Plant breeding, using the combined potential of conventional, molecular and genetically modified technologies, will provide cultivars with greatly enhanced nutrient and water-use efficiency, enhanced tolerance to heat and drought, resistance to diseases and appropriate end-use and nutritional quality, and, possibly most important, increased ability to cope with the increasing extremes in temperature and precipitation occurring at one location over years. Modern crop cultivars developed by seed companies, international crop research centres and national breeding programmes often exhibit very wide geographical adaptation, as well as broad adaptation to the range of environmental and management conditions that occur within and between a target population of environments, or megaenvironments. To identify such cultivars, multi-location testing done by the International Maize and Wheat Improvement Center (CIMMYT) and the International Rice Research Institute (IRRI) remains the most efficient system. International evaluation networks based on exchange of and free access to germplasm and multi-location testing are therefore a cornerstone in the strategies and efforts to develop wheat, rice and maize germplasm that is adapted to the increasingly variable growing conditions encountered due to global climate change. Information from such trials must be combined with information from managed stress trials. Wide performance adaptation is essential to respond to global climate change, to the vagaries of spatial heterogeneity within farmers' fields and their production input management efficacies, and from unpredictable temporal climatic seasonal variability.

Yield potential can be expressed as a product of light interception, radiation use efficiency (RUE), and the partitioning of biomass to grain yield, or harvest index (HI). Traits related to early or late light interception have not been shown to be associated with genetic improvement of spring wheat yield in favourable environments. It is, however, well established that yield improvement is largely a result of increased HI, although the most recent studies comparing genetic progress in HI over time in spring wheat indicate that it has not made any additional progress since the mid 1980s. These observations suggest that future genetic progress in yield will most likely be achieved by focusing on constraints to RUE. Considering the possibility that RUE may be influenced indirectly by sink limitation, it is apparent that biomass may be increased by increasing grain number, for example. Experiments with high yielding spring wheat lines containing the alien translocation 7DL.7Ag showed increased grains m -2 (15%), yield (12%), and biomass (9%) compared with controls. The translocation was also associated with a larger investment in spike mass at anthesis (15%), more grains/spike (10%), and increased flag-leaf photosynthetic rate during grain-filling (20%). The data suggest that increased biomass in 7DL.7Ag lines was due to significantly increased RUE post-anthesis, as a result of a larger kernel number (sink) that increased the demand for photosynthesis during grain-filling. The hypothesis that increased photosynthesis and RUE may respond directly to a larger number of grains/spike was tested experimentally by imposing a light treatment during boot stage. The treatment was associated with a small increase (5%) in the proportion of biomass invested in spike mass at anthesis, reflected by on average three extra grains/spike at maturity. The treatment was associated with 25% more yield and 22% more biomass than controls, while carbon assimilation rate measured on flag-leaves during grainfilling was 10% higher than controls. The results suggest that RUE can be increased indirectly by increasing sink strength and that the current yield limiting process in spring wheat is the determination of kernel number. Experimental data are presented on how spike fertility may be increased through breeding, for example by introgression of the multi-ovary trait to increase grain number per spikelet. In addition, results of analysis of the physiological bases of genotype × year interaction in high yield environments are presented in the context of how such information can provide a focus for genetic studies of sink limitation.

Grain weight is a trait which has hardly been exploited for raising genetic yield potential of wheat. A clearer understanding of physiological determinants of grain weight potential would be useful in establishing the potential value of this trait in future breeding programs. The objective of this study was to improve understanding of how intra-spikelet competition for assimilates pre- and post-anthesis affect grain weight potential, and to evaluate possible mechanisms determining final grain weight in wheat. Two experiments were carried out under field con-ditions. Proximal or distal grains from the two central spikelets of spikes of three synthetic hexaploid lines were detached at heading or 7 d after anthesis. Synthetic wheats were used since they represent a potential source of genetic variability for grain weight potential. Carpel size at anthesis and grain weight during the grain filling period were measured. The de-graining treatment at heading significantly increased grain weight, especially in distal posi-tions. On the contrary, the de-graining treatment carried out after anthesis caused no increase in final grain weight. The largest response to pre-anthesis de-graining occurred in grain positions with the lowest grain mass. In addition, the effect of de-graining prior to anthesis was associated hyperbolically with weight of carpels at anthesis within each grain position. Therefore, carpel weight at anthesis could be partially associated with the regulation of grain weight potential.

