Abstract. Diouri M, Ismaili B, El Halimi R. 2026. Diet composition and plant availability for Atlas deer in Tazekka National Park, Morocco. Biodiversitas 27 (8): d270821. https://doi.org/10.13057/biodiv/d270821. Understanding whether herbivore diets reflect resource availability or selective feeding is essential for evaluating habitat suitability before reintroduction. This study assessed seasonal diet selectivity of Atlas deer (Cervus elaphus barbarus) in the Bab Klati enclosure, Tazekka National Park, Morocco, by comparing fecal micro-histological diet composition with plant availability in autumn and winter. An estimation of the average species relative abundance in the environment (ei) was carried out in 20 and 21 randomly chosen plots in autumn and winter, respectively, in the enclosure. Average relative frequencies of species in the diet (di) were determined by fecal micro-histological analysis in autumn and winter. Fifteen heaps of feces were sampled in each season and analyzed for di. di and ei values were compared within and between seasons, adopting a GAMLSS-Beta-Inflated model for proportion comparison. The Vanderploeg and Scavia electivity index (Ei) was used to assess the preference levels of the enclosure plant species. In autumn, Ulex boivinii was the dominant dietary item, contributing 40.7% of the diet and showing the highest positive electivity value (Ei: 0.67), followed by Poaceae (composed of five species, 28.4%) and Cytisus triflorus (19.0%). In winter, Poaceae became the most consumed item (37.1%) and the most preferred ingredient (Ei: 0.61), followed by C. triflorus (30.1%) and U. boivinii (16.2%). Lavandula stoechas showed a negative selectivity and a moderate intake in both seasons. Fagaceae were available in the environment but absent from the diet, while Pteridium aquilinum showed negative selectivity or complete avoidance. The season effect on di and ei was significant only for C. triflorus (adjusted p-value<0.01). di and ei values were significantly different (adjusted p-value<0.01) for Poaceae, C. triflorus, and L. stoechas. These findings indicate that Atlas deer feeding in the enclosure is shaped by both resource availability and season-dependent plant selection. However, inference remains preliminary because of limited seasonal coverage, modest fecal sampling, unknown population structure, and lack of nutritional or behavioral measurements.