Long-term analysis of bird diversity across urban and non-urban habitats in R
A quantitative ecology research project for BIO 411 analyzing how bird species diversity differs across urban, desert, riparian, and agricultural habitats over time, using long-term CAP LTER bird monitoring data. The work controls for differences between two survey strategies (ESCA broad regional sampling and PASS neighborhood-focused surveys) so that habitat comparisons reflect ecology rather than sampling effort. The central deliverable is a scientific research poster, supported by a series of R-based statistical check-in assignments covering distributions, regression, ANOVA, GLMs, and community ordination. **Highlight:** Species-richness decline over time: R2 = 0.12, p = 0.07
Bird monitoring records from the CAP LTER were cleaned and merged in R, with abundance and richness standardized to account for the different sampling intensities of the ESCA and PASS survey strategies. Habitat differences were assessed with ANOVA after verifying normality and homogeneity-of-variance assumptions, and long-term change was tested with linear regression backed by residual and QQ diagnostics. Results were visualized with ggplot2 and synthesized into a research poster, with multiple regression, mixed-effects models, and machine learning (Random Forest, Gradient Boosting) outlined as future work.
The poster reports a weak, non-significant decline in species richness over time (R2 = 0.12, p = 0.07); urban (PASS) sites showed lower richness but high bird abundance, while riparian habitats held the highest richness, pointing to water and vegetation cover as key drivers. See docs/BIO411Poster.pptx_4_.pdf for the full deliverable.