A new harvest time prognosis tool for forage maize production in Germany

Abstract

Optimum harvest date of silage maize is crucial for exploiting the yield and forage quality potential of the crop, and for minimising losses during silage storage and the feedout phase. Simulation models are a useful tool for determining this optimal date. While numerous models are available for yield and some for quality, no comprehensive model deals simultaneously with yield and quality. The current project aims to fill this gap by (i) adapting the grass model FOPROQ to growth and quality of silage maize, and (ii) calibrating and validating the model for dry matter (DM) yield, DM content, and starch content. This allowed subsequently (iii) to perform a long-term simulation study for quantifying the climatic risk of silage maize grown under the marginal conditions of Northern Germany. Calibration was based on a field trial conducted at four sites over 3 years (1998–2000) in Germany, including six cultivars covering a wide range of maturity classes (early to mid-late) and maturity types (normal, dry-down, stay-green). Samples were taken every 2 weeks over the vegetation period. Validation was based on an independent data set collected during two years (2001, 2002) at two sites, however, with only one cultivar. No changes of algorithms were necessary to adapt the grass model to forage maize. Calibration results showed good agreement between observed and calculated data (RMSE values of 153.8 g DM m −2 for DM yield, 26.8 g DM kg −1fresh weight for DM content, and 33.5 g kg −1 DM for starch content). For severe drought stress situations the validation revealed some discrepancies between observed and simulated starch content. Model refinements by implementing environmental response as a function of developmental stage and by broadening the calibration database are discussed. A 25-year (1979–2003) simulation study was conducted for one mid-early and two early cultivars at Kiel, Northern Germany. The simulation results were in good agreement with corresponding regional mean values for DM content and starch content of maize silage samples. A climatic risk analysis for this long-term simulation showed a substantial likelihood of failure to reach silage maturity, with failure rates between 44 and 64% depending on cultivar. The last decade (1994–2003), characterised by considerable temperature increase, showed somewhat lower failure rates. The reliable results presented in this study support current efforts to develop FOPROQ into a harvest time prognosis tool for Germany.