Extended Data Fig. 1: Fertilisation trial with and without phosphorus.

All fields were fertilised with nitrogen (N) and potassium (K). In a subset of the fields (18 fields in 2018) we also tested whether phosphorus (P) fertilisation influenced inoculation success. Fertilisation alone had no significant effect on MGR (p = 0.3680); yet, we examined a negative interaction effect of field and fertiliser (p = 0.0001). In three fields, inoculation did not have the same effect in the presence and absence of P fertiliser, and MGR was lower in the presence of additional P (highlighted by a bold rectangle; F17: p = 0.0033, F21: p = 0.0352, F10: p = 0.0760). Fertiliser types: NK = nitrogen + potassium, NPK: nitrogen + phosphorus + potassium. The boxes represent the interquartile range (IQR, 25th to 75th percentile). The horizontal line within the box represents the median (50th percentile). The whiskers represent the minima and maxima of the data in comparison to the IQR (±1.5*IQR). Outliers are represented as individual data points.