 we don't really have yet in our breeding materials. So we need to create, we need to look at the magic combinations and other germ pleasant to find those genetic resources. We need to build them into breeding ready products quickly and move to multi environment and are going to be testing as fast as possible. We're making progress in some of these traits. This is just one example on aerobic germination tolerance which we believe is very important for this system so that you have good emergence but at the same time can use also water to control weeds or do seeding into standing water. We have now identified five QTLs, one of them is already being tested, AG1 quite successfully in the first field trials recently in the pyramid of a second one on top of that one has started. So the promise is there, the steps are there and I'm quite hopeful that we can move quickly to new breeding products coming out of this work. Phenotyping is going to be obviously the bottleneck for most of the gene discovery and trait development work. There's no question about this with the cost of genotyping and sequencing in general coming down, I predict that sooner or later every major breed dog is going to have a little sequencing machine on their desk and do their own whatever they wanted to. Just look at what's happening in doctor's offices when you go to a medical doctor and the machinery they're already nowadays have. It's getting smaller, it's getting faster, it's getting cheaper. I think the time of the big sequencing lab is at some point going to be gone. But phenotyping is something that we're not to easily go away. And particularly field development phenotyping that needs to now really change its scope and scale. We wanted to be multi-level to what I've got all over the field. We need to capture phenotypic behavior over time in the field. We wanted to be non-destructive impossible, high throughput and more breeding ready. And we also should not forget about great quality of nutrition traits because of the new consumer drivers. I'm hopeful and I'm convinced that technology will develop fast. The more conceptual question is what are we going to phenotype and how for what purpose? At the moment there has been a lot of efforts on phenotyping genetic diversity panels, which is not easy because of the problems that you can see listed there. And one of them is different phenology in the field. So will we have to invest more in the future going to the right side of this slide? High throughput phenotyping directly embedded closer to the extra breeding applications and using therefore materials that are more breeding ready and also more phenomics friendly, but of course we'll have somewhat less genetic diversity. Finding the right trade-off and the right approach for the different applications is going to be one of the major challenges that we should address as high priority now. And then of course we have to find ways to connect all these databases. I'm sometimes just amazed by the amount of data that we are dealing with and some of it we probably still can't quite imagine yet. So all the way from the genotyping to the phenotyping and making it useful than in breeding information management software. We are in the process of establishing what hopefully will be an international rise informatics consortium, a public and private sector of effort hopefully that will at least try to bring many of the available data resources together in a format and in a platform that will be accessible to the OWA community. We have to establish some rules for this in terms of data access. And we also have to have a financially viable business model behind something like that. We can't be just a loose effort of a loose scientific community. It needs a structure. It needs to be managed because it has to be a long lasting effort because we know we're going to work on this for another 10, 20, 30 years. So later this year we'll share our thinking on that with the community or there may already be some more discussion this week during this conference. And then I think the last example I wanted to highlight is the issue of how can we systematically deploy germplasm to the right places at the right time? And it obviously becomes most important for anything that has to do with biotex stresses in particular diseases and insects where our fundamental problem is that resistance, host plant resistance breaks down very quickly. Jeans break down because pathogens or insects evolve. So the concept that my eerie colleagues are proposing for this is a real time disease management strategy that would include a very high level of gene discovery, targeted gene discovery, disease forecasting and management and then have a resistance toolbox for breeding that makes sure that we have the right combinations of genes in the right germplasm to be deployed in specific parts of the world. So the market and environment segments that I've been talking about. So it obviously requires that we have to invest more into understanding the diversity, genetic diversity of the pathogen populations in key hotspots and we've started working this particularly for blast and bacterial blight in Asia as well as in Eastern Africa. We have to build up, rise microbiome databases like this one, which already is a resource on genomics information on pathogens. And then of course, we also have to improve our prediction tools. This is one example for a simulation of susceptibility to bacterial blight in different parts of Asia. And I think we're able soon to break this down further in terms of analyzing what BLB resistance genes should be deployed when and where. To me, this is an increasingly important area that we should invest in because it is an essential part of maintenance breeding strategies to make sure that all of the other improvements that we make in varieties are not jeopardized by not enough progress in resistance to diseases and insects. And you can put on top of that also scares that people have with regard to climate change, creating new disease or insect problems or making them worse than compared to what we have at present. So I'm optimistic that rados and rice geneticists would have their high school work to do for many decades to come. And you can choose what to work on. You could work on traits that help us closing existing yield or efficiency gaps or reduce the risk. So many of the antibiotic traits, flat tolerance, tart tolerance, salinity tolerance, nitrogen use efficiency part of that category. You can choose to also invest your work in maintenance and value-added breeding and they're in particularly increasingly also the question, what improvements can we make in the rice grain to make it healthier? Also for those consumers who are experiencing not just hunger or malnutrition but actually the opposite problems such as diabetes too. Or you can choose to work on increasing the yield potential. All of that will be needed as part of sustainable development process. Perhaps, and this is my last slide, the biggest challenge remains the same as it has been for the last 40 years. Who is going to be able to really make a breakthrough in yield potential and what will it look like? I don't have the answer to this one. We have it eerie as part of the GRISP, a new Frontiers project that is taking a functional genomics approach to yield potential, introducing genes or the rhombal rises to enhance certain traits. And we hope that this could yield to a new plant that has a higher yield potential. On the right you can see the Christmas card, the New Year's card of Professor Yeol Long Ping last year from China in which he presented his vision of how a new super yielding rice could look like. So he believes that it needs to be a lot taller and stronger than rice looks at present. You can see if you are able to read the scale, the plant on the right is well over one meter 50 tall. So will the rise of the future look like this? Is it possible to keep it standing up like this? I don't know, but this is one of the challenges for you to figure out and I hope you have a very good conference this week. Thank you.