Background Riparian zones play a key role in removing none-point pollution and improving water quality. It is well known that these functions largely depend on the widths of riparian zones. However, due to farm land-use, building construction, road construction and other reasons, some riparian zones have been occupied in some rural areas. It results in that none-point pollution cannot be intercepted efficiently and water quality could be degraded.

Methods We investigated 3 rivers in Baimatang village of Tongxiang city, Zhejiang province, where is attributed as flat rural area. After we divided the 3 rivers into 6 reaches (T1, T2, T3, T4, T5, and T6) according to the land use patterns and cross-section, the removal rate of total nitrogen in each riparian zone were calculated with Riparian Ecosystem Management Model (REMM).

Results 1) The removal rate in T6 was the biggest, 76.03%. But in other 5 reaches, the removal rates of total nitrogen were < 75%. In practice, lands outside the riparian zones were used as farm field or house field along the 5 reaches, the riparian zones were limited and could not been widened. In order to remedy the width deficits, we set 5 experimental blocks with concrete sector bricks, ecological concrete balls, concrete bricks, Wufeng bricks and ecological bags in reach T1, T2, T3, T4, and T5, respectively. 2) Experiments showed that the removal rates by the combined measures of vegetation with ecological concrete balls or ecological bags were improved to 76.07%, 79.71% and their remedied widths were 10.34 m and 11.64 m, respectively. However, the removal rates by other three measures were still < 75% and their remedied widths were not obvious. The combined measures of vegetation with ecological concrete balls or ecological bags presented much better nitrogen removal efficiency. Meanwhile, these measures reduced the width deficits of current widths to satisfy the objectives of buffering the water quality. Among the 5 artificial ecological materials, the ecological concrete ball demonstrated the greatest increase in the nitrogen removal rate per unit width, which was 0.751 9%/m, and the compensation width was >10 m, and the width compensation efficiency was the highest, and the effect of width compensation was the best.

Conclusions In order to further improve the width compensation effect of the riparian zone in the study area, ecological concrete balls can be arranged in the riparian zone to optimize the construction measures of the riparian zone. Additionally, the fitting equations of optimal widths have been presented, it would be helpful to definite the exact width of a riparian zone for river restorations or managements.