HarperCollins. and evolutionary relationships of the Club-winged Manakin Lowen, J. C.; Clay, R. P.; Mazar Barnett, J.; Madroño Ñieto, A.; (Machaeropterus deliciosus). Auk, 117: 465-478. Pearman, M.; López Lanús, B.; Tobias, J. A.; Liley, D. C.; Bradbury, J. W. 1981. The evolution of leks, p. 138-169. In: Brooks, T. M.; Esquivel, E. Z. & Reid, J. M. 1997. New and Alexander, R. D. & Tinkle, D. W. (eds.). Natural selection and noteworthy observations on the Paraguayan avifauna. Bulletin of social behaviour. New York: Chiron Press. the British Ornithologists’ Club, 117: 275-293. Brigham, R. M. & Barclay, R. M. R. 1992. Lunar influence on Manly, B. F. J.; McDonald, L. L. & Thomas, D. L. 1993. Resource foraging and nesting activity of Common Poorwills (Phalaenoptilus selection by animals: statistical design and analysis for field studies. nuttallii). Auk, 109: 315-320. London: Chapman & Hall. Campbell, D. J. 1996. Aggregation and regularity: an inclusive one- Mayfield, H. F. 1961. Nesting success calculated from exposure. tailed nearest-neighbour analysis of small patchy populations. Wilson Bulletin, 73: 255-261. Oecologia, 106: 206-211. Mayfield, H. F. 1975. Suggestions for calculating nest success. Wilson Capper, D. R.; Esquivel, E. Z.; Pople, R. G.; Burfield, I. J.; Clay, Bulletin, 87: 456-466. R. P.; Kennedy, C. P. & Mazar Barnett, J. 2000. Surveys and Mills, A. M. 1986. The influence of moonlight on the behavior of recommendations for the management of Aguará Ñu in the Reserva goatsuckers (Caprimulgidae). Auk, 103: 370-378. Natural del Bosque Mbaracayú, eastern Paraguay. Unpublished Murton, R. K. & Westwood, J. 1977. Avian breeding cycles. Oxford: report. Clarendon Press. Clark, P. J. & Evans, F. C. 1954. Distance to nearest neighbor as Oring, L. W. 1982. Avian mating systems, p. 1-92. In: Farner, D. a measure of spatial relationships in populations. Ecology, 35: S.; King, J. R. & Parkes, K. C. (eds.). Avian biology. New York: 445-453. Academic Press. Clay, R. P.; López Lanús, B.; Tobias, J. A.; Lowen, J. C. & Mazar Perrins, C. M. & Crick, H. Q. P. 1996. Influence of lunar cycle on Barnett, J. 2000. The display of the White-winged Nightjar. laying dates of European Nightjars (Caprimulgus europaeus). Auk, Journal of Field Ornithology, 71: 619-626. 113: 705-708. Clay, R. P.; Mazar Barnett, J. & Esquivel, E. Z. 2014. First Pichorim, M. 2002. Biologia reprodutiva do bacurau-tesoura-gigante description of the eggs and nest-site of the White-winged Nightjar (Macropsalis forcipata, Caprimulgidae) no morro Anhangava, Eleothreptus candicans (Aves: Caprimulgidae). Revista Brasileira de Paraná, Sul do Brasil. Ararajuba, 10: 149-165. Ornitologia, [vol: pp-pp]. Pinheiro, F.; Diniz, I. R.; Coelho, D. & Bandeira, M. P. S. 2002. Cleere, N. 1999. Family Caprimulgidae (Nightjars), p. 302-386. In: Seasonal pattern of insect abundance in the Brazilian cerrado. del Hoyo, J.; Elliott, A. & Sargatal, J. (eds.). Handbook of the birds Austral Ecology, 27: 132-136. of the world. Vol. 5. Barn-owls to hummingbirds. Barcelona: Lynx Pople, R. G. 2003. The ecology and conservation of the White-winged Edicions. Nightjar Caprimulgus candicans. Ph.D. thesis. Cambridge, UK: Cooper, R. J. 1981. Relative abundance of Georgia caprimulgids University of Cambridge. Available online from: http://www. based on call-counts. Wilson Bulletin, 93: 363-371. dspace.cam.ac.uk/handle/1810/225237. Cramp, S. (ed.). 1985. The birds of the western Palearctic. Vol. IV. Pruett-Jones, M. A. & Pruett-Jones, S. G. 1982. Spacing and Oxford: Oxford University Press. distribution of bowers in Macgregor’s Bowerbird (Amblyornis Donnelly, K. P. 1978. Simulations to determine the variance and macgregoriae). Behavioral Ecology and Sociobiology, 11: 25-32. edge effect of total nearest-neighbour distance, p. 91-95. In: Prum, R. O. 1998. Sexual selection and the evolution of mechanical Hodder, I. (ed.). Simulation studies in archaeology. Cambridge, sound production in manakins (Aves: Pipridae). Animal Behaviour, UK: Cambridge University Press. 55: 977-994. ESRI. 1999. ArcView 3.2 GIS software. Redlands: Environmental Ranft, R. & Cleere, N. 1998. A sound guide to nightjars and related Systems Research Institute. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople nightbirds. Robertsbridge: Pica Press and British Library National Sutherland, W. J. 1998. The importance of behavioural studies in Sound Archive. conservation biology. Animal Behaviour, 56: 801-809. Rice, W. R. 1989. Analyzing tables of statistical tests. Evolution, 43: Théry, M. 1992. The evolution of leks through female choice: 223-225. differential clustering and space utilization in six sympatric Rodrigues, F. H. G.; Hass, A.; Marini-Filho, O. J.; Guimarães, manakins. Behavioural Ecology and Sociobiology, 30: 227-237. M. M. & Bagno, M. A. 1999. A new record of White-winged Vilella, F. J. 1995. Reproductive ecology and behaviour of the Nightjar Caprimulgus candicans in Emas National Park, Goiás, Puerto Rican Nightjar Caprimulgus noctitherus. Bird Conservation Brazil. Cotinga, 11: 83-85. International, 5: 349-366. Sick, H. 1993. Birds in Brazil: a natural history. Princeton: Princeton Vos, P.; Hogers, R.; Bleeker, M.; Reijans, M.; Lee, T.; Hornes, M.; University Press. Frijters, A.; Pot, J.; Peleman, J.; Kuiper, M. & Zabeau, M. Snow, B. K. & Snow, D. W. 1979. The Ochre-bellied Flycatcher and 1995. AFLP: a new technique for DNA fingerprinting. Nucleic the evolution of lek behavior. Condor, 81: 286-292. Acids Research, 23: 4407-4414. Sokal, R. R. & Rohlf, F. J. 1995. Biometry: the principles and practice Wilson, M. D. & Watts, B. D. 2006. Effect of moonlight on detection of statistics in biological research. New York: W. H. Freeman & Co. of Whip-poor-wills: implications for long-term monitoring Straneck, R. J. & Viñas, M. J. 1994. Comentarios sobre costumbres strategies. Journal of Field Ornithology, 77: 207-211. y manifestaciones acústicas del Atajacaminos de los Pantanos, Eleothreptus anomalus (Gould, 1838) (Aves, Caprimulgidae). Notulas Faunisticas, 67: 1-4. Associate Editor: Luciano N. Naka Revista Brasileira de Ornitologia, 22(2), 2014 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ornithology Research Springer Journals

Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay

Ornithology Research , Volume 22 (2) – Jun 1, 2014

Loading next page...
