Fig. 2: Annual nitrogen (N) and phosphorus (P) inputs in observational gardens (Sweden) benchmarked for experimental input levels (lines).

Each bar represents one garden in one year (abbreviations on the X-axis show that some gardens were in the study for more than one year). The height of each bar is the cumulative input for N (a) and P (b), with the color breakdown corresponding to types of inputs. The order of bars is ascending from left to right for total N input, and is kept for P (so N and P match by garden). Dotted lines represent experimental (USA) input levels where text colors match Fig. 4. The experimental treatment input levels account for both compost and additional synthetic N applied to meet crop demand. That is, assuming 20% of compost N is available for crop update in a year, the treatments added synthetic N to ensure crop demand was met (average crop N demand 145 kg/ha for the crops selected). This means that the total N inputs depicted are higher than crop demand even if they are designed to meet crop demand in a given year.