April–September desert pocket mouse capture rates fell in four control and four kangaroo rat exclosure plots at Portal in 2024–2025

We tested whether kangaroo rat exclosures retained desert pocket mouse captures during the 2024–2025 decline at Portal, Arizona. We used four long-term plots per treatment and data release 6.117.0. The main comparison used April–September, 61 baseline censuses in 2010–2022, and 12 event censuses. All event data had source quality flag 1. Baseline rates were matched to each event plot and calendar month. Controls fell from 13.16 to 1.96 captures per 100 trap nights. Exclosures fell from 13.15 to 4.04. They retained 14.9% and 30.7% of baseline rates. All eight plots had a decline. The ratio of retained fractions was 2.06. A bootstrap of whole plots and baseline years gave a conditional 95% range of 1.22–3.45. The event years were fixed. An exact plot label test gave p=0.086 and required plot exchangeability. Ten other stable plots gave a smaller ratio of 1.24 against a 2016–2022 baseline. Two 2025 censuses contained repeated source records. Counting raw rows added 33 to the capture count in the eight plots. Removing these censuses gave a retained-fraction ratio of 2.08. Thus, the repeated rows changed capture levels but did not explain the contrast. County rain in water year 2025 was the lowest in 1981–2025. This study does not identify the cause of the decline. Capture rates do not measure true population size. Field reports had already described the decline. This contribution supplies a fixed-release comparison, checks, and a record audit.

desert ecology · species interactions · capture rates · Portal Project · kangaroo rat exclusion · biodiversity monitoring · data quality · precipitation