Abstract: Glyptostrobus pensilis, a monotypic relict tree species endemic to China, has extremely sparse populations in the wild. The world’s largest natural G. pensilis forest is distributed in the Zhuzhou Island Glyptostrobus Forest Nature Reserve, Zhuhai, Guangdong Province. However, artificial plantations of G. pensilis currently exhibit significant decline. To clarify the survival status and dynamic characteristics of G. pensilis populations, we constructed age structure diagrams, compiled static life tables, and applied survival function analysis and time series prediction to analyze population dynamics and driving mechanisms, aiming to provide a scientific basis for conservation and management. The results showed that the artificial G. pensilis population exhibited a “bell-shaped” structure, with fewer juvenile and elderly individuals, and the highest number observed in age class V (20 cm≤DBH<25 cm). Understory natural regeneration was severely limited. The static life table indicated that mortality and disa-ppearance rates initially increased and then decreased, peaking at age class Ⅵ (25 cm≤DBH＜30 cm) and age class Ⅹ (DBH≥45 cm), respectively. Life expectancy declined with increasing age class, and the survival curve aligned with the Deevey-Ⅱ type. Spectral analysis demonstrated significant periodic fluctuations in population dynamics, dominated by the fundamental wave A1 and driven by the third harmonic, with age class V (20 cm≤DBH<25 cm) identified as the critical fluctuation phase. Time series prediction showed that population size increased during age classes Ⅱ-Ⅳ, reaching maximum size at class V, followed by a continuous decline from age classes Ⅵ-Ⅷ onward. Although the population temporarily maintained growth, long-term survival risks arose from insufficient juvenile recruitment, environmental stochasticity, and physiological senescence. To enhance population resilience, the following conservation strategies are recommended, inlcuding artificial propagation, habitat restoration, and invasive species control.

Key words: Glyptostrobus pensilis, population structure, static life table, survival curve, spectral analysis, time series prediction