Mean length changes were uncertain for six fish taxa in matched catches at 168 English sites, 2000–2009 versus 2016–2025

We compared recorded fish lengths in 2000–2009 and 2016–2025 at repeated English freshwater survey sites. We used open Environment Agency records. The main sample has 85,864 fish in 1,432 survey-species records from 1,056 surveys at 168 sites. We matched site, source species group, calendar month and fishing method. Each group had at least two observed years in each period. Each first-pass catch had at least ten fish and a valid individual length for every fish. Six taxa had at least 15 matched sites. We gave sites equal weight within a taxon. Mean length change estimates were −0.2% for brown / sea trout, +10.9% for Atlantic salmon, −17.2% for chub, −6.1% for dace, −6.5% for gudgeon and −8.3% for roach. All six 95% ranges included zero when 50-km spatial blocks and common calendar years were allowed to vary. Shared weights on adjacent two- or three-year blocks gave the same uncertainty result. A further check removed sites with reported coordinate spans above 250 m. In that sample, gudgeon length fell by 8.3%, with a 95% range of −17.7% to −0.6%. We retain both results. An independent calculation checked all selected length values against the source archive and reproduced every primary site estimate. An offline repeat reproduced all 16 checked output files. These results describe conditional catch samples. They do not show stable national fish size, measure growth at a given age, or identify a climate cause. The package preserves source-value extracts, source hashes, code, complete results, figures and repeat instructions.

freshwater fish · body length · Environment Agency · paired surveys · measurement coverage · spatial uncertainty · open data