Behavioral Ecology Advance Access originally published online on February 7, 2008
Behavioral Ecology 2008 19(3):489-501; doi:10.1093/beheco/arn003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Parent–offspring communication in the western sandpiper
a Department of Biological Sciences b Department of Statistics, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA
Address correspondence to M. Johnson, who is now at the US Geological Survey, Forest and Rangeland Ecosystem Science Center, 3200 SW Jefferson Way, Corvallis, OR 97333, USA. E-mail: jedibirdnerd{at}yahoo.com
Received 26 July 2006; revised 13 June 2007; accepted 14 December 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGREFERENCES |
|---|
Western sandpiper (Calidris mauri) chicks are precocial and leave the nest shortly after hatch to forage independently. Chicks require thermoregulatory assistance from parents (brooding) for 5–7 days posthatch, and parents facilitate chick survival for 2–3 weeks posthatch by leading and defending chicks. Parental vocal signals are likely involved in protecting chicks from predators, preventing them from wandering away and becoming lost and leading them to good foraging locations. Using observational and experimental methods in the field, we describe and demonstrate the form and function of parent–chick communication in the western sandpiper. We document 4 distinct calls produced by parents that are apparently directed toward their chicks (brood, gather, alarm, and freeze calls). Through experimental playback of parental and non–parental vocalizations to chicks in a small arena, we demonstrated the following: 1) chicks respond to the alarm call by vocalizing relatively less often and moving away from the signal source, 2) chicks respond to the gather call by vocalizing relatively more often and moving toward the signal source, and 3) chicks respond to the freeze call by vocalizing relatively less often and crouching motionless on the substrate for extended periods of time. Chicks exhibited consistent directional movement and space use to parental and non–parental signals. Although fewer vocalizations were given in response to non–parental signals, which may indicate a weaker response to unfamiliar individuals, the relative number of chick calls given to each type of call signal was consistent between parental and non–parental signals. We also discovered 2 distinct chick vocalizations (chick-contact and chick-alarm calls) during arena playback experiments. Results indicate that sandpiper parents are able to elicit antipredatory chick behaviors and direct chick movement and vocalizations through vocal signals. Future study of parent–offspring communication should determine whether shorebird chicks exhibit parental recognition though vocalizations and the role of chick vocalizations in parental behavior.
Key words: Calidris mauri, call function, parental care, shorebird, vocal signals.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGREFERENCES |
|---|
The ecology of precocial birds is considerably more dynamic during the first weeks of life compared with altricial species. Precocial chicks generally depart the nest shortly after hatch (within a few hours to a couple of days) and forage independently (Clutton-Brock 1991
). Mound builders (Megapodiidae) are the most precocial of all birds with young that receive no posthatch parental interaction or care (Elliot 1994). Parental care among most other precocial birds (Anatidae, Charadriiformes, and Galliformes other than Megapodiidae) typically involves brooding offspring until they reach thermal independence (5–10 days), detecting and distracting predators, leading offspring to food, and in some species presenting or provisioning food to offspring (Carroll 1994; de Juana 1994; del Hoyo 1994; Martinez 1994; McGowan 1994; Porter 1994). In contrast, altricial chicks remain in the nest for longer periods of time (1–2 weeks typically) and are regularly provisioned by one or both parents (Gill 1995). Nests buffer altricial chicks from considerable environmental stochasticity compared with the diverse ecological settings experienced by precocial chicks.
Through behavior, organisms can effectively abate environmental heterogeneity (Brandon 1988), and reliable communication between parents and precocial offspring may increase the probability of offspring survival if parents are able to direct young to resources such as food and shelter and alert them to the presence of predators. Specific calls are given by parents in many precocial or semiprecocial avian species to indicate the presence of food or to attract offspring to a provisioning parent (Collias and Joos 1953; Tinbergen 1960; Lind 1965; Norton-Griffiths 1969; Evans 1970a, 1970b; Beer 1973; Buitron and Nuechterlein 1993). Laughing Gull (Larus atricilla) and Eurasian Oystercatcher (Haematopus ostralegus) parents use varying calls to indicate their distance from chicks (Norton-Griffiths 1969; Beer 1970b, Impekoven 1970), and parents of many bird species give alarm calls that elicit chicks to hide or freeze (Collias and Joos 1953; Tinbergen 1960; Lind 1965; Impekoven 1970, 1976; Beer 1973; Impekoven and Gold 1973). Chicks raised without parents respond differently to different calls in some species (Collias and Joos 1953; Snapp 1969), and young of many avian species can distinguish parental calls from those of other individuals (Tschanz 1968; Beer 1969, 1970a, 1970b; Evans 1970a, 1970b; Impekoven and Gold 1973). Chicks may even begin responding to parental calls before hatch, thereby facilitating rapid development of appropriate behaviors posthatch (Grier et al. 1967; Tschanz 1968; Impekoven 1970, 1976; Gottlieb 1971, 1988; Impekoven and Gold 1973; Heaton and Galleher 1981; Buitron and Nuechterlein 1993). Study of parent–offspring communication among precocial birds can provide insights to the evolution of animal communication systems in general through a better understanding of proximate and ultimate factors influencing production and perception of vocal signals and the responses they elicit. Previous study of parent–offspring communication among shorebirds (Charadriiformes, Charadrii) has primarily been descriptive/observational. This work documents that parents commonly give alarm calls when predators approach their brood (Walters 1982, 1990; Miller 1984, 1985, 1996; Sung et al. 2005), and vocal responses to predators can vary with predator class (mammal, bird, and reptile) and location (terrestrial or aerial, Walters 1990; Sung et al. 2005). Shorebird parents also vocalize when brooding offspring and give calls that appear to elicit chick movement toward the vocalizing adult (Baker 1982; Miller 1985; Sung et al. 2005). Parent or chick recognition has not been demonstrated in a shorebird species, but parental calls may exhibit sufficient variation for individual recognition (Baker 1982). Prenatal and nestling vocalizations have been described in a few plover (Charadrius) species (Tikhonov and Fokin 1979, 1980, 1981; Sung et al. 2005), and parental vocal signals are likely involved in foraging ontogeny in the Eurasian Oystercatcher (Norton-Griffiths 1969). However, hypothesized functions of shorebird parental and chick vocalizations have not been tested experimentally.
Shorebirds are an excellent group to study parent–offspring communication in precocial birds as they have a cosmopolitan distribution, breed in wide range of habitats, and exhibit every described mating system with parental care varying both within and among species. Predation rates on shorebird chicks is thought to be high (Norton 1973; Maher 1974; Safriel 1975; Walters 1982, 1990; Miller 1984; Lanctot and Laredo 1994; Johnson and Connors 1996; Nol et al. 1997; Handel and Gill 2000), and chick loss also may occur when individuals wander away and become lost (Evans and Pienkowski 1984; Gratto-Trevor 1991; Johnson 2002). Shorebird chicks have been observed traveling more than a kilometer in a single day and regularly travel several hundred meters within a 4-h period (Johnson and McCaffery 2004; Ruthrauff and McCaffery 2005). Parental knowledge of surrounding habitat (food and cover availability, predator densities, and activity levels) may influence brood movement and habitat use if parents are able to efficiently direct offspring movements and behavior. We expect selection to have favored a reliable communication system between shorebird parents and chicks to increase the probability that chicks avoid predators, locate areas with rich resources, and maintain brood cohesion during movement.