 
/lp/springer-journals/breeding-biology-of-the-white-winged-nightjar-eleothreptus-candicans-8W3fhYhAQQ
Publisher
Springer Journals
Copyright
Copyright © Sociedade Brasileira de Ornitologia 2014
eISSN
2178-7875
DOI
10.1007/bf03544249
Publisher site
See Article on Publisher Site

Abstract

Revista Brasileira de Ornitologia, 22(2), 219-233 ARTICLE June 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople Department of Zoology, University of Cambridge, Downing Street, Cambridge, CB2 3EJ, UK. Email: rob.pople@cantab.net Current address: BirdLife International, Wellbrook Court, Girton Road, Cambridge, CB3 0NA, UK. Received on 03 September 2013. Accepted on 02 October 2013. ABSTRACT: Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay. I present the first detailed description of the breeding biology of the White-winged Nightjar (Eleothreptus candicans), based on data collected over three breeding seasons during 1998-2001 at Aguará Ñu, Canindeyú, eastern Paraguay. Male nightjars defended small territories situated on the upper slopes of ridgelines. Each territory contained one or more “display arenas” at which the male performed nuptial display flights. Aggregation indices confirmed that the primary display arenas of males were significantly clustered within the survey area. Within their territories, males apparently selected display arenas on the basis of their structural characteristics: mounds used as arenas were significantly lower and broader than random mounds. Males engaged in display activity from late August to early January. On average, males performed 0.54 ± 0.04 display flights per minute during nocturnal focal watches, but there was considerable intra-male variation in display rate. Following a burst of activity immediately after their arrival at display arenas at dusk, male display rate was best explained by ambient levels of moonlight. Males produced a previously undescribed insect-like “tik tik” call when inactive on their territories. Females made two nesting attempts per season, using a different site (360 ± 31 m apart) for each attempt. The clutch of two eggs was laid directly on the ground, adjoining a small access “clearing”, in either campo cerrado or wet grassland vegetation. The first egg was laid within two days of a full moon for all seven clutches for which laying date was confirmed. Incubation, brooding and chick provisioning duties were carried out exclusively by the female. The incubation period was 19 days and the nestling period c. 19-20 days. Overall, my data are consistent with polygyny – and even an exploded lek or “landmark” mating system – in this population of White-winged Nightjars. KEYWORDS: Caprimulgidae; Caprimulgus candicans; cerrado; male display behaviour; nesting biology INTRODUCTION chick (Capper et al. 2000, Clay et al. 2014) and male display behaviour (Clay et al. 2000). Male White-winged Owing largely to their cryptic plumage and crepuscular or Nightjars were found to perform ritualised display flights at small arenas – low anthills or termite mounds nocturnal habits, the Caprimulgidae (true nightjars) are comparatively under-studied. Their breeding biology in – located on the upper slopes of ridgelines in areas of particular is poorly known, and for many species the only open grassland. Although males were present near these published data are anecdotal or based on single nesting arenas throughout the night, display activity was highest attempts. Studies of the better-known species show them under cloudless conditions on moonlit nights. Display flights were accompanied by a dull “tk…grrrrrt” sound to be generally ground-nesting, with clutches of one or two eggs, and an essentially monogamous breeding (recording in Ranft & Cleere 1998), perhaps produced system (Cleere 1999, Holyoak 2001). The Caprimulgidae by movement of air through the outer primaries. are also one of only a few bird families in which lunar Based on the apparent clustering of male display synchrony in reproduction has been documented arenas at Aguará Ñu, Clay et al. (2000) suggested that this population of White-winged Nightjars might exhibit (Murton & Westwood 1977, Holyoak 2001). The White-winged Nightjar (Eleothreptus candicans) a lek or “landmark” mating system. Leks can be broadly is one of the rarest caprimulgids in the Americas, defined as any aggregation of males visited by females considered Endangered by BirdLife International (2013). primarily for the purpose of copulation, whereas landmark Research in Paraguay since the discovery of a population species are characterised by the use of a specialised habitat as encounter sites for mating (Höglund & Alatalo 1995). in 1995 at Aguará Ñu, Mbaracayú Forest Nature Reserve (Lowen et al. 1997), has clarified some aspects Four criteria were proposed by Bradbury (1981) to of its breeding biology, including the nest site, eggs and distinguish “classical” leks from other lek-like mating Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople systems: (1) lack of male parental care; (2) aggregation attached to certain individuals, including all adult females of males at an arena to which females come for mating; captured (see Pople 2003 for details). (3) arenas containing no resources required by females (e.g. nest sites, food) except males themselves; and (4) Male territories and display arenas free mate choice by females visiting an arena. Lekking has been suggested to occur in at least three other species of I use the term “territory” to describe the relatively small caprimulgid (see Holyoak 2001). However, no nightjars area within which a male nightjar conducted all his were included on a list of 97 lek-mating bird species breeding-season display activity (as per Clay et al. 2000). compiled by Höglund & Alatalo (1995). Each territory contained a primary “display arena” at Here I present the first comprehensive description which the male carried out most display activity, but some of the breeding biology of the White-winged Nightjar, also contained one or more secondary arenas, which were on the basis of data collected during a three-year study used more sporadically, for example when the male was in Paraguay, including male display behaviour, the disturbed from its primary arena (Clay et al. 2000). For distribution of male territories and details of nesting simplicity, the location of each territory was defined by biology based on data from multiple nesting attempts. its primary display arena. Radio-telemetry showed that c. I use these and other data to assess whether the study 95% of nocturnal breeding-season fixes for radio-tagged population exhibits a polygynous or lekking mating males fell within 100 m of their display arenas (Pople system, and consider the implications for the species’s 2003). In occasional cases when the primary display arena conservation. of a territory changed between or, more rarely, within seasons, the display arena used for longer was selected for subsequent analyses. METHODS The study site was surveyed for displaying males each season between August and December. Males were Fieldwork was conducted over three breeding seasons initially located by the “tk…grrrrrt” sound produced (September-January) between October 1998 and January during display flights, which was often audible from up to 2001 at Aguará Ñu (24°10’S, 55°17’W), a c. 5500-ha c. 200 m. Surveys were mainly conducted on dry, relatively area of open-country habitats within Mbaracayú Forest still, moonlit nights, when male display activity was at its Nature Reserve, Canindeyú department, eastern Paraguay. highest (Clay et al. 2000). Once a displaying male was Aguará Ñu is a low plateau, 170-270 m above sea level, detected, it was observed for 10-15 minutes to identify its which forms a natural island of cerrado vegetation, primary display arena, the precise location of which was flanked by forest to the south and west, and isolated from recorded using a Magellan 2000 XL GPS receiver. It was Paraguay’s main areas of cerrado in the departments of then captured and fitted with an individually numbered Concepción and Amambay to the north-west (Jiménez metal tarsus ring, a coloured plastic ring, and a small, & Knapp 1998). uniquely shaped piece of reflective tape, affixed to the Most fieldwork was focused on an area of c. 400 ha in dorsal side of one of the non-central tail feathers. This southern Aguará Ñu, where a series of shallow valleys drain tail mark was unobtrusive when birds were at rest during southwards into the Arroyo Guyrá Kehá. The seasonally the day, but conspicuous when illuminated with a torch wet grasslands of the valley bottoms and lower slopes at night. Territories were repeatedly revisited to study are dominated by grass and sedge species, and woody male display behaviour. During these visits, the identity vegetation is almost entirely absent. The drier soils of the of the male and the location of the display arena(s) were upper slopes and ridge-tops support “campo cerrado” confirmed. vegetation, consisting primarily of grasses and herbs, but To investigate aggregative tendencies in displaying with a scattering of shrubs (e.g. Cochlospermum regium, males, nearest-neighbour analysis (Clark & Evans 1954, Caryocar brasiliense), Yata’i palms (Butia paraguayensis), Krebs 1989) was used to quantify the spatial distribution saplings and the occasional fully grown tree. Palm density of territories (see also Höglund & Stöhr 1997). The is lowest at the tips of the ridges in the south, where