 which is so different I once asked one of the breeders in Australia I was working more in trait development back then I asked him what is your first priority trait and I was now I was working on several disease and quality traits I was waiting for my turn to come and he said you said okay what's your second priority trait he also said you okay what's your third priority trait he also said you and I said well you know when do you start to metric he said I made me about 11 top 10 about you so I just wanted to show you that that the philosophy in other crops is so different to what we are currently thinking now and and I think I'm not saying this is right I'm just saying there's so many alternative skin and following to that point with I've obviously taken a lot of that philosophy into our new design but yield is used as the ultimate predictor of a combination of traits if you test some new varieties in a in a disease prone area you'll your yield measurements also reflect your disease resistance and other trait resistances too and that's just again just a point to to to me to size there's a lot of changes to our breeding trial structure too many to describe this afternoon but I'll just focus on the observation we yield try with completely redesigned the sort of the thing that the underlying thinking behind it and also just the composition but the main changes those with keen eyes can already see that we typically now leave gaps between our plots and that's the better logistical thing because our brilliant trials are subsequently used to send seed to our NAS partners and by using so we use them for maintenance for trait scoring and all that stuff so that's just that's just the minor one but really apart from the whole operation the second dot point is that we are now also fed by other research teams and what do I mean by that as you saw in the earlier slide we have all these trait teams which are now separated from the breeding activities and so we have a lot of other scientists who can contribute material and include them alongside our irrigated material so just as an example we've got a lot of material from Zal San Chen's population improvement program with but from various but from swammy and Chen for various activities that they're involved with some might be asking well why swammy because he's doing biofortification indeed he is but he may have material that is high yielding and fits our requirements but maybe doesn't make the cut for his biofortification activity so we have a whole suite magic has got a lot of material over the last few season and even some others from the trait teams especially where in my case or in our case where the traits that people are working with are highly relevant to irrigated ecosystems so for example I put set these photos there because she's got a lot of sub one material and also some anaerobic germination material as well the other big difference is this has become a feeding ground for the other breeders particularly those at the hubs so I've just got shown pictures here I've got particularly so Rameshya has a lot of he's based at the Indian hub and he requires a lot of material to be sent from here and the other reason to that is because due to germplasm restrictions within India they cannot he cannot send material to Bangladesh and all other countries so they must be done from headquarters here so and again in line with our sort of philosophy that we want to minimize the selection for local conditions and so that by following an RGA based system followed by sort of with minimal selection and entry into the OYT we feel this pait us for the needs of our other breeders and so our breeders from the Africa hub also would come and cherry pick their promising lines that they want to test back home. I'll also mention this some cross fertilization with the rainfed Southeast Asia and Yawichi Kato has been involved with some of our activities as well as Dr RK Singh. So there is a lot of new linkages here with as a result of this new structure. I think everyone in PBGB may know this but for those who may not be familiar with breeding is you might be asking well what's the big deal about multi-location trials and I'll put this slide in because I'm sure to bring a smile to the pathologist here because this is IR64 three years ago and this is IR64 at one of our new testing locations just was it last early last year last dry season so the IR64 just gets absolutely smashed by the last disease which is not present here the races are different I presume and but it is absolutely it's just not we're barely harvested grain from that plot at that location so breeders often talk about something called genotype by environment interactions or different performances based on different locations this is just an example so in order to more thoroughly test some of our promising entries we now have some multi-location trials about OYT and PYT material this is very fresh very new we've only really just we slowly started last year and really only this season have we sort of pushed it along even more many of you would have heard of the MET system the multi-environment trial testing system which Dr. Edward Dunner was leading just before this is what I would describe as being upstream it's specifically for irrigated testing but it really flows into the current meat system so this is just what we're exploring with at the moment and so we for want of a better term we call them pre-MET trials we for that sort of covers explains what this testing is about and as I said they're very new and there's a whole suite of challenges that we've experienced I mean from the basic things like just letters of agreement making sure that they're paid on