The warmth slope of usable moth sample availability increased on common English and Scottish traps from 1992-2003 to 2004-2015

We tested a change in the link between daily warmth and usable moth samples. We used the same seven English and two Scottish trap units in 1992-2003 and 2004-2015. The panel had 25,392 candidate dates in selected summer months. Models controlled trap-year-month and separate weekday and day-of-month patterns in each period. The late minus early slopes were +1.86 percentage points per C in England and +2.74 in Scotland. Two-test intervals were +1.03 to +2.69 and +0.90 to +4.57. English site omissions kept positive intervals. The Scottish contrast was not clear without Glensaugh. Availability also fell between periods. Thus a change on the probability scale need not mean a change in trap sensitivity or staff choices. This is a test of the observation process, not moth population change. Selected months, coarse weather, few sites and possible dependence between years limit inference. The package supplies a verified source panel, a raw-input audit, code, results and a figure.

moths · observation bias · sample availability · temperature · period contrast · ECN · E-OBS