The sustainability of artificial seed banks for three riparian herbs under flooding stress

To provide evidence for the feasibility of artificial seed banks for ecological restoration of fluctuation zone, a flooding experiment was conducted in which the variation in seed viability, protein content, soluble sugar content, and starch content for Cynodon dactylon, Hemarthria sibirica, Pycreus globosus artificial seed banks was tested. The flooding experiment included 0, 5, 6, 7, 8, and 9 month flooding treatments that simulated the flooding rhythms at different altitudes in the Three Gorges Reservoir region. The results showed that all three artificial seed banks had good sustainability under short-term submersion. Seed viability of the three artificial seed banks remained relatively high after submersion for 5~6 months. Seed activity of the C. dactylon seed bank ranked the highest among the three seed banks. However, seed activity of all artificial seed banks decreased dramatically after submersion for 7 months and decreased further with increasing submersion time. The seed activity of C. dactylon, H. sibirica, and P. globosus declined by 34.2%, 43.0%, and 63.8%, respectively, when submersed for 9 months. Submersion also induced significant reductions in protein, starch, and soluble sugar content of seeds (P<0.05). A dramatic decline occurred when seed banks were submersed for 7 months. The C. dactylon seed bank had the best sustainability, followed by H. sibirica and P. globosus. An additional correlation analysis indicated that there were significant negative linear correlations between submersion time and seed viability for H. sibirica and P. globosus seed banks. Seed nutrients also exhibited a negative linear correlation with the viability of seed banks, with the degree of correlation depending on the species.