We describe several key elements of parent–offspring communication in a Nearctic breeding shorebird, the western sandpiper (Calidris mauri). Western sandpipers are small (25–30 g), highly migratory shorebirds that predominately breed in western Alaska and winter along the Pacific coast from California to Peru and the Atlantic coast from New Jersey to Surinam (Connors et al. 1979; Wilson 1994). Western sandpipers are socially and genetically monogamous (i.e., extrapair paternity is rare, Blomqvist et al. 2002) and exhibit biparental care of eggs and young; however, either sex, usually the female, may desert its mate and brood shortly after hatch (Holmes 1971, 1973). Females initiate 4-egg clutches from mid-May through late June that are incubated for approximately 21 days (Holmes 1971, 1973; Sandercock 1997). Precocial young leave the nest shortly after hatch and fledge at 13–16 days posthatch (Holmes 1971, 1973). Posthatch parental behaviors include brooding offspring until they reach thermoindependence (5–7 days) and leading and defending the brood.
We have noted at least 4 distinct calls, produced by parents, that are apparently directed toward their chicks. Parents gave the "brooding call" (Figure 1C) just prior to or while brooding offspring; chicks responded to the brooding call by moving directly to the vocalizing parent. The "gather call" (Figure 1A) was regularly given by attendant parents when they were not brooding offspring. Chicks responded to the gather call by moving toward the vocalizing parent. The "alarm call" (Figure 1B) was given when predators were in the immediate vicinity of or approaching the brood (10–50 m). Parents often alarm called within approximately 5–10 m of predators. Parents gave the "freeze call" (Figure 1D) when predators were directly beside or above chicks. Chicks responded to the freeze call by crouching on the substrate and remaining motionless.
|
Chick responses to parental calls just described are correlative and may result from cues other than parental vocalizations such as visual signals. We therefore experimentally isolated the effect of vocalizations using playback experiments in an arena setting to test hypotheses about the function of adult vocalizations. We hypothesized that 1 function of the gather call is to elicit chick movement toward the vocalizing parent. As parents often gave the alarm call relatively close to predators, we hypothesized that 1 function of the alarm call is to direct chick movement away from the vocalizing parent. We hypothesized that the freeze call functions to elicit hiding or concealment behavior in chicks, and we also predicted that both the alarm and freeze calls function to reduce the number of chick vocalizations, potentially offering auditory concealment from predators. To determine the generality of chick responses to adult vocalizations and control for any bias associated with individual recognition, we performed 2 experiments. Methods were identical in both experiments except that we presented chicks with parental call signals produced from their assumed parents in one and parental call signals produced from apparently nonrelated adults in the other. We also provide quantitative descriptions of these adult vocalizations and 2 chick vocalizations (chick-contact and chick-alarm calls, Figure 1E,F).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGREFERENCES |
|---|
Study site and general field methodology
We studied parent–offspring communication in the western sandpiper at the Yukon Delta National Wildlife Refuge's Kanaryarmiut Field Station, Yukon-Kuskokwim River Delta (YKD), Alaska (61°22'N, 165°07'W). On the YKD, western sandpipers inhabit upland tundra habitat that is typically a mosaic of patches, some of which contain graminoid species and some of which do not, with intermingled wet, low-lying areas comprised of sedge and grass (lowland moist low scrub community, Jorgenson and Ely 2001
; Johnson and McCaffery 2004). In all, 2–4 observers surveyed a 36-ha study plot daily from early May through late July for banded birds, nests, and broods (2004–2005). Adults and chicks were marked at the nest with a US Fish and Wildlife Service identification band as well as with unique UV-stable color band combinations, and the location and behavior of banded birds was recorded daily. Nest locations were mapped and nests were monitored through hatch, predation, or abandonment. After hatch, sandpiper parents and broods were resighted (brood location mapped and parent and chick behaviors recorded using scan sampling [Altmann 1974]) twice daily until they fledged, were depredated, or were abandoned.
Acquisition and description of parental vocalizations
During 2004–2005, we recorded parent–offspring vocalizations from 15 sandpiper families using a HHB minidisc recorder (MDP-500) and a Sennheiser microphone (ME 62) mounted in a 58.4-cm parabolic dish. All recordings of parental calls were made approximately 10–20 m from vocalizing birds, on calm days, with little to no extraneous background noise. We digitized recordings via RAVEN software (Cornell Lab of Ornithology) by sampling at 44.1 kHz with 16-bit accuracy. We also employed RAVEN software to produce spectrograms that were used to describe parental and chick vocalizations and to create playback signals for experimentation (frame length 512 points, Hanning window, time grid resolution 5.8 ms with 50% overlap, Fourier transformation size 512 points 3-db filter bandwidth 124 Hz).
To describe parental vocalizations, we visually inspected and compared all spectrograms. We then randomly selected 20 sandpiper parents (selection process occurred separately within each sex; 10 males, 10 females) from the 15 family units that were vocally sampled and extracted 10 examples (subsamples) of both the gather and alarm call for each parent. We used RAVEN software to measure the following call features and calculated means for each parent based on the 10 subsamples: delta time (difference between begin and end time, s), low frequency (lower frequency bound, Hz), high frequency (upper frequency bound, Hz), delta frequency (difference between low and high frequency, Hz), maximum frequency (frequency at which maximum power occurred, Hz), and maximum amplitude (dB).
The brooding call was difficult to record because of its low amplitude; further, parents appeared to hide themselves and their chicks from observers when brooding. Thus, attempting to record the brooding call caused disturbance of sandpiper families. Similarly, because parents only gave the freeze call when a threat to offspring was eminent, recording the freeze call either required prohibitive amounts of time or required us to approach the brood closely enough to be perceived as an eminent threat. We therefore only recorded the brooding call serendipitously (n = 2) and limited acquisition of the freeze call to 1 of the 2 parents from 12 broods (number of freeze calls per parent ranged from 1 to 6).
Arena experiment
To demonstrate parental vocalization function, we exposed 27 sandpiper chicks from 15 broods (1–4 chicks per brood) to parental vocalizations (gather, alarm, and freeze calls) in 2 experiments conducted on the field site between 19 June and 9 July 2004–2005. In 2004, we presented chicks (n = 13) with parental call signals produced from their assumed parents. In 2005, we replicated the experiment and presented chicks (n = 14) with parental call signals that were produced from apparently nonrelated adults. We used non–parental signals in 2005 to remove any bias associated with individual recognition and to determine the generality of parental call functions. We used previously recorded vocalizations to produce 141 call signals that we employed in both experiments (60 gather call, 60 alarm call, and 21 freeze call). In all, 8–12 signals were produced for each sandpiper family (2 alarm and 2 gather call signals for each parent and 0–4 freeze call signals per brood). We created individual signals for each parent by splicing together 2–3 of their previously recorded calls (0.1–0.6 s each) and repeating these calls for 30 s with 2 s of silence between successive calls. Parental and non–parental call signals were used only once during experiments to avoid pseudoreplication.
Chicks were removed from banded broods one at a time and individually tested in an open-bottom arena constructed of plastic tubing and tarpaulin (1.5 x 1.5 x 0.5 m), with a viewing window (60 cm2) cut into each of the arenas' walls. During playback experiments, the arena was positioned on level upland tundra habitat approximately 200 m away from the brood about to be tested. Two speakers were centered just outside the arena facing inward along 2 opposing walls, and we controlled both speakers from a position near a viewing window in 1 of the remaining 2 arena walls (Figure 2). Signal amplitude was standardized at 60 ± 3 dB at 4 m from the speaker during signal presentation simulating the typical volume of vocalizing parents. Each experimental subject (individual chick) was exposed to 4 consecutive parental call signals within 1 of 2 treatment groups that differed in the order of presentation of parental call signals (gather call–alarm call–gather call–freeze call or alarm call–gather call–alarm call–freeze call, example of former treatment group provided in Figure 2) then returned to its family unit. Six chicks were only exposed to 3 parental call signals during experimental trials in 2004 because we were unable to procure a freeze call from their respective parents. Experiments were conducted on calm days, chick age ranged between 7 and 10 days, and total handling time per chick was approximately 5 min.
