Disparate data streams together yield novel survival estimates of Alaska‐breeding Whimbrels.

Survival estimates are critical components of avian ecology. In well‐intentioned efforts to maximize the utility of one's research, survival estimates often derive from data that were not originally collected for survival assessments, and such post hoc analyses may include unintentional biases. We estimated the survival of Whimbrels captured and marked at two breeding sites in Alaska using divergent data streams that in isolation were subject to methodological biases. Although both capture sites were chosen to study the migration ecology of Alaska‐breeding Whimbrels, maximizing the conservation value of the data we collected was obviously desirable. We used multi‐year telemetry information to infer survival from one site (Colville River) and mark–resight techniques to estimate survival from a second site (Kanuti River). At Colville River, we could not feasibly include a control group of birds to assess potential survival effects of externally mounted transmitters, and at Kanuti River we were unable to account accurately for potential emigration events because we used resightings alone. We integrated these datasets in a Bayesian hierarchical framework, an approach that permitted insights across sites that moderated methodological biases within sites. Using telemetry enabled us to detect permanent emigration events from breeding sites in two of 10 birds, results that informed estimates for birds without tracking devices. These datasets yielded point estimates of true survival of Whimbrels from Colville River equipped with solar‐powered satellite transmitters that were higher (0.83) than true survival estimates of Whimbrels from Kanuti River marked with leg flags alone (0.74) or equipped with surgically implanted satellite transmitters (0.50), but the 95% credible intervals on these estimates overlapped across groups. For species such as Whimbrels that are difficult and costly to study, combining information from disparate data streams allowed us to derive novel demographic estimates, an approach with clear application to other similar studies. [ABSTRACT FROM AUTHOR]