Diving into the past: Recovering historic dive records for decade-to-decade comparison of diving behavior in Weddell seals

Over the past four decades, time depth recorders (TDRs) have become a useful tool to research the previously unexplored diving behavior of marine organisms. Of the early TDRs invented in the 1970’s, the Kooyman-Billups TDR (KBTDR) was the first placed on a free-ranging diving animal and it documented behavior using a pressure-sensitive arm that moved an LED light across a rolling window of film. With the advancements of TDRs from film-based instruments to computerized tags and biotelemetry devices, long-term comparisons of diving behavior have been exceptionally difficult. Here, I described a novel computational method for recovering dive records from film-based KBTDRs using data from the Weddell seal (Leptonychotes weddellii). This recovery process involved record scanning, image processing, and bias correction such that the historic data are comparable with dive data from modern devices. The efficacy of this method was assessed through a comparison with a previous analysis of the same records done by hand in 1992. Following recovery, I also explored long-term changes in dive metrics and diving patterns by comparing the 1970s records with dive data from the 2010s. I hypothesized that seals are diving deeper, longer, exceeding their aerobic dive limit more often, and spending more time at the bottom phase of their dive as large, energy-dense prey have likely become less abundant due to commercial fishing. From my decade-to-decade comparisons on the same population of seals, I found that seals appeared to be diving shallower in the mid/early austral summer, but deeper, longer, and more frequently in the late austral summer. Seals also appeared to be spending a higher percentage of their time at the bottom phase of their dive and foraging more efficiently since the 1970s. These behavioral changes could represent a shift in Weddell seals from the 1970s to 2010s towards foraging on a higher volume of smaller prey items, but more data are needed to explore this possibility. Considering climate change and other kinds of anthropogenic influence alter ecosystems over decadal time scales, this project highlights the importance of recovering and maintaining long-term datasets for conservation and management decisions.