Nest predation is the most frequent cause of nest failure in birds such as the Red-winged Blackbird (Agelaius phoeniceus) that nest on or near the substrate. Nestlings should therefore exhibit adaptations to reduce the risk of nest predation. We tested the nestling antipredator hypothesis by examining the begging responses of Red-winged Blackbird nestlings to vocalizations of: (1) an important nest predator (American Crow, Corvus brachyrhynchos), (2) a predator that rarely preys on nestlings (Cooper’s Hawk, Accipiter cooperii), and (3) a nonpredator (Northern Flicker, Colaptes auratus). We performed playbacks with: (1) both parents present at the nest, (2) male at the nest, and (3) neither parent present. Following playback, we measured duration of nestling begging after the parent departed (begging persistence), bouts of otherwise normal begging when no parent was present (parent-absent begging), and calling without postural components of begging (nonpostural begging). When the male or both parents were present during playback, adults responded with alarm calls and nestlings significantly reduced parent-absent begging following American Crow and Cooper’s Hawk playbacks. Nonpostural begging was significantly reduced following Cooper’s Hawk playback, but there were no significant differences in the other begging variables. When neither parent was present, we found no significant differences in nonpostural begging in response to the three playback types, but parent-absent begging was significantly reduced following American Crow and Cooper’s Hawk playbacks when compared to Northern Flicker playbacks. These results show that nestlings suppress their vocal begging in response to calls of predators including Cooper’s Hawks even though they are not common nest predators.

We conducted our playbacks at Newark Road Prairie (42^o32’ N, 89^o08’ W), 13 km northwest of Beloit in south-central Rock County, Wisconsin, USA. This site is a wet-mesic remnant prairie and sedge meadow habitat maintained by regular burning. Early in the breeding season we located nests by systematically searching likely locations where we consistently observed female activity. Later in the breeding season when vegetation was too dense to allow effective systematic searching, we found nests by observing females carrying nesting material and giving nest-associated vocalizations when visiting specific locations. Once nests were discovered, we marked their locations with vinyl flagging and checked their contents daily to determine whether they failed or produced fledglings. Playbacks were performed 16 May through 6 July 2016, 24 May through 21 July 2017, and 30 April through 13 July 2018. In 2016 and 2017 we did playbacks at active nests in locations that allowed us to video-record the nestlings. We performed playbacks between 0600 and 1400 hours CDT to nestlings 7–8 days after hatching. Previous studies have shown that Red-winged Blackbird nestlings of this age beg vigorously (Bernath-Plaisted and Yasukawa 2011). We obtained 10–13 samples of American Crow caws (Verbeek and Caffrey 2002), Cooper’s Hawk cak-cak-cak calls (Curtis et al. 2006), and Northern Flicker long calls (Wiebe and Moore 2017) from the website xeno-canto (http://www.xeno-canto.org) and edited each to 6–7 seconds of calling with Audacity v 2.2.0 (https://audacityteam.org/). We used an Ecogear ECOXBT (Grace Digital Audio, Peterborough, Ontario, Canada) speaker with a frequency response of 20–20000 Hz to perform playbacks. The three calls were broadcasted in random order at the speaker’s maximum volume of about 80–85 dB SPL at 1 m. The vocalizations were played from an iPhone (Apple, Inc., Cupertino, California, USA) connected to the speaker by a 30-m cable. Specific samples of each call were chosen using the shuffle feature of the iPhone. We performed playbacks under three conditions.

In Condition 1 (2016): both parents present, playbacks were conducted when the female was at the nest and the male was nearby. In Condition 2 (2017): male present, playbacks were performed with the male nearby as soon as the female left the territory after feeding the nestlings. An observer controlled the playback from a position 40 m from the nest. An observer at this distance had no obvious effect on male or female behavior throughout the duration of the playback. Playback sessions comprised at least six provisioning visits by the female (in one case the male provisioned the nestlings). Additional visits were inadvertently obtained at several nests when we were unable to see the female feeding the nestlings. For each of the playback types we first observed nestling responses following the parent’s visit without performing a playback (Control), and then after a visit with playback (Playback). In sessions with extra visits, we used observations from the control period immediately before the relevant playback period. Control and playback periods for each playback call type enabled us to use a matched-pairs design to control for among-nest differences such as brood size, sex ratio, and begging intensity. Video recordings of nestling response to playbacks at nests less than 1.5 m above the ground were made with a Sony HDR XR520 camera (Sony Corp. of America, New York, New, USA) mounted on a tripod placed within 1 m of the nest. For nests 1.5–2 m above the ground, a GoPro Hero+ (GoPro, Inc., San Mateo, California, USA) attached to a 2.1-m pole was placed next to the nest. We used the video recordings to note the times at which the female left the territory, returned to the territory, and fed the nestlings. We used these data to calculate the time between each visit and rates of begging behavior.

For Condition 3 (2018): no parent present, we removed a nestling that begged in response to light tapping on the rim of the nest and carried it to a vehicle parked at the study area. The nestling was placed in a nest collected earlier that breeding season and allowed 5 min to acclimate. We then played the three types of calls in random order and observed the response of the nestling for 5 min after each playback. This abbreviated schedule allowed us to return the nestling to its nest within 30 min and we observed no changes in provisioning or nest guarding by the parents and no cases of brood reduction in the days following nestling removal and replacement.

We used observations of nestling begging to test the nestling antipredator adaptation hypothesis. In Conditions 1 (both present) and 2 (male present), we used the video recordings to measure: (1) begging persistence, (2) parent-absent begging, and (3) nonpostural begging. Begging persistence was the duration of begging (s) once the female departed. As the intervals between feeding visits varied considerably, we calculated the rates (number/min) for parent-absent and nonpostural begging. We noted the number of bouts of parent-absent begging as well as the duration of each control and playback period. A “bout” of parent-absent begging was defined as a series of begging calls in quick succession in the absence of a care-giving adult; a silence of at least 3 s indicated the end of a bout of begging. We used the rate (number/min) of calling in the absence of any postural components of begging behavior as a measure of nonpostural begging. Nestlings typically remained motionless at the bottom of the nest during such calling. As both parents were not present in Condition 3, we only recorded the nonpostural begging rate and bout rate of parent-absent begging during the 5-min observation periods.