Additive, synergistic, or antagonistic? Relationships between sowing date and seeding rate affecting yield of winter wheat

Abstract Delayed sowing can reduce winter wheat ( Triticum aestivum L.) yield and increasing seeding rate is commonly proposed to compensate for some of the negative effects, but tillering plasticity, shortening of critical period, and environmental conditions make the yield response uncertain. We analyzed worldwide published data from 40 experiments that factorially combined sowing dates and seeding rates to (i) evaluate whether the relationship between seeding rate and delayed sowing in winter wheat was additive, synergistic, or antagonistic and (ii) evaluate the impact of sowing date on the critical seeding rate, that is, the rate to achieve 95% of peak yield. The compiled database contained 658 mean yields from 0.3 to 10.5 Mg ha −1 . The relationships between sowing date and seeding rate were classified as either positive or negative additive, antagonistic, or synergistic. Out of 169 sowing date × seeding rate relationships, 40% were negative additive and 25% were negative antagonistic, independent of the yielding environment. Negative additivity suggests that in many environments, the usual seeding rate was already close to or above the critical seeding rate required to maximize yield at delayed sowing. Negative antagonistic relationships indicate that higher seeding rates partially compensated for the yield reduction with late sowing. The remaining relationships were mostly positive additive or antagonistic, with synergistic effects being rare. Michaelis–Menten models were fitted to estimate critical seeding rate, which increased from 292 seeds m −2 under the usual sowing date to 463 seeds m −2 for ≥30‐day sowing delay. In environments where yield was reduced due late sowing, higher seeding rates partially offset yield losses. Our results reinforce that site‐specific seeding rate recommendations should account for the time of sowing.