Radiation use efficiency (RUE) of a crop is a function of several interacting physiological phenomena, each of which can be tackled independently from the point of view of genetic improvement. Although wheat breeding has not raised RUE substantially, theoretical calculations suggest room for improvement. Selection for higher rates of leaf photosynthesis at saturating light intensities (Amax) has not resulted in improved RUE of crops, perhaps in part because most leaves in a canopy are not light-saturated. However, higher Amax may be observed as a pleiotropic effect of other yield-enhancing genes (e.g. genes for reduced height). Genetic transformation of Rubisco to double its specificity for CO2 would theoretically increase Amax by perhaps 20%, and some evidence suggests that photosynthesis at sub-saturating light intensities would also be improved. However, photo-protection may be jeopardized if capacity for oxygenase activity is impaired. Photosynthetic rate of the whole eanopy can be enhanced by manipulation of leaf angle, which is under relatively simple genetic control, and possibly by manipulating leaf-N distribution throughout the canopy. Genetic diversity for adaptation of lower canopy leaves (e.g. changes in chlorophyll a:b ratio) to reduced light intensity observed in some crops needs to be investigated in wheat. Improved RUE may be achieved by increasing sink demand (i.e. kernel number) if excess photosynthetic capacity exists during grain filling, as suggested by a number of studies in which source-sink balance was manipulated. Some evidence suggests that improved sink strength may be achieved by lengthening the duration of the period for juvenile spike growth. Balancing source- and sink-strength is a complex genetic challenge since a crop will change between source and sink limitation as conditions vary during the day, and with phenological stage. Improved RUE will be partly a function of a genotype's ability to buffer itself against changes in its environment to match the demand imposed by its development. Analysis of the physiological basis of genotype by environment interactions may indicate avenues for genetic improvement. The genetic control of photosynthetic regulation may be elucidated in the future through the application of genomics. However, given a lack of specific knowledge on the genetic basis of RUE, empirical selection is currently the most powerful tool for detecting favourable genetic interactions resulting from crosses between lines with superior photosynthetic traits and other high yielding characteristics. Selection for superior segregants can be accelerated using rapidly measured physiological selection traits, such as stomatal conductance or canopy temperature depression.

This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 2 0274-6638/20©2020IEEE IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE MONTH 2020 C rop yields need to be improved in a sustainable manner to meet the expected worldwide increase in population over the coming decades as well as the effects of anticipated climate change. Recently, genomics-assisted breeding has become a popular approach to food security; in this regard, the crop breeding community must better link the relationships between the phenotype and the genotype. While high-throughput genotyping is feasible at a low cost, highthroughput crop phenotyping methods and data analytical capacities need to be improved. High-throughput phenotyping offers a powerful way to assess particular phenotypes in large-scale experiments, using high-tech sensors, advanced robotics, and imageprocessing systems to monitor and quantify plants in breeding nurseries and field experiments at multiple scales. In addition, new bioinformatics platforms are able to embrace large-scale, multidimensional phenotypic datasets. Through the combined analysis of phenotyping and genotyping data, environmental responses and gene functions can now be dissected at unprecedented resolution. This will aid in finding solutions to currently limited and incremental improvements in crop yields. BACKGROUND Worldwide demand for food will increase through 2050 and beyond due to the increasing global human population. This represents a huge challenge to crop researchers and agricultural policymakers because current yield gain rates will not be sufficient for the demands of population growth, while climate change will make the difficulty even greater. Today's DNA sequencing, marker-assisted breeding, transgenic technology, genome-wide association study (GWAS) approaches, and quantitative trait loci (QTL) identification have been applied, to a limited extent, to improve crop yields [1]-[4]. While it is now relatively easy to select for monogenic traits, current genome sequence datasets have not been sufficiently mined for more genetically complex (multigenic) performance characteristics, at least in part because of the lack of crop phenotypic information collected from real-world field situations. Furthermore, traditional crop growth analysis often involves destructive sampling that is time-consuming and prone to measurement error. At High-Throughput Estimation of Crop Traits A review of ground and aerial phenotyping platforms