Skip to main content
Dryad logo

Data from: Large trees and forest heterogeneity facilitate prey capture by California spotted owls

Citation

Zulla, Ceeanna J et al. (2022), Data from: Large trees and forest heterogeneity facilitate prey capture by California spotted owls, Dryad, Dataset, https://doi.org/10.5061/dryad.4b8gthtfn

Abstract

Predators are among the most threatened animal groups globally, with prey declines contributing to their endangerment. However, assessments of the habitat conditions that influence the successful capture of different prey species are rare, especially for small, cryptic predators. Accordingly, most predator conservation plans are based on the relative importance of habitats inferred from coarse-scale studies that do not consider habitat features contributing to hunting success, which can vary among prey species. To address this limitation, we integrated high-resolution GPS tracking and nest video monitoring to characterize habitat features at prey capture locations during the nestling provisioning stage for the Spotted Owl (Strix occidentalis) a small, cryptic predator that has been at the center of a decades-long forest management conflict in western North America. When all prey species were considered collectively, males provisioning nests tended to capture prey: (i) in areas with more large-tree forest, (ii) in areas with more medium trees/medium canopy forest, and (iii) at edges between conifer and hardwood forests. However, when we considered the owl’s two key prey species separately, males captured woodrats (Neotoma fuscipes) and Humboldt flying squirrels (Glaucomys oregonensis) in areas with markedly different habitat features. Our study provides clarity for forest management in mixed-ownership landscapes because different prey species achieve high densities in different habitat types. Specifically, our results suggest that promoting large trees, increasing forest heterogeneity, and creating canopy gaps in forests with medium trees/high canopy cover could benefit Spotted Owls and their prey, which has the ancillary benefit of enhancing forest resilience. Combining high-resolution GPS tagging with video-based information on prey deliveries to breeding sites can strengthen conservation planning for small predators by more rigorously defining those habitat features that are associated with successful prey acquisition.

Methods

Please see the Methods section below for the specific methods (copied from the manuscript).

Overall, these data include 4 shapefiles detailed below. Please note that data were spatially shifted due to the sensitive nature of the California spotted owl.

  • Nest_locations_shifted
    • Point file showing the location of the spotted owl nest
    • Attributes include:
      • Site – unique site identifier for each tagged bird
      • P_KDE_publc - estimated proportion of public land in that bird's KDE
  • GPS_points_shifted
    • Point file showing the location of all GPS tag locations
    • Attributes include:
      • Site – unique site identifier for each tagged bird
      • Year – year that the GPS point was recorded
      • Month – year that the GPS point was recorded
      • Day – year that the GPS point was recorded
      • Hour – year that the GPS point was recorded
      • Minute – year that the GPS point was recorded
  • Capture_polygons_shifted
    • Polygon file showing the used and available capture polygons and their associated prey, split by vegetation type
    • Attributes include:
      • Site – unique site identifier for each tagged bird
      • PolyID – the capture polygon identifier (unique to each used capture polygon, and serving as a link to all available polygons)
      • Rand_ID – a used/random identifier
      • Area_sqm – the area of each polygon in square meters
      • Veg_ID – the numeric vegetation type ID
      • VegClass – a text description of the numeric Veg_ID
      • Prey_comm – the common name of the prey associated with that polygon
      • Prey_sci – the scientific name of the prey associated with that polygon
  • Buffered_capture_polygons_shifted
    • Polygon file showing the 50 m buffered used and available capture polygons and their associated prey, split by vegetation type
    • Attributes include:
      • Site – unique site identifier for each tagged bird
      • PolyID – the capture polygon identifier (unique to each used capture polygon, and serving as a link to all available polygons)
      • Rand_ID – a used/random identifier
      • Area_sqm – the area of each polygon in square meters
      • Veg_ID – the numeric vegetation type ID
      • VegClass – a text description of the numeric Veg_ID
      • Prey_comm – the common name of the prey associated with that polygon
      • Prey_sci – the scientific name of the prey associated with that polygon

 

Methods – copied directly from the manuscript

2.2 | GPS tagging

We located territorial breeding Spotted Owls as part of our routine monitoring surveys conducted annually on the EDSA and SSA (e.g., Hobart et al., 2019b; Roberts et al., 2017). Briefly, owls were located during call-based surveys at night and found during dawn/dusk surveys the following day to ascertain their reproductive status and locate nests (Franklin et al. 1996). Owls were fed live mice during follow-up surveys, which breeding owls then typically delivered to nest sites (Franklin et al. 1996).

