Recorded total biomass had a weaker richness–stability relation than biomass of listed plants at Cedar Creek in 2010–2021

We compared temporal stability of total live biomass with biomass of plants in the planting records at Cedar Creek Biodiversity II. We used the same sorted samples from 154 plots and 12 annual harvests in 2010–2021. Assigned richness was 1, 2, 4, 8 or 16 species. Stability was temporal mean divided by standard deviation (SD). The slope of log stability per richness doubling was 0.074 for total biomass and 0.156 for listed plants. The paired difference was −0.082, with a 95% plot bootstrap interval of −0.110 to −0.054. The total-biomass slope was 52% smaller. An exact decomposition gave a mean contribution of −0.122 and an SD contribution of +0.040. Other biomass raised relative mean biomass more in low-richness plots. The SD change partly offset this effect. The difference stayed negative with common year-block sampling, removal of a linear trend, and alternative planting definitions. The source audit found planting information outside the 18 membership columns. Removing all ambiguous other-biomass groups left a small difference whose interval included zero. Assigning common fractions of ambiguous groups to listed plants retained smaller negative differences. We preserve both planting definitions, raw data, code, figures and all results. This is an arithmetic comparison at one managed field site. It does not estimate the effect of removing other plants or establish future climate resilience.

biodiversity · temporal stability · grassland · Cedar Creek · biomass · planting records · reproducible research