Benchmarking beta-diversity measures and transfer functions for sedimentary ancient DNA

Area of Science:

• Paleoecology
• Molecular Ecology
• Environmental DNA Analysis

Background:

• Analyzing past ecosystems aids understanding of biodiversity and environmental change.
• Sedimentary ancient DNA (sedaDNA) offers insights into past biodiversity beyond fossils, enabling reconstruction of ancient environments and ecosystem functions.
• Transfer functions calibrate modern data to reconstruct past environmental parameters from biodiversity data, but sedaDNA presents challenges due to limited quantity and fragmentation, leading to noisy datasets.

Purpose of the Study:

• To test the impact of sedaDNA noise on beta-diversity measures and transfer function performance.
• To identify robust analytical workflows for sedaDNA data.
• To provide a framework for optimizing postprocessing in sedaDNA studies.

Main Methods:

• Simulated sedaDNA reads count matrices from synthetic and empirical datasets.
• Tested 464 combinations of counts transformations, beta-diversity indices, and ordination methods.
• Assessed performance in separating ecological signal from noise and predicting environmental conditions.

Main Results:

• Commonly used workflows are sensitive to sedaDNA noise.
• Combinations of methods incorporating recent ordination techniques are robust to sedaDNA noise.
• These robust methods yield improved transfer functions for environmental reconstruction.

Conclusions:

• The choice of analytical methods significantly impacts the reliability of sedaDNA-based environmental reconstructions.
• Recent ordination methods offer a promising solution for handling noise in sedaDNA data.
• The study provides a framework for developing effective postprocessing strategies for sedaDNA analysis.