Description:
The data column "total nitrogen input" accounts for the combined nitrogen inputs from manure and commercial fertilizers. Until the mid 1920's, spreading manure was the primary means of adding nitrogen to agricultural fields. The amount of nitrogen from manure for 1865 to 1925 was back calculated from available statistics on the number of livestock units. By knowing the number of livestock, researchers in this study were able to estimate the amount of manure produced. Interestingly though, the nitrogen content of manure changed during this period of time. To account for this variability, these researchers found data on the amount of milk produced, and used a correlation between milk produced and nitrogen content to estimate the amount of nitrogen in the manure. The correlation they used was "roughly for every 1000 kg increase in milk production, nitrogen in manure increased by 20 kg". Hoffman et. al. also accounted for the loss of nitrogen (in the form of ammonia) as it transfers from manure to soil. Manure data after 1925 was obtained from Statistics Sweden. Data on the use of fertilizer nitrogen was based on sales statistics from Statistics Sweden as described in the journal article.
In the late 19th century it was realized that demand for manure nitrogen would exceed the supply. Just before World War I, the German researcher Fritz Haber discovered the commercial process (now referred to as the Haber-Bosch process) to convert ammonia gas into fertilizer. The heavy use of commercial nitrogen today is a two edged sword: modern levels of agricultural production could not be sustained without it, yet studies demonstrate negative impacts on water quality and ecosystems, especially as nitrogen leaches from land to water sources. As described in the reference article, "water quality data from major European rivers, such as the Rhine and the Daugava, indicate that a major increase in riverine loads of N coincided with the increased use of commercial fertilizers during the first decades of the post-war period".
It's interesting to compute the harvest-nitrogen ratios between 1865 and the present. These ratios (and the concavity of the graph) indicate that harvest totals are increasing at a slower rate as nitrogen input increases. Other increases in agricultural science (genetic engineering, irrigation, use of pesticides, etc.) should make the harvest/nitrogen ratio increase over time. Perhaps a different nutrient is in short supply, or soil quality is being reduced. There is a good power function fit to the data (very close to a square root function).
