Resprouting sustains cork oak seedlings while recent grazing limits resprouting and survival

Regeneration failure is a major bottleneck in Mediterranean oak woodlands. Cattle can hinder or promote recruitment, depending on the spatial, temporal, and intensity patterns of grazing. Herbivory theory predicts that repeated defoliation and trampling deplete seedling reserves, whereas resprouting can extend survival; yet field studies rarely separate intensity from recency or pair long-run grazing records with individual fates and microhabitat/climate context.

We test how management-driven heterogeneity shapes cork-oak ( Quercus suber ) seedling survival and resprouting, combining 12 years of paddock-level grazing records with individual tracking of 8431 seedlings across 24 rotational paddocks. We fitted Bayesian mixed-effects survival models relating oak seedling lifespan to grazing pressure × history (moderate ≤150; high >150 LSU·ha ^-1 ·days·yr ^-1 ), with covariates (initial height, resprouting status, shrub distance, cattle dung counts as a proxy of very recent grazing, and 1-month SPEI as recent water balance), and Bayesian logistic mixed models relating resprouting probability to grazing treatments.

Survival was consistently lower in grazed than ungrazed paddocks and declined along management gradients: median lifespan fell from 460 (moderate grazing) to 256 days (high), and from 460 (old grazing; two-year exclusion) to 199 days (recent). A two-year exclusion raised survival under moderate pressure but was insufficient where pressure remained high, indicating legacy effects and that recovery windows must scale with pressure.

Resprouting was the dominant axis of persistence: resprouters lived >5× longer than non-resprouters (2351 vs 460 days). Taller seedlings lived longer, shrub proximity conferred a modest benefit.

Drought exacerbates mortality, whereas wetter recent periods (higher SPEI) markedly boosted survival. Cattle reduced the odds of resprouting, with the strongest penalty under recent use.