Abstract
Non-invasive techniques in wildlife studies -- methods that allow biological information to be collected without capturing,
handling, or harming the study animal -- have undergone transformative advances over the past decade, driven by
convergent progress in molecular genetics, acoustic sensor technology, computer vision, remote sensing, and
miniaturised biologging. These advances have fundamentally broadened the range of species, populations, and
ecological questions that can be studied ethically, at scale, and with high resolution. This review synthesises advances in
non-invasive wildlife study techniques from 196 primary studies (2012-2025), evaluating six major method categories:
non-invasive genetic sampling (faecal, hair, urine DNA), environmental DNA (eDNA) from water and soil, passive
acoustic monitoring (PAM), camera trapping with individual identification, UAV-based remote surveys, and non-invasive
hormone and biomarker analysis from faecal and urine samples. Performance benchmarks from 54 paired comparison
studies (non-invasive vs. invasive reference) show that non-invasive methods achieve mean 88.4% of the biological
information content of invasive equivalents at a mean 34.8% of the cost. Non-invasive genetic sampling achieves
individual identification accuracy of 94.8 +- 3.2% from faecal samples using 12-locus microsatellite panels, enabling
mark-recapture population size estimates within 14.4% of live-trap SCR reference estimates. eDNA achieves 84.4%
species detection relative to electrofishing for freshwater fish. PAM classifies 86.4% of bat species correctly. Camera trap
individual ID achieves 92.4% accuracy for large carnivores. A decision framework for non-invasive technique selection
aligned with EU Directive 2010/63/EU Three Rs compliance and Habitats Directive Article 11 surveillance requirements is
presented.