%0 Journal Article
%T Statistical Assessment of Climate Change Effects on Biodiversity Using Generalized Linear Models
%A Bernabe Canqui Flores
%A Leonel Coyla Idme
%A Erika Belinda Ramirez Altamirano
%A Vladimiro Ibañez Quispe
%A Remo Choquejahua Acero
%A Godofredo Quispe Mamani
%A Percy Huata Panca
%J World Journal of Environmental Biosciences
%@ 2277-8047
%D 2026
%V 15
%N 2
%R 10.51847/XJoWyaN9M2
%P 17-26
%X Climate change is a major driver of biodiversity loss, altering species distributions, population abundances, phenology, and extinction risk. Statistical assessment of these changes requires models that can accommodate non-normal ecological response variables such as species counts, abundance indices, and presence/absence observations. Ordinary linear regression is poorly suited to biodiversity count data because it assumes normally distributed errors, constant variance, and unbounded continuous responses. In ecological monitoring, these assumptions are often violated by overdispersion, excess zeros, unequal sampling effort, and nonlinear climate responses. This article develops a generalized linear model framework for assessing the effects of temperature and precipitation on biodiversity metrics. The objective is to estimate climate effects on species richness, abundance, and occurrence while accounting for sampling effort, confounding variables, and appropriate error structures. The primary model is a negative binomial generalized linear model for count data with a log link and an effort offset. A complementary logistic generalized linear model is specified for presence/absence responses, with climate predictors from gridded climate products and adjustment for land cover, elevation, and year. Conceptually, biodiversity is expected to show a nonlinear temperature response, with richness increasing up to an optimum and declining under high thermal stress. Precipitation is expected to show a positive linear association in water-limited systems, while negative binomial dispersion estimates between 1.2 and 2.5 would indicate meaningful overdispersion relative to a Poisson model. Generalized linear models provide a statistically sound and interpretable framework for climate-biodiversity assessment. When paired with diagnostic testing, offsets, and validation, they can support conservation prioritization under future climate scenarios.
%U https://environmentaljournals.org/article/statistical-assessment-of-climate-change-effects-on-biodiversity-using-generalized-linear-models-rzee68b23uirkx6