Abstract:Abies faxoniana coniferous forests are distributed in the eastern Tibetan Plateau and the upper reaches of the Yangtze River. These forests play important roles such as conserving water and soil and maintaining habitat stability for the giant panda (Ailuropoda melanoleuca). A. faxoniana seedling and sapling banks determine the population size and strongly influence the succession, regeneration, and restoration of the coniferous forests. Gaps created by disturbance in the forest canopy are important sites for tree regeneration from seeds. The seedling structure of A. faxoniana can reflect plant development and population dynamics, as well as the relationship among the population, environment, and each species' position in the coniferous forests. Comprehensive studies of A. faxoniana seedling population structure under forest canopy or in canopy gaps where habitat is regenerating can reveal population size, survival conditions, and biotope suitability. The objective of this study was to elucidate seedling population structure and quantitative dynamics, living status, and development trends, and evaluate the effects of canopy gaps on A. faxoniana regeneration. We investigated seedling (H ≤ 100 cm) population structure and natural regeneration of A. faxoniana in the eastern Tibetan Plateau. Three strip transects (300 m×2 m), including forest canopy and canopy gaps, were established in a subalpine coniferous forest in the Wanglang National Nature Reserves, Pingwu County, western Sichuan Province. The size structure of the A. faxoniana seedling populations were tabulated and analyzed for canopy gaps (CP population), under forest canopy (FC population), and the whole population. Based on these data, we estimated the values of population parameters, including the survival curve, mortality rate, killing power, survival rate, mortality density, and hazard rate. The results showed that the age structures of the three populations were spindle shaped, suggesting an increasing population for A. faxoniana seedlings; however, there were fewer individuals in age classes 1-3 than in age class 4. The survival curves of CG, FC and the whole population tended to be the Deevey II type. The observed life expectancies were the highest in 1-year olds, and were higher in the FC than the CG population. The mortality rate, vanish rate, cumulative mortality rate, and hazard rate increased with age class, and the survival rate and mortality decreased. In the same age classes, the CG population had higher values for cumulative mortality rate and hazard rate than those in the FC population. The ChenXiaode's quantitative index indicated that the three populations were growing, and the dynamic index of population size structure for the whole population was Vpi=0.1059. The sensitive index of population structure dynamic (Pmax) and dynamic index of population size structure (Vpi') under random disturbance were the highest in the CG population. This indicated that the CG population was more sensitive to environment disturbance and better able to increase under random disturbance than the FC population. Spectral analysis of three populations showed that there was a marked periodic fluctuation in the process of natural regeneration. Given the roles of the canopy in A. faxoniana population dynamics and the quantity of A. faxoniana seedlings, random disturbance enhanced the capacity for population increase. Therefore, our results supported the creation of appropriate canopy gaps, which would facilitate successful A. faxoniana population regeneration.