the distance from each primary arena to its nearest neighbour campo cerrado is relatively open. Termite mounds, 15-90 (i.e. the measure used herein to represent distances cm in height, are present throughout the campo cerrado between territories) was derived from GPS locations, and and wet grassland, and low anthills occur occasionally in the average nearest-neighbour distance (NND) within the campo cerrado. each season was calculated. Estimates of expected NND I identified breeding seasons by their principal year, were then obtained (Clark & Evans 1954), assuming a so the season from September 2000 to January 2001 is random distribution of males within the survey area. termed the “2000 season”. Adult White-winged Nightjars The latter was defined by a convex polygon (of c. 400 were captured, sexed and ringed, and lightweight (1.3- ha) encompassing the zone surveyed most thoroughly 2.5 g; <5% of adult body weight) radio-transmitters were Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople for territories (see Results). Since the small sample sizes m away. Nautical twilight (the time at which the sun is in the current study prevented the use of a boundary 12° below the horizon) was used as the point delineating strip, estimates of expected NND were corrected using dusk or dawn from true “night” (Brigham & Barclay Donnelly’s (1978) adjustment for edge-effect bias. An 1992). Until this point, lunar illumination is exceeded index of aggregation (R) was obtained for each season by by residual solar light (Austin et al. 1976). Seventy-one dividing the mean observed NND by the mean expected per cent (64 of 90) of focal watches were begun before NND. When R = 1.0 the observed spatial distribution the end of nautical twilight, and thus termed “dusk does not deviate from random, whereas values of R watches”. The remaining 26 watches were conducted between 1 and 0 indicate an increasing degree of clustering under truly nocturnal conditions and termed “night (Krebs 1989). watches”. The gross topographical characteristics of male During dusk watches, the observer arrived at the territories and an equal number of random points were observation point shortly after sunset, and the watch measured using a digital version of a 1:50000 Paraguayan commenced when the focal male arrived at the display Instituto Geográfico Militar map, imported as a layer into arena. The number and time of all displays were then a geographical information system (GIS) in ArcView 3.2 recorded, as well as details of any interactions with other (ESRI 1999). The following variables were recorded for individuals. A “display” was defined as any flight up from each site: (i) gradient, to nearest 1°; (ii) altitude, to nearest the display mound accompanied by both the “tk” and 5 m above sea level; (iii) altitude difference, from nearest “grrrrrt” sounds. Following the completion of the watch, “spot height” (peak local altitude); (iv) distance to spot the identity of the focal male was confirmed based on its height, to nearest 25 m; (v) distance to stream, to nearest 25 reflective tail mark and/or colour rings. m; (vi) aspect, allocated to one of four quadrants (north- Potential environmental correlates of male display east, south-east, south-west and north-west). rate were investigated using a multiple regression model, Qualitative observations suggested that males with the following variables: (i) time, in minutes, in did not select display arenas at random, even within relation to nautical twilight; (ii) temperature, minimum territories. To determine how mounds used as display recorded during the watch; (iii) moon phase, proportion arenas differed from available mounds, the structural of the moon face illuminated (MFI); (iv) moon height, characteristics of nine display arenas used during the estimated degrees above the horizon; (v) moonlight, 2000 season were measured and compared to those from estimated lunar illuminance (see below). Ambient a sample of random mounds. The arenas included seven temperature was measured every 15 minutes with an primary arenas and two (regularly used) secondary arenas electronic logger. Details of astronomical phenomena of seven males. For each mound I measured: (i) maximum (i.e. MFI, times of sunrise/sunset, moonrise/moonset and height (including any “towers”), to nearest 5 cm; (ii) modal twilights) were obtained from astronomical almanacs. height, i.e. height of the majority of the mound (excluding Moon phase and height significantly influence overall any “towers”), to nearest 5 cm; (iii) basal “circumference”, lunar illuminance (Austin et al. 1976), but neither factor to nearest 0.1 m; (iv) maximum basal chord, across the shows a simple linear relationship with illuminance longest axis of the mound, to nearest 0.1 m; (v) presence/ (R. Willstrop in litt.), so their combined effects were absence of “towers”, i.e. free-standing vertical structures estimated, using information in Austin et al. (1976), and projecting above the modal level of the mound (see treated as a separate measure: “moonlight”. Figure 3a). For each display arena, four random mounds Tape recordings of male display sounds and other were selected by walking 20 m from the display arena on “vocalisations” were made at the display arenas of marked each of the four cardinal points of the compass, and then males, from distances of c. 5-10 m, during the 1999 and identifying the nearest mound. On two occasions it was 2000 breeding seasons. Recordings were later digitised not possible to locate a mound, hence the total number and spectrograms of male display sounds generated in of random mounds measured was 34. Avisoft using the following settings: FFT-length = 128; frame size = 100%; window = Blackman; overlap = Male display behaviour 75%. These settings provided sufficiently fine temporal resolution (8 ms) to distinguish the separate elements To collect quantitative and qualitative data on male of the “grrrrrt” sound. The following temporal measures display behaviour, 30-minute focal watches were were obtained using the on-screen cursors (Figure 1): conducted on dry, relatively still evenings at the primary interval between “tk” and “grrrrrt” components (INT); display arenas of territory-holding males. Focal males length of “grrrrrt” component (GTLEN); number of were monitored from c. 10-20 m away with the aid of elements constituting “grrrrrt” component (GTNUM). a torch set on diffuse beam. Individuals did not appear Frequency measures were not taken, however, given the to be disturbed by the presence of an observer, and on relatively poor resolution (31 Hz) and limited frequency other occasions continued to display when I was just 5 range of display sounds. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople Hatching success was calculated as the percentage of eggs laid that hatched successfully. This method can bias values for hatching success upwards, as it takes no account of nests failing before they are found (Beintema INT GTLEN/GTNUM & Müskens 1987). However, my method of nest location reduced this risk, and the sample size was insufficient to use the daily exposure method (Mayfield 1961, 1975). Nests were visited every two to four days after hatching, and chicks were weighed (to nearest 0.5 g) to obtain data on their growth and development. Chicks were ringed FIGURE 1. Spectrogram of male White-winged Nightjar display at c. 10 days old, when their tarsus width was similar to sound showing “tk” and “grrrrrt” components and the three temporal variables measured for each recording (see text for definitions). adult birds. Younger chicks were individually marked with non-toxic white correction fluid on their feet. During the nestling period, nocturnal focal watches Nesting biology of 50-255 minutes were conducted at nests to characterise the division of chick brooding and provisioning duties Nest sites were located by monitoring radio-tagged by the parents. Nest sites were monitored from distances females and checking their daytime “roosts” for eggs every of c. 10-20 m, using a head-torch with diffuse beam to c. 2 days during the breeding season. The “initiation date” obtain eye-shine from adults without disturbing them. for a nesting attempt was defined as the date on which The duration (to nearest five seconds) of each visit was the first egg was laid: presence of eggs was the only factor recorded, along with the sex of the visiting adult. Radio- reliably distinguishing nest sites from roost sites. Once a tracking equipment was occasionally used to confirm the nest was initiated, its precise location was recorded using identity of tagged birds, but monitoring was primarily a GPS receiver. Gross topographical characteristics of nest visual. The data collected were used to calculate mean sites and an equal number of random sites were measured visit rates and average times spent at, or away from, as described above for male territories. To ensure that the nest site. A small number of daytime spot-checks random sites reflected those available to, but not used by, were conducted to assess parental allocation of diurnal nesting females, they were selected to fall within 400 m brooding duties. of the nest sites with which they were paired. Two radio- tagged females monitored during the 2000 breeding Statistical analysis season ranged up to 403 ± 33 m from the centre of their breeding ranges (pers. obs.). All data were tested for normality and homoscedasticity, The following variables were