Mechanisms driving breeding dispersal are complex but are of high interest because dispersal strongly links individual fitness to population dynamics. We examine the relative importance of personal information, neighborhood effects, and structural habitat characteristics in determining an individual’s propensity for breeding dispersal. We attempted to identify relevant cues for breeding dispersal of a North American territorial migratory bird species, the Black-capped Vireo. We color marked and radio-tagged males in Southwestern Oklahoma and used a conditional inference tree analysis to evaluate 11 variables that individuals could use as predictors of emigration. We used the correlation between arrival date and habitat structure to determine habitat preference. Breeding dispersal propensity among Black-capped Vireos depended mostly on their personal breeding experience, but also on reproductive information gleaned from their neighbors. Older and younger age classes that reproduced successfully did not disperse, but younger age class individuals that failed to reproduce were more likely to disperse than older individuals within the breeding season. Dispersal events among young males were significantly related to the proportion of their neighbors that successfully reproduced with more dispersal from neighborhoods of fewer, less successful neighbors. Vegetation structure within a territory was not identified as a significant cue.

We conducted our study on the Fort Sill Military Installation in Southwest Oklahoma, USA, from April to September in 2017 and 2018. The Fort Sill Military Installation, roughly 38,000 ha, with adjacent Wichita Mountain National Wildlife Refuge adding 23,885 ha total, with about 10,000 ha suitable vireo habitat (Diamond and Elliot 2015), forms a contiguous track of the Wichita Mountain ecoregion in Oklahoma. Two common vegetation structures in vireo habitat included patches of short stature (one to three meters in height) black-jack and post oaks (Quercus marilandica, Quercus stellata, respectively), skunkbush (Rhus triolobata), flame-leaf sumac (Rhus lanceolata), and tall stature oak woodland (3 to 10 meters in height) with hackberry (Celtis occidentalis) common in the understory (Diamond and Elliot 2015). During the two years of this study, Fort Sill and the wildlife refuge controlled for Brown-headed cowbirds, a frequent nest parasite of black-capped vireos, by removal through trapping in the spring and summer.

We selected two main study sites that represented these two dominant vegetation structures used by Black-capped Vireos. The 49-hectare Quanah study site consisted predominantly of tall stature oak woodland habitat. The Sherman study site was 42 hectares and located in the central parcel of Fort Sill and was predominantly short stature oak habitat. We attempted to monitor breeding activities all territorial males within the boundaries of the study sites during the breeding period from April to July. To achieve sufficient sample sizes of vireo age classes, we also monitored selected territorial males of a certain age class in the area adjacent to the Sherman study area boundary.

Field Methods

We attempted to capture and mark all territorial males within the monitored study areas with a USGS issued aluminum numerical band and unique combination of colored leg bands. We captured Black-capped Vireos using six-meter length, 30mm mesh mist-nets while broadcasting Black-capped Vireo songs, scolds, heterospecific scolds, and Eastern Screech Owl calls. We determined sex, age, wing length, and mass. Age is most accurately determined by feather wear between the greater and primary coverts, but there are a high proportion of SY males with gray napes compared to black napes typical of ASY males (Pyle 1997, Cimprich 2018). Plumage characteristics are also diagnostic for sex as Black-capped Vireos are the only members in their family that are sexually dimorphic, but we also checked for full brood patches for females. Males that we could not catch were aged as after hatching year (AHY), though we note that most of these males had fully black napes.

From the first week of April to the first week in May, we recorded arrival dates for territorial males by surveying the study sites at least four days in each seven-day period to detect new individuals. Observers spent 20 minutes in areas to detect individuals. We either visited both study sites in a single day or regularly alternated visiting sites to reduce effort bias.

We gathered presence/absence and location data of monitored males from April to July by visiting their territories once at least every three days. A male was considered territorial if we found the same individual within 25 meters of a previous location where it had been singing or displaying other acts of area defense for three consecutive visits. We followed territorial males and recorded their locations using GPS (Garmin Rhino 650). Observers followed individuals only as closely needed to sight leg bands; usually no closer than 10 meters. We marked the GPS location of the vireo only after the vireo had moved voluntarily to avoid biasing movements based on the observer. GPS locations were taken no less than five minutes apart with no more than five locations per day to capture the full size of the breeding territory while the vireos were actively nesting.

We monitored a total of 130 males in 2017 (Sherman n = 38, Quanah n = 36) and 2018 (Sherman n = 30, Quanah n = 26) within the study sites. This count includes birds of three age classes: SY, ASY, and AHY. We excluded the age class of AHY from our analyses because these birds could be either SY or ASY but had not been captured to confirm their age. Within the study sites, 15 of 130 (11.5 %) of the males were SY males. Eighteen additional territorial SY males were located and monitored outside of the main study sites. We sought out these additional 18 territories of SY males to increase our sample size of SY males.

Males were monitored for two weeks after nesting activity ceased in that territory, which included actively building nests or incubating behavior. Nests were located using behavioral cues. We considered a male successful if it produced at least one fledgling. Reproductive success differed between the sites, with 13.8 % more territories producing offspring on the Sherman site than on the Quanah site across both years (Table 1). Both Sherman and Quanah males experienced more success in 2017 than 2018. Compared to 2017, 27 % fewer males were successful at Sherman and 17 % fewer males were successful at Quanah in 2018. An early April frost in 2018 likely explains the difference in success between years. Budding oak leaves were killed and did not fully redevelop until late April, which, delayed nesting and also likely reduced time for subsequent nesting following failed attempts (PMC unpubs. data). We monitored 73 nests in 2017 and 83 nests in 2018 that had at least one egg. Nest failure was mainly due to predation, where all eggs or nestlings would be absent with the nest intact likely due to snake depredation (82 nests), or nests that had been removed from the shrub likely due to a mammalian predator (11 nests). Eight nests were abandoned with eggs while five nests had eggs destroyed by having holes poked in them. Cowbirds were controlled through trapping at Fort Sill and the Wichita Mountain Wildlife Refuge in both years, and we only found one nest parasitized on the study sites. We chose ordinal date 181 (June 30) as a cut off for the primary breeding season because our nest data indicate 98% (n = 172, includes nests in building stages) of nests were completed either as successful or failed, and territories were not initiating new nests.

Dispersal of territorial males during the breeding season was determined in two ways. First, we resighted color-banded males by searching territories at least once every three days to determine presence or absence. Second, we radio-tagged (JDJC corps, 260 mg) individuals (2017 n = 38: 7 ASY and 31 SY, 2018 n = 45: 11 ASY and 34 SY) beginning mid-May through July using a modified leg-loop harness method with an elastic thread degradable within 30 to 60 days (Rappole and Tipton 1991) in addition to colored leg bands and a USGS band. Not all transmitted males were territorial when initially caught. We located radio-tagged males using a hand-held three-element Yagi antenna and receiver (Model R4000, Advanced Telemetry Systems, Inc., Isanti, MN) and employed the homing method. Transmitters weighed < 4 % of an individual’s body mass and had a battery life that ranged from 28 to 50 days for radio-tagged males that we were able to relocate for the first three days. We checked for radio-tagged males every day until they disappeared, the radio tag fell off and was recovered, or the tag battery died. If the male was territorial, we mapped their territory with up to five locations in one that were no less than five minutes apart. If the male abandoned their territory , we attempted to relocate the radio-tagged bird at least twice to get a location point no less than two hours apart to determine if they returned, settled elsewhere, or kept moving.