The need for agricultural cropland to meet future food demand is a highly debated issue. We took a pragmatic approach to assess whether or not the current agricultural cropland is able to sustain future food demand growth. The premise of our approach was the assumption that, in 2005, food sup­ply equalled food demand; both were indexed at 100%. For this analysis, we did not analyse current or future food distribution patterns that might lead to local food shortages. We only assumed that there is no shortage of food production at the global level for which we would need to set aside additional land and that there is no overuse of cropland at the global level which could be taken out of food production without affecting supply.

We forecast the evolution of food demand and supply to 2050 as fol­lows:

We extrapolated the growth in food demand using the following step­wise approach:

1. We started with current per capita calorie values of ~ 10 MJ caput-1 (~2400 kilocalories caput-1) of plant product [20] and 1.46 MJ ca­put-1 (350 kilocalories caput-1) and 3.97 MJ caput-1 (950 kilocalo­ries caput-1) animal product in non-OECD and OECD regions [21], respectively. Animal product calories were converted into crop equivalents with conversion factors based on the crop feed intake necessary to produce them. The basis for these factors were feed intakes of ~17 kg kg-1, ~2.4 kg kg-1 and ~1.7 kg kg-1 of feed per produced amount of meat, eggs and dairy respectively. For meat, this factor was derived from literature values for feed efficiencies per animal type [2] and current distribution of consumption of meat per animal type [22], which is a mixed diet of bovine, ovine, pig and poultry products. For eggs, literature values from [23] were used. For dairy, literature values from [2], [24] and [25] were used. The feed is assumed to have an energy content of19 MJ kg-1 of dry matter [2].

2. We calculated a “business-as-usual” (BAU) per capita diet in the period 2005-2050 differentiated between OECD and non-OECD countries. This was done based on existing diet projections [22].

3. We then assumed that total animal product consumption worldwide will be constrained to a growth of no more than ~65% between 2005 and 2050, which means that the average animal product con­sumption per capita (in crop equivalents) increases by about 10% over the same timeframe, given population projections. In prac­tice, it could be desirable to divide the global 10% increase across regions in a non-equal manner, e. g. a reduction in animal product consumption in OECD countries and a significant increase in non- OECD countries.

4. We then multiplied the constrained per capita diets with population growth numbers used in [1] to get a total growth in food demand in crop equivalents. This was indexed against 2005.

We extrapolated the growth in food supplied by the current agricultural cropland by using a yield increase of 1% per year. This value is an inter­mediate value in a range of yield increase projections of 0.4-1.5 % found in literature ([7], [8], [22], [26] and [27]). The impact of climate change on yield projections was not explicitly considered in this analysis. However, by choosing the intermediate value of yield increase projections we have tried to be moderate in our assumptions. This yield increase was applied to the indexed value in 2005 of 100%.

The results for the extrapolated food demand and food supply are pre­sented in Fig. 4. For reference, Fig. 4 also contains the indexed yield de­velopment of coarse grains over the last 50 years, which has been higher than the 1% assumed in this work.

From the graph in Fig. 4 it can be observed that, based on our assump­tions the current agricultural cropland is projected to be able to supply the entire food demand in 2050. However, in intermediate years this is not

always the case. We have calculated that the maximum shortage of food from current agricultural cropland occurs in 2035 and amounts to about 4% of current agricultural cropland. This totals 630,000 km2 (63 million hect­ares).

Although the identified 630,000 km2 (63 million hectares) is the largest amount of additional land needed for meeting food demand in any given year to 2050, we choose to exclude it from the potential over the entire pe­riod in our analysis. This reduction is additional to the exclusion of current agricultural cropland based on IIASA data.