The efficient use by crops of nitrogen from manures is an agronomic and environmental issue, mainly in double-annual forage cropping systems linked to livestock production. A six-year trial was conducted for a biennial rotation of four forage crops: oat-sorghum (first year) and ryegrass-maize (second year) in a humid Mediterranean area. Ten fertilization treatments were introduced: a control (without N); two minerals equivalent to 250 kg N ha-1 yr-1 applied at sowing or as sidedressing; dairy cattle manure at a rate of 170, 250 and 500 kg N ha-1 yr-1 and four treatments where the two lowest manure rates were supplemented with 80 or 160 kg mineral N ha-1 yr-1. They were distributed according to a randomized block design with three blocks. The highest N mineral soil content was found in the summer of the third rotation, in plots where no manure was applied. The yearly incorporation of manure reduced, in successive cropping seasons, the amount of additional mineral N needed as sidedressing to achieve the highest yields. Besides, in the last two years, there was no need for mineral N application for the manure rate of 250 kg N ha-1 yr-1. This amount always covered the oat-sorghum N uptake. In the ryegrass-maize sequence uptakes were as high as 336 kg N ha-1 yr-1. In the medium term, the intermediate manure rate (250 kg N ha-1 yr-1) optimizes nutrient recycling within the farming system, and it should be considered in the analysis of thresholds for N of organic origin to be applied to systems with high N demand.

Core Ideas • In intensive forage systems 170 kg N ha-1 just from manures limits C sequestration • Biennial rotations of four crops maximize nutrient recirculation in calf-rearing farms • Double-annual forage crop rotation reduces P surpluses in crop-livestock systems

The use of cattle manure (CM) for fertilization presents challenges for optimizing nitrogen (N) use. Our work aimed to assess N efficiencies, in a 6‐year experiment with three biennial rotations of four crops: oat–sorghum (first year) and ryegrass–maize (second year) in a rainfed humid Mediterranean area of Spain. Fertilization treatments included the following: control (no N), 250 kg mineral N ha−1 year−1 (250MN), three CM rates (supplying 170, 250 and 500 kg N ha−1 year−1) and four treatments where the two lowest CM rates were complemented with either 80 or 160 kg mineral N ha−1 year−1. Treatments were distributed randomly in each of three blocks. Maximum dry‐matter yield (~44–49 t ha−1 rotation−1) was achieved in the third rotation, and only the control and the 170CM yielded significantly less. Within the limitations of the EU Nitrate Directive, the N steady state supply of 170CM always requires a complement of mineral N (80 kg N ha−1) to maximize N agronomic efficiency. The maximum N‐fertilizer replacement value (250CM vs. 250MN) was 0·67, without significant differences between the two treatments in other N‐related efficiency indexes, which indicates that plants took advantage of residual‐N effects. Nitrogen losses by leaching in the 250CM treatment were around 5–7% of the N applied. This reinforces the sustainability of manure recycling in long cropping seasons.