|
When chicks were first placed in the experimental arena, they were kept under a cloth cylinder (12 x 20 cm) on top of a square piece of cardboard (361 cm2) at the center of the arena. The cloth cylinder did not keep chicks in total darkness but did prevent any visual bias during signal presentation. Two pieces of metal tubing were lashed perpendicular to each other to form a "+" and rested atop the arena. Metal tubing served as a pulley system to release chicks, via rope, from under the cloth cylinder. After 15 s of silence, while chicks were still under the cloth cylinder, we presented them with a randomly selected signal of either the parental alarm or gather call for 30 s from a randomly selected speaker (gather call, e.g., in Figure 2, trial 1). After 30 s of signal presentation, chicks were kept under the cloth cylinder for an additional 15 s of silence before releasing them. After release, we recorded the initial direction of chick movement away from the arena center and the location where chicks first made contact with an arena wall. During the first 30 s after chick release, we also recorded the number of chick vocalizations and time spent on the signal presentation half of the arena (Figure 2, shaded area trial 1). We then randomly selected and presented a signal of the call that was not used during trial 1 for 30 s from the opposite speaker (alarm call, e.g., in Figure 2, trial 2). During trial 2, we recorded the number of chick vocalizations and time spent on the signal presentation half of the arena (Figure 2, shaded area trial 2). We then initiated a third signal presentation (trial 3) broadcasting from the same speaker used in trial 2, but presenting a different signal of the call broadcast during trial 1 for 30 s (gather call, e.g., in Figure 2, trial 3). We recorded the number of chick vocalizations and time spent on the signal presentation half of the arena (Figure 2, shaded area trial 3). After trial 3, we presented chicks with a 30-s signal of the freeze call from a randomly selected speaker (Figure 2, trial 4). We recorded the number of chick vocalizations, number of times chicks crouched motionless on the substrate, and total amount of time chicks spend crouching on the substrate. Chicks were returned to their family units immediately after trial 4. No birds were harmed during this study, and experimentation did not appear to impact chick survival to independence (26/27 experimental subjects successfully fledged).
Acquisition and description of chick vocalizations
We captured some chick vocalizations while recording parental vocalizations; however, chicks predominately foraged during these recording sessions, and when chicks were vocalizing, parental calls regularly overlapped their vocalizations. Therefore, we made additional recordings of western sandpiper chicks during arena experiments by positioning a microphone just outside an arena wall (n2004 = 4, n2005 = 13). We digitized recordings of arena trials and produced spectrograms in the same manner previously described. We visually inspected and compared all spectrograms from arena trails to describe chick vocalizations. Through this process, we discovered 2 apparently distinct chick calls (chick-contact and chick-alarm calls, Figure 1E,F). Data from all recorded chicks (n = 17) were used to examine the relationship between chick call type (chick-contact, chick-alarm) and parental call signal (gather and alarm calls). Fourteen sandpiper chicks produced at least 10 calls of 1 type (chick-contact or chick-alarm) and were included in spectrogram analyses. We extracted 10 examples (subsamples) of either the chick-contact call or chick-alarm call from these 14 sandpiper chicks to describe chick vocalizations. We used RAVEN software to calculate the same call features that we examined during parental call analyses and calculated means for each chick based on the 10 subsamples. For consistency, we only included chick-alarm calls that contained harmonics when calculating call features.
Statistical methods
Describing parental and chick vocalizations
We performed factor analyses (SAS PROC FACTOR) to examine differences in mean call features (delta time, low frequency, high frequency, delta frequency, maximum frequency, and maximum amplitude) between parental (gather and alarm calls) and chick (chick-contact and chick-alarm calls) vocalizations (Sharma 1996). We used a principal component solution with orthogonal rotation (varimax) after factor extraction to obtain factor loadings. Residuals of mean call features were normally distributed, and data were checked for linearity among variables prior to analyses.
Initial direction of chick movement and initial contact with an arena wall
We used the Hodges–Ajne test to examine whether initial direction of movement away from the arena center was uniformly distributed for each treatment group because this test does not assume any specific underlying distribution (Hodges et al. 1955; Ajne 1968; Bhattacharyya and Johnson 1969). We tested whether initial direction of movement away from the arena center varied between the treatment groups (those presented with the alarm call signal first and those presented with the gather call signal first) and within each treatment group between the 2 experiments (2004, parental signals; 2005, non–parental signals) using Watson's 2-sample U2-test for nonparametric data with ties (Watson 1962). We performed the Batschelet test for circular uniformity (Batschelet 1981) using a 2-tailed binomial test (Zar 1999) to determine whether the initial direction of movement away from the arena center was concentrated near the signal source for subjects presented with the gather call signal first or concentrated along a trajectory opposite the signal source for those presented with the alarm call signal first. We also used a 2-tailed binomial test to determine whether the probability that subjects initially made contact with the arena wall nearest the signal source or the arena wall opposite the signal source was random or not.
Chick responses to the gather and alarm calls throughout each trial
In both experiments (2004, parental signals; 2005, non–parental signals), we presented each subject (chick) with 4 consecutive trials within 1 of 2 treatment groups (gather–alarm–gather–freeze call signals or alarm–gather–alarm–freeze call signals). For this analysis, we compared chick responses (time spent on signal presentation half of the arena and number of chick vocalizations) to the gather and alarm calls during the first 3 trials using a split-plot analysis of variance (ANOVA) with an augmented Latin rectangle design at the split-plot level (Cochran and Cox 1992; Littell et al. 2002). This was a completely randomized design at the individual chick level, with year (parental signals, non–parental signals) as a fixed effect. Each chick was 1 block of the Latin rectangle with the consecutive experimental trial being the other block. The design is an augmented Latin rectangle because there were 3 experimental trials with 2 treatment groups appearing in an alternating fashion.
Time spent on signal presentation half of the arena
We performed an analysis using SAS PROC MIXED to determine whether chicks presented with the gather call signal spent significantly more time out of each 30-s trial on the signal presentation half of the arena compared with chicks presented with the alarm call signal. This analysis compared chicks that initially heard the gather call with chicks that heard the gather call following an alarm call, and compared chicks initially hearing the alarm call and chicks hearing the alarm call after a gather call. We also contrasted time spent on the signal presentation half of the arena between the gather call and alarm call during each trial separately. The overall comparison between the gather and alarm calls was made using least-square means.
To determine whether chicks hearing the gather call spent significantly more time on the signal presentation half of the arena than chance and whether chicks hearing the alarm call spent significantly less time on the signal presentation half of the arena than chance (time spent on signal presentation half <15 s), 15 s was subtracted from the observed time chicks spent on the signal presentation half of the arena. Using this measure (time spent, 15 s), we repeated the analysis above. The tests of whether the least-square means were zero are then tests that determine whether time spent on the signal presentation half of the arena varied significantly from chance. The step-down Bonferroni adjustment was used to account for the multiple testing of the means.
Number of chick vocalizations
To determine whether the number of chick vocalizations varied between chicks hearing the gather and alarm call signal, we performed an analysis similar to the analysis of time spent on the presentation half of the arena. The only differences were that responses (number of chick vocalizations) were square root transformed to satisfy assumptions for the analysis. We report the back-transformed means and 95% confidence intervals for the gather and alarm call signal groups.
Variation in chick call type
We asked whether chick call type (chick-contact and chick-alarm calls, Figure 1E,F) varied with parental call signal presentation using a twice-nested ANOVA (split–split-plot design). This analysis was similar to those used for the amount of time chicks spent on the signal presentation half of the arena and the number of chick vocalizations (see above section, Chick responses to the gather and alarm calls throughout each trial) with an additional split level (chick call type) added to the analysis. The P values for the differences between means were adjusted using Tukey's adjustment.
