The 20-year change in species dissimilarity did not clearly differ between nearby and distant pairs of 1,709 Swedish forest plots

We tested whether the 20-year change in recorded species dissimilarity differed between nearby and distant Swedish forest plots. We used 1,709 fixed locations with 100 square metre surveys on productive forest land in 1999–2004 and 2019–2024. Each plot had visits exactly 20 years apart and within 30 days of the same day of year. We compared 193 monitored species in the same 104,819 nearby pairs (10 to less than 100 km) and 546,514 distant pairs (at least 500 km). Mean Simpson dissimilarity fell by 0.02518 nearby and by 0.01997 at distance. The distant change minus the nearby change was +0.00521, with an approximate 95% interval of -0.00350 to +0.01393 from joint deletion of 61 spatial blocks. Thus, lists became more similar in both groups, but the test did not establish a change in the gap between them. Equal focal weights, larger blocks, closer visit dates, vascular species, and same-year spatial pairs also gave intervals that include zero. Without the date limit, a larger sample gave a clear positive contrast. The main result does not establish equal changes or correct detection error. The package has all compact source inputs, code, data provenance, block-pair sums, figures and independent checks. A fresh run reproduced all nine numerical and selection files exactly.

forest vegetation · spatial similarity · biodiversity · Swedish National Forest Inventory · distance decay · reproducibility · null result