time the whole logistical thing becomes more and more challenging when you've got multi-locations and of course we have to travel there to inform I'd just like to thank my field general there Vittler Penard he's smiling that photo but I don't know if he was smiling after all the travels he's done in the last few months of a transplant but anyway and also I must acknowledge a lot of our collaborators here and people who've been involved in this process Jean Du from Bohol I've got some fuel rice collaborators and a lot of the Erie collaborators as well and particularly Ed Rodonia and Glenn Bagorio who really facilitated a lot of this this this multi-location trial testing another thing that is going really dr. R. K. Singh will describe this some more but my in Mars is being considered and has actually started this season as more than just a single MET location and we were there just last month and and so were some other stuff and this really starting from scratch as well and for people from senior management here is quite a challenge to get something from zero when you don't speak the language and the resources that are existing already is so minimal if there are many challenges but so I can see really and the team that has been going there to establish that the nurseries really deserves a lot of credit for pushing that along okay just some wrapping up this is the previous scheme I think I've sort of summarized it there probably the key points to note here number one the entry into MET would take by my rough calculations about eight years and what we refer to as a breeding cycle so when we start to make new crosses generally takes at least eight nine years typically in our new scheme we believe number one we've shortened the time it takes to develop lines to be nominated for the MET testing and that's by using RGA but also re revamping our trial testing stage but something a bit more subtle is we believe that by using parents earlier then we were previously done because we've got multi location trials here we're actually shortening our breeding cycle and that's that's critical if Erie or or or scientists want to increase the rate of genetic gain so I showed you that linear one percent line we want to we want the gradient to be much steeper and this has also been this foundation here has been really built with the view to use a new genomics technology which Mike Thompson and Guru Yee will talk about in the future using something called genomic selection where we can really increase our precision and select the best crosses to make and really shorten that breeding cycle even more quickly before I move on because our time is marching on I just like to acknowledge a few people who really helped to shape this it didn't happen overnight I can assure you I think I spoke to virtually every breeder here at Erie I was on the phone to call breeders in my network and calling breeders back in Australia but these three guys here really I really am grateful for them to bounce a lot of ideas off and they've had an input into this what we think is a very effective scheme now okay moving on the last part of our redesigning really involves efficiency and effectiveness this is my old reading shed and this is one of my old staff members and what he's doing it's just it's very beautifully simple he has got barcodes on the harvest samples he's got a barcode scanner connected to the Wayne balance which goes to the laptop so within just a few seconds he can scan the sample and we've got computerized data collection the point I'd like to mention is that this technology was established in the 1980s probably in the 1970s but we're not quite there yet at Erie at least not in the breeding programs not the irrigated breeding program so this has been a high priority and Vic LaPena again has been really pushing this as something that we just must get fully operational as soon as possible it requires a lot of small-scale investment in equipment but it is so important that we just have to focus on it just one example so it's another short video this computerized station rethresher I think really encapsulates what I believe the TRB projects about it's a new beast which was purchased by the experiment station and hats off to Lee Vile for having the courage to buy this beast so we're threshing and sorting a sample so this was harvested in the field brought to the just outside the area so they're just they're just emptying the sample bag to this station rethresher I think many breeding teams and even the agronomy teams have already started to use this it's a it's a real beauty so what's happening that the the clean sample is now it's it's it's clean and we've got it and the weight and moisture test is fed into the computer so what used to take days and days and weeks this video runs about one minute 40 so per sample we're talking days to you know or weeks to days or even hours to process the same level of of samples and this to me is just where you really needs to go so it's got a barcode but the important thing that the data is all stored in that in that computer part there so again thanks to the ES team for really investing in such a beautiful thing okay again another snapshot from another country another crop just something probably many of you realized but just to highlight it that in other crop species all breeding operations at least in developed countries even public sector ones they're all mechanized and that's another direction where I believe we need to go in and we've started these things don't happen overnight they take several seasons to really get things going but the potential for increasing the efficiency of the breeding operations is just huge there's certainly