Interdisciplinary Uses and Kernel-Based Enhancements of Line Transect Sampling
Omar M Eidous
Line transect sampling is a widely used statistical framework for estimating population densities and abundances, particularly in wildlife and ecological studies. Traditionally relying on parametric detection functions, the methodology has evolved through nonparametric, kernel-based, histogram, and semi-parametric approaches, enhancing flexibility and robustness, especially when classical assumptions such as the shoulder condition are violated. This review synthesizes theoretical foundations, methodological developments, and interdisciplinary applications of line transect sampling, with emphasis on kernel-based innovations pioneered by Eidous and collaborators.Applications extend beyond ecology to reliability analysis, linear degradation models, environmental monitoring, epidemiology, industrial inspection, astronomy, and statistical approximation. Recent advances, including adaptive bandwidth selecion, double-kernel methods,bias correction, and two-parameter detection functions, highlight ongoing developments and future research directions, demonstrating that line transect sampling constitutes a general framework for density estimation across the sciences.
