Robust abundance estimation of harvested populations from low quality time series data: A red deer case study

• Adequate population management requires reliable estimates of population size. • Novel methodology for estimating population abundance from noisy time-series data. • Requires few assumptions about initial population and the data collection process. • Validated by use of Norwegian time-series data on red deer reported by hunters. • May facilitate effective management of harvested populations. Reliable estimates of the size and composition of harvested populations over time are key to designing adequate population management plans, regardless of management objectives. In Norway, a national system for collecting and analysing hunter-reported data on red deer ( Cervus elaphus ) has been operational for about 20 years. The system was expected to provide population metrics that would substantially improve deer population management routines at the municipal level. This has proven to be challenging when using existing state-of-the-art estimation methodology. The main reasons are that the variation in the observation data is generally much larger than population abundance variability, and that one does not have a clear understanding of the stochastic process generating the observation data. Here, using hunter-reported observation data and harvest data from six Norwegian municipalities collected in the period 2007–2023, we show that a straightforward estimation methodology based on population modelling can produce robust abundance estimates despite frequent low quality of the observation data. Its major assets are that it does not involve strong assumptions about the stochastic processes underlying the observation process and that it does not involve assumptions about initial population size and structure in terms of prior statistical distributions. We anticipate that the method can be applied in several other population management contexts, and we think that the results offer fresh perspectives on to what extent noisy citizen-collected time series data can be used to inform management decisions.