Combination of observational and functional trait-based approaches in developing a polyculture design tool

Models of species composition in diversified cropping sys-tems utilize mostly functional trait-based or observation -based approaches. We argue that a combination of these two approaches makes polyculture design tools more robust. We assessed quantity, quality, and complementarity of information from multiple sources for designing diversi-fied cropping systems with vegetables and spice crops for cold temperate climate. Trait and observational data were integrated from: (i) two grower-oriented and one academic crop database, (ii) a survey of farmers practicing community-supported agriculture, and (iii) a systematic literature review on the use of spice crops in vegetable farming. Survey results reveal that the farmers were capable of achieving medium to good levels of their main goals, but failed to reach desired multifunctionality with their poly-cultures, which can be potentially improved with computational tools. None of the analyzed data sources provided a comprehensive dataset for all target crops and functional traits. However, source combination allowed for design from known crop companions (farmers survey and grower-oriented databases), to addressing specific pest problems (literature review), and increasing functional complementarity and facilitation by trait matching (academic and grower-oriented trait databases). Integrating information from dif-ferent sources increased the number of crop combination options but also planning and management complexity.