Chick responses to the freeze call
We predicted that chicks would crouch on the substrate and remain relatively silent during presentation of the freeze call. After a square root transformation, we contrasted the number of chick vocalizations between the third experimental trial (either gather or alarm call signal presented) and the fourth experimental trial (freeze call signal) for both experiments (2004, parental signals; 2005, non–parental signals) using a paired-sample t-test (Zar 1999). We report the mean (±standard deviation [SD]) number of times chicks crouched motionless on the substrate during presentation of the freeze call and the mean (±SD) amount of time chicks spent crouching (out of 30 s).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGREFERENCES |
|---|
Parental and chick vocalizations
Spectrogram analyses indicated that call features differed among the 4 parental vocalizations and between the 2 chick vocalizations (Figure 1, Table 1). The first 2 factors that we extracted explained 87% and 97% of the variance in parental (gather and alarm calls) and chick (chick-contact and chick-alarm calls) vocalizations, respectively (Table 2). Factor 1 corresponded to call length, frequency bounds, difference between frequency bounds, and maximum amplitude, and factor 2 corresponded with the frequency at which maximum amplitude occurred (Table 2). During initial factor extraction for chick calls, maximum amplitude loaded on both factors. Therefore, we reran this analysis without maximum amplitude. The parental gather call (Figure 1A) is a trill that lacks harmonic content and modulates between 2.1 and 4.0 kHz, whereas the parental alarm call (Figure 1B) is a series of notes with several harmonics ranging in frequency from 1.6 to 18.5 kHz (Table 1, Figure 3). The alarm call is longer in duration than the gather call, with greater amplitude and a higher frequency at which maximum amplitude occurs (Table 1, Figure 3). From limited sampling (8 and 10 calls recorded from 2 broods, 1 male, 1 female), the brooding call (Figure 1C) appears similar in structure to the gather call. The brooding call is a trill of short duration compared with either the gather or alarm calls (0.1 ± 0.01 s) with lower amplitude (92 ± 0.9 dB) and a lower frequency at which maximum amplitude occurs (2.1 ± 0.05 kHz). The freeze call (Figure 1D) is a single note containing 3 harmonics with a frequency range and maximum amplitude similar to the alarm call (delta time = 0.14 ± 0.01 s, low frequency = 1,580 ± 62 Hz, high frequency = 18,891 ± 144 Hz, delta frequency = 17,311 ± 179 Hz, maximum frequency = 3,962 ± 39 Hz, maximum amplitude = 122 ± 0.09 dB, n = 12 individuals). The chick-contact call (Figure 1E) is a trill modulating between 4.3 and 5.7 kHz, whereas the chick-alarm call (Figure 1F) is a single note that may completely lack harmonic components or contain 1 or 3 harmonics ranging in frequency from 2.3 to 11.6 kHz (Table 1, Figure 3). The chick-alarm call is shorter in duration than the chick-contact call and occurs at higher frequencies with a greater difference between low and high frequency (Table 1, Figure 3).
|
|
|
Initial direction of chick movement and initial contact with an arena wall
Our data met the minimum sample necessary (n = 8) to test for circular uniformity at the 10% significance level for only 1 call type in either year (2004: ngather call = 7, nalarm call = 6; 2005: ngather call = 6, nalarm call = 8). We therefore used Watson's 2-sample U2-tests to compare chick responses to the gather and alarm call between years. We were unable to reject the null hypotheses that initial chick movement did not vary between years (gather call, U7,62 = 0.033, P > 0.5; alarm call, U6,82 = 0.066, P > 0.5). We then tested for circular uniformity for both years combined and found that initial direction of movement away from the arena center was not uniformly distributed (Figure 4; gather call, m0.05,13 = 1, 0.10 > P
0.05; alarm call, m0.05,14 = 1, 0.10 > P 0.05).
|
We rejected the null hypothesis that initial chick movement between the 2 treatment groups (gather vs. alarm call) was from the same population or 2 populations having the same direction (2004, U7,62 = 0.318, P < 0.01; 2005, U6,82 = 0.345, P < 0.005; Figure 4). Chicks presented with the gather call signal concentrated initial movement toward the signal source (2004–2005, C0.05(2),13 = 1, 0.05 > P
0.02), whereas initial movement was concentrated along a trajectory opposite the signal source among chicks presented with the alarm call signal (2004–2005, C0.05(2),14 = 1, 0.02 > P 0.01). Chicks presented with the gather call signal made initial contact with the arena wall nearest the signal source more often than chance (2004, C0.05(2),7 = 0, 0.05 > P 0.02; 2005, C0.05(2),6 = 0, 0.10 > P 0.05; Figure 5), and chicks presented with the alarm call signal made initial contact with the arena wall opposite the signal source more often than chance (2004, C0.05(2),6 = 0, 0.10 > P 0.05; 2005, C0.05(2),8 = 0, 0.05 > P 0.02; Figure 5).
|
Time spent on signal presentation half of the arena
There was a significant effect of parental call signal type (gather call vs. alarm call) on the amount of time chicks spent on the signal presentation half of the arena during the first 3 experimental trials (Figure 6). Chicks hearing the gather call signal spent twice as much time on the presentation half of the arena compared with chicks hearing the alarm call signal (t1,50 = 9.03, P < 0.001; time spent on signal presentation half of arena averaged across all 3 trials, gather call signal = 20.6 ± 0.4 s, alarm call signal = 10.3 ± 0.5 s). This pattern was the same in both experiments (t24 = –0.27, P > 0.79; 2004, parental signals and 2005, non–parental signals; Figure 6) and consistent across all 3 trials (trial 1, t50 = –6.3, P < 0.001; trial 2, t50 = 5.1, P < 0.001; trial 3, t50 = –4.2, P = 0.001). We found no evidence for presentation sequence effect as the amount of time chicks spent on the signal presentation half of the arena was similar across all 3 trials for both treatment groups (all P > 0.8; i.e., gather call [trial 1] = gather call [trial 2] = gather call [trial 3], and alarm call [trial 1] = alarm call [trial 2] = alarm call [trial 3]). Chicks hearing the gather call spent significantly more time on the signal presentation half of the arena than chance (t76 = 6.9, P < 0.001, time on signal presentation half >15 s), and chicks hearing the alarm call spent significantly less time on the signal presentation half of the arena than chance (t76 = –5.8, P < 0.001, time spent on signal presentation half <15 s).
|
Number of chick vocalizations
There was a significant effect of parental call signal type (gather call vs. alarm call) on the number of chick vocalizations during the first 3 trials (Figure 6). The number of chick vocalizations varied as a function of year (2004, parental signals; 2005, non–parental signals), treatment group (gather–alarm–gather–freeze call signals or alarm–gather–alarm–freeze call signals), and trial number, and there were significant interactions between year and treatment group and treatment group and trial number (Table 3). Chicks hearing the gather call signal vocalized more often than chicks hearing the alarm call signal (t1,49 = –6.37, P < 0.0001; number of chick vocalizations averaged across all 3 trials, gather call signal = 22.1 ± 3.0, alarm call signal = 12.5 ± 2.7). This pattern was significant during the first 2 trials (trial 1, t49 = –3.8, P = 0.006; trial 2, t49 = –3.6, P = 0.01; Figure 6); however, by the third trial chicks produced 17–21 vocalizations regardless of signal presentation (t49 = –1.1, P = 0.88).
|
We found no evidence for presentation sequence effect for the gather call signal, as the number of chick vocalizations given was similar across all three trials (all P > 0.2; i.e., gather call [trial 1] = gather call [trial 2] = gather call [trial 3]). However, chicks presented with the alarm call signal in trial 3 produced more vocalizations compared with chicks hearing the alarm call in trial 1 (t49 = –5.8, P < 0.001). The number of vocalizations given in response to the alarm call signal was not significantly different between trials 2 and 3 (t49 = –1.0, P = 0.9), and there also was no difference in the number of vocalizations given in response to the alarm call signal between trials 1 and 2 (t49 = –2.5, P = 0.16; Figure 6). The number of vocalizations given varied between years with chicks vocalizing less during 2005 when non–parental signals were presented compared with 2004 when parental signals were presented (t24 = –4.0, P < 0.001; total number of chick vocalizations averaged across all 3 trials, 2004, 24.2 ± 14.7; 2005, 10.5 ± 7.1).
