Related Experiment Video
Updated: Oct 18, 2025
15:31
Tissue Collection of Bats for -Omics Analyses and Primary Cell Culture
Published on: October 23, 2019
12.5K
Optimizing noninvasive sampling of a zoonotic bat virus
John R Giles1,2, Alison J Peel2, Konstans Wells3
1Department of Epidemiology Johns Hopkins University Bloomberg School of Public Health Baltimore MD USA.
Ecology and Evolution
|October 1, 2021
Summary
Common bat virus sampling methods overestimate prevalence due to pooled samples. Optimizing under-roost sheet design can improve accuracy for bat-borne zoonoses surveillance.
Area of Science:
- Disease Ecology
- Epidemiology
- Wildlife Health
Background:
- Bat-borne zoonoses are a growing concern, necessitating accurate ecological and epidemiological studies.
- Current field methods often use pooled bat excreta samples, introducing bias and complicating viral dynamic analysis.
Purpose of the Study:
- To investigate bias in spatial sample pooling for bat-borne viruses using Hendra virus as a case study.
- To assess the accuracy of different under-roost sampling designs and identify optimal strategies.
Main Methods:
- Utilized generalized additive models and field data from individually captured bats and pooled urine samples.
- Employed theoretical simulation models of bat density and under-roost sampling to analyze bias drivers.
- Compared commonly used sampling designs with proposed stratified random designs.
Main Results:
- The most common sampling design overestimated viral prevalence by 3.2 times and exhibited 5-7 times higher positive bias compared to other designs.
- Spatial autocorrelation and bat roosting behavior significantly contributed to sampling bias.
- A stratified random design with 30-40 pooled samples from 80-100 sheets (0.75-1 m²) is proposed to minimize bias and false negatives.
Conclusions:
- Widely used under-roost sampling methods are sensitive but lack specificity, limiting insights into viral dynamics.
- Minor adjustments to sampling designs, including reducing sheet size and increasing sheet number/distribution, can improve true prevalence estimation.
- Findings offer critical insights into optimizing spatial sample pooling for disease ecology studies influenced by pathogen prevalence, host density, and aggregation patterns.
