Yield potential of tropical irrigated rice: achieved levels, current plant types, and ways forward

By Michael Dingkuhn
Senior Scientist II
Crop Physiology and Phenotyping Group, CESD
IRRI/CIRAD

18 September 2014, Thursday
1:15-2:15 p.m.
Havener Auditorium
IRRI

Abstract:

There is consensus on the need to increase the genetic yield potential (YP) of irrigated rice, for a host of humanitarian, economic, and ecological reasons. Although breeders assert that they strictly select for grain yield (GY), explicit plant type concepts have been guiding them since the beginnings of the Green Revolution. The fact that YP has been stagnating for many decades urges us to revisit these ideotypes. To realize new crops, molecular tools now give us access to the entire available genetic diversity.

After critically exploring the meaning of YP and yield ceilings, this paper proceeds to evaluate YP levels currently achieved on the basis of a fresh survey. We then zoom in to the performance, component traits, trait-trait compensations, and radiation-use efficiency (RUE) of 12 elite genotypes experimentally grown at PhilRice and IRRI sites. The results direct our attention to the heading-to-maturity period. During this period, rice, being a determinate grass and C3 crop, faces physiological tradeoffs between biomass growth and the reallocation of nitrogen (N) and carbon reserves to grains. These tradeoffs hinge on two main factors, terminal senescence and the maintenance respiration burden, both of which are directly related to N uptake.

A simple modeling exercise, drawing in part from data resources generated at IRRI 20 years ago, is used to quantify these tradeoffs and explore ideotypes that would perform better during reproductive and ripening periods. We inevitably end up with new research questions, but on a basis that confirms the main components of IRRI’s NPT and China’s Super Rice concepts, which are similar, while adding some new elements to these ideotypes. Substantial gains in YP seem to be possible. We also ask how breeding for the often complex trait and trait combinations could be accelerated with new phenotyping and molecular breeding approaches.