a lot of technical adjustments we need to make that the settings of how we lay out the land how we manage it and you know just getting the spacing right on the cedar etc and when we really start to exploit the gains from mechanization is when we also use the combine harvester to harvest the samples that have been direct seed so this is something there's been a lot of input from a lot of particularly the agronomists and ES fundings here the hybrid rice breeders also use this and we've we've had several discussions and and benoit clausier as well who's been doing a lot of work on DSR so this is something again there's a long way to go but I think this is a must for for iwi's future and has actually been written into the TRB project as well I'll very briefly talk about some new statistical methods I don't want to bore you with with some gory biometrics especially not so late in the afternoon but but just to give you an example and I I'm not so much of a geek to understand this and explain it to you properly but but just to illustrate the benefits 2012 wet season we had a replicated yield trial which was a randomized complete block with three reps this was the field footprint this was the area heritabilities it's a data quality measure of of the quality of the the yield data that was used it was pretty high that's pretty good we tried an alternative design there was no extra effort required except maybe half an hour to an hour on the computer preparing a new design we had only two replications but they were arranged in blocking in two directions and by doing that you can account for spatial variation in your field so by using two reps we saved our field field footprint was was lower I mean the plot sizes were a bit different as well but you can see that we reduced by that point six and the bottom line was with only a minor difference in in heritability we saved several thousand dollars from one trial alone just by using alternative design I'm not saying this is going to work on every single trial there are many other types of trials as well but but I think we need to think of adopting some of these innovative new new methods and seeing what they will do for us one thing iro nasilla has really grinded into the breeding operations of what are called metrics and I think it's long overdue it's something again private certainly private and many public programs in developed countries do is they monitor several aspects of their operations all the time it's continuous so things like the just what the annual budget is and costs for for all their their expenditure the size of the operation number of plots number of lines and they also are very big on monitoring their progress which in in a simple way can simply be benchmark by having your your checks and then you know measuring your your yield advantage over the best checks and so and of course data quality is also part of that as well so again it's an it's it makes sound so simple but iri hasn't been doing it before and I really think all of the pipelines will benefit by doing this moving on because i'm running out of time also connected to trb many new tools for characterizing our environments and i'll refer to this as breeding awareness for want of a better term and this example actually i stole this from akim's report a few years ago but it just shows what the gis people can do with their their beautiful maps that and this is so important that iri where we're breeding for multiple targets that we have information about what's required in each of our target environments so this this is a simple very nice map for amulose content which is a key determinant of grain quality so and this is just really i think it's quite exciting i think this is just getting started in terms of what ssd and the gis teams can really do for the breeders i think it's quite exciting there are so many that the potential is enormous and again thanks to particularly Alice and Andy who have had a big input to some of our activities already and this is part this is one of the objectives of the trb project as well based on that we have now developed and iro has has really preached this of the importance of this just like you would read a product specification for a cell phone we now also consider them for varieties and it's just really helps us just target our activities to make them really product focus i won't go into it there's too much there but it's just simply a specification that we really need to to breeders need to be aware of it's there plain and simple in black and white about the the key parameters we need to to do to to develop a new irrigated variety and i think we can take this even one step further that uh we perhaps need to go even beyond that and well actually it was kind of set in that last product profile but but there are kind of like extra features again like a cell phone with a you know a wi-fi act you know capability or a nice camera or something we want more features i think with some of our varieties and particularly irrigated apart from just high yielding and good quality i think there is a need to even further delineate certain markets and this will come out of the ssd work about the other traits and profiles that are needed so to wrap it up take home messages the irrigated program has been redesigned it's still ongoing but we sincerely believe that it's now quicker more efficient more effective and and importantly cheaper and i hope i've convinced you that that at least some of that is true if not all there's still a long way to go we are certainly still building and i think it'll take us at least another one year to to to keep going and and fully make the the foundation or the pipeline flow we