Seeing spots: Using Great Barracuda (Sphyraena barracuda) to assess abundance metrics and replicate independence for...

Baited remote underwater videos (BRUVs) are increasingly used to study fish assemblages. Therefore, understanding the biases and limitations of this methodology is critical, including questions about the derived metrics of abundance and independence among replicates. However, assessing these biases requires identifying individual fish, which is typically challenging. Here, we take advantage of the distinct spot patterns on Great Barracuda (Sphyraena barracuda) to identify individuals on BRUVs to compare the most common BRUVs abundance metric (MaxN) to an alternative method based on identified individuals (IND). We also assessed whether individual S. barracuda were seen on multiple cameras to determine if a commonly used spacing (500 m) was sufficient to assume independence between replicate BRUVs. S. barracuda were identified based on spot patterns on >2000 videos at 43 sites across 12 countries. MaxN and IND were highly positively correlated (0.93), but there were significant differences between the measures (p < 0.001) that became particularly apparent when MaxN was >4. IND abundances were, on average, 1.4 times greater than MaxN. Five sites had at least one S. barracuda appear on multiple BRUVs (<4% of videos), but only one occurrence during the same deployment on two cameras separated by >500 m. Thus, 500 m is a reasonable standard for independent sampling units for species with similar mobility to S. barracuda. While identifying individuals is time-consuming and not feasible for many species, it does improve the accuracy of abundance estimates, and we recommend its use when possible and where estimates of true abundance on videos are important.