Variation in chick call type
There was a significant effect of parental call signal (gather call vs. alarm call) on the number of each type of chick vocalization (chick-contact, chick-alarm) given during the first 3 experimental trials (t31 = –3.03, P = 0.005; Figure 7). The number of each type of chick vocalization varied as a function of year (2004, parental signals; 2005, non–parental signals) and treatment group (gather–alarm–gather–freeze call signals or alarm–gather–alarm–freeze call signals), and there were significant interactions between call type and year and call type and trial number (Table 3). The chick-alarm call was given more often during 2004 when parental signals were presented compared with 2005 when non–parental signals were presented (t1,47 = 4.93, P < 0.001; mean number of calls per chick, 2004, 39.8 ± 7.6; 2005, 14.5 ± 3.5); however, there was no difference in the number of chick-contact calls given between years (t1,47 = –0.45, P = 0.97). The chick-alarm call was given more often during the second and third trials compared with the chick-contact call (trial 2, t1,47 = 5.9, P < 0.001, chick-alarm = 13.3 ± 3.0, chick-contact = 1.8 ± 0.8; trial 3, t1,47 = 6.6, p < 0.001, chick-alarm = 14.2 ± 3.1, chick-contact = 1.5 ± 0.7). There was no significant difference in the number of each call type given during the first trail (t1,47 = –0.82, P = 0.97), but the chick-contact call was given more often during the first trial compared with the second or third trials (trial 1 vs. trial 2, t1,47 = 3.1, P = 0.046; trial 1 vs. trial 3, t1,47 = 3.28, P = 0.022; mean number of chick-contact calls per chick, trial 1 = 7.4 ± 1.9). During the first trial, the chick-contact call was predominately given after presentation of the gather call signal (number calls postgather call signal = 11.5 ± 2.6, number of calls postalarm call signal = 3.5 ± 2.4, Figure 7).
|
Chick responses to the freeze call
Chicks always responded to the freeze call by squatting motionless on the substrate (mean ± SD number of time chicks squatted in the substrate = 2.1 ± 1.1, n = 21). The amount of time chicks spent motionless on the substrate during presentation of the freeze call varied (mean ± SD amount of time spent motionless of 30 s = 14.0 ± 8.1, range 2–30 s), yet 71% of chicks were motionless for more than a third of the time when presented with the freeze call. In contrast, chicks never squatted motionless on the substrate when presented with a gather or alarm call signal. Compared with the preceding trial, chicks vocalized less often when presented with the freeze call (2004, P < 0.001; 2005, P < 0.001; mean ± SD number of vocalizations during freeze call trial and preceding trial, respectively, 4.8 ± 4.2, 18.2 ± 12.9).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGREFERENCES |
|---|
Parental care of precocial offspring, in the form of brooding, is vital to chick survival during the first days posthatch. Continued care through the attainment of flight likely increases the probability of offspring survival, and parental vocalizations appear to play a key role during extended brood attendance. Precocial chicks experience diverse ecological settings as they regularly travel several hundred meters a day, but parents may ameliorate some environmental stochasticity by directing chick movements and habitat use and through predator detection/distraction. Several factors likely influence brood movement and habitat use (prey availability, thermoregulatory environment, predator distribution, and vegetative cover), but the relative importance of each factor is unclear and factors explaining brood behavior are not mutually exclusive (Johnson and McCaffery 2004
). Regardless of the proximate causes underlying brood movement and habitat use, our study indicates that western sandpiper parents, through vocal signals, are able to elicit antipredatory chick behaviors and direct chick movement and vocalizations. Contextual use and quantitative descriptions of parental and chick vocalizations have been reported for several shorebird species; however, functions of specific vocalizations have not previously been demonstrated. We quantitatively described several parental and chick vocalizations given by western sandpipers during brood rearing and experimentally tested hypothesized parental call functions and found chick behavioral responses consistent with proposed call functions.
Gather, alarm, and freeze calls
Our results indicate that the function of the gather call is to elicit chick movement toward the vocalizing parent and that the alarm call functions to direct chick movement away from the vocalizing parent. The gather and alarm calls elicited consistent directional chick movement during arena experiments. Initial movement and initial contact with an arena wall were along the same trajectory for both the gather and alarm calls (Figures 4 and 5). Chicks hearing the gather call initially moved toward the signal source and initially made contact with the arena wall nearest the signal source. Concomitantly, chicks hearing the alarm call initially moved away from the signal source and initially made contact with the arena wall opposite the signal source. Directional movement may function to move chicks toward an area rich in resources (e.g., high food availability or vegetative cover) or away from exposed sites or predators.
Similar to their opposing effects on chick movement, the gather and alarm calls elicited contrary effects on chick vocal behavior. Chicks hearing the gather call were consistently more vocal compared with chicks hearing the alarm call (Figure 6). A duel response by chicks to the gather call, directional movement and increased vocalizations, is advantageous if chick vocalizations assist parents in locating individual offspring and facilitate gathering the brood. The alarm call appears to function not only to elicit directional movement of chicks away from the signal source but also to reduce the number of chick vocalizations, potentially offering auditory concealment from predators while retreating from the area. Chicks are probably not the only intended receiver of the alarm call. Parents increase their ability to attract a predator's attention by calling while moving toward it, and sandpiper parents typically attempt to lead predators away from the brood while giving the alarm call in conjunction with a variety of injury-feigning and distraction displays (Brown 1962; Holmes 1973).
Precocial avian offspring are known to exhibit directional movement toward a vocalizing parent. In many Galliformes and waterfowl (Anatidae) species, female vocalizations (e.g., assembly, exodus, family, guiding, and leading calls) attract offspring (Carroll 1994; de Juana 1994; del Hoyo 1994; Martinez 1994; McGowan 1994; Porter 1994). However, data on parental calls eliciting directional movement among shorebirds are more limited. Most shorebird parents are known to give contact calls, which are suspected to facilitate brood cohesion (Cramp and Simmons 1983), but the functions of these calls are ambiguous. Specific parental vocalizations given just prior to and during brooding appear to elicit chick movement to the vocalizing parent (Baker 1982; Miller 1985; Sung et al. 2005), and Eurasian oystercatcher parents appear to use provisioning calls to lead chicks from the nest (Norton-Griffiths 1969). Observational study near the nest suggests that piping plover (Charadrius melodus) parents give calls that function to elicit directional movement in chicks (A9 and A10 calls, Sung et al. 2005); however, these calls also were used in other contexts, throughout the breeding season, when chicks were not present. In contrast, western sandpiper parents were only observed giving the gather call in the presence of chicks.
Parents of many bird species, including shorebirds, give alarm calls that elicit chicks to hide or freeze (Miller 1985; Walters 1990; Sung et al. 2005). Although call function has not previously been demonstrated among shorebirds, it is reasonable to assume that parental alarm calls serve to signal the presence of predators and likely increases the probability that chicks avoid predators. However, western sandpiper chicks never responded to the alarm call by hiding or freezing. Rather, what we termed the western sandpiper alarm call elicits directional movement away from the signal source coupled with a reduction in chick vocalizations. This is in contrast to the freeze call which does function to elicit hiding or concealment behavior in western sandpiper chicks. Chicks always squatted motionless on the substrate when presented with the freeze call in arena experiments, and chicks also vocalized relatively less often during presentation of the freeze call signal compared with the gather or alarm call signals. Calidridine sandpipers are known to give 2 or more types of alarm calls while performing predator distraction displays (Cramp and Simmons 1983; Miller 1985; Sung et al. 2005), and observational study of southern lapwings (Vanellus chilensis) reported that chicks responded to one parental alarm call by crouching motionless on the substrate while a different alarm call elicited evasive behavior in chicks (Walters 1990). We have demonstrated that western sandpiper chicks exhibit differential responses to distinct parental alarm calls.