recorded for each site: and then analysed using standard parametric or non- (i) gradient, to nearest 1°; (ii) altitude, to nearest 5 m parametric univariate tests, as appropriate. If appropriate, above sea level; (iii) distance to nearest stream, to nearest 25 data were subsequently also analysed using multivariate m; (iv) distance to nearest male display arena, to nearest 10 techniques. When the dependent variable was continuous m; (v) aspect, allocated to one of four quadrants (north- (e.g. male display rate), multiple linear regression was east, south-east, south-west and north-west); (vi) habitat, used to investigate the effects of potential covariates. classified as one of two dominant types (campo cerrado Analyses were conducted using a forward stepwise and wet grassland). Habitat type was determined from a procedure with entry and removal probabilities set at P = simplified habitat map of the study site, added as a layer 0.05. At each step, the variable with the most significant to the GIS (see Pople 2003 for details). score statistic was entered into the model, provided that The mass of freshly laid eggs was measured to the its inclusion significantly improved model fit: significance nearest 0.5 g. Measures of length and maximum breadth was tested using partial F-tests (Hair et al. 1995). This of eggs were taken to the nearest 0.1 mm. Egg volume process was repeated until no further variables met the was estimated following Hoyt (1979), using the equation: criteria for entry or removal, at which point the model volume = 0.51 × length × (maximum breadth) . During was considered final. the egg stage, nests were visited every three to five days When the dependent variable could be allocated to confirm the sex of any adults present. These “spot- to one of two discrete categories (e.g. male territory checks” were carried out at various times of day and or random site, display mound or random mound), night, independent of ongoing radio-telemetry fixes binary logistic regression was used to identify which taken to characterise the home range and movements combination of variables best distinguished between of individuals. Although radio-tracking equipment was the two categories. Analyses were conducted using a sometimes used to confirm an adult’s identity, the bird’s forward stepwise procedure as outlined above, but with location was not known prior to any spot-check. significance tested using likelihood ratio tests in which Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople the difference in deviance (-2 × log likelihood) between arenas were significantly clustered within the survey area models with and without the variable was treated as in all three seasons (Table 2). ², with degrees of freedom (df ) equal to the number In total, 13 distinct territories were identified of parameters being added (Manly et al. 1993). When between 1998 and 2000: five of these were occupied in all comparing display mounds and random mounds, the three breeding seasons, and a further three were occupied inclusion of secondary mounds for two males raised in at least two. The ownership of certain territories the possibility of pseudoreplication. To address this, changed both between and, less frequently, within parameter estimates produced by the model were jack- seasons, probably owing to the death or emigration of knifed following Sokal & Rohlf (1995), taking the male the original territory-holding male. In nine of 11 such as the sampling unit. Jack-knifed parameter estimates changes in territory ownership, the original male was did not differ significantly from those of the original never recorded again; in two instances the male moved to model, suggesting that pseudoreplication was not a major a territory vacated by another “missing” male. problem. Male territories were located almost exclusively on Categorical data were analysed using the G-test of the upper slopes of ridgelines running perpendicular to independence. A form of the G-test for goodness of fit, the main Arroyo Guyrá Kehá valley (running north-west adjusted using the Williams correction for the two-cell to south-east across the lower half of the maps in Figure 2). case (Sokal & Rohlf 1995), was used to compare the Preliminary analysis of gross topographical characteristics observed sex ratio of adult captures with that expected using univariate tests suggested that territories differed on the basis of a 1:1 sex ratio. Statistical significance was significantly from random points for only one variable set at P < 0.05 unless otherwise specified, and means are (Table 3). A binary logistic regression model, explaining given ± one standard error (SE). Since the repeated use of c. 27% of the deviance in site use, also revealed that a statistical test increases the probability of committing a Type I error, the sequential Bonferroni technique was employed to adjust the initial a level when making many simultaneous comparisons (Rice 1989). RESULTS Forty-nine White-winged Nightjars (34 adult and 15 young birds; Table 1) were captured and ringed between 1998 and 2001. The sex ratio of captured adults was biased towards males in all three field seasons, and the overall ratio of 3.25:1 male:female differed significantly from parity. The four females captured in 1998 were all caught late in the breeding season, and none of them was confirmed to (re-)nest that season; two individuals captured in late December had already begun post- nuptial moult. Hence, all data on nesting biology were obtained from females caught during the 1999 and 2000 breeding seasons. Male territories Breeding-season territories were identified for 10 males in 1998 and eight males in each of 1999 and 2000. One territory, discovered late in the 2000 breeding season, was excluded from the nearest-neighbour analysis as it was located outside the normal survey area (Figure 2). The remaining territory-holding males occurred at a mean density of one male per 50 ± 5 ha (n = 3 seasons) within the area surveyed. The closest male display arenas FIGURE 2. Location of male White-winged Nightjar territories were just 70 m apart (in 1998), but the average nearest- (closed circles) during: (a) 1998; (b) 1999; and (c) 2000 breeding neighbour distance across all three seasons was 254 ± 8 seasons. Pale brown lines are contours, dashed lines are watercourses. m. Aggregation indices confirmed that primary display The polygon represents the c. 400-ha survey area. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople “distance to stream” was the only significant predictor deviance in mound use, showed “modal height” to be (likelihood ratio test: ² = 9.89, df = 1, P < 0.005): on the only significant predictor (likelihood ratio test: ² = average, male territories were further from streams than 11.01, df = 1, P < 0.001): on average, display mounds random points. The model correctly reclassified 69% of were lower than random mounds. The model correctly sites overall, with an equal degree of classification success reclassified 91% of random mounds, but only 22% of for territories and random sites. Aspect was not entered mounds used as display arenas, resulting in an overall into the model, but univariate tests suggested that male correct reclassification rate of 77%. territories were more likely to have a westerly than easterly aspect compared to random points (11 of 13 versus 5 of Male display behaviour 13; G-test: G = 5.80, df = 1, P < 0.05). At a finer scale, preliminary analysis of the structural In both 1999 and 2000, sporadic display activity was characteristics of mounds used as display arenas in 2000 noted on certain nights leading up to the August full suggested that they differed significantly from random moon, but male display activity did not commence in mounds for a number of variables (Table 4). However, a earnest until September. Activity levels remained high binary logistic regression model, explaining 25% of the through to December, but subsided soon after: the latest date on which display activity was noted was 6 January. Although an attenuated version of the mechanical “grrrrrt” sound was occasionally heard at the start of foraging sallies, the full “tk… grrrrrt” display sound was only ever noted during male display flights at arenas. Of 132 display flights by five males in 2000, the majority (63%) were initiated from a nearby perch, but 23% immediately followed the return of the male from a foraging sally, and the remaining 14% involved males already sitting on their display arena. In these latter cases, the male was observed to jump up and down on the display arena to produce the “tk” (probably by contact with the mound), before flying up with the “grrrrrt” sound as per usual. Display activity was not solely restricted to males in “definitive” plumage. A young male moulting into definitive plumage, but still retaining five or six (brown) juvenile outer primaries, was observed to display on two nights in early November 1998 (see below). Recordings of 8-19 display sounds were obtained for eight marked males during the 1999 and 2000 breeding seasons. Spectrograms showed that display sounds had a relatively fixed structure, with the “tk” and “grrrrrt” components separated by 500 to 700 ms, but with broadly similar frequency ranges (c. 100-400 Hz; Figure 1). The display sounds produced by the pre-definitive male in 1998 were of a noticeably higher frequency, but a recording of its display was not obtained for direct comparison. Two of the three