Other core quadrats recorded the same species after 47% of annual 1 m² plant record losses at 45 NEON sites, 2013–2024

We tested how often a plant species lost from an annual 1 square metre quadrat record was still recorded elsewhere in the same NEON plot. The main sample had 82,924 species-quadrat losses in 6,284 consecutive-year plot pairs at 45 sites in 2013–2024. We held six core quadrats fixed, used one survey bout per site, and matched dates within 30 days. With equal site weights, the same species was recorded in another core quadrat after 47.0% of losses. The 95% whole-site bootstrap interval was 44.7–49.4%. A planned subset had full 400 square metre surveys at the next visit. In its 50,138 losses, the core fraction was 47.4% and the full-plot fraction was 72.5% (site interval 69.3–75.5%). The fraction depended on initial occupancy: 15.2% for species first recorded in one core quadrat and about 75% for species first recorded in five or six. Date, cover, and identification checks gave similar core fractions. Shared-year block checks conditioned on site coverage supported the survey-area contrast. Independent table joins from the source matched every primary loss identity and all eligible pairs. The package has the unchanged open source data, provenance, code, all results, uncertainty draws, figures, and offline repeat instructions. These are changes in records at fixed sampling extents. They do not establish plant movement, extinction, detection probability, or an environmental cause.

plant distributions · NEON · temporal turnover · spatial scale · nested sampling · recorded loss · reproducibility