A geospatial approach to assessing erosion-induced Dermochelys coriacea nest loss on dynamic coastlines
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background Seven species of marine turtles are recognized globally, six of which are listed as endangered or threatened under the IUCN Red List. Leatherback turtles ( Dermochelys coriacea ) exhibit the lowest hatching success rate among these species, with nest failure strongly influenced by nesting site characteristics and geomorphological processes. Trinidad hosts some of the highest-density leatherback nesting sites worldwide, yet coastal erosion, exacerbated by sea-level rise, poses a growing threat to nest viability. Despite recognition of this vulnerability, quantitative assessments of erosion-induced nest loss remain scarce in the Caribbean. Methods We developed a geospatial framework to assess nest exposure and loss due to erosion at Matura Beach, Trinidad. High-resolution drone imagery was used to generate digital surface models (DSMs) and orthomosaics, which were combined with spatial nest distribution data collected across three nesting seasons (2023–2025). Stable nesting areas were identified and evaluated across annual erosion–accretion cycles. Results Analysis revealed that shoreline retreat accounted for between 15% and 45% of total annual nest loss during the study period. The integration of drone-derived DSMs with nest distribution data provided fine-scale insights into spatial vulnerability and highlighted dynamic shifts in nesting habitat stability. Conclusion This framework demonstrates the utility of drone-based monitoring for dynamic coastal habitats and offers a replicable method for conservation planning across similarly vulnerable nesting sites. Findings support adaptive management strategies, including targeted nest relocations and erosion mitigation, to safeguard critical nesting areas for the northwestern Atlantic leatherback population.