Understanding changes in abundance is crucial for conservation, but population growth rates often vary over space and time. We use 40 years of count data (1979–2019) and Bayesian state-space models to assess the African penguin Spheniscus demersus population under IUCN Red List Criterion A. We deconstruct the overall decline in time and space to identify where urgent conservation action is needed. The global African penguin population met the threshold for Endangered with a high probability (97%), having declined by almost 65% since 1989. An historical low of ~17,700 pairs bred in 2019. Annual changes were faster in the South African population (−4.2%, highest posterior density interval, HPDI: −7.8 to −0.6%) than the Namibian one (−0.3%, HPDI: −3.3 to +2.6%), and since 1999 were almost −10% at South African colonies north of Cape Town. Over the 40-year period, the Eastern Cape colonies went from holding ~25% of the total penguin population to ~40% as numbers decreased more rapidly elsewhere. These changes coincided with an altered abundance and availability of the main prey of African penguins. Our results underline the dynamic nature of population declines in space as well as time and highlight which penguin colonies require urgent conservation attention.

These data are counts of occupied nests of African penguins at 18 colonies in South Africa and 4 colonies in Namibia. Counting methods are detailed in the publication with which this dataset is assocaited. The following is text from that publication: "Briefly, counts were undertaken by teams of people walking through a penguin colony and counting occupied nest sites. Larger colonies were broken down into predefined census areas, each of which was counted separately. Counts in South Africa were predominately made between February and September each year, those in Namibia were made monthly (Crawford et al., 2011; Crawford et al., 2013). At some small and difficult to access localities counts made outside the main breeding seasons were used if no other count was available for that year. Where more than one count was made at a locality in a year, the highest count was taken to represent the number of pairs breeding that year (Crawford et al., 2011). In Namibia, an occupied site was a penguin nest containing fresh eggs or chicks. In South Africa, an occupied site was considered active if it contained fresh eggs or chicks, or was defended by an adult penguin (or pair of penguins) that were not in moult, and considered potential if it was not active but showed recent signs of use, e.g. the presence of substantial fresh guano or nesting material, the recent excavation of sand from a burrow nest, the presence of many penguin footprints in its vicinity, or a combination of these factors. Breeding by African penguins is not always synchronous (Crawford, Shannon, & Whittington, 1999), so potential nests were counted as they may be occupied by pairs that have recently finished breeding or that are about to breed (Crawford et al., 2011). Groups of unguarded chicks (crèches) were divided by two to estimate the number of nest sites they represented (mean clutch size is ~1.8 eggs; Crawford et al., 1999; Shannon & Crawford, 1999), with remainders taken to represent an additional site, e.g. crèches of five and six chicks would both be taken to represent three nests (Shelton et al., 1984)."

NA are years in which counts were not conducted, either for logistical reasons or because the colony had gone extinct (Lambert's Bay after 2005) or not yet been (re)colonised (Robben Island before 1983, Stony Point before 1982, Simonstown before 1985). The 0.1 entered at the start or end of the time-series at Lambert's Bay, Robben Island, Stony Point and Simonstown are for cases where we know that the real popualtion size was zero at the either end of the time-series. They were entered to 'anchor' the state-space model used in the publication with which this dataset is associated to avoid knowingly overestimating the population at these sites. They do not represent real count data and should be omitted if using these data in some other framework.