Portal records for 2019–2024 did not show that kangaroo-rat capture rates were within 0.8–1.25 times control rates after the 2015 change

We tested whether rodent recovery after a 2015 treatment change persisted at Portal, Arizona. We used PortalData 6.117.0 for three former kangaroo-rat exclusion plots, three former total-rodent removal plots, and four long-term controls. The 2019–2024 sample had 46 matched censuses and 4,054 granivore captures. Kangaroo-rat capture-rate ratios to controls were 0.747 (95% plot-level interval 0.523–1.068) and 0.865 (0.651–1.148). Neither contrast met a stated equivalence range of 0.8–1.25, also with a standard 90% interval test. Equal year weights gave a deficit on former kangaroo-rat exclosures, so evidence for a deficit depended on weighting. Species composition had excess Bray–Curtis distances of 0.090 and 0.093 above control-to-control distances. Equal-capture rarefaction gave similar estimates, but plot-label reference tests gave weak evidence for a group difference. The later records therefore establish neither practical rate equivalence nor a lasting treatment effect under the main model. Limits are three plots per former treatment, one site, sparse pandemic sampling, and capture rates without a detection correction. The package preserves source data, provenance, code, figures, all results, and reproduction checks.

ecological recovery · Portal Project · rodent communities · equivalence test · species composition · long-term experiment · reproducibility