Understanding the genetic basis of CO2 responsiveness under controlled conditions: design challenges in the genomics era

Pinkard EA1, Brookhouse M2, Shimono H3, Bush D1 and Farquhar G2

  1. CSIRO Land and Water.
  2. Research School of Biology, ANU.
  3. Faculty of Agriculture Iwate University.

The move towards large genomic experiments, and the substantial phenotyping required to support these experiments, challenges our capacity to conduct reproducible experiments, and correctly identify key traits. If not adequately addressed this can affect the interpretation of results and potentially lead to flawed investment decisions regarding genetic selection for superior performance. In this work, we identify potential problems associated with phenotyping large G x CO2 experiments, and suggest a possible solution. We address the following matters. (1) In fast-growing, young plant material, small differences in environmental conditions or initial plant size can substantially affect outcomes. (2) Though working with clonal material is attractive in that an identical genotypes can be placed in different environments, propagation difficulties and plagiotropic growth can be challenging (3) Size of controlled environment facilities rapidly becomes limiting for large plants, and controlling within-glasshouse spatial variation is an issue, though modern statistical analytical techniques can assist. (4) In genome-wide association experiments, there is often a need to include a large number of genotypes, implying low replication within genotypes. This reduces statistical power for some genetic parameter estimation and requires correction for glasshouse spatial variation (5) The large numbers of plants associated with genome-wide association experiments makes trait assessment difficult to standardise in time, particularly for fast-growing tree species. Many traits are time-consuming to measure and ongoing growth or trait changes due to plant developmental stage during the measurement period can confound results. We discuss the application of pre-screening as a tool in large genome x e[CO2] experiments, to reduce costs and overcome issues of access to suitable CO2-enrichment facilities. In this approach a surrogate to e[CO2] is used to select cultivars for more detailed analysis under e[CO2] conditions. It provides a potential mechanism to screen large numbers of plants at relatively low cost, although to date there has been only limited testing.