have a lot of challenges including just optimizing the system and fine-tuning certain components as you saw there's a lot of coordination now required between our other breeders our other trade teams and and and other uri scientists in fact and i think we will certainly need to consolidate and establish or rekindle some of the previous linkages as a result of these these changes and certainly we'll need to communicate this hence me being up here today as a first step just to give you a plug pbgb will be having some seminars more on this topic exactly when uncertain at this date but probably august september and there will certainly be a special session during the rice congress in thailand in october one other goal we we certainly had and this is also written into the trb project is we want to establish partnerships with our nas partners uh within trb we will be establishing formal partnerships with south asia but also we hope to do that with a lot of other partners as well and particularly fuel rice we we they've expressed a lot of interest in what we're doing already and we really hope as part of the next gen project which uh there'll be opportunities to sort of to to cross pollinate from from the tib into into that program our goal is certainly that we hope to become a role model breeding operation for nas to follow and based on metrics we hope to show that based on quantifiable information our efficiency is at a high level but that it is undoubtedly one of our our main targets i think this is my one of my last slides um i started with some history and i'll end on some this is a table from a really nice review paper by shelving pen and dr kush and it just sort of summarizes some of the uh it's sort of just nicely summarizes some of the breeding highlights in the last 50 years so you can see the first decade was the semi dwarfing gene that was what the first decade was about the second decade 1970s it was about disease and insect resistance particularly the third decade grain quality was a priority uh the fourth there was highlights about the hybrids new plant type and yield i think pbgb has a lot of history writing to do i've given it just a first go i think the last decade will be i think really uh remembered for the stress tolerant varieties particularly for the rainfield environments this is where we want to go now we certainly want higher yielding varieties with a faster rate of genetic gain that is undoubtedly our target for the urgated activities and i think we need one more column and that really refers to how we derive those varieties it's not just how it's not just yes we've delivered it's how we deliver and i think we really want to add that information we really want to be highly efficient and and very precise last but not least uh i've tried to acknowledge all of the many colleagues that i've been working with but most of all i want to thank my team uh it's not easy to uh to change something that's been running for 50 years let me tell you and they've been very brave that i've dragged them completely out of their comfort zone and they've had to be re-skilled and everything but they've done a brilliant job in a short time and so thanks to to them particularly but uh there's many people i hope i've i've acknowledged particularly for this particular work i think i might end on that note now we are open for questions uh the barcode in so can you comment a little bit but i mean it's really interesting do you have that ready out? well i suppose fortunately a lot of the barcodes have been produced by the the iris team here for a few years but they've never been connected to the actual data flow um i think it like it's so it's already started and i notice i think some other groups that they're probably well more advanced than we are but the barcodes was the first step with the trb project we've got funding for some small-scale equipment which is also needed when you want barcode standards and a few sort of weighing balances and things like that um i really i think we just need to just simply just to do it and uh i we're implementing it now but i think there'll be a lot of trickle down and flow effects to to others who are not currently using it yeah do you answer your question? can you expect the first part of the new philosophy? good question um i have to i have to check some of our crossing notes and when i would say in a few years time three three four years time given that we've started we've got RGA derived populations in the field this season. It also depends how you define product. I'm pretty confident that we'll know that we've got something good or not after, you know, OYT or PYT stages. So 18 months, certainly within two years, we'll know that we've got something promising in that time. Thank you very much. I'd like to just suggest my opinion about the yield selection. I could not hear from you that how to determine the good combination. So it's very easy for us to enlarge our population, cross-combination. At every season, we can make about 500 cross-combination or like that. But what is difficult to us is how to reduce the population. It's one of the key ones. So it's just my opinion is not something scientific way, but I'd like to just suggest that let's try to use of the heteroses in early generation, in F1 generation or after generation. If the yield gain comes from the heteroses, it will be disappeared in the next generation. But some of the transgressive or well-recombined, it will be remain. It's from my experience that the early generation selection is quite important. So just make use of it in your way. It's just my opinion. It will be effective. Or the generation selection. Thank you. Yeah, thank you Dr. Yo.