Taxon temperature terms gave positive catch allocation scores in four British moth samples in 1993-2015

We tested whether taxon night-temperature terms improve catch allocation beyond a baseline proportional to total core moth catch. We rebuilt counts from public ECN files and used verified night weather. Four tests covered England and Scotland in 1993-2003 and 2004-2015. Each model trained only in the other period. Mean annual gains were 0.00226 to 0.00425 nats per selected moth. Conditional four-test intervals were above zero in both English samples and later Scotland. Three-day groups kept clear English gains, but not the later Scottish gain. Equal-taxon weighting gave a slightly negative later Scottish score. Three-year block intervals left only early England clear of zero. These are small conditional prediction gains, not detection estimates or causal climate effects. Few sites, source selection, shared traps and omitted training uncertainty limit transfer claims. The package gives code, aggregate evidence, source hashes, retrieval routes and numerical checks. External raw data and nightly record tables remain outside the bundle.