QTL analysis of ergot resistance in sorghum

Sorghum ergot, caused predominantly by Claviceps africana Frederickson, Mantle, de Milliano, is a significant threat to the sorghum industry worldwide. The objectives of this study were firstly, to identify molecular markers linked to ergot resistance and to two pollen traits, pollen quantity (PQ) and pollen viability (PV), and secondly, to assess the relationship between the two pollen traits and ergot resistance in sorghum. A genetic linkage map of sorghum RIL population R931945-2-2 x IS 8525 (resistance source) was constructed using 303 markers including 36 SSR, 117 AFLP , 148 DArT and two morphological trait loci. Composite interval mapping identified nine, five, and four QTL linked to molecular markers for percentage ergot infection (PCERGOT), PQ and PV, respectively, at a LOD >2.0. Co-location/linkage of QTL were identified on four chromosomes while other QTL for the three traits mapped independently, indicating that both pollen and non pollen-based mechanisms of ergot resistance were operating in this sorghum population. Of the nine QTL identified for PCERGOT, five were identified using the overall data set while four were specific to the group data sets defined by temperature and humidity. QTL identified on SBI-02 and SBI-06 were further validated in additional populations. This is the first report of QTL associated with ergot resistance in sorghum. The markers reported herein could be used for marker-assisted selection for this important disease of sorghum.

Modelling the growth and water uptake function of plant root systems: a review

Crop models have been intensively used as a tool to analyse the performance of cropping systems under variable climate in terms of productivity, profitability, and off-site impact. The importance of modelling the function of plant roots in water and nutrient uptake from the soil is becoming increasing clear with the expanding application areas of crop models. This paper reviews the approaches and assumptions used in growth and uptake modelling of plant roots, and how the responses of plant root system to internal and external factors are captured in the widely used crop models. Most modelling approaches are based on one of the following assumptions: (i) that plant roots are uniformly distributed in homogenous soil layers and all roots have the same ability for uptake, or (ii) that plant root length is always sufficient for resource uptake in rooted soil layers. In structured soils, an overestimation of water uptake is likely to be expected. Further studies on root growth, distribution, and function in structured soils will require quantification of soil structures and root distribution patterns; and for non-uniformly distributed plant populations, spatial distribution of plant roots and non-uniform uptake need to be modelled. Root architecture modelling may help to address such issues. However, in order for the model to be useful at the field production level, simplified approaches that require easily measurable inputs need to be developed. Some examples are given. The oversimplification of root response to soil drying and hardness is likely to lead to overestimation of root growth and water uptake in dense soils. A soil strength factor needs to be incorporated so that the improved model can help evaluate the effect of subsoil compaction on production and resource use. Responses of root growth and uptake to soil salinity, boron toxicity, and extreme pH need to be further investigated if models are to be used for evaluation of crop performance in such environments. Effect of waterlogging also needs to be added for use of the model on heavy clay soils under irrigation or concentrated rainfall. There is an urgent need for joint efforts of crop physiologists, agronomists, breeders, and soil scientists to integrate interdisciplinary knowledge and to collect data that better describe the crop root system and its growth and uptake ability, to quantify plant process level responses, and for better soil quantification. Such knowledge and data are essential for improvement of model performance and successful applications.