Parental use of the alarm and freeze calls
Behavioral studies indicate that western sandpiper parents exercise context specific vocal responses to predators (Johnson M, personal observation). Parents gave the freeze call when predators were on the ground within approximately 10 m of a chick or when aerial predators flew within approximately 200 m of the brood. When predators were observed on the ground at greater distances from a brood, parents flew toward the predator and gave the alarm call. Parents appeared to give the alarm call when chicks could avoid predator detection by remaining relatively quiet and moving away from the predator. As previously mentioned, parental distraction displays and predator leading behaviors given in conjunction with the alarm call may increase the probability that offspring are able to move away from a predator's location and avoid detection. In contrast, when the best method of avoiding predators was to hide and remain silent, parents typically gave the freeze call before squatting motionless on the tundra themselves. Chick responses to the alarm and freeze calls during arena experiments support these observations.
Many bird, rodent, and primate species exhibit antipredator signaling systems with distinct alarm calls that encode semantic information on predator attributes (referential signaling, Cheney and Seyfarth 1988) and/or provide information on the level of threat perceived by the signaler (response urgency signaling, Blumstein 1995). Functionally referential alarm call signals exhibit production specificity (i.e., species or class specific), and in the absence of the stimulus that normally elicited them, alarms call signals produce appropriate antipredator responses in individuals hearing them (Blumstein 1999). Alarm call signaling systems also may exhibit aspects of both referential and response urgency-based signaling. The black-capped chickadee (Poecile atricapilla) gives one call to denote a predator in flight and another to indicate a stationary predator with variation in the latter reflecting the size, or perceived threat, of the predator. Although we lack quantitative data on the stimuli that elicit western sandpiper alarm and freeze call signals, field observations indicate that western sandpiper parents employ response urgency-based signaling. Assessment of risk is the most important component of an animal's antipredator behavior (Lima and Dill 1990), and western sandpiper parents are able to reduce context-specific predation threats to offspring by directing their chicks to take immediate evasive action (freeze call) or a more gradual escape (alarm call) when either is most advantageous.
Parental versus non–parental signals
Total number of calls given by chicks varied between the 2 arena experiments with chicks vocalizing less during 2005 when non–parental signals were used. The reduction in chick vocal responses to non–parental call signals may indicate a weaker response to unfamiliar individuals. However, chicks did exhibit consistent directional movement and space use during both experiments, and although fewer vocalizations were given in response to non–parental signals, the relative number of chick calls given to each type of parental call signal was consistent across years. Young of many avian species can distinguish parental calls from calls of other individuals (Tschanz 1968; Beer 1969, 1970a, 1970b, Evans 1970a, 1970b, Impekoven and Gold 1973), and our results suggest this possibility for western sandpipers. However, as recently hatched chicks responded similarly to playbacks of parental and non–parental signals, chick responses are likely innate with brood cohesion/isolation further maintained through adult behaviors toward conspecifics. Western sandpiper parents were regularly observed chasing adult conspecifics away from their brood (35% of behavioral observations in 2005), and parents always attacked (pecked and threw) chicks from neighboring broods when they approached within approximately 20 m of their own offspring (4% of behavioral observations in 2005, n = 381; Johnson M, unpublished data). Attacks on unrelated chicks never resulted in chick death; rather, 2–10 s after an attack was initiated, 1 of the chicks' parents arrived and led the chick away from the attacker and her/his brood. Our experiment was designed to determine the generality of chick responses to adult vocalizations and control for any bias associated with individual recognition but not to test for parental or individual recognition. Future study of western sandpiper parent–offspring communication should determine whether chicks exhibit parental recognition though vocalizations and if so to what extent parental recognition is learned.
Chick vocalizations
The type of call given by chicks (chick-contact or chick-alarm, Figure 1E,F) during arena experiments was contingent on whether a parental call signal was being broadcast at the time and which parental call signal (gather or alarm call) was previously broadcast. The chick-contact call was predominately given during the first experimental trial after subjects were presented with a gather call signal (Figure 7). No parental call signals were being broadcast during this time period. In contrast, chicks rarely gave the chick-contact call during the initial trial after being presented with an alarm call signal. During the second and third experimental trials when parental call signals were being broadcast, the chick-alarm call was commonly given and the chick-contact call was rarely given regardless of parental signal type. We labeled parental calls based on demonstrated functions, but we refer to chick calls by names based on the context when each call was given and do not imply function.
We noted the chick-contact call only during arena experiments. The chick-alarm call also was observed in the field, specifically when non–parental adults chased chicks from their nest site or brood location or when observers inadvertently separated a brood while walking the study plot. Parents responded to the chick-alarm call by approaching the vocalizing chick. The chick-alarm call may represent a graded signal than takes the form of a single note lacking a harmonic component or a note with one or more harmonics (Figure 1F). It is doubtful that the presence or absence of harmonics noted in the chick-alarm call is an artifact of signal intensity or distance from the subject during recordings as recordings were made when ambient noise was minimal and audio quality was high. Further, subjects were never more than 2.1 m from the microphone during arena experiments and generally <10 m from the microphone during in situ recordings. We hypothesize the addition of harmonics to the chick-alarm call indicates increased disturbance to the chick. This hypothesis is based on anecdotal observation: Chicks commonly gave a single note version of the chick-alarm call when a brood was mildly disturbed, such as when an observer stood relatively close to a brood. However, when an observer chased an individual chick to capture it, chick-alarm calls were produced with one or more harmonics.
Physical structure of vocalizations
The physical structure of a vocal signal affects a receivers' ability to locate its source whether the receiver is intended or not. Typically calls used to attract or locate other individuals are composed of short notes with a broad frequency range as broader frequency range within notes provides greater information concerning both direction and distance (Marler 1955). Mobbing calls are typically of short duration with a broad frequency band that is easy to locate and attracts other birds to the site. The structure of the chick-alarm call (Figure 1F), when containing harmonics, is similar to mobbing calls of other species and likely functions to draw the attention of a tending parent, and field observations support this hypothesis. We broadcast chick-alarm calls to parents actively tending broods during preliminary experimentation (n = 4) and found that nearly every adult within approximately 200 m responded by approaching within 10 m of the speaker. In fact, when broods were difficult to locate, we successfully used this method to draw a tending parent toward the speaker, thereby revealing its location (n = 3).
Physical structure also determines the distance a sound will travel and how much distortion it will sustain before reaching the receiver. Interference, absorption, and scattering of sound waves by vegetation, the ground, and air progressively distort a sound, with low frequency sounds traveling further than high-frequency sounds (Chappuis 1971; Morton 1975; Wiley and Richards 1982). The lower frequency bound of the chick-alarm call is nearly half that of the chick-contact call which would facilitate its use to attract parents from considerable distances. Sounds that are rich in temporal structure with complex frequency modulations are advantageous in open habitats because simple sustained notes tend to be distorted by temperature gradients and air turbulence (Chappuis 1971; Morton 1975). The chick-contact call (Figure 1E) does have considerable temporal structure and complex frequency modulation compared with the chick-alarm call but does not have the frequency range nor as extreme a lower frequency bound as the chick-alarm call. We hypothesize that the chick-contact serves in short-distance communication between parent and offspring. The chick-contact call may represent a compromise between providing too much information about an individuals' location that may be intercepted by unintended receivers (predators) but enough information for a parent to effectively locate its source over relatively short distances.