temporal measures (GTLEN and GTNUM) were highly correlated (Spearman rank correlation: r = 0.96; n = 103; P < 0.001). Only GTNUM was considered further, as it was judged to be less susceptible to measurement error. Males showed significantly more inter- than intra-individual variation for both INT (Kruskal-Wallis test: H = 57.82, df = 7, P < 0.001) and GTNUM (H = 60.40, df = 7, P < 0.001), suggesting consistent inter-individual differences in these temporal measures. A total of 90 focal watches was conducted at male display arenas over the course of three breeding seasons FIGURE 3. Typical male display arena with “towers” (a), eggs (b) and chicks aged 14 days old (c) of White-winged Nightjar at Aguará Ñu. (1998: n = 10; 1999: n = 20; 2000: n = 60). No significant Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople TABLE 1. Sex and age composition of White-winged Nightjars captured. Season Males Females Young birds Total Sex ratio (males/female) G P 1998 20 4 3 27 5.00 11.41 < 0.001 1999 14 3 5 22 4.67 7.50 < 0.01 2000 9 3 9 21 3.00 3.01 ns Overall 26 8 15 49 3.25 9.89 < 0.005 (a) Column totals do not equal “Overall” total due to recapture of certain individuals between seasons. TABLE 2. Mean ± SE nearest-neighbour distances (NND) and degree of aggregation for breeding-season territories of male White-winged Nightjars. Season No. male Observed Expected Aggregation index Significance territories NND (m) NND (m) (R) 1998 10 253 ± 97 372 ± 69 0.68 z = -1.71, P = 0.044 1999 8 241 ± 35 425 ± 89 0.57 z = -2.07, P = 0.019 2000 7 268 ± 34 460 ± 103 0.58 z = -1.87, P = 0.031 (a) Assuming random distribution of males within 406 ha survey area (see Methods). (b) P-values from one-tailed z-tests (Campbell 1996). TABLE 3. Mean ± SE values of topographical variables for male White-winged Nightjar breeding-season territories and an equal number of random points. Variable Male territories Random points Significance (n = 13) (n = 13) Gradient (°) 2.2 ± 0.6 4.5 ± 0.7 W = 220, P = 0.023 Altitude (m) 232 ± 4 214 ± 6 t = -2.51, P = 0.019 Altitude difference (m) 11 ± 3 29 ± 7 W = 220, P = 0.023 Distance to spot height (m) 406 ± 77 683 ± 110 t = 2.06, P = 0.050 Distance to stream (m) 567 ± 74 310 ± 46 W = 122, P = 0.007 Aspect – number of sites facing: NE (0-90°) 0 (0%) 0 (0%) SE (90-180°) 2 (15%) 8 (62%) SW (180-270°) 6 (46%) 2 (15%) NW (270-360°) 5 (38%) 3 (23%) (a) From unpaired t-tests or Mann-Whitney U-tests. Values underlined indicate table-wide significance at an adjusted initial a level of 0.01 (Rice 1989). TABLE 4. Mean ± SE values of the structural characteristics of male White-winged Nightjar display arenas used during 2000 breeding season compared to a sample of random mounds. Variable Display arenas Random mounds Significance (n = 9) (n = 34) Maximum height (m) 0.42 ± 0.10 0.61 ± 0.04 W = 831, P = 0.014 Modal height (m) 0.24 ± 0.03 0.51 ± 0.05 W = 842, P = 0.006 Basal “circumference” (m) 6.0 ± 0.8 3.7 ± 0.5 W = 665, P = 0.013 Maximum basal chord (m) 2.1 ± 0.3 1.1 ± 0.1 W = 654, P = 0.005 “Tower(s)”: present 3 (33%) 8 (24%) G = 0.32, ns absent 6 (67%) 26 (76%) (a) From Mann-Whitney U-tests or G-test. Values underlined indicate table-wide significance at an adjusted initial a level of 0.01 (Rice 1989). Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople differences in display rate were apparent between seasons was subsequently recorded on several occasions during (Kruskal-Wallis test: H = 1.18, df = 2, P > 0.05), so data display watches (see Figure 5 for spectrogram), and were pooled for subsequent analyses. On average, males was only observed from males perched in or near their performed 0.54 ± 0.04 displays per minute during focal display arena. The thin, high-pitched nature of the call watches, although considerable variation in display rate made it difficult to detect under certain conditions, but was apparent (range: 0.00-1.37 displays/min; n = 90). it was apparently produced at rates of 1.24 ± 0.20 “tik However, when comparing the six males for which five tiks” per minute (range: 0.00-3.33 calls/min; data from or more focal watches were conducted, no evidence was 26 display watches). There was no significant correlation found for consistent inter-male differences in display rate between the rate of “tik tik” production and male display (H = 5.95, df = 5, P > 0.05). rate during display watches (Pearson correlation: r = 0.31, On average, males arrived at their display arenas 26.6 n = 26, P > 0.05). ± 0.6 minutes after sunset and 27.9 ± 0.7 minutes before Aggressive interactions between males were the end of nautical twilight. On arrival, males showed a observed during 14 focal watches (30% of watches for burst of high display activity. The mean display rate in the which details were noted) and on various occasions first half of dusk watches was more than twice that in the during ad hoc observations at display arenas. Interactions second half (0.75 ± 0.06 versus 0.35 ± 0.04 displays/min; typically comprised rapid chases, sometimes involving Wilcoxon signed ranks test: T = 1742, n = 64, P < 0.001), up to three males, and occasionally also physical contact. whereas there was no significant difference between the A soft, liquid “gurgling” call produced during chases was two halves of night watches (0.50 ± 0.07 versus 0.54 ± noted on a number of occasions. Aggressive interactions 0.06 displays/min; paired t-test: t = -1.01, P > 0.05). were observed most frequently at one particular display When considering night watches alone, “moonlight” arena (11 of 14 watches), and this bias was significant was the only environmental variable that explained a when allowing for the distribution of watches among significant amount of the variation in display rate (overall display arenas (G-test: G = 7.64, df = 1, P < 0.01). There fit of model: R = 0.166, F = 4.77, P < 0.05). Display was a non-significant trend for higher rates of display 1,24 rate was positively related with this estimate of overall activity during watches that included male–male chases lunar illuminance, reflecting a tendency for males to compared to during those that did not (0.74 ± 0.10 display more when the moon was fuller and higher in versus 0.53 ± 0.06 displays/min; unpaired t-test: t = the sky (Figure 4). However, even when conditions were -1.76, P = 0.085). not suitable for display activity, males were generally still It was not possible to record female visits to present – either sallying or resting – on their territories. male display arenas reliably, owing to difficulties in There was no significant relationship between display rate distinguishing unmarked females from juveniles or from and sallying rate during focal watches in which the latter females of other similar-sized species (e.g. Little Nightjar was measured (Spearman rank correlation: r = -0.26; n = Setopagis parvulus). However, during one dusk watch in 36; P > 0.05). October 1998, five nights prior to full moon, a female During one display watch in November 1998, the visit and presumed copulation were documented. Over focal male produced a previously undocumented, insect- the course of c. 20 minutes, the female landed on the like “tik tik” call whilst perched in the vicinity of its display display arena at least twice, and three presumed copulation arena. This vocalisation (or a single “tik” equivalent) attempts by the focal male were witnessed. Prior to one 1,4 1,2 1,0 0,8 0,6 0,4 0,2 0,0 0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,0 'Moonlight' FIGURE 5. Spectrogram of previously undescribed, insect-like “tik tik” call given by male White-winged Nightjars when inactive FIGURE 4. Male White-winged Nightjar display rate in relation to in vicinity of display arenas. Note background insect noise at c. 3.5 “moonlight”, an estimate of lunar illuminance. Data from “night” and 5.0 kHz. Recordings made by M. C. Velázquez at Aguará Ñu in watches only (Pearson correlation: r = 0.41, n = 26, P < 0.05). December 2000. Revista Brasileira de Ornitologia, 22(2), 2014 Male display rate (displays/min) Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople attempt, the male appeared to display to the female whilst Nesting biology they were both perched on the arena, by spreading its wings and cocking and splaying its tail. When the female I documented nine breeding attempts by four different left the display arena after copulation, the male followed females : five in 1999 and four in 2000 (Table 5). A fifth and flew parallel to her, with wings raised, in a slow female captured in early December 2000 was heavily buoyant flight, conspicuously different to the rapid flight gravid, but not subsequently relocated. Eggs were laid typical of male–male encounters. On another occasion, directly on the ground, in campo cerrado (six nests) or wet in December 2000, a male was observed to conduct a grassland (three nests) vegetation, within a small “tunnel” similar “escort” flight, with raised wings, for a presumed between tussocks of grass or herbaceous plants and facing female flying through his territory. onto a small area of bare earth (c. 25-50 cm in diameter), FIGURE 6. Distribution of White-winged Nightjar nest sites in relation to male display arenas during: a) 1999 and b) 2000 breeding seasons. Nest sites displayed as filled circles; multiple nesting attempts distinguished by “A”, “B”, etc. Display arenas displayed as open circles; labels refer to the territory-holding male. Note the territorial fidelity in successive seasons shown by males M001, M052, M055, and M057. