On approaches and applications of the Wageningen crop models

The Wageningen group has a long tradition in developing and applying crop models in its agro-ecological research program, based on the pioneering work of C.T. de Wit. In the 1960s and 1970s the main aim of these modelling activities was to obtain understanding at the crop scale based on the underlying processes. De Wit and co-workers at the Department of Theoretical Production Ecology¹ of Wageningen University, and the DLO Research Institute for Agrobiology and Soil Fertility² developed the model BACROS and evaluated components of the model (such as canopy photosynthesis) with especially designed equipment and field experiments (De Wit et al., 1978, Goudriaan, 1977, Van Keulen, 1975, Penning de Vries et al., 1974). These modelling approaches have served as the basis and inspiration for modelling groups around the world. A pedigree of the Wageningen crop models has been extensively described by Bouman et al. (1996).

In the 1980s a wide range of scientists in Wageningen became involved in the development and application of crop models. The generic crop model SUCROS for the potential production situation was developed (Van Keulen et al., 1982, Van Laar et al., 1997), which formed the basis of most recent Wageningen crop models such as WOFOST (Van Keulen and Wolf, 1986, Boogaard et al., 1998), MACROS (Penning de Vries et al., 1989), and ORYZA (Bouman et al., 2001). A simplified approach with respect to simulation of dry matter accumulation was developed by Spitters and Schapendonk (1990) based on the light use efficiency (LUE) as introduced by Monteith (1977): the model LINTUL. For water- and nitrogen-limited production situations, model components were added to SUCROS resulting in models such as ARID CROP, SAHEL and PAPRAN (Van Keulen et al., 1981, Seligman and Van Keulen, 1981). These developments were associated with development-oriented research programmes such as ‘Actual and potential production of semi-arid grassland’ (Van Keulen et al., 1981) in Israel and ‘La Productivité des Pâturages Sahéliens’ (PPS; Penning de Vries and Djitèye, 1982) in West Africa. Major efforts were directed to obtaining full understanding of crop performance under a wide range of conditions. These included modelling of perennial species in forest systems (Mohren, 1987, Kramer, 1996), and the effects of yield-reducing factors, such as weeds (Spitters and Aerts, 1983, Kropff et al., 1984, Kropff, 1988), pests (Fransz, 1974, Rabbinge, 1976, Van der Werf, 1988), diseases (Rabbinge, 1976, Van der Werf, 1988, Rossing, 1993, Bastiaans, 1993a) and air pollution (Kropff, 1989). Practical applications were realised in decision support systems for pests and diseases (Rabbinge and Rijsdijk, 1984, Rossing, 1993). Kropff et al. (1995a) have reviewed the approaches and applications of coupling crop and pest models.

In the 1990s the Wageningen group focused more on applications in research, agronomic practice and policy making. In a major project (Simulation and Systems Analysis for Rice Production, SARP) of Wageningen, the International Rice Research Institute (IRRI) and 15 NARS in Asia, interdisciplinary teams of Asian scientists were trained in the development and application of simulation models (Kropff et al., 1994a, Ten Berge and Kropff, 1995). Along with this program, a wide range of issues was studied using crop models, such as: mixed cropping (e.g. Baumann et al., 2001), relay cropping in upland rice (Akanvou et al., 2002), effects of climate change (Nonhebel, 1993, Wolf, 1993, Goudriaan, 1996, Rodriguez et al., 1999), breeding applications (Kropff et al., 1995b, Bindraban, 1997, Aggarwal et al., 1997, Yin et al., 2000), yield gap analyses (e.g. Casanova, 1998), and water and nitrogen management (e.g. Ten Berge et al., 1997a, Ten Berge et al., 1997b, Farré et al., 2000).

In the 1990s crop models also found their application in studies at the higher levels of integration, i.e. farm and regional scale. In these studies, crop models were used to quantify a broad range of land use systems; subsequently these land use systems were aggregated to farm or regional scale using various techniques (e.g. linear programming) or procedures. Studies on designing environmentally friendly systems for arable (Ten Berge et al., 2000), dairy (Van de Ven, 1996) and flower bulb farms (Rossing et al., 1997) were conducted, also enabling analysis of trade-offs between economic and environmental objectives. Land use studies were carried out with a focus on interactive exploration of different strategies for the European Union, Mali, Costa Rica and South-east Asia (e.g. Rabbinge and Van Latesteijn, 1992, Breman and Sissoko, 1998, Bouman et al., 2000, Roetter et al., 2000). Finally, crop models were used to explore limits for food production capabilities at global scale (Penning de Vries et al., 1995).

The philosophy of the Wageningen modelling group has been based on open exchange of information and a strong linkage with teaching. To facilitate this, models were published in books and reports with full code and describing the scientific basis. The high citation intensity of these books indicates the value of such a publication medium in addition to publishing short articles in refereed international scientific journals. The Wageningen crop modelling community currently works on a wide range of projects using models in many different ways, but with a joint conceptual framework of production-defining, -limiting and -reducing factors (Section 2.1). A joint modelling framework, however, has not been developed so far, and evaluation standards and software checking procedures have not (yet) been implemented in a standardised way.

In this paper we give an overview of currently available crop modelling approaches and examples of applications in research, education, agricultural practice and policy making. We pay special attention to the existence of models of varied degrees of complexity for different purposes. Section 2 describes the three criteria used throughout this paper to classify the Wageningen models. Section 3 presents the major summary and comprehensive modelling approaches for potential production, water- and nitrogen-limitation and yield reductions, further detailed in Section 4. Then, we discuss issues on model testing and data requirements (Section 5) and on software implementation, its documentation and availability (Section 6). Section 7 presents examples of applications in three domains and we conclude with reflections and prospects (Section 8).