Crop Specific Application Rates of Nitrogen for Soil-Grown Crops in Greenhouses to Protect the Groundwater

In a desk study, nitrogen (N) fertilizer application rates were studied and Nconcentrations at one meter depth or at the drain were estimated for Chrysanthemum grown in twelve greenhouse cropping systems, differing in soil type, supplemental lighting and transmissivity. Two N-saving strategies were also studied of which one estimated the N-concentration in the irrigation water to comply with the EU target value of 11.3 mg N L. N-surplus varied between 6 and over 709 kg ha yr. The Nsaving strategies reduced the N-surplus but none of the estimated N-concentrations were below the target value. Growers affect both the N-surplus and the irrigation surplus. Predefined concentrations are therefore not met if only N-fertilizers are regulated through crop specific application rates. Guidelines on irrigation applications need to be provided as well. INTRODUCTION In order to comply with the European target value of 11.3 mg N L in the shallow groundwater, crop specific N-application rates are introduced by the Dutch government. Specific application rates are developed to ensure maximum nitrate-N surpluses which leach to the groundwater. Under steady state conditions, the nitrate concentration in the percolating soil solution is calculated as this N-surplus divided by the annual winter precipitation surplus, and meets the target value in the shallow groundwater (Schröder et al., 2004). The crop specific application rates are defined for every crop grown outdoors. Among other measures, crop specific application rates contribute to reduce nitrate concentrations for outdoor crops (Jørgensen, 2004). However, questions can be raised if this approach is applicable to soil-grown greenhouse crops as irrigation and fertilization strategies, and conditions of nitrate leaching differ considerably with outdoor crops and there is obviously no precipitation surplus in the autumn/winter period. It is therefore not clear whether the methodology applied to outdoor crops can be used for soil-grown greenhouse crops. A desk study using cut Chrysanthemum (Dendranthema) as a model crop was therefore performed. MATERIALS AND METHODS Cropping Systems Twelve greenhouse cropping systems for soil-grown Chrysanthemum, representing the majority of systems in the Netherlands, were defined in this study: dry and wet sandy soils, clay and peaty clay soils were considered, greenhouses with low and high light-transmissivity (65 and 83%) and with or without use of supplemental lighting (130 μmol m s, 10 to 20 hours per day) (Table 1). For steady state conditions with respect to soil fertility aspects (no change in soil organic matter content), a (nitrate) NProc. XXVII IHC-S6 High-Qual. Crop Prod. under Protect. Cultiv. Ed.-in-Chief: D.J. Cantliffe Acta Hort. 761, ISHS 2007 396 surplus (Nsurplus, kg ha yr) was calculated for each greenhouse cropping system according to: d rem org fert surplus N N N N N − − + = (1) where Nfert is the applied N-fertilizer (kg ha yr), Norg the amount of N mineralized from organic applications (kg ha yr), Nrem is the N removal by the harvested crop and crop residues at the end of the cropping period (kg ha yr) and Nd is the N lost through denitrification (kg ha yr). Nfert was based on the fertilizer recommendations of 8 to 11 mmol L N in the irrigation water multiplied by the applied irrigation amount (see formula 2). Nrem was deduced from measurements of commercial greenhouses throughout the Netherlands from 1994 to 2003 and ranged from 750 to 1000 kg N ha (4.5 crops/year, Table 2). Although some cropping systems were not investigated, soil type has a minor effect on Nrem. The standardized value of 750 kg N ha for systems 3, 6, 9 and 12 may seem high compared to the measured value of 500 kg N ha. However, these data were found in 1994 and 1995 and corrected for increased productions for the last decade. The estimated denitrification was taken from Heinen (2006) for Dutch conditions: T S N p d f f f N N * * * = (2) in which the potential denitrification rate Np (kg ha year) is reduced by dimensionless reduction functions for the nitrate concentration in the soil (fN), the degree of saturation (fS) and the temperature (fT). The temperature was assumed to be 20°C resulting in fT = 1. The potential denitrification rate was soil specific (Heinen, 2006). As denitrification is extremely sensitive for fS, two soil specific sets of parameter values resulting in a high and a low value for fS (Heinen, 2006) were used to indicate a range of Nd. The N-concentration (Nconc, mg L) in the percolating soil solution was calculated as: surplus surplus conc I N N = (3) where Isurplus is the irrigation surplus (applied irrigation minus the evapotranspiration). The standard evapotranspiration of 721 mm year for Chrysanthemum (independent of soil type; Voogt et al., 2002) grown in a greenhouse with low transmissivity of light and no supplemental lighting (cropping system 3, 6, 9 and 12; Table 1) was increased by 18% when the transmissivity was high. In addition, the standard evapotranspiration was increased by 115 mm year to correct for additional lighting whereas no corrections were made for soil type. Standard Fertigation Strategy Differences in the total amount of N applied by fertilizers Nfert usually evolve through different irrigation strategies as N-concentrations in the irrigation water are in agreement with recommendations and irrigation amounts are adjusted to the crop’s needs. The irrigation in the standard strategy resembles irrigation in practice and was proportional to the evapotranspiration. In all systems irrigation was 15% higher than estimated evapotranspiration in order to ensure an even distribution of irrigation water. Irrigation was adapted for soil type, +29% for the dry sandy soil, +12.5% for the wet sandy soil, -20% for the peaty clay soil and +8% for the clay soil (Table 1). In case supplemental lighting was applied, irrigation was increased by 115 mm. N-Saving Strategies Two N-saving strategies were studied. The first strategy involved a reduction of the irrigation amount by 15%, which subsequently resulted in a reduced Nfert. The second strategy was based on reaching the target value of 11.3 mg N L in the percolating soil solution. The allowed N-concentration of the irrigation water was calculated from the regulatory limit, provided the irrigation was performed as in the standard strategy.