Estimated American wigeon latitude increased for all fixed mean latitudes within 50 survey regions, 1974–1993 to 2000–2019

We tested whether changes in estimated duck latitude depend on the locations assigned to large survey regions. We used public North American waterfowl estimates and polygon boundaries for 50 fixed regions. We compared mean annual regional shares in 20 complete surveys from 1974–1993 with 18 from 2000–2019. We calculated exact latitude bounds for unknown mean locations within each region. With the source estimates held fixed and each region mean fixed between periods, American wigeon had a northward change of +1.18 to +2.19 degrees for every allowed location choice. The lower endpoint had year-block resampling percentiles of +0.48 to +1.90 degrees. If each region mean could change between periods, zero first entered the bounds at an allowed change of 1.51 degrees. This threshold varied from 0.65 to 1.66 degrees in planned period, area and weighting checks. With only the polygon limits, wigeon bounds were −1.22 to +4.58 degrees. Northern pintail, the other planned focal taxon, allowed both signs even with fixed means. The study supplies an exact bound method and an audited open-data application. It does not estimate individual movement or test a climate cause. The bounds condition on published regional estimates. The resampling omits survey estimate error and does not give simultaneous confidence statements. The package preserves unchanged source files, provenance, the local plan, all results, figures, code, numerical checks and offline repeat instructions.

species distribution · waterfowl · American wigeon · northern pintail · spatial uncertainty · latitude bounds · open data · reproducibility