Sign up for email alerts of new Fluid Journal issues!
Fluid Journal : Summer 2016
15 The Fluid Journal Summer 2016 years were due equally to breeding and improved management, then we have multiple opportunities to close the yield gap. To this end, we introduce the concept of the interaction of G x E x M as a foundation for moving forward to feed the future world. The rationale for a departure from the classic G x E interaction is to highlight the effects of climate variability on the environmental factor, and the opportunities for management to enhance performance of genetic resources under varying environmental conditions. An example of this approach was undertaken by Martin et al. (2014) in which they evaluated these interactions for winter wheat from Denmark. They found that current annual grain yield improvements of 0.3 to 1.2 Mg ha-1 would be insufficient to keep pace with demand and improved management could potentially add to 1.8 Mg ha-1 to yield, and genetic improvements with a greater sensitivity to climate could add another 3.8 Mgha-1 of yield. This type of analysis combining climate scenarios with genetic and phenotypic improvements and management scenarios provides more realistic yield projections and identification of viable solutions. Simulation models can be effective in describing the genetic x environment interactions as demonstrated by Yin and Struik (2010) and Gu et al. (2014). This approach is the motivation for the international Agricultural Model Inter-comparison and Improvement Project (AgMIP), as described by Rosenzweig et al. (2013). AgMIP seeks to use the most advanced and robust crop simulation models to project future crop production and enhance development of adaptation strategies to cope with climate change. AgMIP is verifying that the current generation of crop simulation models inadequately account for soil-crop- atmosphere interaction responses to the wide variability of temperature and precipitation accompanying changing climate, as noted by Hatfield et al. (2011). Solutions to yield reductions from non-optimal soil water, soil, and air temperatures, and N are most often addressed independently during research. The following sections examine these three limitations of crop production from the perspective of G x E x M and offer evidence that advocates for an integrated approach. Soil water The greatest challenge is non-optimal water supply and thus there is need for management strategies that conserve and provide adequate soil water to meet crop water demands for rain- fed agriculture. Hatfield et al. (2001) showed that improved soil management can increase WUE, and supplying more available water to the plant benefits production and ensures the YF is closer to YP for a given site. The growing uncertainty of precipitation and rising air temperatures causing an increase in atmospheric demand are among the challenges posed by changing climate leading to crop drought stress. Genotypic variation of crop response to water stress offers insight into these interactions. For rice, Pantuwan et al. (2002) utilized a drought response index (DRI) calculated as the ratio of the (Yact –Y est)/ SE of Yest , where Yact is the actual grain each for each individual genotype, Y est is the estimated yield for each genotype, and SE of Yest is the standard error of all entries. The Yest was an estimate of the potential grain yield under non-limited conditions. They found drought stress before flowering delayed flowering and the delay was negatively associated with grain yield. Genotypes with delayed flowering continued to use soil water, had higher water deficits, and had larger yield reductions due to drought. The authors proposed that screening genotypes for drought resistance could be done with DRI or flowering response. Gu et al. (2014) combined simulation models with quantitative genetics to develop a genotype to phenotype approach for screen rice under drought stress, and this approach has the potential to provide new insights into the physiological factors limiting yield under stress. Kumar et al. (2012) proposed using a GGE biplot (genetics x genetics--environment interactions) to screen rice germ- plasm across multiple environments and identified stable genotypes across a wide range of environments. The GGE biplots quantify the genotype and genotype x environment interactions as two sources of genotypic evaluation (Yan et al., 2000) as potential tools for cultivar screening. Zhang et al. (2013) found that environments could be separated by year (Y) and location (L) and using a factor analytic model partitionedintoGxY,GxL,andG x Y x L interactions and applied this approach to canola (Brassica napus subsp. napus). They found phenology was an important factor for adaptation to specific environments. Abdolshahi et al. (2015) utilized selection criteria on 40 bread wheat varieties and observed heritability for secondary traits was significantly higher than for grain yield. Ten secondary traits include water use, relative ionic leakage, leaf length, root length, grain number, awn length, above-ground biomass, yield potential, days to flowering, and grain filling period, and could significantly discriminate high and low yield genotypes under drought stress conditions. Recent analysis by Razaei et al. (2015) observed shifts in winter phenology were able to offset the effects of increased temperatures, and earlier flowering reduces the likelihood of exposure to high temperatures at flowering. There is a need to implement the available tools and quantify genotypic responses to different environmental conditions to screen germplasm for G x E x M interactions. There are other indices besides phenology and yield that may be suitable for quantifying germplasm responses to water deficits. Carcova et al. (1998) used the crop water stress index (CWSI) with three maize hybrids and found that the CWSI changes were consistent across hybrids with no variation among hybrids. Earlier, Hatfield et al. (1987) found that cotton germplasm could be screened using canopy temperatures as a method of quantifying water conservation among lines. Bandyopadhyay et al. (2014) found water and N stresses in wheat could be quantified using remotely sensed spectral indices and proposed a normalized water stress index. The advantage of this method was that grain yield could be accurately predicted at the milk stage of wheat, providing a forecast of the potential drought effects (Bandyopadhyay et al., 2014). Further refinement of reflectance and thermal indices may advance our ability to screen germplasm for their response to water stresses. Drought stress also affects grain quality and is of concern to achieving