Changes in grey seal pup survey system: implications for pup production estimates

Colonies encompassing the majority of UK grey seal pup production (~95%) are monitored by SMRU aerial surveys. Classed (whitecoat versus fully moulted) counts from 4 to 5 surveys across a season are combined with life history and observation parameters to estimate pup production. Image quality and conditions impact these observation parameters. Indeed, the change from film aerial surveys (up to 2010) to digital (Hasselblad; H4D) aerial surveys (2012-2022) led to a clear change in these parameters. The SMRU aerial survey programme now utilises a new digital Phase One (PAS) camera system. This system was used, for the first time, to survey the colonies of East Scotland, Northeast England and Southeast England SMUs between late October and mid-December 2023. To facilitate comparison of the H4D and PAS systems, during one of the survey rounds two planes were used to survey colonies with both systems concurrently. The aim of this study was to assess the evidence for different observation parameters between the two methods. This would facilitate interpretation of the PAS-based pup production estimates and inform the potential range of these parameters for use in the current redevelopment of the pup production model. This was realized through two objectives: (1) colony-level comparisons of the classed counts between the two systems, and (2) estimation of the feasible values for the observation parameters for each system using pup-level comparison between counts from each system and assumed truth (drone and/or ground surveys). The pup production model is currently being redeveloped, partly to allow incorporation of a more complex observation process. This study provided prior information, for both the H4D and PAS, systems which will be utilised in the redeveloped model. Currently pup production is estimated using the observation process, and parameters, used for the digital survey time-series (H4D and PAS). The outputs of the two objectives suggest that the change in system will likely be associated with only modest changes in the observation parameters, and that any bias (compared to H4D) in estimates from the current pup production model would likely to be slightly upwards. The updated time-series of pup production estimates which included both H4D and PAS based estimates indicated that the PAS-based estimates were not at odds with recent trends.