Abstract:The Yellow River Delta wetland is the largest wetland ecosystem in the warm temperate zone of China and the most active region of land-ocean interaction among the large river deltas in the world, which is of typical continental monsoon climate with distinctive seasons. In recent years, low flow of the Yellow River has led to a decrease in water supply to the wetlands. As a result, Phragmites australis wetland ecosystem in the Yellow River Delta has deteriorated significantly. To prevent further degradation of the wetland and improve the wetland functions, an ecological restoration project was implemented in the region governed by Dawenliu Management Station in the Yellow River Delta Nature Reserve in July 2002 by supplementing abundant freshwater to the degraded wetlands. So far, enormous changes have occurred in the wetland ecosystem in the past 12 years following restoration project. A monitoring study was implemented along with the restoration project. By using the method of time-space mutual substitution, ecological traits of P. australis, such as height, density, litter production, and aboveground biomass were studied in non-restoration wetland (R0), restoration wetland for five years (R2007), and restoration wetland for ten years (R2002) of the Yellow River estuary. Results showed that the height of P. australis in the growing season generally increased in different restoration phases; the increase in R2002 differed from that in R2007 and R0 (P < 0.01). The density of P. australis had also obvious seasonal dynamics in different restoration phases. Besides, the density of P. australis in R0 differed significantly from those in R2002 and R2007. The aboveground biomass and its components in different restoration phases were in the order of R2002 > R2007 ≈ R0, which could be described by the parabola model (y = b0+ b1t + b2t2, R2 ≥ 0.90, P < 0.01). The proportion of aboveground parts of P. australis during the growing season showed no significant differences among different restoration phases. The percentage of stem biomass in the aboveground biomass was the highest, which in different restoration phases was in the order of R2002 > R2007 > R0. The variations of litter production could be explained by an exponential model (y = A0ebT + B0, R2 > 0.95, P < 0.01), but no significant differences were found among different restoration phases. In general, the litter production in these restoration phases was in the order of R2002 > R2007 > R0. This study also found that wetland hydrological regime (water depth and water quality) were the most important factors influencing the different ecological traits and the aboveground biomass of P. australis in different restoration phases. Therefore, it was suggested that, in the next step, the wetland restoration project should adopt the approach of less but more frequent replenishment. Furthermore, the restoration project should avoid the period of poor water quality .The results of this paper provide basic information for evaluating the effects of the ecological restoration project and improving the management of wetland restoration in the future. However, long-term monitoring is still needed to determine the inter-annual variations of ecological traits of P. australis within the restoration wetlands, which is essential to evaluate the effects of wetland restoration projects.