Ecology and Evolution (May 2019)

Resolving the heading‐direction ambiguity in vertical‐beam radar observations of migrating insects

  • Zhenhua Hao,
  • Vincent Alistair Drake,
  • John R. Taylor

DOI
https://doi.org/10.1002/ece3.5184
Journal volume & issue
Vol. 9, no. 10
pp. 6003 – 6013

Abstract

Read online

Abstract Each year, massive numbers of insects fly across the continents at heights of hundreds of meters, carried by the wind, bringing both environmental benefits and serious economic and social costs. To investigate the insects' flight behavior and their response to winds, entomological radar has proved to be a particularly valuable tool; however, its observations of insect orientation are ambiguous with regard to the head/tail direction, and this greatly hinders interpretation of the migrants' flight behavior. We have developed two related methods of using wind data to resolve the head/tail ambiguity, and we have compared their outputs with those from simply assigning the heading direction to be that which is closer to the track direction. We applied all three methods to observations of Australian plague locust migrations made with an insect monitoring radar. For the study dataset, some of the headings selected by the simpler method are shown to be clearly incorrect. The two new methods generally agree and reveal a significantly different, and presumably more accurate, relationship of heading direction to track direction. However, use of these methods leads to quite a large proportion of the sample being lost because the wind values, which derive from a regional‐scale numerical model, are shown to be incompatible with the radar observations. This exploratory study has moreover demonstrated that locusts are frequently oriented at a large angle to their track and that quite often their movement is at least slightly tailfirst. Both new methods appear to be a significant improvement on the simpler method. As well as providing an accurate representation of migratory flight behavior, they allow occasions when the model wind values are unreliable to be eliminated from the data sample.

Keywords