Working package 2 consists of two columns: soil core sampling of vegetated filter strips (VFS) and an experimental approach. For the latter, we conducted artificial runoff experiments using grassland plots (2 x 5 m) that were subjected to artificial runoff. The water was applied through an overflow tank and was spiked with bromide, to be able to distinguish it from autochthonous water in the soil, as well as phosphate, to mimic nutrient enriched agricultural runoff. Our main interest was on how concentrated runoff (i.e., due to flow convergence in the field or at the field edge) affects flow characteristics and VFS performance. To this end, we used three different widths of the overflow tank and analysed total water budget and flow velocities, as well as bromide and phosphate concentrations at the end of the plots.
The outdoor experiments are completed, currently the data is processed and analysed. We expect that flow concentrations have a significant effect on runoff characteristics, e.g., less infiltration and, in turn, a higher amount of runoff water, accompanied by higher flow velocities and less time for soil/water interaction processes. All these aspects would have direct consequences for VFS performance. However, flow concentration is rarely accounted for in VFS design guidelines and recommendations.
Together with colleagues from the UK, we published an article in Frontiers in Environmental Science in which we make a case for a more holistic approach in vegetated filter strip (VFS) research, design, policy, and implementation.
The paper can be downloaded for free at: https://www.frontiersin.org/articles/10.3389/fenvs.2022.764333/full
Focussing on phosphorus (P), we describe the downfalls of current approaches and ways for improvement. In a VFS, the amount of incoming P must not exceed the amount of P that the soil can retain and the amount of P that can be removed, e.g., by harvesting the vegetation. This requires comprehensive VFS designs in accordance to actual runoff and erosion patterns and a more flexible positioning in line with local conditions. Designs and evaluations of VFS that match the complexity of the processes involved are crucial for the effective and long-term protection of surface waters.
We hope that this article stimulates a much needed discussion on the potential and limits of state-of-the-art VFS.
In addition to the two sites that were sampled last year, we took further soil core samples at six sites across Western Lower Austria. For the additional sites we used an adapted sampling array, focussing on samples from within the area of concentrated runoff (and erosion) and outside. A convergence of the flow, e.g. due to topography (thalweg), tillage, or other factors, is rather the norm than the exception, causing a concentration of runoff and erosion—coupled with nutrients and other pollutants. With the data from the soil cores, we aim at a better understanding of nutrient retention processes in real-life buffer strip soils and, eventually, improved planning and design recommendations for buffers that are truly effective in protecting surface water from agricultural inputs.
Artificial runoff experiments are a part of working package 2, in which we seek to analyse the response of buffer strip soils to different runoff scenarios. To improve the meaningfulness of the data, we plan to combine undisturbed soil monoliths to larger soil plots, taking advantage of both the flexibility of indoor experiments and the realism of undisturbed soils. To our knowledge, this has not been done before. Therefore, we started a preliminary trial to ascertain that combining blocks of soil does not interfere with the runoff characteristics.
To this end, we took six undisturbed soil monoliths from a buffer strip. Half of them were cut in the middle and then recombined. The monoliths were then used for a runoff experiment, during which the blocks received a constant flow from a tank, with water spiked with phosphorus (P) and salt tracers. Most water left the monoliths as surface runoff, but substantial amounts were also recorded as drainage water (passing through the soil body, probably due to the natural macropore network) and bypass water (e.g. water that leaves the monolith at its sides). Interestingly, we noticed an enrichment of the surface runoff with P, which means that the buffer soil acted as a P source at it surface, rather than a sink.
We found no significant differences between cut and uncut monoliths. In fact, it was apparent that the inherent variability between the monoliths (due to the inevitable spatial heterogeneity of the soil) was much larger than any effect that the cutting could have had. We conclude that a careful combination of soil monoliths is a valid procedure and plan to further pursue this approach for the main runoff experiment starting next year.
The results from the soil cores of the first sampling site are back from the laboratories and show some interesting trends and gradients along all three dimensions. Final results will be available next year, after the sampling and analysis is complete, however, preliminary results suggest that P levels are substantially lower in buffer strips throughout all P pools. Especially, deeper layers appear to not only have a higher capacity for P uptake (sorption) but also a lesser degree of P saturation, highlighting the potential of these sub-surface areas for P retention and the importance of infiltration and, thus, the often-neglected vertical dimension.
In November, we started with the soil sampling for work package 2. For the first two sites, we have chosen an intensive sampling scheme, using soil cores along transects from the field to the buffer strip, as well as inside and outside of the area of concentrated runoff. Together with samples from different depth classes from the soil cores, we get a 3D representation of the field and buffer strip. This labour-intensive approach is rarely seen in buffer strip research, although it provides high quality data and ample opportunities for in-depth analyses. The samples will then be analysed for various physical and chemical parameters, with a focus on phosphorus (e.g. different P pools, degree of P saturation, P sorption index). Further sites will get sampled next year.