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople the latter apparently serving as a “runway” for the adult asymmetry in growth rate was observed in two broods: bird (as at roost sites; pers. obs.). There was no evidence in one (circular symbols in Figure 7), the larger chick to suggest that the gross topographical characteristics of (26.3 g) weighed more than twice its sibling (10.9 g) by nest sites differed significantly from those of random sites day 10 (see also Figure 3c). The latter had disappeared available to nesting females (Table 6). by day 15. Initiation date was confirmed for seven clutches, Nocturnal focal watches conducted at six nests (of and consistently fell within two days of a full moon four females) showed that chick brooding and provisioning (1.3 ± 0.3 days; range: 0-2). The earliest initiation date was undertaken solely by the female. In almost 25 hours observed was 25 September (in 1999), although two of monitoring (mean watch length: 113.6 ± 17.3 min, n chicks discovered on 13 October 2000 were estimated to = 13), no male visits were observed, whereas 67 female be approximately 10 days old, and had probably hatched visits were recorded, at a mean rate of 3.04 ± 0.80 visits/ from eggs laid around the full moon of 14 September. hour (range: 1.20-4.94; n = 4 females). Females spent Females made two nesting attempts per season, using a on average 7.7 ± 1.3 minutes (range: 0.1-46.0 min; n = different site for each attempt (mean distance between 58) at the nest during each visit, and 11.8 ± 1.3 minutes nest sites: 360 ± 31 m, range: 296-461 m, n = 4; Figure (range: 1.5-59.9 min; n = 55) away from the nest between 6). The female (F067) captured in late November 1998, visits. It was not possible to confirm whether chicks whilst brooding a single chick, was probably making a were provisioned on every visit. Time spent at the nest second breeding attempt (Table 5). during nocturnal focal watches was negatively correlated Clutch size was two for all seven clutches discovered with minimum ambient temperature (Spearman rank at the egg stage. It was not possible to check each nest correlation: r = -0.81, n = 13, P < 0.005). In 100% (10 daily, but the second egg was generally laid within of 10) of daytime checks during the chick period, the 24 hours of the first, although in one case there was a female was present at the nest site, either brooding or delay of at least 30 hours. Eggs were pale cream-beige, immediately adjacent to the chicks. with variable dark brown and mauve-grey speckling and Females displayed a range of anti-predator blotching (Figure 3b). Differences in egg patterning were behaviours whilst incubating or brooding. When initially observed within most clutches, with one egg being more approached, they relied on their cryptic plumage, densely and uniformly covered in fine markings, and crouching motionless with eyes closed to slits. If the the other showing fewer, larger markings, often with a intruder approached to within 0.5-2.0 m, they would subapical ring of grey blotches at the obtuse pole. Egg shuffle forward into the adjacent clearing in preparation morphometrics are summarised in Table 7. Both eggs for take-off. On one occasion, a female incubating a of a clutch weighed on day 1 and day 18 of incubation single, recently laid egg performed a presumed threat decreased in mass by 0.7 g (11% and 12%) during the display: drooping the wings slightly, puffing up the body intervening period. Incubation began with the laying of the first egg, and was carried out exclusively by the female: no male 45 was ever found at the nest site. Thirty-four spot-checks carried out at seven nests during the incubation period found the female in attendance during 100% (20 of 20) of daytime and 57% (8 of 14) of nocturnal checks. During nine daytime checks, the female was present but sitting slightly behind one (n = 3) or both (n = 6) of the eggs. There was no significant difference in ambient air temperature on these occasions compared to 11 occasions when the female was incubating both eggs (unpaired t-test: 23.4 ± 1.8°C versus 23.7 ± 1.3°C; t = -0.13; P > 0.05). Although I could not confirm the hatching date for every clutch, the modal incubation period was 19 days, and chicks probably hatched within 24 hours of 0 5 10 15 each other. Age of chick (days) Hatching success was 86% (12 of 14 eggs) for the seven nests for which initial clutch size was known. At FIGURE 7. Growth rate of eight White-winged Nightjar chicks hatching, chicks had open eyes and were covered in a between hatching and 15 days of age. Each brood represented by different shape; open and closed symbols represent two chicks within pale buff-coloured down, with greyish legs and pale- same brood. Four chicks weighed on day 0 had masses of c. 5 g (square tipped greyish bill. The average mass of four chicks (from and diamond symbols). two clutches) on day of hatching was 5.2 ± 0.1 g. Clear Revista Brasileira de Ornitologia, 22(2), 2014 Mass of chick (grams) Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople feathers, raising and slightly splaying the tail, whilst pattern. After about 10 days, the developing flight feathers rocking the body and tail from side-to-side. When began to project from their feather sheaths. Although I eventually flushed (off eggs or chicks), females typically could not confirm the exact age of fledging, extrapolation performed an injury-feigning distraction display, flying of observed growth rates suggests that chicks attain mean 5-10 m away from the nest with a low, laboured flight, adult mass (c. 50 g) at c. 19 days. The sole surviving chick before landing clumsily in the vegetation. If approached of one clutch present on day 18 could not be relocated again, the female would fly off normally, although on on day 20. most occasions she was subsequently found to have Owing to uncertainty over the fate of most broods, remained near the nest site. I could not estimate fledging success or overall nesting At one week old, chicks were already beginning to success. I recorded two instances of predation (once of lose their natal down and attain the juvenile plumage eggs, once of chicks; Table 5), but could not identify TABLE 5. Summary of White-winged Nightjar nesting attempts during 1999 and 2000 breeding seasons. First nesting attempt Second nesting attempt Season Female Date first Incubation Hatching Chicks last seen Date first Incubation Hatching Chicks last seen egg laid period date at nest site egg laid period date at nest site a b † 1999 F063 26 September – – – 25 October 19 days 13 November 1 December c † d 1999 F066 25 September 19 days 14 October 25 October 21 November 19 days 10 December 14 December e † 1999 F067 – – – – – – – 10 December f g † † 2000 F010 – – – 18 October 12 November 19 days 1 December 14 December h † 2000 F066 11 October 17 days 28 October – 13 November 19 days 2 December 17 December (a) Eggs taken by unknown predator within one week of laying. (b) Smaller of two chicks disappeared during afternoon of 28 November; larger chick no longer present by 3 December. (c) Nest not checked for five days; no sign of chicks on 31 October. (d) Both chicks still present – no further fieldwork in 1999. (e) Single chick discovered on 23 November; estimated by mass and stage of feather development as approximately 10 days old. (f ) Two chicks discovered on 13 October; estimated by mass and stage of feather development as approximately 10 days old. (g) Female and chicks found to have moved to new site by 20 October; not seen subsequently. (h) Chicks disappeared, presumably taken by a predator, within one week of hatching. (†) Denotes chick(s) believed to have fledged successfully. TABLE 6. Mean ± SE values of topographical variables for White-winged Nightjar nest sites and an equal number of random sites. Nest sites Random sites Variable Significance (n = 9) (n = 9) Gradient (°) 5.6 ± 1.2 4.8 ± 0.9 t = -0.52, ns Altitude (m) 218 ± 5 217 ± 5 t = -0.07, ns Distance to nearest stream (m) 306 ± 61 319 ± 54 t = 0.17, ns Distance to nearest male display arena (m) 227 ± 49 208 ± 56 t = -0.25, ns Aspect – number of sites facing: NE (0-90°) 1 (11%) 0 (0%) SE (90-180°) 1 (11%) 1 (11%) SW (180-270°) 5 (56%) 6 (67%) NW (270-360°) 2 (22%) 2 (22%) Habitat – number of sites in: Campo cerrado 6 (67%) 8 (89%) Wet grassland 3 (33%) 1 (11%) (a) From unpaired t-tests. