Management and conservation plans often lack adequate data about population estimates, habitat selection, reproductive ecology, and even migration ecology of migratory species. Migratory and understudied birds, like Northern Harriers (Circus hudsonius), are no exception. The Northern Harrier is a widespread but declining North American raptor species that specializes in wetland and grassland habitats. They are a secretive raptor species with a unique combination of behavioral adaptations, like ground nesting, avoidance of urban development and human disturbance, and avoidance of traditional raptor trapping techniques, that make them challenging to study. Recent advances in technology, like the miniaturization of GPS/GSM transmitters, can greatly improve management and conservation of Northern Harriers and help fill knowledge gaps about their ecology and behavior across their range. The first challenge to studying Northern Harriers was finding efficient trapping techniques to capture an adequate sample size of individuals for transmitter deployment. In Chapter 1, we developed modified and novel trapping techniques for wintering and breeding Northern Harriers. First, we reviewed and tested successful trapping techniques from other Northern Harrier breeding populations. The most widely used technique in other populations was the use of a live or taxidermy mounted owl near a nest surrounded by mistnets or dho-gazas, which facilitates capture by eliciting an aggressive response by the breeding pair. Surprisingly, none of the published techniques were successful in our population, leading to the development of a new technique. We had the highest success flushing adult females into two dho-gazas placed in a “V” formation around the nest, a method that was modified from capturing nesting American Bitterns (Botaurus lentiginosus). Most females did not show behavioral responses to net placement at the nest and continued provisioning nestlings without visible stress or alarm calling. Unlike the use of an owl lure, our technique allows for passive capture with minimal disturbance to nesting birds. However, because this technique does not elicit an aggressive response, adult males are not amenable to capture with this technique. Winter trapping is typically not attempted for Northern Harriers because they rarely respond to traditional methods like bal-chatri traps baited with live lure birds or small mammals. Noose carpets can work well, but also result in incidental capture of multiple birds or different raptor species simultaneously, leading to potential injuries of captured birds. Knowing that Northern Harriers are scavengers of dead waterbirds in winter wetland habitats, we used remote-triggered bow nets baited with waterbird carcasses to capture wintering birds. This novel trapping technique was overwhelmingly successful, with nearly 80 individual birds captured across three winter trapping seasons. This technique also allows for targeted capture of individual birds, eliminating the risk of capturing multiple birds or non-target species. Our research suggests that trapping techniques are not necessarily universal across populations, and careful consideration of behavior and life history traits can improve targeted research for species like Northern Harriers.
Equipped with the trapping techniques developed in Chapter 1, we studied the habitat selection of breeding adult female Northern Harriers across multiple spatial scales in Suisun Marsh, California, in Chapter 2. We measured fine-scale microhabitat characteristics at nests, extracted macrohabitat characteristics across the landscape, and examined breeding home range habitat selection in adult female Northern Harriers captured at the nest and equipped with GPS/GSM transmitters. At the microhabitat scale we found that Northern Harriers select tall emergent and terrestrial vegetation, which is consistent with habitat selection studies in populations across their range. This result also suggests a selection for both marsh habitat and upland habitat for nest sites. Specific to Suisun Marsh, we found that Northern Harriers also have a high probability of selecting nest sites with California Rose (Rosa californicus). Macrohabitat selection consisted of nest placement ~ 100 m from water and revealed an avoidance of shrub vegetation. These results combined indicate a selection for vegetation structure and nest placement in vegetation that provides protection from mammalian predators. A subset of nests on the Grizzly Island Wildlife area avoided nest placement near all-terrain vehicle (ATV) tracks, and spatially clustered nests revealing additional adaptations for predator and disturbance avoidance as well as semicolonial nesting behavior. Small mammal densities had no effect on nest site selection, despite the known importance of small mammals, particularly voles, on Northern Harrier reproductive success. Lastly, we found that adult females had a significantly higher probability of wetland habitat use over any other habitat type even as home ranges expanded, and females foraged farther from nest sites across the nesting season. Our results inform habitat managers to maintain tall, undisturbed vegetation in both upland and marsh habitats while ensuring dry areas for nest sites are available with complex vegetation and habitat structure, like California Rose and nearby water, to provide protection from mammalian predators.
Chapter 3 builds on the research from Chapter 2 and examines the effects of multi-scale habitat characteristics on nest survival using logistic-exposure models. We found apparent nest success was relatively low (40%) compared to nest survival in populations across the Northern Harrier range, but Mayfield’s nest survival was similar to nests in Suisun Marsh three decades ago. Despite similar survival rates, the number of nests and nest density has declined over the past three decades, revealing a declining population with already low nest success. High spring small mammal densities had a positive effect on nest survival, and high summer small mammal density increased the number of fledglings. However, the California Vole (Microtus californicus) population is alarmingly low in Suisun Marsh compared to historical observations. Voles are an important prey species and improving their abundance could improve harrier nest survival and increase the breeding population in Suisun Marsh. Our results also show that a high proportion of live vegetation and the presence of residual vegetation at the nest are important microhabitat characteristics that influenced nest survival. Further, nests closer to California Rose also had higher nest survival, which is consistent with nest site selection determined in Chapter 2. Lastly, nest survival was highest in managed marsh habitat, despite potential selection for tidal marsh habitat. Lower nest survival in tidal marsh habitat may result from inappropriate habitat structure and extreme flooding events that could be exacerbated by sea level rise in the future. Restoring tidal marsh to large, contiguous habitat patches with numerous smaller channels to better diffuse the effects of high tides across the tidal plane could reduce nest flooding for harriers and other sensitive tidal marsh species in Suisun Marsh.
Chapter 4 focuses on migration ecology of Northern Harriers wintering in Suisun Marsh. Using the winter trapping techniques developed in Chapter 1, we marked adult females with GPS/GSM transmitters in this first study of Northern Harrier migration ecology and habitat selection across their annual cycle. We recorded a total of 18 spring and 11 fall complete (round-trip) and partial (one-way) migrations for 14 individual Northern Harriers and identified nest sites across five Western United States (AK, CA, ID, OR, WA). We also recorded the three longest-distance migrations for any Northern Harriers to date across two individuals breeding in Alaska, ranging from 13,000 to nearly 20,000 km traveled roundtrip. Of the 11 fall migrations recorded, all birds returned to Suisun Marsh, highlighting its importance in Northern Harrier wintering ecology in Western North America. Mean spring migration was shorter than mean fall migration by nearly two months, with fewer stopovers and a faster migration speed (~ 200 km d-1) suggesting strong selection pressure to reach the breeding grounds early to secure a mate and territory, and to increase reproductive success. Migration timing is generally consistent with known timing for this species from raptor migration monitoring stations across North America. Migration routes were primarily along central and eastern corridors through California, Washington, and Oregon, and generally continued along inland intermountain regions through British Columbia and into Alaska for long-distance migrants. Wetland habitat was the most consistently selected habitat type across the annual cycle, with grassland and shrubland habitat also selected at stopover locations, and cultivated habitat also selected during the winter. Though habitat selection varied across the annual cycle, many breeding areas and stopover locations occurred on protected state and federal lands, which provide a unique opportunity for focused management and conservation efforts for migratory Northern Harriers. Raptor migration monitoring and banding stations located throughout Western North America may be misaligned with Northern Harrier migration, leading to low detections and population estimates because migration occurred along corridors where migration monitoring and banding stations are not located. Focusing migration monitoring along wetland habitat corridors and increasing breeding population research could improve Northern Harrier management and conservation efforts.
