Structured monitoring enhances regional trend and abundance estimates for Pygoscelis penguins around the Antarctic Peninsula

Census counts of Pygoscelis penguins on the Antarctic Peninsula and Scotia Arc are conducted both regularly around field stations and opportunistically from ships with the goal of assessing the breeding abundances and trends of these indicator species. Regional trends and abundances are often more policy-relevant than population-specific metrics, but survey patterns in the area have not been planned to optimize regional inference. We evaluated the ability of several monitoring schemes to describe regional trends and abundances of Pygoscelis penguins. Using a ‘pseudo-true’ time series as a starting point, we simulated time series of ‘observed’ census data for the period of 2000–2024. We then fit a hierarchical Bayesian population model to each dataset to assess how well each monitoring scheme recovered the known regional trends and abundances. Simulations were repeated across combinations of survey efforts, variances in inter-population abundances and census uncertainties. We found that implementing structured monitoring schemes would improve our ability to recover regional trends and abundances compared to the actual distribution of surveys conducted during the 2000–2024 period. For scenarios characterized by low variance in abundance such as the gentoo penguins on the central-west Antarctic Peninsula, we could recover trend and abundance estimates with considerably less survey effort if we deployed a monitoring scheme that cycled through all populations at an equal interval. For scenarios characterized by high variance in abundance such as the Adélie penguins on the north-east Antarctic Peninsula, surveys were optimized by splitting effort between satellite monitoring of the largest populations and cycling through the other populations. These findings are strongly influenced by the rapid increase in uncertainty regarding population abundance in the seasons after a census count, highlighting the importance of prioritizing sites where abundances are highly uncertain due to long intervals since the last survey.