An analytical framework for estimating aquatic species density from environmental DNA

Area of Science:

• Ecology
• Environmental Science
• Molecular Biology

Background:

• Environmental DNA (eDNA) analysis is a powerful tool for aquatic species detection and distribution.
• Current limitations exist in using eDNA for precise species density estimation.
• Inferring animal density from eDNA is the next frontier in ecological monitoring.

Purpose of the Study:

• To develop and assess a modeling approach for inferring animal density from eDNA data.
• To combine eDNA data with animal count data for robust density estimation.
• To evaluate the performance of different statistical distributions for modeling eDNA variability.

Main Methods:

• A modeling approach was developed combining eDNA and animal count data from selected sites.
• Cross-validation was performed using experimental data with known fish densities (carp).
• Field data from a stream salamander study were used to evaluate the model in natural settings.
• Two distributions (Normal and Negative Binomial) were assessed for modeling eDNA concentration variability.

Main Results:

• The Negative Binomial model significantly outperformed the Normal distribution for estimating carp density in controlled experiments.
• Application to field data showed greater imprecision but still yielded useful density estimates for stream salamanders.
• The Negative Binomial model proved more accurate for overdispersed eDNA data.

Conclusions:

• The developed modeling approach shows promise for inferring animal density from eDNA.
• The Negative Binomial distribution is recommended for modeling eDNA concentration variability.
• Further research on model development and sampling design optimization is crucial for advancing eDNA-based density estimation.