Detection probability of the Pileated Woodpecker (Dryocopus pileatus): Implications for developing habitat use models

As old-growth forest ecosystems become increasingly scarce in North America, the need to accurately and efficiently survey, monitor, and model old-growth specialists and keystone species, such as the Pileated Woodpecker (Dryocopus pileatus), becomes increasingly important. Little is known about the behaviour and habitat associations of the Pileated Woodpecker at the northern edge of its range in Alberta, Canada. Attempts at modeling Pileated Woodpecker habitat use for this region by the Alberta Biodiversity Monitoring Institute (ABMI) have high uncertainty, particularly for vegetation types. One explanation is that the data used to build these habitat selection models comes from surveys done in June, which may not capture the peak of Pileated Woodpecker detectability and breeding behaviour. As a large-bodied bird with an expansive home range, it is unknown how or if occupancy (traditionally used as the response variable in many habitat models) can effectively represent Pileated Woodpecker habitat associations. To model Pileated Woodpecker habitat, I first determined periods of peak detectability and evaluated how different methods of measuring and estimating use influence habitat models. I explored temporal variation in Pileated Woodpecker behaviour using passive acoustic monitoring methods. Peak detection periods for Pileated Woodpeckers were near sunrise in early April. Mean daily temperature and day length were the most influential environmental variables that affected the drumming of Pileated Woodpeckers. Based on these findings, I provide minimum recommendations for future survey efforts regarding the timing and number of surveys required to ensure accurate data collection for the Pileated Woodpecker in Alberta, Canada. These guidelines can be used to plan future surveys and methods to utilize existing non-optimized surveys to ensure the accuracy of Pileated Woodpecker site occupancy. Using these guidelines, I optimized data collection and built regional habitat models for the boreal forest in Alberta, Canada. These models evaluated how different response metrics, land cover data sources, and scales affect Pileated Woodpecker habitat associations and the predictive accuracy of models. I compared two response metrics, occupancy and intensity of use, to determine Pileated Woodpecker habitat use. Biomass and canopy closure were important environmental variables for both response metrics. However, these models were not particularly predictive, possibly due to errors in land cover data, the nature of the acoustic sampling strategy, the species' biology, or a combination of these factors. Additionally, I determined that land cover data sources can greatly affect model predictive capacity and the number of reliable predictors identified. Furthermore, I determined that broad-scale land cover data, which may represent the environment at the landscape level, may be more predictive in determining Pileated Woodpecker habitat use than local definitions of land cover. From these results, I outlined considerations regarding sampling and modeling techniques for future Pileated Woodpecker studies and identified important habitat characteristics for the Pileated Woodpecker in Alberta, Canada.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.