Baylis, S. M. | Search bioRxiv

bioRxiv2026

kinference: Pairwise kinship detection for Close-Kin Mark-Recapture

AO_SCPLOWBSTRACTC_SCPLOWClose-Kin Mark-Recapture (CKMR) is a statistical framework for estimating demographic parameters of wild populations. Instead of recapturing individuals, it relies on the identification of closely-related pairs such as parents and offspring, or siblings. By measuring how often such close-kin are "recaptured" among sampled animals (whether alive or dead), scientists can estimate demographic parameters such as census size, mortality rates, and connectivity. CKMR is starting to change fisheries and wildlife management by giving more reliable demographic information, even for many species that resist conventional approaches. Here we introduce the kinference R package, which provides a set of tools for finding close-kin pairs among thousands of samples each genotyped at thousands of SNPs, and for associated quality control. The CKMR context implies different requirements and assumptions to many other kinship programs. In particular, kinference accounts empirically for linkage without requiring a genome assembly, is able to estimate and control false-negative and false-positive probabilities, and can cope with null alleles. The package has been developed and used in numerous CKMR projects since 2017. This paper documents the assumptions, statistical algorithms, and intended workflow for kinference.

genetics↗

bioRxiv2017

Drones count wildlife more accurately and precisely than humans

Ecologists are increasingly using technology to improve the quality of data collected on wildlife, particularly for assessing the environmental impacts of human activities. Remotely Piloted Aircraft Systems (RPAS; commonly known as drones) are widely touted as a cost-effective way to collect high quality wildlife population data, however, the validity of these claims is unclear. Using life-sized seabird colonies containing a known number of replica birds, we show that RPAS-derived data are, on average, between 43% and 96% more accurate than data from the traditional ground-based collection method. We also demonstrate that counts from this remotely sensed imagery can be semi-automated with a high degree of accuracy. The increased accuracy and precision of RPAS-derived wildlife monitoring data provides greater statistical power to detect fine-scale population fluctuations allowing for more informed and proactive ecological management.

ecology↗