Register now After registration you will be able to apply for this opportunity online.
eDNA stratified sampling design with remote sensing
The project aims to develop an open-source tool using satellite images and machine learning to optimize environmental DNA sampling in rivers. The tool will help researchers identify diverse ecological conditions, design effective sampling campaigns, and enhance biodiversity monitoring. The innovative approach promises to revolutionize eDNA sampling and inform conservation strategies.
- **Environmental DNA** (eDNA) analysis from rivers shows promise to become a great biodiversity observatory. However, a large fraction of eDNA studies are not making proper sampling plan before going to the field. An appropriate sampling methodology is the key to getting most information we can gather from eDNA sampling campaigns. In river environments, it is common that eDNA studies sample locations at equal intervals or sample based on samplers' insights, which may lead to sub-optimal sampling as regard to the species information collected given that the number of samples is generally limited.
- In this thesis project, we investigate how fine-resolution **satellite images** can help guide the sampling of biodiversity in large rivers globally, be supporting the planning of eDNA sample collection. We calculate **essential ecological variables** to determine the environmental conditions of the habitat of the species through the implementation of clustering algorithms and predefined thresholds. By utilizing **clustering unsupervised machine learning algorithms**, we aim to **develop an open-source tool capable of optimizing stratified eDNA sampling strategies** at a large scale.
- Our methodology utilizes satellite images to **identify and classify diverse ecological conditions** within river habitats. By systematically mapping ecological preference classes, we empower researchers to design sampling campaigns that capture the full spectrum of species diversity present in a given ecosystem.
- **Environmental DNA** (eDNA) analysis from rivers shows promise to become a great biodiversity observatory. However, a large fraction of eDNA studies are not making proper sampling plan before going to the field. An appropriate sampling methodology is the key to getting most information we can gather from eDNA sampling campaigns. In river environments, it is common that eDNA studies sample locations at equal intervals or sample based on samplers' insights, which may lead to sub-optimal sampling as regard to the species information collected given that the number of samples is generally limited.
- In this thesis project, we investigate how fine-resolution **satellite images** can help guide the sampling of biodiversity in large rivers globally, be supporting the planning of eDNA sample collection. We calculate **essential ecological variables** to determine the environmental conditions of the habitat of the species through the implementation of clustering algorithms and predefined thresholds. By utilizing **clustering unsupervised machine learning algorithms**, we aim to **develop an open-source tool capable of optimizing stratified eDNA sampling strategies** at a large scale.
- Our methodology utilizes satellite images to **identify and classify diverse ecological conditions** within river habitats. By systematically mapping ecological preference classes, we empower researchers to design sampling campaigns that capture the full spectrum of species diversity present in a given ecosystem.
Through the integration of satellite imagery and advanced data analysis techniques, our tool promises to revolutionize eDNA sampling campaigns. By enhancing the efficiency and effectiveness of biodiversity monitoring, researchers can gain deeper insights into river ecosystems and inform evidence-based conservation strategies.
Through the integration of satellite imagery and advanced data analysis techniques, our tool promises to revolutionize eDNA sampling campaigns. By enhancing the efficiency and effectiveness of biodiversity monitoring, researchers can gain deeper insights into river ecosystems and inform evidence-based conservation strategies.
Shuo Zong shzong@ethz.ch
Dr. Hansjörg Seybold hseybold@ethz.ch
Prof. Dr. Loic Pellissier loic.pellissier@usys.ethz.ch
Shuo Zong shzong@ethz.ch Dr. Hansjörg Seybold hseybold@ethz.ch Prof. Dr. Loic Pellissier loic.pellissier@usys.ethz.ch