Landscape complexity is associated with crop yields across a large temperate grassland region

• Published in: Agriculture, ecosystems & environment ecosystems & environment, 2020-03, Vol.290, p.106724, Article 106724
• Main Authors: Galpern, Paul, Vickruck, Jess, Devries, James H., Gavin, Michael P.
• Format: Article
• Language: English
• Subjects: Canadian prairies; Crop yield; Ecosystem services; Landscape complexity; Perennial vegetation; Semi-natural areas
• ISSN: 0167-8809; 1873-2305
• DOI: 10.1016/j.agee.2019.106724

•Landscape complexity was positively associated with crop yields at the district-level in four crops.•Farmers may be able to receive a small crop productivity benefit from enhancing landscape complexity.•Potential negative effects of non-crop land covers (e.g., as pest or weed reservoirs) were not dominant.•Increasing semi-natural vegetation may present few risks for crop productivity in Canadian Prairie annual croplands. Establishing semi-natural areas within annual croplands can provide habitat for beneficial organisms and ecosystem services to crops through a spillover effect. However, this approach to increasing landscape complexity may have no effect on crops, or it may promote pests, weeds and other disservices that reduce productivity. An argument for changing landscape complexity may be more persuasive if it is associated with higher crop yields. Here, we examine regions that vary in their landscape complexity and, therefore, may also naturally differ in the potential for ecosystem services, disservices and crop yields. Specifically, we examine crop-growing districts in the Canadian province of Alberta to test whether the presence of more non-crop land covers has increased crop yields. Our dataset covered about one-quarter of the seeded area in Canada between 2012 and 2017 consisting of 10,069 records representing average field-level yields reported to a crop insurance provider. In total, we analyzed summary data for 250,000 km2 of seeded area for seven grain crops. Using a functional regression approach, we found evidence for a plausibly positive association between yield and the non-crop land covers found within and near fields in four of seven crops. Landscape complexity, therefore, represented a measurable yield benefit for farmers, although the variance in yield explained by the landscape was small. These findings suggest there may be a low risk of disservices to crops from non-crop land covers in this region. Our study adds support at a broad geographic extent for initiatives that restore perennial and other semi-natural vegetation in annual cropping systems and suggests that, in this temperate grassland region, their promotion (e.g., as carbon stores or as biodiversity refugia) may have no adverse effects for crop production.