Environmental Engineering Reference
In-Depth Information
The majority of data collected at night was detected by radar only, but
sometimes species where identified at night when an observer was able to hear
flock calls. Assumed flock size and height for nocturnal trails was determined
by using average flock size and flight height recorded during diurnal movements
for that season. It was suspected that many resident birds also flew through the
site at night, such as ducks and Canada geese
Branta canacensis
. However, as
trails could not identify species at night, all trails that followed broadly the same
migratory path as the observed diurnally detected internal migratory shorebirds
were assumed to be SIPO flocks for the purposes of inputting into the strike
model.
Results of Monitoring
Migratory trail data from the migratory shorebird surveys were analysed in Arc GIS
using a spatial analyst tool.
The key parameters used in the analysis were:
•
Population field: Number of potential and confirmed shorebird trail polylines for
the duration of the survey;
•
Search radius: 500 m;
•
Area Units: km
2
; and
•
Output cell size: 50 m.
Three main attributes of migratory trails were established as shown in the raw
trail maps and the trail density maps (Figs.
3
,
4
,
5
,
6
,
7
, and
8
):
• On the northbound migration (summer) birds travelled in a north, northwest, or
north-easterly direction;
• On their southbound migration (winter) birds travelled in a west, south-west,
south or south-easterly direction; and
• Migratory birds travelled in a broadly straight line, i.e. they were flying in
a directional manner rather than constantly changing direction along their
path.
The figures show radar trails which were verified by field observer data to be
internal migrant shorebird movements, as well as initially unidentified radar trails
which were determined to be internal shorebird movements based on trial character-
istics. Note that we have not presented the summer 2009 density map as the data are
not as complete as in other years. Further, the data from summer 2009 monitoring
period were not included in the Collision Risk Model.