We captured 15 nesting males (5 in 2019 and 10 in 2020) for GPS tagging using noose poles and “hand capture” (Wood et al. 2021). Females were not tagged because they spend most of their time incubating eggs and brooding nestlings while nesting and relatively little time foraging (Forsman et al. 1984). Males were selected opportunistically for tagging based on the accessibility of the nest for video-monitoring (see below) throughout the nesting months (May – July) and the likelihood of recapture to remove transmitters. We affixed GPS tags (Alle-300, Ecotone, Poland) weighing 10 grams and with remote downloading capability as tail mounts following methods described in Kramer et al. (2021) (Fig. 1). We programmed tags to collect locations at two-minute intervals to characterize Spotted Owl nocturnal movements and ultimately identify prey capture locations (described below). We recaptured five individuals for a second deployment to increase the number of observations of prey captures. Following the final deployment, we attempted to recapture all owls to remove GPS tags; two individuals that were not recaptured were expected to molt during that season or the following season, thus shedding the GPS tag. 

2.3 | Nest video monitoring

We monitored prey deliveries using infrared (IR) video cameras placed at the nest sites of the 15 GPS-tagged males concurrent with the collection of GPS locations. To do so, we climbed a nearby adjacent tree (10-50 m from the nest tree) using a single rope technique and secured a video camera across from the nest tree (Fig. 1). We monitored nests using AXIS Q1786 – LE 4 megapixel outdoor infrared video cameras that continuously recorded high quality video throughout the nocturnal foraging period that coincided with GPS tag data collection (2000 to 0630 Pacific Daylight Time). We reviewed each video to detect and identify the different prey species delivered to each nest.

 2.4 | Identifying prey capture locations

We located successful foraging sites (hereafter ‘prey capture locations’) by visually identifying tight clusters of GPS points followed immediately by a straight flight path back to the nest tree (Marsh et al. 2014, Wood et al. 2021). We defined a prey capture location as a cluster of up to ten GPS points (within the 20 minutes prior to movement back to the nest), all within 250 meters of one another based on that likely being the maximum hunting radius of a perched owl (Whitmore, personal observation). If the two farthest points were greater than 250 meters apart, we removed the oldest point (the first point that defined the prey capture location) and continued to remove the points until they were all within 250 meters of one another. Each cluster of points was transformed into a minimum convex polygon (mean = 0.38 ha, range = 0.002 - 2.713 ha) surrounding the cluster and assumed to represent the prey capture location. Although this approach, including the criterion of <250 m among pairs of points, was somewhat subjective in that it was based on field observations, it was repeatable and typically yielded tight, defined clusters of points (compared to the distribution of owl GPS points overall) followed by distinct straight-line movements back to the nest site (Fig. 2). We then matched the time stamp of the video with the time stamp of the GPS return to the nest to link the species of prey to its putative capture location (Fig. 2). The area contained by the minimum convex polygon of these GPS points that matched with a prey delivery caught on video was defined as a prey capture location. Three prey capture locations were excluded from the analysis because they likely represented the retrieval of prey from a cache site. We suspected these deliveries to be from a cache location because of their close proximity to the nest site (within 50 m) and the early timing of the delivery (all three instances were the first delivery of the night and occurred before sunset), which both indicated a cache delivery (Whitmore, personal observation).

 2.5 | Vegetation classification  

We visually interpreted aerial imagery from the National Agriculture Imagery Program (NAIP) collected in 2018 and 2020 to characterize vegetation conditions within prey capture locations (as well as random locations, see below) following methods developed and described by Tempel et al. (2014) that typically results in >80% classification accuracy. Specifically, we considered ten possible vegetation classes based on species composition, canopy cover, and the size of the dominant trees that was largely based on the California Wildlife Habitat Relationships system (Mayer and Laudenslayer 1988) (Table 1). Areas classified as hardwood typically represented patches of black oaks, with tanoaks in the understory typically being obscured by overstory conifer trees and thus not measured. We digitized polygons around relatively homogenous areas of vegetation with a minimum mapping unit of 20 m2 and then classified the area within the polygon as one of the ten vegetation classes.

Usage Notes

ArcGIS Pro was used to make the files, but they can be opened with any GIS software such as QGIS.

Funding

Funding was provided by the US Forest Service Region 5 and Pacific Southwest Research Station, and Sierra Pacific Industries.