The evolutions of irrigation demand and harvest amounts have been produced for 1951 to 2100. An
example illustrated in Fig. 5 is the harvest and irrigation evolution of all genotypes from 1990 to 2100.
The genotype “Melody” offers the biggest harvest in the future, but its irrigation amount is also high.
There are two levels of irrigation demand. Most genotypes are in the smaller one. So it’s possible to find a
genotype with the small level irrigation, but with a comparably big harvest. The genotype “Euroflor” fits
the standard. It has the second largest harvest. Although this amount is less than the one from “Melody”,
its large irrigation saving could make up for the loss. Especially in drought regions, it could be a good
selection for water sustainability reasons. Generally speaking, the harvest evolution is decreasing, and the
irrigation evolution has a slight tendency to increase in simulations. This is not in agreement with reality
as the sunflower harvest has increased in the last few decades, because the simulation ignores genotype
evolution and other positive factors. This statement of harvest evolution is made by only considering
climate change in long time range. It predicts a negative influence of future climate in both harvest and
irrigation. While harvest has been widely recognized to face potential declines because of water shortage
and potential increase of farming land, this conclusion puts more pressure. Our irrigation optimisation
strategy on appropriate genotype and farming region selection is one way to mitigate it.