Varieties for droughts meet testing and doubts

There have been lots of interesting papers in the literature lately. More than I can keep up with for reading, much less to blog about. For example, some very nice work looking at land saving effects of productivity, an improved understanding of rainfall changes in the Sahel, and work on farmer perceptions of climate change. Plus the AgMIP efforts are starting to generate publications, such as this comparison of 27 wheat models.

But two I wanted to highlight concern the evidence on whether new “drought tolerant” varieties are outperforming other varieties. They especially caught my attention because we are currently analyzing some datasets for the US, partially to look at this issue. One of the papers, by Jason Roth and colleagues in Agronomy Journal, field tested some of DuPont Pioneer’s “Aquamax” varieties vs. other Pioneer hybrids without the special drought genes. The tests were done in Indiana for 2011, a pretty normal year in this location, and 2012, which was very dry and hot. They found no statistical difference between grain yields (GY), and also very little difference in terms of other outcomes like photosynthesis rates or transpiration. So their basic message is that the label of “drought tolerant” did not translate to any real differences in performance, although they emphasize that “Conclusions regarding the lack of superiority of drought-tolerant hybrids during the drought year are pertinent only to the specific environmental conditions encountered in the particular location tested.”

To me, there are a couple of possible ways to interpret this. One is that the newer varieties being marketed by companies are not really much better in general. Or these results might indicate that the types of droughts the newer varieties were designed for are somehow different than the type of droughts they were exposed to in this experiment. In particular, as we’ve discussed in prior posts, 2012 was a drought characterized by very high temperatures and vapor pressure deficits, the kind of droughts that one expects more of with climate change.

How much does the type of drought matter? Another interesting new study by Cairns et al. in Crop Science sheds some light on that question. They tested varieties developed for drought tolerance in eastern and southern Africa across multiple sites around the world. What was particularly novel was that they tested them not only in “drought”, but also in “heat” treatments (usually by planting late), and also for combined “drought+heat” treatments. The table below summarizes their results by showing the correlation between variety performance in different treatments (they also include a “well-watered” treatment intended to look at yield potential).

What’s really interesting is how remarkably low the correlation between performance in “drought” and “drought+heat” is (0.08). Like with the Roth study, it’s important not to extrapolate this too quickly beyond the particular sites and kind of treatments used in the study. But it certainly does provide support for the notion that very hot droughts require a different kind of variety. (What exactly that variety should look like is something I plan to focus on over the next year, including (I hope) a stint in Australia.)