The physical structures of adult vocalizations also support their hypothesized functions. The parental alarm call (Figure 1B) is a series of short notes with a broad frequency range that is typical of calls used to attract or locate individuals (Marler 1955). Such qualities are in accord with call function if parents give the alarm call to attract/distract predators from the brood and elicit directional movement of chicks away from the parents' location. The parental gather call (Figure 1A) may represent a structural compromise similar to that described for the chick-contact call. If parents give the gather call to draw chicks toward their location, temporal structure and complex frequency modulation within the gather call would serve this purpose over relatively short distances without broadcasting the parent and brood location to unintended receivers at greater distances.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGREFERENCES |
|---|
Using observational and experimental methods in the field, we described and demonstrated the form and function of parent-chick communication in the western sandpiper. We documented 4 distinct calls produced by parents that are directed toward their chicks and potentially other receivers (brooding call, gather call, alarm call, and freeze call). We discussed how these calls may be used by parents to elicit antipredatory chick behaviors, and direct chick movement and vocalizations, and described 2 distinct chick vocalizations (chick-contact and chick-alarm calls). Future study of western sandpiper parent–offspring communication should determine whether chicks exhibit parental recognition though vocalizations and the role of chick vocalizations in parental behavior. The inexpensive field assay technique we developed will likely serve equally well for study of parent–offspring communication in other precocial and semiprecocial species, being less intrusive and potentially more realistic than typical laboratory study.
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGREFERENCES |
|---|
US Fish and Wildlife Service (Yukon Delta National Wildlife Refuge) and the Harold F. Bailey Fund at Virginia Polytechnic Institute and State University.
|
ACKNOWLEDGEMENTS |
|---|
We thank Brian McCaffery and the entire staff of the Yukon Delta National Wildlife Refuge for supporting this research. We also thank Jesse Conklin, Branden Locke, Peter Laver, and Lew Oring for assistance in the field and S. Haig, T. Jenssen, and R. Greenberg for comments on an early draft of this manuscript.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSFUNDINGREFERENCES |
|---|
Ajne B. A simple test for uniformity of a circular distribution. Biometrika (1968) 55:343–354.
Altmann J. Observational study of behavior: sampling methods. Behaviour (1974) 48:227–265.
Baker MC. Individuality of vocalizations in Dunlin: a possible acoustic basis for recognition of parent by offspring. Auk (1982) 99:771–774.[Web of Science]
Batschelet E. Circular statistics in biology. (1981) New York: Academic Press.
Beer CG. Laughing gull chicks: recognition of their parents' voices. Science (1969) 166:1030–1032.
Beer CG. Individual recognition of voice in the social behavior of birds. Adv Study Behav (1970a) 3:27–74.
Beer CG. On the response of laughing gull chicks (Larus atricilla) to the calls of adults II. Age changes and responses to different types of call. Anim Behav (1970b) 18:661–677.[CrossRef][Web of Science]
Beer CG. A view of birds. Minneapolis symposium of child psychology. (1973) Vol. 7. Minneapolis (MN): University of Minnesota Press. 47–86.
Bhattacharyya GK, Johnson RA. On Hodges's bivariate sign test and a test for uniformity of a circular distribution. Biometrika (1969) 56:446–449.
Blomqvist DB, Kempenaers RB, Lanctot BK, Sandercock BK. Genetic parentage and mate guarding in the arctic-breeding western sandpiper. Auk (2002) 119:228–233.[CrossRef][Web of Science]
Blumstein DT. Golden-marmot alarm call. I. The production of situationally specific vocalizations. Ethology (1995) 100:113–125.[Web of Science]
Blumstein DT. The evolution of functionally referential alarm communication. Evol Comm (1999) 3:135–147.
Brandon RN. The levels of selection: a hierarchy of interactions. In: The role of behavior in evolution—Plotkin HC, ed. (1988) Cambridge (MA): MIT Press. 51–71.
Brown RGB. The aggressive and distraction behaviour of the western sandpiper Ereunetes mauri. Ibis (1962) 104:1–12.[CrossRef]
Buitron D, Nuechterlein GL. Parent-young vocal communication in eared grebes. Behaviour (1993) 127:1–20.[CrossRef]
Carroll JP. Family Odontophoridae. In: Handbook of the birds of the world, vol. 2. Hoatzin to auks—del Hoyo J, Elliott A, Sargatal A, eds. (1994) Barcelona (Spain): Lynx Edicions. 412–433.
Chappuis C. Un exemple de l'influence du milieu sur les émissions vocals des oiseaux: L'évolution des chants en forêt équatoriale. Terre Vie (1971) 118:183–202.
Cheney DL, Seyfarth RM. Assessment of meaning and the detection of unreliable signals by vervet monkeys. Anim Behav (1988) 36:477–486.[CrossRef][Web of Science]
Clutton-Brock TH. The evolution of parental care. (1991) Princeton (NJ): Princeton University Press.
Cochran WG, Cox GM. Experimental designs (1992) Hoboken (NJ): John Wiley & Sons, Inc.
Collias NE, Joos M. The spectrographic analysis of sound signals of the domestic fowl. Behaviour (1953) 5:175–188.[Medline]
Connors PG, Myers JP, Pitelka FA. Seasonal habitat use by Arctic Alaska shorebirds. In: Behavior of marine animals, current perspectives in research, vol. 5—Burger J, Olla BL, eds. (1979) New York: Plenum Press. 101–111.
Cramp S, Simmons KEI. The birds of the Western Palearctic. In: Waders to gulls (1983) vol. 3. Oxford: Oxford University Press.
de Juana E. Family Tetraonidae. In: Handbook of the birds of the world, vol. 2. Hoatzin to auks—del Hoyo J, Elliott A, Sargatal A, eds. (1994) Barcelona (Spain): Lynx Edicions. 376–411.
del Hoyo J. Family Cracidae. In: Handbook of the birds of the world, vol. 2. Hoatzin to auks—del Hoyo J, Elliott A, Sargatal A, eds. (1994) Barcelona (Spain): Lynx Edicions. 310–363.
Elliot A. Family Megapodiidae. In: Handbook of the birds of the world, vol. 2. Hoatzin to auks—del Hoyo J, Elliott A, Sargatal A, eds. (1994) Barcelona (Spain): Lynx Edicions. 378–309.
Evans RM. Imprinting and mobility in young ring-billed gulls, Larus delawarensis. Anim Behav Monogr (1970a) 3:193–248.
Evans RM. Parental recognition and the "mew call‘ in black-billed gulls (Larus bulleri). Auk (1970b) 87:503–513.[Web of Science]
Evans PR, Pienkowski MW. Implications for coastal engineering projects of studies, at the Tees Estuary, on the effects of reclamation of intertidal land on shorebird populations. Water Sci Technol (1984) 16:347–354.
Gill FB. Ornithology. (1995) New York: WH Freeman and Company.
Gottlieb G. Development of species identification in birds. (1971) Chicago (IL): University of Chicago Press.
Gottlieb G. Development of species identification in ducklings: XV. Individual auditory recognition. Dev Psychobiol (1988) 21:509–522.[CrossRef][Medline]
Gratto-Trevor CL. Parental care in semipalmated sandpipers Calidris pusilla: brood desertion by females. Ibis (1991) 133:394–399.[CrossRef]
Grier JB, Counter SA, Shearer WM. Prenatal auditory imprinting in chickens. Science (1967) 155:1692–1693.
