Publication: Influence of climate variability on the potential forage production of a mown permanent grassland in the French Massif Central
Full text at PDC
Advisors (or tutors)
Elsevier Science BV
Climate Services (CS) provide support to decision makers across socio-economic sectors. In the agricultural sector, one of the most important CS applications is to provide timely and accurate yield forecasts based on climate prediction. In this study, the Pasture Simulation model (PaSim) was used to simulate, for the period 1959–2015, the forage production of a mown grassland system (Laqueuille, Massif Central of France) under different management conditions, with meteorological inputs extracted from the SAFRAN atmospheric database. The aim was to generate purely climate-dependent timeseries of optimal forage production, a variable that was maximized by brighter and warmer weather conditions at the grassland. A long-term increase was observed in simulated forage yield, with the 1995–2015 average being 29% higher than the 1959–1979 average. Such increase seems consistent with observed rising trends in temperature and CO_(2), and multi-decadal changes in incident solar radiation. At interannual timescales, sea surface temperature anomalies of the Mediterranean (MED), Tropical North Atlantic (TNA), equatorial Pacific (El Niño Southern Oscillation) and the North Atlantic Oscillation (NAO) index were found robustly correlated with annual forage yield values. Relying only on climatic predictors, we developed a stepwise statistical multi-regression model with leave-one-out cross-validation. Under specific management conditions (e.g., three annual cuts) and from one to five months in advance, the generated model successfully provided a p-value < 0.01 in correlation (t-test), a root mean square error percentage (%RMSE) of 14.6% and a 71.43% hit rate predicting above/below average years in terms of forage yield collection. This is the first modeling study on the possible role of large-scale oceanic–atmospheric teleconnections in driving forage production in Europe. As such, it provides a useful springboard to implement a grassland seasonal forecasting system in this continent.
© 2019 The Author(s). We thank the National Centers for Environmental Prediction, the European Center for Medium-Range Weather Forecasts and Météo-France/Hymex/MISTRALS for providing the NCEP, ERA-40, ERA-Interim and SAFRAN re-analyses. We thank the UK Met-Office Hadley Center for the HadSST database. This study was supported by the project MACSUR - Modeling European Agriculture with Climate Change for food Security (FACCE-JPI), funded by Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA), the meta-program ACCAF (Adaptation of agriculture and forests to climate change) of the French National Institute for Agricultural Research (INRA) and the project PRE4CAST (CGL2017-86415-R) of the Spanish Ministry of Economy and Competitiveness (MINECO). We also thank the Laqueuille site as part of the SOERE-ACBB project (http://www.soere-acbb.com), funded by the French National Infrastructure (https://www.anaee-france.fr). Iñigo Gómara was supported by MINECO (Juan de la Cierva-Formación contract; FJCI-2015-23874) and Universidad Politécnica de Madrid (Programa Propio – Retención de Talento Doctor). Iñigo Gómara's research stay at INRA UREP (September-December 2017) was also funded by UPM (Programa Propio - Ayudas al personal docente e investigador para estancias breves en el extranjero). We would like to thank Olivier Darsonville, Rémi Perrone, Iris Lochon, Katja Klumpp, Catherine Picon-Cochard (INRA UREP), Pere Quintana-Seguí (Observatori de l'Ebre) and Roberto Suárez-Moreno (Columbia University) for their helpful cooperation during the progress of this study. Finally, we are indebted to the two anonymous reviewers, whose pertinent comments and suggestions have contributed to substantially improve this manuscript.