Search for: All records

Using tags within a mark-recapture framework allows researchers to assess population size and connectivity. Such methods have been applied in coastal zone habitats to monitor salt marsh restoration success by comparing the movement patterns of Mummichogs (Fundulus heteroclitus) between restored and natural marshes. Visible Implant Elastomer (VIE) tags are commonly used to tag small fish like Mummichogs, though the retention and survival of small fish using this method varies between studies, producing uncertainty during mark-recapture-based approaches. To address this, we conducted a laboratory experiment to determine the rate of tag loss and mortality of VIE tags on Mummichogs of two size classes (greater or less than 61 mm) and across different taggers. Tag loss and mortality increased over time, and the latter significantly varied between taggers. We then developed a predictive model, R package ‘retmort’, to account for the effect of this increase on mark-recapture studies. When adapted to a series of published works, our model provided rational estimates of tagging error for multiple species and tagging methods. Of the case studies the model was applied to (n = 26), 15 resulted in a percent standard error greater than 5%, signaling a significant percent of error due to uncounted, tagged animals. By not accounting for these individuals, recapture studies, particularly those that assess restoration efforts and coastal resilience, could underestimate the effects of those projects, leading to superfluous restoration efforts and erroneous recapture data for species with low tag retention and high mortality rates. 

Oysters are a vital keystone species in coastal ecosystems, providing significant economic, environmental, and cultural benefits. As the importance of oysters grows, so does the relevance of autonomous systems for their detection and monitoring. However, current monitoring strategies often rely on destructive methods. While manual identification of oysters from video footage is non-destructive, it is time-consuming, requires expert input, and is further complicated by the challenges of the underwater environment. To address these challenges, we propose a novel pipeline using stable diffusion to augment a collected real dataset with photorealistic synthetic data. This method enhances the dataset used to train a YOLOv10-based vision model. The model is then deployed and tested on an edge platform; Aqua2, an Autonomous Underwater Vehicle (AUV), achieving a state-of-the-art 0.657 mAP@50 for oyster detection.