Prior hot days gave uncertain fixed-size mosquito richness slopes in four US samples in 2016-2019 and 2021-2024

We tested whether prior hot days had a negative link to mosquito species richness at a fixed count of 20 identified females. We used NEON RELEASE-2026 records and Daymet V4 R1 weather. Four samples cover the eastern and western conterminous United States in 2016-2019 and 2021-2024. Final models used 1,199 site-date means from 5,703 traps at 32 sites. Controls included mean prior warmth, rain, capture weather, sampling conditions, site-month and site-year effects. Slopes per five days at or above 30 C ranged from -0.453 to -0.073 species. All four adjusted intervals included zero. Only 5,964 of 17,963 usable traps met the 20-female rule before fixed-effect selection. Controls left 2-8 percent of the original hot-day variation. Small western samples and four year groups per fit give weak precision. One secondary two-way variance estimate was negative and is not estimable. The evidence does not establish a repeated heat effect or its absence. Count selection, capture and identification bias, interpolated weather and an unaudited curated source limit inference. The package gives original code, aggregate evidence, two figures, exact source hashes and an external-data rebuild. It tests no air pollution effect or causal human climate effect.