Mark-Resight Data. From 1979 to 2021, SRW were annually surveyed in October by helicopter between Nature’s Valley and Muizenberg, False Bay along the south coast of South Africa. All individuals sighted were noted, and callosity patterns of females with calves were carefully photographed for identification purposes. Photos were compared to a long-running photo-identification catalog and matched. Over the 42-year period, this resulted in a mark-recapture data set of 2476 unique females, of which 139 have known birth years. Survey methods and photo identification details are available in (1) and (2).

Mark-recapture records for NARW spanned approximately the same period. Starting in the mid-1970s a photographic catalog of all NARWs began, and the observation effort intensified and became more organized in the 1980s. Unlike the SRW mark-recapture data, observations for NARW are assembled by a large group of collaborating scientists working across the species’ entire range. The photographic catalog and data are maintained by the North Atlantic Right Whale Consortium, which made the mark-recapture records of females available for this analysis. All NARW encountered by members of the consortium at any time during the year and in any geographic region, were photographed. These photos are compared to the master catalog maintained by the consortium and matched to individuals on file. This much more intense survey effort results in many observations per year for most individuals. Here, we analyze only observations from females.

Following (3), these multiple detections per year were reduced so that the detection period was simply annual, and individuals in a given year were either detected (1) or not detected (0). In total, this data set spanned 1974 to 2020 and included 328 unique females, of which 205 had known birth years. Further details on data collection and the photo-identification catalog can be found in (3) and (4).

One notable difference between the NARW and SRW mark-recapture time series is the size and trajectory of the respective populations that were observed. As the SRW population is large and growing, new individuals have steadily been incorporated into the SRW dataset since the study’s inception, most of which do not have known birth years. Individuals with unknown birth years detected after 1990 also have short observation periods (relative to lifespan), and thus provide minimal information about the shape of the survival function. Therefore, we limited the analysis of SRW data to individuals with known birth years or individuals first observed (but without a known birth year) prior to 1985. The resulting SRW dataset included 361 individuals, 139 with known birth years and 222 with unknown birth years but observation periods >35 years. Similar subsampling was not required for NARW because most of the individuals in the dataset have known birth years and nearly all individuals added to the mark-recapture database without known birth years were added before 1990 and thus have long observation periods relative to their lifespans. Both the SRW and NARW mark-recapture datasets have had relatively steady observation efforts over their respective time series with no missing observation years. Thus the analyzed samples for SRW and NARW were of similar size, length, and quality and, as they both included hundreds of individuals, were more than adequate for estimation of longevity.

References