The effects of air temperature, wind vectors and nocturnal illumination on the behavior of moths at mercury-vapour light-traps

Abstract

Field experiments, with 125 watt Mercury-Vapour light-traps, are described in this Thesis. Catches of moths in water traps arranged in a circular mosaic pattern around the light-trap were distributed bivariate normal. Comparisons between the bivariate mean moth vector and the mean wind vector suggested that the wind vector determined the distribution of moths around the light-trap. In light winds, less than 1 metre/second, moths were randomly distributed around the trap. In stronger winds the direction of moth displacement from the trap was associated with the direction component of the mean wind velocity. A second experiment with light and water traps was used to estimate the effective light-trap radius. Two light-traps, surrounded by water traps, were separated by up to 25 metres on different evenings. When one light-trap was downwind of another the catch was significantly greater in the downwind trap. However, in crosswind conditions the two catches were similar. Bivariate methods were used to estimate that the effective trap radius was between 10 and 25 metres. A further experiment, using multivariate correlation and regression analysis, found that air temperature, windspeed and moonlight affected light-trap catches. An algorithm is described to enable the calculation of moonlight illuminance. Increased windspeed and increased moonlight illuminance tended to decrease the light-trap catch, however, increased air temperature was associated with an increase in light-trap catch. These effects varied depending on family and species. A final experiment used a remote sensing technique to estimate light-trap efficiency. A 10 cubic metre volume around the light-trap was observed using a video camera. Moth tracks recorded were classified into New Arrivals, Passers By and Local Flights. The number of these tracks was compared to the lighttrap catch so that efficiency estimates could be obtained. Results suggested that at ground level, 125 watt Mercury-Vapour light-traps are only 20 % efficient within this 10 cubic metre volume.