Ideotype breeding and tradeoffs? What tradeoffs?


Near the beginning of the question period for this recent lecture at the University of Minnesota, I suggested that:

1) nobody has done a good comparison of ideotype breeding with breeding for yield, and
2) many plant breeders who use the word "ideotype" ignore tradeoffs.

The main point of Donald's 1968 paper, which coined the term, "ideotype" was that there are often tradeoffs between individual-plant competitiveness and the collective performance of plant communities, so we can improve the latter by sacrificing the former. That's a major theme of my book, as well.

But both my numbered points above turn out to be wrong, at least partly.

Yuan et al. (2011) compared ideotype breeding with breeding for yield. I criticized some of their choices for "ideotype traits" in my third lecture at the International Rice Research Institute, but it's still an impressive study.

And, rereading Rasmusson's 1984 paper on ideotype breeding, I find extensive discussion of tradeoffs, though he doesn't explicitly mention the tradeoff between competitiveness and yield potential hypothesized by Donald (1968).

I am correcting these errors in an perspective I'm writing for the journal, Evolution.


While in graduate school in the eighties it was my impression that 'ideotype' had fallen out of favor. Rasmusson's barley work does help redeem the concept, but I'm still persuaded that most plant breeders today who even acknowledge the concept are not enthused. And to be honest - I have to include myself in this number.

What may be worth noting in the discussion of ideotype breeding is that even if we choose to ignore the concept, what the vast majority of our effort boils down to is an ideotype approach. There was a time when narrow leaf soybean was a commercially available type (narrow leaf phenotype controlled by a single gene - ln). So you might consider narrow leaves an ideotype. In today's marketplace you'll only find narrow leaves on a few soybean lines grown for natto. Improved germplasm with narrow leaves exists, but I won't be holding my breath for it making a comeback. Different leaflet number germplasm also exists in soy (three leaflets considered wild type... 5 and 7 leafleted types exist). And the tradeoff in yield among some 5 and 7 leaflet types is arguably small enough to set one wondering why they haven't received more attention. My hunch? The normal type soybean leaf - three ovate leaflets per trifoliolate leaf is still at this time sufficient for our purposes. But this is indeed an ideotype... so even though we don't refer to it as such, we are still ideotype breeders.

Donald really had two independent hypotheses:
1) there are tradeoffs between individual-plant competitiveness and plant-community performance (e.g., yield), and
2)we can identify yield-enhancing ideotype traits

If hypothesis 1 is false, then hypothesis 2 could be either true or false. (See Clem's comments.) But if hypothesis 1 is true, then the choice of ideotype traits should be guided by an analysis of tradeoffs.

Leave a comment