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople the predator on either occasion. One female and her nightjar species (e.g. Jackson 1985, Mills 1986, Vilella chicks were found to have moved to a new clearing, c. 1995, Perrins & Crick 1996, Pichorim 2002), and authors 10 m from the original nest site, when the chicks were typically emphasise the advantages of moonlight for about 10 days old. Assumptions about the fate of other foraging during the first two weeks of the nestling period. broods, based solely on the disappearance of chicks from The modal incubation period of 19 days resulted in chicks the original nest site, therefore seem inappropriate. Of hatching between the new moon and first quarter, and 11 chicks ringed at the nest site, I did not recapture any hence during a period of increasing moonlight. However, (post-fledging) before the end of my fieldwork in June with male investment in reproduction apparently limited 2001. However, one chick ringed in October 2000 was to genes, it might also be argued that females use nights subsequently recaptured as a yearling male in December with high moonlight to assess the quality of males before 2001 (J. Mazar Barnett in. litt.). securing copulations (although these two considerations are not incompatible). Male White-winged Nightjars displayed for a TABLE 7. Summary of morphometrics of 14 White-winged Nightjar eggs from seven clutches. relatively prolonged period between late August and early January, broadly coinciding with the vocal activity of other nightjar species breeding at the study site (e.g. Measure Mean ± SE Range Common Pauraque Nyctidromus albicollis, Little Nightjar, Spot-tailed Nightjar Hydropsalis maculicauda and Rufous Mass (g) 6.7 ± 0.1 5.8-7.5 Nightjar Antrostomus rufus). The considerable variation in display rate within this period was likely due more to Length (mm) 27.7 ± 0.3 26.7-29.9 environmental conditions than to consistent differences Maximum breadth (mm) 21.0 ± 0.1 19.8-21.7 between males. The usually rather short-lived high levels of activity observed when males first arrived at display Volume (cm ) 6.2 ± 0.1 5.4-7.1 arenas at dusk were perhaps analogous to the dawn peak of song activity noted for many diurnal birds. Following (a) Mass provided only for eggs weighed within two days of this initial burst, male display rate was best predicted by laying (n = 10). ambient levels of moonlight. Increases in vocal activity under conditions of increased moonlight have been noted DISCUSSION for various nightjar species (e.g. Cooper 1981, Mills 1986, Wilson & Watts 2006), so it is not surprising that During the breeding season, male White-winged Nightjars a similar relationship should exist for a species with such a strong visual component to its display. The “tik tik” call defended small territories on the upper slopes of low ridgelines in the southern sector of Aguará Ñu. The patterns produced by males while in the vicinity of their display of territory use and ownership I observed supported the arenas seems likely also to play a role in mate attraction suggestion that male territories are traditional (Clay or territory defence. et al. 2000). Territories were significantly aggregated My results support previous suggestions that the “tk” and “grrrrrt” components of male display sounds are within the survey area, with the primary display arenas of neighbouring males separated on average by distances both mechanical, rather than vocal, in origin (Clay et al. of c. 250 m. Males played no part in the parental care 2000). Careful observation indicated that the “tk” sound of eggs or chicks, and females were only rarely observed was not produced by clapping of the wings below the near display arenas, with anecdotal evidence suggesting body, as postulated by Clay et al. (2000), but probably by contact with the display mound. However, there was little that they mainly visited around full moon when assessing males and seeking copulations. Nest sites were on average doubt that the “grrrrrt” sound was produced by the wings 227 m from the nearest display arena, with no evidence during the male’s near-vertical ascent from the display to suggest that they fell within the boundaries of male mound. Although not as highly modified as those of the territories. Chick paternity analyses using the amplified Sickle-winged Nightjar Eleothreptus anomalus (which is known to produce mechanical wing sounds; Straneck fragment length polymorphism (AFLP) technique (Vos et al. 1995) suggested that the nearest territory-holding & Viñas 1994), the rigid, curved outer primaries of male was not always the father of a brood, and that one definitive male White-winged Nightjars probably play male sired the chicks of at least two different females a key role in generating the “grrrrrt” (Clay et al. 2000, during 1999 (K. Dasmahapatra in litt.). Pople 2003). The production of mechanical sounds is relatively rare in birds (Bostwick 2000), but its evolution The timing of White-winged Nightjar nesting attempts showed remarkable synchrony with the full is often associated with acrobatic, polygynous courtship moon. A relationship between nest initiation date and displays (Prum 1998), as occur in the White-winged moon phase has been demonstrated for a number of Nightjar system. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople Overall, my results indicate that the population management of its habitats. The sex-ratio bias of adult of White-winged Nightjars at Aguará Ñu employs a captures is likely to have been an artefact of the greater polygynous mating system. Although the breeding conspicuousness of males (particularly during the biology of most nightjar species remains poorly breeding season), rather than a true reflection of a skewed known, polygyny is apparently relatively uncommon population sex ratio. However the finding that the study in the Caprimulgidae, and has only been demonstrated population probably exhibits a polygynous, possibly even convincingly for two other species: the Standard-winged lekking, mating system does have potential consequences Nightjar Macrodipteryx longipennis and Pennant-winged for reproductive skew and effective population size Nightjar M. vexillarius (Fry et al. 1988, Holyoak 2001, (Sutherland 1998). Furthermore, if male display arenas are Jackson 2004). Male emancipation from parental care located at traditional sites, these areas could be of special is an important prerequisite for any form of polygyny importance to the population, with limited possibilities (Höglund & Alatalo 1995), but is unexpected in altricial for relocation if they are destroyed (Clay et al. 2000). bird species if fledging success is limited by the amount As with most species of grassland bird at the study site of food brought to the nestlings (Oring 1982). Snow & (pers. obs.), the White-winged Nightjar nests between Snow (1979) proposed three ecological attributes that September and December, on the ground amongst campo favour the evolution of male emancipation in altricial cerrado or wet grassland vegetation. Any wildfires during bird species: frugivory; the ability to regurgitate food; and the nesting period would have a major impact on breeding roofed nests (in rainy season breeders). Nightjars are well success, and extensive burns in the latter half of August known to regurgitate food for their chicks (e.g. Cramp could also disrupt nesting if vegetation did not regrow 1985, Sick 1993, Cleere 1999), and this ability probably sufficiently before the onset of breeding activity. Wildfires facilitates maximal provisioning efficiency per trip away are a regular occurrence at Aguará Ñu, particularly during from the nest. In addition, for species living in savanna the austral spring (e.g. August and September), so my habitats, such as the White-winged and Standard-winged findings add weight to recommendations elsewhere for Nightjars, the mass hatches of winged termites and ants more active fire management within the few protected that occur during summer rains (e.g. Jackson 2000, areas where this globally threatened species is known to Pinheiro et al. 2002) could be analogous to the periodic persist (e.g. Rodrigues et al. 1999, Capper et al. 2000, super-abundance of food associated with frugivory. Pople 2003). Although the population of White-winged Nightjars at Aguará Ñu appeared to fulfil all four of the criteria stipulated by Bradbury (1981) for lek-mating species, the ACKNOWLEDGEMENTS exact nature of the mating system remains uncertain. The inter-male distances I recorded far exceed those typical This work was carried out as part of a Ph.D. funded by the Biotechnology and Biological Sciences Research of so-called “classical” lekking species, but were within the range observed for other “exploded” or “dispersed” Council and supervised by Michael Brooke at the avian leks (e.g. Théry 1992, Alvarez Alonso 2000, Gray Department of Zoology, University of Cambridge, UK. et al. 2009). Moreover, the question of whether males Permission to work within the Mbaracayú Forest Nature were aggregated for sexual reasons, or solely as a result of Reserve was kindly granted by the Comité de Asuntos Científicos of the Fundación Moisés Bertoni (FMB). the patchy availability of suitable display habitat, remains unanswered. If the latter is true, and males were spaced Thanks are also due to several other FMB staff (current regularly within the available habitat (e.g. Pruett-Jones & and former), particularly Alberto Yanosky, René Palacios, Pruett-Jones 1982), the study population might best be Claudia Mercolli, Tito Fernández, Myriam Velázquez (for described as exhibiting a “landmark” mating system. This help with fieldwork and sound-recordings) and Rafael González (for assistance with male display watches). term has traditionally been reserved for insect species in which males aggregate on hilltops or in forest clearings Additional help with fieldwork was also kindly provided (e.g. Alcock 1981), but is applicable to any taxon using a by Alejandro Bodrati, Osvaldo Carrillo, Hugo del Castillo, specialised habitat to provide encounter sites for mating Rob Clay, Nigel Cleere, Alberto Esquivel, Estela Esquivel, (Höglund & Alatalo 1995). Although the evolutionary Lies Hiers, Mario Ledesma, Arne Lesterhuis, Alfredo Stroessner, and Sergio Villanueva. In the UK, particular causes of aggregation in landmark species differ from those for classical lekking species, females of the former thanks are due to Kanchon Dasmahapatra (for chick do also visit male aggregations primarily for