At 867 English river sites, fish communities changed by about one third of the amount that the between-site relation to invertebrate ASPT predicts (1990–2004 to 2012–2025)

Question: River invertebrate ASPT (a pollution-sensitivity score) increased in England after 1990. Did fish communities change more toward clean-water species at sites where ASPT increased more, and was the change as large as the between-site (spatial) relation predicts? Data: Environment Agency open fish data (41,826 electric-fishing surveys, one per site and year, May-October, 1990-2025) linked to EA invertebrate samples at the nearest site in the same WFD water body (1 km or less). The primary panel has 867 sites in 263 catchments with data in 1990-2004 and 2012-2025. Method: For 16 common taxa and a fish community quality index (CQI, the mean ASPT optimum of the species present, with optima from 3,214 other sites), we compared the temporal slope (change in fish per unit ASPT change, with region and effort terms) with the spatial slope (fish per unit ASPT between sites in 1990-2004). We corrected both slopes for ASPT noise with spring-autumn split-half reliability. Intervals come from a catchment bootstrap. Results: CQI increased more where ASPT increased more: 0.082 per ASPT unit (95% CI 0.034-0.130). The spatial slope was 0.239 (0.215-0.261). The tracking ratio was 0.34 (0.15-0.54), and 0.41 (0.18-0.64) with habitat terms. Across taxa the pooled ratio was 0.29 (0.12-0.45). In seven sensitivity analyses (other windows, link rules, stable effort, early-ASPT covariate) the ratio was 0.25-0.50 and every upper limit was 0.84 or less. Atlantic salmon had the strongest spatial slope but no temporal response (ratio about 0). Brown trout responded at 0.35 (0.06-0.61) of its spatial slope. Roach loss was near the spatial expectation (0.78, 0.27-1.21). Bullhead, minnow, stone loach, lamprey and stickleback increased widely, but not more where ASPT increased more. Limitations: The data are observational. Spatial slopes can contain habitat effects that the habitat terms do not remove, so a ratio below 1 does not prove a time lag. Fish presence depends on effort; a stable-effort subset (311 sites) gave 0.25 (-0.04 to 0.53). We did not test barriers, stocking, flow, temperature or marine survival.

freshwater fish · river invertebrates · WHPT ASPT · water quality recovery · space-for-time · Environment Agency · England · electric fishing · Atlantic salmon · brown trout · community index · ecological recovery