Methods for studying stopover ecology of migrating landbirds with weather surveillance radar
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Populations of numerous migratory landbird species in the eastern United States are
declining and these populations may be limited during their migratory journey.
Weather surveillance radar is a useful tool for monitoring large scale movements of
birds during migration and particularly for mapping stopover distributions of
migratory landbirds because it detects birds low to the ground as they initiate
nocturnal migratory flight. This approach is sensitive to the time when flight exodus is
sampled because the number of birds in the air at this time changes rapidly. Thus, in
order to use radar to map densities of migrant birds on the ground, an empirical
determination is needed to identify an unbiased method to sample migrant density in
the air. I assessed the relationship between seasonal mean migrant bird ground
densities and seasonal mean radar reflectivity, an estimate of emigrant bird density
aloft, sampled at a series of sun elevation angles ranging from 1.5° to 10° below the
horizon at 26 sites in Delaware, Maryland, and Virginia within 80 km of the Dover,
Delaware (KDOX) and Wakefield, Virginia (KAKQ) WSR-88D stations during fall
2013 and 2014. Additionally, because the timing of flight exodus varied among nights
within and among radars, I fit a logistic growth curve to the change in mean
reflectivity through time during the onset of nocturnal flight to determine the sun angle
at the inflection point of the curve (i.e., at the maximum growth rate) for each
sampling night by radar. I computed correlations between ground bird densities and
mean reflectivity among the series of radar sampling times and among a series of
times relative to the inflection point of daily exodus curves. Sampling radar at the
inflection point of daily exodus curves provided a consistent moderate to strong
correlation and this approach is likely robust to broad spatio-temporal changes in the
timing of exodus that would not be accounted for by using an absolute sun angle.
Placing stopover sites for migratory landbirds into a functional framework based on
intrinsic and extrinsic factors may be a key to conserving declining populations.
Landbirds typically use numerous stopover sites during migration, which vary in
usefulness regarding replenishment of energetic resources. To classify stopover sites
across a broad spatial scale, I determined relative stopover duration at study sites
mentioned above combined with data collected using similar protocols during a
previous study in Alabama and Louisiana by integrating ground transect data with
weather surveillance radar data. Functional types within the function framework
initially included “fire escape,” “convenience store,” and “service hotel”, but
clustering resulted in four distinct groups, which I redefined as “coastal fire escape,”
“inland rest stop,” “convenience store,” and “full service hotel.”, a novel designation
for landbirds. ☐ I incorporated hardwood forest within 5 km, distance to the coast, and insect density
into the analysis as potential drivers of stopover duration. One third of our study sites
were deemed as full service hotels, making the majority of our study sites coastal fire
escapes, inland rest stops, or convenience stores, which typically receive less attention
in conservation planning. There were regional differences, where the mid-Atlantic
lacked full service hotels and the Gulf Coast lacked coastal fire escapes. Using a
system of functional types facilitates the prioritization of stopover sites because I can
evaluate sites within each functional type rather than across functional types. Each
functional type serves a purpose and all are necessary in conservation, but all sites
cannot be protected, so using a functional type system allows us to prioritize sites
more easily and efficiently. Using weather surveillance radar and ground surveys
allowed me to assess stopover use at a broad spatial scale, which is difficult to do with
more traditional methods.