Name: Peixoto S

Biodeserts supervisors: Velo-Antón G

Co-supervisor:  Jarman S

Title: Improving eDNA methodologies to detect amphibians in turbid waters

Institution: University of Porto

Status: Completed




Amphibians are a focus of conservation programs worldwide. Among all vertebrate groups, amphibians are the most threatened, with about 40% of species facing extinction. The large diversity of this group and their ecological importance requires accurate knowledge of species distributions in order to develop conservation and management actions that can mitigate amphibian decline. This requires an active engagement between amphibian specialists and researchers from other areas in order to improve survey and monitoring strategies. Environmental DNA has emerged as a powerful, non-invasive tool for species detection, which will potentially overcome the limitations associated with traditional biodiversity surveys.
Despite its increasing application in recent years, most eDNA surveys in aquatic systems have focused on low-turbidity waters and the efficacy of eDNA methods in turbid systems remains understudied. Regardless of the high sediment loads, turbid aquatic environments can harbour a large diversity of amphibian species. The challenges of biodiversity assessment using eDNA in turbid environments should be better understood, and the comparison of currently available sampling methods as well as the development of new methods needs to be addressed. In the first chapter of this thesis, three capture methods (precipitation, filtration with disc filters and filtration with capsules) were evaluated across a turbidity gradient. Sampling was conducted in an amphibian meta-population system composed of temporary puddles and small ponds at Mindelo Ornithological Reserve, northern Portugal. The performance of each method was evaluated based on both total captured eDNA and detection rates of a single target species - Salamandra salamandra – with two approaches (qPCR and high-throughput sequencing). Compared to the two filtration methods, precipitation underperformed both in terms of total eDNA captured and species detection rates. Capsules captured a higher quantity of total eDNA than disc filters, but species detection was similar among the two methods. The lower price and the possibility to detect species in highly turbid environments seem to favour the use of disc filters for sampling in aquatic turbid ecosystems, however, more research is needed in the future to validate this conclusion. Although the sensitivity of the qPCR assay was higher than that of high-throughput sequencing, for eDNA captured using disc filters, no difference between assay sensitives was noted otherwise.
The efficiency of eDNA methods in aquatic systems can be affected by several biotic and abiotic factors. Understanding the processes limiting eDNA detection is crucial in order for accurate eDNA-based surveys. Accounting for those factors by optimizing sampling designs will provide better estimates of species distribution and increase the reliability of eDNA biodiversity monitoring. In the second chapter of this thesis, the probability of detection of amphibian species was assessed according to the volume of water filtered and several environmental variables (pond area and depth, water clarity, conductivity, pH and temperature). Sampling was conducted in a natural park on the coastal plain of southwest Portugal. A significant influence on species detection was observed for most of the variables measured, but not in a consistent pattern among species. Generally, increasing pond depth and water conductivity decreased species detection, while increasing pond area and pH increased species detection. The only variable influencing detection in more than one species was water clarity, exhibiting opposite patterns.
Overall, this study brings new insights into eDNA research in aquatic environments. The work developed in this study provides accurate comparisons between three sampling methods, offering important elucidations regarding the best method for biodiversity assessment in turbid aquatic environments as well as the effects of environmental variables on amphibian eDNA detection in Mediterranean temporary ponds. This work has also helped to identify new research areas and highlight important aspects that have been overlooked by previous studies, potentially leading to erroneous conclusions. This will provide the development of optimized sampling designs in the future that can increase the efficiency and reliability of eDNA biodiversity monitoring in aquatic habitats and help combat global amphibian declines effectively.