Dynamics of runoff and sediment trapping performance of vegetative filter strips: Run-on experiments and modeling

- The performance of vegetative filter strips (VFSs) is species- and time-specific.
- Both soil physical properties and vegetation characteristics affect VFS performance.
- Flume experiment and VFSMOD was applied for white clover and cocksfoot VFS.
- Neither soil physical properties nor vegetation characteristics is negligible.
- VFSMOD explained the dynamics of VFS performance well.

Vegetative filter strips (VFSs) are a labor-saving and cost-effective agricultural best management practice to trap water runoff and sediment from the source areas. They also provide forage and/or fuel and are therefore potentially profitable for land owners. VFSs are however a dynamic system: the runoff delivery ratio (RDR) and sediment delivery ratio (SDR) vary with growth stage and vegetation types. The impacts of vegetation characteristics as well as soil physical properties modified by vegetation growth, on the RDR and SDR of VFS were evaluated by a flume experiment. Two plant species (cocksfoot (Dactylis glomerata L.) and white clover (Trifolium repens L.)) were tested at three stages in the growing season of 2016 (May, July, and August). The measured RDR and SDR were compared with the simulated results from Vegetative Filter Strip Modeling System (VFSMOD). In the early stages of the growing season, the cocksfoot formed a dense network of stems with high strip Manning's roughness faster than white clover. The runoff and sediment trapping effects of the white clover VFS were greater than that of cocksfoot VFS in all the three stages (lower RDR and SDR). This is likely attributed to strongly tillering, creeping stem posture and high infiltration capacity of the white clover VFS. VFSMOD simulated the RDR and SDR reliably except under low vegetation coverage conditions (white clover in May). The results suggested that (1) both soil physical properties and vegetation characteristics should be considered for the species-specific, temporally variable performance of VFS; and (2) when using VFSMOD inform the VFS design, modelers should take the dynamics of vegetation, mainly through vertical saturated hydraulic conductivity, stem spacing and strip Manning's roughness into account, and select parameters that reflect the actual field conditions.

544