Handel CM, Gill RE. Mate fidelity and breeding site tenacity in a monogamous sandpiper, the black turnstone. Anim Behav (2000) 60:471–481.[CrossRef][Medline]
Heaton MB, Galleher E. Prenatal auditory discrimination in the bobwhite quail. Behav Neural Biol (1981) 31:242–246.[CrossRef][Medline]
Hodges JL Jr, Ramsey PH, Wechsler S. A bivariate sign test. Ann Math Stat (1955) 26:523–527.[CrossRef]
Holmes RT. Density, habitat and the mating system of the western sandpiper (Calidris mauri). Oecologia (1971) 7:191–208.[CrossRef][Web of Science]
Holmes RT. Social behavior of breeding western sandpipers Calidris mauri. Ibis (1973) 115:107–123.[CrossRef]
Impekoven M. Prenatal experience of parental calls and pecking in the laughing gull (Larus atricilla L.). Anim Behav (1970) 19:475–480.[CrossRef][Web of Science]
Impekoven M. Responses of laughing gull chicks (Larus atricilla) to parental attraction- and alarm-calls, and effects of prenatal auditory experience on the responsiveness to such calls. Behaviour (1976) 56:2150–278.
Impekoven M, Gold PS. Prenatal origins of parent-young interactions in birds: a naturalistic approach. Gottlieb G, ed. (1973) New York: Academic Press. 325–356.
Johnson M. Are nest exclosures an effective tool in plover conservation? and plover parental care when either parent deserts [dissertation]. (2002) Reno (NV): University of Nevada.
Johnson M, McCaffery BJ. Use of upland tundra habitats by western and rock sandpipers during brood-rearing on the Yukon-Kuskokwim Delta, Alaska. Wader Stud Group Bull (2004) 103:36–39.
Johnson OW, Connors PG. American golden-plover (Pluvialis dominica), Pacific golden-plover (Pluvialis fulva). In: The birds of North America, no. 201–202—Poole A, Gill F, eds. (1996) Philadelphia (PA). The Academy of Natural Sciences; Washington (DC): The American Ornithologists' Union.
Jorgenson T, Ely C. Topography and flooding of coastal ecosystems on the Yukon-Kuskokwim Delta, Alaska: implications for sea-level rise. J Coast Res (2001) 17:124–136.
Lanctot RB, Laredo CD. Buff-breasted Sandpiper (Tryngites subruficollis). In: The birds of North America, no. 91—Poole A, Gill F, eds. (1994) Philadelphia (PA): The Academy of Natural Sciences. Washington (DC): The American Ornithologists' Union.
Lima SL, Dill LM. Behavioral decisions made under the risk of predation: a review and prospectus. Can J Zool (1990) 68:619–640.
Lind H. Parental feeding in the oystercatcher (Haematopus o. ostralegus (L.). Dan Ornithol Foren Tidsskr (1965) 59:1–31.
Littell RC, Milliken GA, Stroup WW, Wolfinger RD. SAS system for mixed models. (2002) Cary (NC): SAS Institute Inc.
Maher WJ. Ecology of pomarine, parasitic, and long-tailed Jaegers in northern Alaska. Pac Coast Avifauna (1974) 37:103–124.
Marler P. Characteristics of some animal calls. Nature (1955) 176:6–7.[CrossRef]
Martinez I. Family Numididae. In: Handbook of the birds of the world, vol. 2. Hoatzin to auks—del Hoyo J, Elliott A, Sargatal A, eds. (1994) Barcelona (Spain): Lynx Edicions. 554–567.
McGowan PJK. Family Phasianidae. In: Handbook of the birds of the world—del Hoyo J, Elliott A, Sargatal A, eds. (1994) vol. 2. Hoatzin to auks. Barcelona (Spain): Lynx Edicions. 434–553.
Miller EH. Communication in breeding shorebirds. In: Behaviour of marine animals, current perspectives in research—Burger J, Olla BL, eds. (1984) vol. 5. New York: Plenum Press. 16–24.
Miller EH. Parental behavior in the least sandpiper (Calidris minutilla). Can J Zool (1985) 63:1593–1601.
Miller EH. Acoustic differentiation and speciation in shorebirds. In: Ecology and evolution of acoustic communication in birds—Kroodsma DE, Miller EH, eds. (1996) Ithaca (NY): Comstock Publishing Associates. 241–257.
Morton ES. Ecological sources of selection on avian sounds. Am Nat (1975) 109:17–34.[CrossRef][Web of Science]
Nol E, Sullivan Blanken M, Flynn L. Sources of variation in clutch size, egg size, and clutch completion dates of semipalmated plovers in Churchill, Manitoba. Condor (1997) 99:389–396.
Norton DW. Ecological energetics of calidrine sandpipers breeding in northern Alaska [dissertation]. (1973) Fairbanks (AK): University of Alaska.
Norton-Griffiths M. The organization, control and development of parental feeding in the oystercatcher(Haematopus ostralegus). Behaviour (1969) 34:55–114.[CrossRef]
Porter WF. Family Meleagrididae. In: Behaviour of marine animals, current perspectives in research—Burger J, Olla BL, eds. (1994) Vol. 5. New York: Plenum Press. 364–375.
Ruthrauff DR, McCaffery BJ. Survival of western sandpiper broods on the Yukon-Kuskokwim Delta, Alaska. Condor (2005) 107:597–604.
Safriel UN. On the significance of clutch size in nidifugous birds. Ecology (1975) 56:703–708.[CrossRef][Web of Science]
Sandercock BK. Factors affecting the breeding demography of western sandpipers (Calidris mauri) and semipalmated sandpipers (C. pusilla) at Nome, Alaska [dissertation]. (1997) Burnaby (Canada): Simon Fraser University.
Sharma S. Applied multivariate techniques (1996) Inc: New York: John Wiley and Sons.
Snapp BD. Recognition of maternal calls by parentally naïve Gallus gallus chicks. Anim Behav (1969) 17:440–445.[CrossRef][Web of Science]
Sung H-C, Miller EH, Flemming SP. Breeding vocalizations of the piping plover (Charadrius melodus): structure, diversity, and repertoire organization. Can J Zool (2005) 83:579–595.[CrossRef]
Tikhonov AV, Fokin SY. Acoustic signaling and behaviour of shorebirds in early ontogenesis. I. Prenatal stages of development. Biol Nauki (1979) 10:33–40.
Tikhonov AV, Fokin SY. Acoustic signaling and behaviour of shorebirds in early ontogenesis. II. Signalization and the behaviour of nestlings. Biol Nauki (1980) 10:45–54.
Tikhonov AV, Fokin SY. Acoustic signaling and behaviour during the nesting period in shorebirds. Byull Mosk O-va Ispyt Prir Otd Biol (1981) 86:31–42.
Tinbergen N. The herring gull's world. (1960) New York: Basic Books.
Tschanz B. Trottellummen. Z Tierpsychol Beih (1968) 4:1–103.
Walters JR. Parental behavior in lapwings (Charadriidae) and its relationships with clutch sizes and mating systems. Evolution (1982) 36:1030–1040.[CrossRef][Web of Science]
Walters JR. Anti-predatory behavior of lapwings: field evidence of discriminative abilities. Wilson Bull (1990) 102:49–70.
Watson GS. Goodness of fit test on a circle. II. Biometrika (1962) 49:57–63.
Wiley RH, Richards DG. Adaptations for acoustic communication in birds: sound transmission and signal detection. In: Acoustic communication in birds, vol. 1: production, perception, and design features of sounds—Kroodsma DE, Miller EH, eds. (1982) New York: Academic Press. 132–181.
Wilson WH. Western sandpiper (Calidris mauri). In: The birds of North America, no. 90—Poole A, Gill F, eds. (1994) Philadelphia (PA): The Academy of Natural Sciences. Washington (DC): The American Ornithologists' Union.
Zar JH. Biostatistical analysis. (1999) Englewood Cliffs (NJ): Prentice Hall.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||