Cedar Creek field B showed greater plant composition recovery with annual fire after high fertilizer input stopped, in 2015–2022

We tested plant composition after 32 years of fertilizer input stopped at Cedar Creek. The primary sample had 18 plots at three high rates in field B. Each plot had stopped and continued halves. Nine plots had annual spring fire from 1992. We compared the halves with untreated controls from the same year and fire group. The data cover six sampling years in 2015–2022. The primary measure was Bray–Curtis distance on relative live herbaceous biomass. Its scale is 0 to 1. The paired distance decrease was 0.150 with annual fire and 0.031 without annual fire. The difference was 0.119, with a 95% plot bootstrap interval of 0.059 to 0.178. The bootstrap included the shared control plots. Large differences remained. Stopped annual-fire halves had a mean distance of 0.825 from controls. Distinct controls had a mean distance of 0.314 from each other. These values give 23% closure of the composition gap above control variation. A presence-only measure gave 46% closure. In an exploratory exact breakdown, little bluestem supplied 90.5% of the biomass-distance improvement with annual fire. The result held under changes to the taxon filter, sample years, and control reference. This extends an earlier SCI richness analysis with composition tests. It does not show full restoration. Limits include one field, three controls per fire group, small moving strips, unequal half widths, and only eight seasons after cessation. Fire began before cessation. All fertilizer inputs stopped together. The package preserves source data, provenance, code, figures, and independent arithmetic checks.

grassland · plant composition · nutrient cessation · fire · Cedar Creek · ecological restoration · Bray-Curtis · reproducibility