Estimation of Parental Abundance Using Hierarchical Bayesian Modeling With Data Augmentation.

Area of Science:

• Ecology
• Population Genetics
• Computational Biology

Background:

• Pedigree-based methods estimate population dynamics by tracking genetic inheritance from parents to offspring.
• Pedigree accumulation estimators infer parental abundance within a cohort, useful for species difficult to sample across life stages.

Purpose of the Study:

• To evaluate a novel hierarchical Bayesian estimator for population dynamics using pedigree data.
• To compare its performance against existing estimators like Chao1 and iChao.

Main Methods:

• Utilized simulated data with varying sample sizes and sex ratios.
• Employed a hierarchical Bayesian framework with data augmentation.
• Assessed estimator robustness to errors in pedigree reconstruction, including false negatives.

Main Results:

• The Bayesian estimator demonstrated comparable accuracy and superior precision to Chao1 and iChao.
• It showed greater robustness to pedigree errors, particularly false negatives.
• The model provides insights into reproductive ecology via an explicit observation process.

Conclusions:

• Hierarchical Bayesian pedigree accumulation models are a powerful tool for population dynamics estimation.
• They offer advantages in precision, robustness, and ecological insight over existing methods.
• The parametric nature allows for information pooling and uncertainty propagation in complex models.