BackgroundPremature leaf senescence induced by external stress conditions, e.g. drought stress, is a main factor for yield losses in barley. Research in drought stress tolerance has become more important as due to climate change the number of drought periods will increase and tolerance to drought stress has become a goal of high interest in barley breeding. Therefore, the aim is to identify quantitative trait loci (QTL) involved in drought stress induced leaf senescence and drought stress tolerance in early developmental stages of barley (Hordeum vulgare L.) by applying genome wide association studies (GWAS) on a set of 156 winter barley genotypes.ResultsAfter a four weeks stress period (BBCH 33) leaf colour as an indicator of leaf senescence, electron transport rate at photosystem II, content of free proline, content of soluble sugars, osmolality and the aboveground biomass indicative for drought stress response were determined in the control and stress variant in greenhouse pot experiments. Significant phenotypic variation was observed for all traits analysed. Heritabilities ranged between 0.27 for osmolality and 0.61 for leaf colour in stress treatment and significant effects of genotype, treatment and genotype x treatment were estimated for most traits analysed. Based on these phenotypic data and 3,212 polymorphic single nucleotide polymorphisms (SNP) with a minor allele frequency >5 % derived from the Illumina 9 k iSelect SNP Chip, 181 QTL were detected for all traits analysed. Major QTLs for drought stress and leaf senescence were located on chromosome 5H and 2H. BlastX search for associated marker sequences revealed that respective SNPs are in some cases located in proteins related to drought stress or leaf senescence, e.g. nucleotide pyrophosphatase (AVP1) or serine/ threonin protein kinase (SAPK9).ConclusionsGWAS resulted in the identification of many QTLs involved in drought stress and leaf senescence of which two major QTLs for drought stress and leaf senescence were located on chromosome 5H and 2H. Results may be the basis to incorporate breeding for tolerance to drought stress or leaf senescence in barley breeding via marker based selection procedures.Electronic supplementary materialThe online version of this article (doi:10.1186/s12870-015-0524-3) contains supplementary material, which is available to authorized users.

Bread wheat (Triticum aestivum) is a major staple food and therefore of prime importance for feeding the Earth's growing population. Mycorrhiza is known to improve plant growth, but although extensive knowledge concerning the interaction between mycorrhizal fungi and plants is available, genotypic differences concerning the ability of wheat to form mycorrhizal symbiosis and quantitative trait loci (QTLs) involved in mycorrhization are largely unknown. Therefore, a diverse set of 94 bread wheat genotypes was evaluated with regard to root colonization by arbuscular mycorrhizal fungi. In order to identify genomic regions involved in mycorrhization, these genotypes were analyzed using the wheat 90k iSelect chip, resulting in 17 823 polymorphic mapped markers, which were used in a genome-wide association study. Significant genotypic differences (P < 0.0001) were detected in the ability to form symbiosis and 30 significant markers associated with root colonization, representing six QTL regions, were detected on chromosomes 3A, 4A and 7A, and candidate genes located in these QTL regions were proposed. The results reported here provide key insights into the genetics of root colonization by mycorrhizal fungi in wheat.

Elevated CO2 (eCO2) concentrations can stimulate crop growth, but little is known about intraspecific variability in the response to eCO2 and the underlying genetics in cereals. Field experiments over two years with 98 barley genotypes were conducted in open‐top chambers (OTCs) under ambient CO2 (400 ppm) and eCO2 (700 ppm) concentrations. At crop maturity, different fractions of aboveground biomass (AGB) were measured, and genome‐wide association studies (GWASs) were conducted to identify quantitative trait loci (QTL). Averaged across all genotypes, eCO2 significantly enhanced AGB by 15%, while the increase in culm and ear biomass alone was not significant. The AGB response to eCO2 of the individual genotypes ranged from c. −36% to +95% compared with ambient CO2 (aCO2), showing a large variability of growth responses. In GWAS, 51 associations between SNP markers and the relative changes (eCO2/aCO2) in biomass were detected on different chromosomes. Loci potentially involved in biomass alterations under eCO2 were identified. The wide range of variability in responses might be exploited by marker‐based breeding for climate‐resilient barley.

No abstract

We propose a novel training methodology---Concept Group Learning (CGL)---that encourages training of interpretable CNN filters by partitioning filters in each layer into \emph{concept groups}, each of which is trained to learn a single visual concept. We achieve this through a novel regularization strategy that forces filters in the same group to be active in similar image regions for a given layer. We additionally use a regularizer to encourage a sparse weighting of the concept groups in each layer so that a few concept groups can have greater importance than others. We quantitatively evaluate CGL's model interpretability using standard interpretability evaluation techniques and find that our method increases interpretability scores in most cases. Qualitatively we compare the image regions which are most active under filters learned using CGL versus filters learned without CGL and find that CGL activation regions more strongly concentrate around semantically relevant features.