Multi‐decadal land use impacts across the vast range of an iconic threatened species

Abstract

Aim: To explore spatiotemporal changes in Hyacinth Macaw Anodorhynchus hyacinthinus distribution and the impacts of land use change over 25 years, across its vast range in central/eastern South America.Location: Brazil, Bolivia and Paraguay, South America, covering almost 3 million km 2 . Methods:We use a novel, multi-temporal species distribution model, to combine both year-specific occurrence records and land use/cover data in a single model that is subsequently projected over a land cover time series. We investigate changes in geographic range over 25 years and potential drivers of range extent at multiple spatial scales (10 and 30 km) and time-lags (current and 10 years previously). We also assess protected area coverage and impacts from degazettement within the remaining range and highlight priority areas to search for undiscovered populations. Results:The modelled range of the species increased by 75,000 km 2 to over 500,000 km 2 between 1995 and 2019, during which period agriculture and pasture increased by 600,000 km 2 within the species' extent of occurrence. Habitat suitability was influenced most strongly by distribution of palms, forest cover and changes in pasture over 10 years and usually by predictors measured at larger spatial scales.Just 8% of the macaw's modelled range falls within protected areas, and more than 20,500 km 2 of protected areas overlapping with the range have been degazetted in the last 40 years. We highlight key areas to search for undiscovered populations in under-sampled and remote areas, especially led by community citizen science initiatives involving indigenous groups and protected area staff.Main conclusions: Novel modelling methods, combining multi-temporal occurrence records and land cover data, appropriate for small sample sizes per year, have revealed habitat dynamics and changes to the range of this threatened species over 25 years. The method may have wide applicability for a range of species, including elusive and poorly studied species.

Decomposing the spatial and temporal effects of climate on bird populations in northern European mountains

“…Here, we would expect the same relationship between the species' distribution or abundance and climate, regardless of whether the species‐climate relationship is in space or time or is describing the imminent or more distant future (Figures 2 and 3 ). Fast species responses can be modeled with dynamic covariates, in which covariates represent variation at fine temporal resolutions, such as seasons or years, and the resulting forecasts can incorporate rapid changes in distributions or abundances (Figures 2 and 3 ; Briscoe et al, 2021 ; Devenish et al, 2021 ). Thus, the underlying approaches of SDMs with static and dynamic covariates are different, with rapid effects of climate conditions omitted when using static covariates, but included via dynamic covariates.…”