Agroforestry and cover crops for improving surface water and groundwater quality

Abstract

Unsustainable agricultural practices and the rapid expansion of agricultural lands to satisfy the food demand of the growing world population have caused significant sediment, and nutrient losses from agricultural fields, harming the environment and human health. Conservation practices prevent water pollution by reducing runoff and retaining sediment and nutrients in agricultural watersheds. Since the environmental effects of agricultural pollution are severe and cost plenty of resources to countries, a complete understanding of the mechanisms to reduce non-point source pollution (NPSP) from agricultural fields is critical to preserve nature, human health, and economic resources. The objectives of this study were to (i) Determine the long-term effects of agroforestry buffers on runoff, sediment, and nutrient losses from agricultural watersheds under a corn (Zea mays L.)-soybean [Glycine max (L.) Merr.] rotation. (ii) evaluate the effects of cover crops on terraces with subsurface drainage on water quality under a corn (Zea mays L.)-soybean [Glycine max (L.) Merr.] rotation. (iii) determine the effects of agroforestry and grass buffers on shallow groundwater on a hillslope landscape under grazing management and (iv) simulate the benefits of agroforestry buffer scenarios for three different widths on groundwater nitrate-nitrite (NN) discharge to a lake. The first study was on three paired watersheds under a corn-soybean rotation with buffers consisting of tree + grass, grass-only, and no-buffers (control) had significant reductions in some non-point source pollution contaminants when the values were compared to predictions from a calibration period. The buffers were established in 1997 with the tree + grass buffers containing the following grasses: Redtop (Agrostis gigantea Roth), brome grass (Bromus inermys Leyss.), and birdsfoot trefoil (Lotus corniculatus L.) and trees [Pin Oak (Quercus palustris Muenchh.), Swamp White Oak (Q. bicolor Willd.), and Bur oak (Q. macrocarpa Michx.), located in the center of the grass-legume strips planted at 3 m apart]. Water samples were collected after runoff events. The tree + grass treatment reduced total suspended solids (TSS), NN, total nitrogen (TN) and total phosphorus (TP) losses in runoff by 62 percent, 25 percent, 64 percent, and 23 percent, respectively, while grass-only buffers reduced TSS, NN, and TP by 71 percent, 14 percent, and 33 percent. The second study consisted of two paired watersheds with tile-drained terraces and surface inlets under a corn-soybean rotation. The soils in the watershed had a shallow claypan; thus, the need to drain the watersheds with subsurface pipes. The instruments in the field were placed on November 2017. The treatment watershed had cover crops (CC) (Secale cereale L.) since November 2019, when the treatment period started. Water samples were collected after subsurface flow events and were stored at 4?C until analysis. The two watersheds had corn in 2019 and 2021 and soybean in 2018, 2020, and 2022. The CC reduced tile flow, TSS, orthophosphate (OP), and TP losses in tile flow by 27 percent, 26 percent, 67 percent, and 74 percent compared to no CC. Study 3 consisted of two adjacent small watersheds, each instrumented with a transect of three wells representing summit, backslope, and footslope positions. One of the watersheds had a tree + grass buffer with cottonwoods (Populus deltoides Bortr. ex Marsh.) and grasses [tall fescue (Schedonorus phoenix (Scop.)) Holub, red clover (Trifolium pretense L.), and lespedeza (Lespedeza Michx)]. The other watershed had a grass-only buffer with the same grasses in the tree + grass buffer. Four rows of trees were planted in 2001 at 3-m between and within row spacing. The six wells were monitored from November 2019 to January 2022. Tree + grass buffers reduced NN and TN in groundwater by 99 percent and 85 percent, respectively, while grass-only buffers reduced NN and TN by 94 percent and 62 percent compared to no-buffer. The air temperature significantly affected groundwater NN concentrations in footslope wells in both watersheds. Warmer temperatures were correlated with lower NN concentrations in groundwater, suggesting that NN plant uptake and denitrification might be the main processes influencing NN concentration. Study 4 consisted of a numerical model for hydraulic heads and concentrations in the same watersheds as Study 3. First, a steady-state model was calibrated in MODFLOW for groundwater flow and MT3DMS for solute transport. Hydraulic conductivity, specific yield, porosity, and evapotranspiration rates were adjusted at this stage. The hydraulic conductivity was first estimated through a slug test, and the porosity and specific yield were determined from soil cores in the laboratory. The steady-state model was used to build a 10-year transient flow and solute transport model to simulate the watersheds under four main buffer scenarios: tree + grass, tree-only, grass-only, and no-buffer with three buffer widths 7.5-, 15-, and 30-m. The tree + grass, tree-only, and grass-only scenarios with buffers 15-m wide reduced the total mass of NN discharged from the study unit to the lake by 99.14 percent, 98.51 percent, and 97.43 percent, respectively, compared to the no buffer scenario. Overall, increasing the buffer width from 15-m to 30-m decreased the NN discharge to the lake by 20-fold. The 7.5-m buffer had the greatest NN exports to the lake.