ABSTRACT
Managed forests in southern U.S. are a potential source of lignocellulosic biomass for biofuel production. Changes in management practices to optimize biomass production may impact the quality of waters draining to nutrient-sensitive waters in coastal plain regions. We investigated shallow groundwater quality effects of intercropping switchgrass (Panicum virgatum L.) with managed loblolly pine (Pinus taeda L.) to produce bioenergy feedstock and quality sawtimber in a poorly drained soil of eastern North Carolina, U.S.A. Treatments included PINE (traditional pine production), PSWITCH (pine-switchgrass intercropped), SWITCH (switchgrass monoculture) and REF (mature loblolly pine stand). Each treatment was replicated three times on 0.8ha plots drained by parallel-open ditches, 1.0-1.2m deep and 100m apart. Water samples were collected monthly or more frequently after fertilizer application. Water samples were analyzed for organic nitrogen (ON), ammonium N (NH4+- N), and nitrite+nitrate N (NO3-+ NO2-- N), ortohophosphate phosphorus (OP), and total organic carbon (TOC). Overall, PSWITCH did not significantly affect shallow groundwater quality relative to PINE and SWITCH. ON, NO3-+ NO2-- N, and TOC concentrations in PSWITCH, PINE and SWITCH were substantially elevated during the two years after tree harvest and site establishment. The elevated nutrient concentrations at the beginning of the study were likely caused by a combination of rapid organic matter decomposition of the abundant supply of post-harvest residues, warming of exposed soil surfaces and reduction of plant nutrient uptake that can occur after harvesting, and pre-plant fertilization. Nutrient concentrations returned to background levels observed in REF during the third year after harvest.
ABSTRACT
The movement of the herbicide metolachlor [2-chloro-N-(2-ethyl-6-methylphenyl)-N-(2-methoxy-1-methylethyl)acetamide] via runoff and leaching from 0.21 ha plots planted to corn on Mississippi River alluvial soil (Commerce silt loam) was measured for a 6-year period, 1995-2000. The first three years received normal rainfall (30 year average); the second three years experienced reduced rainfall. The 4-month periods prior to application plus the following 4 months after application were characterized by 1039 +/- 148 mm of rainfall for 1995-1997 and by 674 +/- 108 mm for 1998-2000. During the normal rainfall years 216 +/- 150 mm of runoff occurred during the study seasons (4 months following herbicide application), accompanied by 76.9 +/- 38.9 mm of leachate. For the low-rainfall years these amounts were 16.2 +/- 18.2 mm of runoff (92% less than the normal years) and 45.1 +/- 25.5 mm of leachate (41% less than the normal seasons). Runoff of metolachlor during the normal-rainfall seasons was 4.5-6.1% of application, whereas leaching was 0.10-0.18%. For the below-normal periods, these losses were 0.07-0.37% of application in runoff and 0.22-0.27% in leachate. When averages over the three normal and the three less-than-normal seasons were taken, a 35% reduction in rainfall was characterized by a 97% reduction in runoff loss and a 71% increase in leachate loss of metolachlor on a percent of application basis. The data indicate an increase in preferential flow in the leaching movement of metolachlor from the surface soil layer during the reduced rainfall periods. Even with increased preferential flow through the soil during the below-average rainfall seasons, leachate loss (percent of application) of the herbicide remained below 0.3%. Compared to the average rainfall seasons of 1995-1997, the below-normal seasons of 1998-2000 were characterized by a 79% reduction in total runoff and leachate flow and by a 93% reduction in corresponding metolachlor movement via these routes. An added observation in the study was that neither runoff of rainfall nor runoff loss of metolachlor was influenced by the presence of subsurface drains, compared to the results from plots without such drains that were described in an earlier paper.
