Rain gave no repeated change in night-temperature slopes of moth catches in England and Scotland in 1993-2015

We tested whether rain changed the night-temperature slope of recorded moth catches. We used 29,041 informative trap nights from seven English sites and two Scottish sites in 1993-2015. Four separate models controlled each trap-week catch level, wind, cloud conditions, moonlight and rain amount. Errors were grouped by year. The wet-minus-dry slope was -0.024 log catch per C in England in 2004-2015. Its four-test interval was -0.046 to -0.002. The other three intervals included zero. The later English interval also included zero with a curved temperature term or three-day controls. Thus, a negative rain effect did not repeat across regions and periods. This does not show that rain has no effect. The Scottish samples were small. Catch data cannot separate flight activity, attraction to light, emergence and local abundance. The study did not test air pollution, extreme rainfall or a causal climate-change effect. The package includes the fixed input panel, source hashes, code and checks against public catch data and a separate model implementation.

moth · observation bias · rainfall · temperature · light trap · regional replication