Optimising site-specific potato seeding rates for maximum yield and profitability

Abstract

Current site-specific seeding (SSS) recommendations are based on two approaches, whose comparative performance for maximising yield and profit is unknown. These are: the Kings and Robin Hood approaches recommending the highest seed density for the highest and poorest fertility zones, respectively. This study compared the agronomic and economic performance of these two SSS recommendations for potato in comparison with uniform rate seeding (URS). Soil pH, organic carbon, available P, K, Mg, Ca, Na, moisture content and cation exchange capacity predicted with an on-line visible and near-infrared reflectance spectroscopy, and crop normalised difference vegetation index retrieved from Sentinel-2 images were used for delineating management zones (MZ) using K-means clustering for four agricultural fields in Belgium. Soil and crop attributes together with geo-referenced yield observations were used to delineate MZ maps, whose classes were ranked according to the MZ-specific average yield. A strip experiment with five seed-to-seed spacings was implemented, comparing two SSS recommendation approaches. Results showed the highest yield was obtained for the Kings SSS (59.33–14.50 t ha−1) in all fields, whereas the Robin Hood approach provided higher yield than URS in two of four fields [e.g., field 1 (43.69 t ha−1) and field 4 (31.10 t ha−1)]. Compared to the URS gross margin (735.0–5806.75 € ha−1), the Kings approach led to higher relative gross margin (147.50–596.88 € ha−1) than the Robin Hood (−113.30 to 205.16 € ha−1) treatment. Therefore, it is suggested to adopt the Kings as the profitable recommendation approach for map-based SSS of potato production.