the purpose paternity analyses) and Roderick Willstrop (for advice of mating, and hence the phenomenon still fits the broad on astronomical phenomena). Mike Brooke, Ian Burfield definition of lekking proposed by Höglund & Alatalo and Andy Radford provided valuable feedback on an early version of the manuscript. Rob Clay and Nigel Collar (1995). My results have various implications for the provided constructive input and encouragement during conservation of the White-winged Nightjar and the preparation of the final manuscript, which was further Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople Fry, C. H.; Keith, S. & Urban, E. K. 1988. The birds of Africa. Vol. improved by the review comments of Mark Brigham III. London: Academic Press. and Des Jackson. Finally, special mention must be made Gray, T. N. E.; Hong, C.; Collar, N. J. & Dolman, P. M. 2009. of the late Juan Mazar Barnett, whose discovery of the Sex-specific habitat use by a lekking bustard: conservation population of White-winged Nightjar at Aguará Ñu in implications for the Critically Endangered Bengal Florican 1995 set in motion the series of events that ultimately led (Houbaropsis bengalensis) in an intensifying agroecosystem. The Auk, 126: 112-122. to my study. Hair, J. F.; Anderson, R. E.; Tatham, R. L. & Black, W. C. 1995. Multivariate data analyses with readings. Englewood Cliffs: Prentice-Hall. REFERENCES Höglund, J. & Alatalo, R. V. 1995. Leks. Princeton: Princeton University Press. Höglund, J. & Stöhr, S. 1997. A non-lekking population of Black Alcock, J. 1981. Lek territoriality in the tarantula hawk wasp Grouse Tetrao tetrix. Journal of Avian Biology, 28: 184-187. Hemipepsis ustulata (Hymenoptera: Pompilidae). Behavioral Holyoak, D. T. 2001. Nightjars and their allies: the Caprimulgiformes. Ecology and Sociobiology, 8: 309-317. Oxford: Oxford University Press. Alvarez Alonso, J. 2000. The breeding system of the Orange-crowned Hoyt, D. F. 1979. Practical methods of estimating volume and fresh Manakin. Condor, 102: 181-186. weight of bird eggs. Auk, 96: 73-77. Austin, R. H.; Phillips, B. F. & Webb, D. J. 1976. A method for Jackson, H. D. 1985. Aspects of the breeding biology of the calculating moonlight illuminance at the earth’s surface. Journal of Fierynecked Nightjar. Ostrich, 56: 263-276. Applied Ecology, 13: 741-748. Jackson, H. D. 2000. Food of the nightjars in Zimbabwe. Ostrich, Beintema, A. J. & Müskens, G. J. D. M. 1987. Nesting success of 71: 404-407. birds breeding in Dutch agricultural grasslands. Journal of Applied Jackson, H. D. 2004. Courtship displays of the Pennant-winged Ecology, 24: 743-758. Nightjar, Semeiophorus vexillarius. Ostrich, 75: 1-4. BirdLife International. 2013. Species factsheet: Eleothreptus candicans. Jiménez, B. & Knapp, S. 1998. Palms of the Reserva Natural del http://www.birdlife.org/datazone/speciesfactsheet.php?id=2390 Bosque Mbaracayú (Mbaracayú Natural Forest Reserve). Principes, (access on 3 August 2013). 42: 65-79. Bostwick, K. S. 2000. Display behaviors, mechanical sounds, Krebs, C. J. 1989. Ecological methodology. New York: HarperCollins. and evolutionary relationships of the Club-winged Manakin Lowen, J. C.; Clay, R. P.; Mazar Barnett, J.; Madroño Ñieto, A.; (Machaeropterus deliciosus). Auk, 117: 465-478. Pearman, M.; López Lanús, B.; Tobias, J. A.; Liley, D. C.; Bradbury, J. W. 1981. The evolution of leks, p. 138-169. In: Brooks, T. M.; Esquivel, E. Z. & Reid, J. M. 1997. New and Alexander, R. D. & Tinkle, D. W. (eds.). Natural selection and noteworthy observations on the Paraguayan avifauna. Bulletin of social behaviour. New York: Chiron Press. the British Ornithologists’ Club, 117: 275-293. Brigham, R. M. & Barclay, R. M. R. 1992. Lunar influence on Manly, B. F. J.; McDonald, L. L. & Thomas, D. L. 1993. Resource foraging and nesting activity of Common Poorwills (Phalaenoptilus selection by animals: statistical design and analysis for field studies. nuttallii). Auk, 109: 315-320. London: Chapman & Hall. Campbell, D. J. 1996. Aggregation and regularity: an inclusive one- Mayfield, H. F. 1961. Nesting success calculated from exposure. tailed nearest-neighbour analysis of small patchy populations. Wilson Bulletin, 73: 255-261. Oecologia, 106: 206-211. Mayfield, H. F. 1975. Suggestions for calculating nest success. Wilson Capper, D. R.; Esquivel, E. Z.; Pople, R. G.; Burfield, I. J.; Clay, Bulletin, 87: 456-466. R. P.; Kennedy, C. P. & Mazar Barnett, J. 2000. Surveys and Mills, A. M. 1986. The influence of moonlight on the behavior of recommendations for the management of Aguará Ñu in the Reserva goatsuckers (Caprimulgidae). Auk, 103: 370-378. Natural del Bosque Mbaracayú, eastern Paraguay. Unpublished Murton, R. K. & Westwood, J. 1977. Avian breeding cycles. Oxford: report. Clarendon Press. Clark, P. J. & Evans, F. C. 1954. Distance to nearest neighbor as Oring, L. W. 1982. Avian mating systems, p. 1-92. In: Farner, D. a measure of spatial relationships in populations. Ecology, 35: S.; King, J. R. & Parkes, K. C. (eds.). Avian biology. New York: 445-453. Academic Press. Clay, R. P.; López Lanús, B.; Tobias, J. A.; Lowen, J. C. & Mazar Perrins, C. M. & Crick, H. Q. P. 1996. Influence of lunar cycle on Barnett, J. 2000. The display of the White-winged Nightjar. laying dates of European Nightjars (Caprimulgus europaeus). Auk, Journal of Field Ornithology, 71: 619-626. 113: 705-708. Clay, R. P.; Mazar Barnett, J. & Esquivel, E. Z. 2014. First Pichorim, M. 2002. Biologia reprodutiva do bacurau-tesoura-gigante description of the eggs and nest-site of the White-winged Nightjar (Macropsalis forcipata, Caprimulgidae) no morro Anhangava, Eleothreptus candicans (Aves: Caprimulgidae). Revista Brasileira de Paraná, Sul do Brasil. Ararajuba, 10: 149-165. Ornitologia, [vol: pp-pp]. Pinheiro, F.; Diniz, I. R.; Coelho, D. & Bandeira, M. P. S. 2002. Cleere, N. 1999. Family Caprimulgidae (Nightjars), p. 302-386. In: Seasonal pattern of insect abundance in the Brazilian cerrado. del Hoyo, J.; Elliott, A. & Sargatal, J. (eds.). Handbook of the birds Austral Ecology, 27: 132-136. of the world. Vol. 5. Barn-owls to hummingbirds. Barcelona: Lynx Pople, R. G. 2003. The ecology and conservation of the White-winged Edicions. Nightjar Caprimulgus candicans. Ph.D. thesis. Cambridge, UK: Cooper, R. J. 1981. Relative abundance of Georgia caprimulgids University of Cambridge. Available online from: http://www. based on call-counts. Wilson Bulletin, 93: 363-371. dspace.cam.ac.uk/handle/1810/225237. Cramp, S. (ed.). 1985. The birds of the western Palearctic. Vol. IV. Pruett-Jones, M. A. & Pruett-Jones, S. G. 1982. Spacing and Oxford: Oxford University Press. distribution of bowers in Macgregor’s Bowerbird (Amblyornis Donnelly, K. P. 1978. Simulations to determine the variance and macgregoriae). Behavioral Ecology and Sociobiology, 11: 25-32. edge effect of total nearest-neighbour distance, p. 91-95. In: Prum, R. O. 1998. Sexual selection and the evolution of mechanical Hodder, I. (ed.). Simulation studies in archaeology. Cambridge, sound production in manakins (Aves: Pipridae). Animal Behaviour, UK: Cambridge University Press. 55: 977-994. ESRI. 1999. ArcView 3.2 GIS software. Redlands: Environmental Ranft, R. & Cleere, N. 1998. A sound guide to nightjars and related Systems Research Institute. Revista Brasileira de Ornitologia, 22(2), 2014 Breeding biology of the White-winged Nightjar (Eleothreptus candicans) in eastern Paraguay Robert G. Pople nightbirds. Robertsbridge: Pica Press and British Library National Sutherland, W. J. 1998. The importance of behavioural studies in Sound Archive. conservation biology. Animal Behaviour, 56: 801-809. Rice, W. R. 1989. Analyzing tables of statistical tests. Evolution, 43: Théry, M. 1992. The evolution of leks through female choice: 223-225. differential clustering and space utilization in six sympatric Rodrigues, F. H. G.; Hass, A.; Marini-Filho, O. J.; Guimarães, manakins. Behavioural Ecology and Sociobiology, 30: 227-237. M. M. & Bagno, M. A. 1999. A new record of White-winged Vilella, F. J. 1995. Reproductive ecology and behaviour of the Nightjar Caprimulgus candicans in Emas National Park, Goiás, Puerto Rican Nightjar Caprimulgus noctitherus. Bird Conservation Brazil. Cotinga, 11: 83-85. International, 5: 349-366. Sick, H. 1993. Birds in Brazil: a natural history. Princeton: Princeton Vos, P.; Hogers, R.; Bleeker, M.; Reijans, M.; Lee, T.; Hornes, M.; University Press. Frijters, A.; Pot, J.; Peleman, J.; Kuiper, M. & Zabeau, M. Snow, B. K. & Snow, D. W. 1979. The Ochre-bellied Flycatcher and 1995. AFLP: a new technique for DNA fingerprinting. Nucleic the evolution of lek behavior. Condor, 81: 286-292. Acids Research, 23: 4407-4414. Sokal, R. R. & Rohlf, F. J. 1995. Biometry: the principles and practice Wilson, M. D. & Watts, B. D. 2006. Effect of moonlight on detection of statistics in biological research. New York: W. H. Freeman & Co. of Whip-poor-wills: implications for long-term monitoring Straneck, R. J. & Viñas, M. J. 1994. Comentarios sobre costumbres strategies. Journal of Field Ornithology, 77: 207-211. y manifestaciones acústicas del Atajacaminos de los Pantanos, Eleothreptus anomalus (Gould, 1838) (Aves, Caprimulgidae). Notulas Faunisticas, 67: 1-4. Associate Editor: Luciano N. Naka Revista Brasileira de Ornitologia, 22(2), 2014

Journal

Ornithology ResearchSpringer Journals

Published: Jun 1, 2014

Keywords: Caprimulgidae; Caprimulgus candicans; cerrado; male display behaviour; nesting biology

References