A Conversation with Dr. Sarah McNeil
Dr. Janet Lewis, the BTF’s vice president, reached out to Dr. Sarah McNeil (Battenfield), an accomplished modern-day scholar, breeder, and former CIMMYT in-field trainee. Janet wanted to learn more about Sarah’s background and hear her thoughts on the relevance of in-field training for current and future students. Read on to learn about a modern-day breeder’s path to where she is today!
Until December 2020, Sarah was Syngenta’s lead hybrid wheat breeder for the Central Plains of the U.S. as well as Syngenta’s genetics project lead for North America. Sarah is a Jeanie Borlaug Laube Women in Triticum Early Career awardee and was awarded a Monsanto Beachell-Borlaug International Scholars fellowship to conduct her Ph.D. work.
During her Ph.D., Sarah worked on one of the earliest studies of genomic selection in wheat for grain quality, and her research also required extensive in-field phenotyping. Her M.S. work was conducted in the wheat breeding program of Dr. Art Klatt at Oklahoma State University.
Dr. McNeil looking at wheat demo varieties with Dr. Art Klatt in Ciudad Obregon, Sonora, MX.
Janet: Sarah – Thank you for meeting with me today! You have a terrific combination of experiences in current-day cutting-edge breeding and field phenotyping. Can you tell us about your journey to becoming a wheat breeder?
Sarah: I actually come from a wheat farming family background. My dad's a wheat farmer in Oklahoma, and I grew up as a little child riding around in the combine. When I got a little older, I was helping my grandma with cooking for the harvesting crew. As I got older, I was driving the combines, pulling the headers, and being involved in the harvest. Wheat has been in my family for a long time and it’s really important to me.
Early generation selection: Dr. McNeil in her father's wheat field in northwest Oklahoma.
Dr. McNeil discussing hybrid wheat in the field with KSU graduate students from AgriPro breeding fields.
I originally was getting a bachelor’s degree in another area, but part-way through I realized that agriculture was where my heart was. It was pretty late into my bachelor's degree that I went to talk to the advisor at Oklahoma State. He said, “Well, you know, your grades are good so as long as you get a job, and you get one of the professors to vouch for you - we'll take you in a master's program.” He continued, “Art Klatt is looking for someone to emasculate wheat today - let's go see what he has to say.” I ended up working for Art Klatt and eventually doing my M.S. with him. Art was a great teacher of the “Why” question.
J: I love the “Why” question. I don’t think enough people spend enough time on the “Why.” I think if you don’t figure out the “Why” then the “How” may be irrelevant.
S: Art Klatt was really focused on the need to bring adult plant rust resistance into wheat adapted to the Central Plains. He taught me a lot about looking - looking at the pustules, looking at the infection type, seeing adult plants, seeing juvenile plants. I spent a lot of time looking hard at the symptoms, not just collecting data and crunching numbers. Getting to know the plants and getting to know the pedigrees in addition to crunching the data - looking at the whole package from multiple angles and talking these things over in the field - these are the types of things that Norman Borlaug spoke about. So, I had a really cool experience as an undergrad learning a lot from Art Klatt. We had our small room where we did huge numbers of single-head threshes. We had a small team that was getting germplasm from everywhere - from other programs in the plains like Texas A&M and Kansas State University and CIMMYT’s programs like Alex Morgunov’s winter wheat program and Ravi Singh’s spring wheat program. We were working to pull those APR [adult plant resistance] genes into germplasm adapted to the Central Plains.
“What I really wanted to do with my Ph.D. was marry the phenotypic breeding with the new data technologies that were coming through so that I had a leading-edge kind of project where I could do the hands-on phenotyping.”
When I was considering getting my Ph.D. I thought Kansas State University was one of the places that I might like to go to. Art said, “Why don't you go visit Allan [Fritz] while you're going to see your trials in Kansas?” When I met with Allan, I didn’t know that Art had already talked to him. Allan asked me “When would you like to start?” I thought the opportunity was great, but I also gathered up all my courage and said, “Starting is great, but I'd really like to do a Ph.D. study in collaboration with CIMMYT if I could.” Allan said, “If you're willing to travel, then we'll see.” Oh, wow! Sometimes it took all the courage I had to ask if I could do something. Chances are, if I asked, somebody would say, “Well, you're willing to go do it – so, ok.”
What I really wanted to do with my Ph.D. was marry the phenotypic breeding with the new data technologies that were coming through so that I had a leading-edge kind of project where I could do the hands-on phenotyping. I really wanted to learn the breeder-type skills in the field, and then also develop some good data skills like those used in genomic selection. The project we picked was the genomic selection for wheat quality. Collectively, I was traveling abroad for probably seven months during my Ph.D., but I was going specifically during phenotyping times and I was getting hands-on experience in the fields. In one year, I actually had four field seasons to oversee in different places - Obregon and El Batan [Mexico], Kansas [USA], and Njoro [Kenya]. I was tagging plants, looking at pedigrees, understanding what people selected for - all of those kinds of things. In the winter I was back in Kansas doing genomic selection for wheat quality. By the time that I was wrapping up my Ph.D. I had the genomic selection models and we were doing the genotyping in time so that when the breeders were preparing to make selections, they had a preview of the quality data through the quality predictions. At the end of my Ph.D., the wheat genome became publicly available, so we could map all these genes that I had been using in predictive models for a few years.
“Sometimes it took all the courage I had to ask if I could do something. Chances are, if I asked, somebody would say, “Well, you’re willing to go do it – so, ok.” ”
What we actually did was stack up individual GWAS [Genome-Wide Association Study] models for each year. I had taken this inspiration from type-two diabetes mapping in humans. We applied a meta-GWAS analysis. I actually worked with the people who develop the JMP-Genomics software (from SAS). It was a really cool experience to work with them on this powerful CIMMYT data and bring all this information together. And these were data that the breeding program was already generating, so they weren't having to add more labor to what they were doing, but they got this extra piece of information out of it that essentially said ‘These are the underlying factors of the quality traits that we're selecting for.’ That was a pretty neat experience.
Sterile wheat gaping for hybrid production
Overall, I got to come out of that experience with heavy data analytics and predictive modeling background married with a heavy phenotypic selection type of experience. From there, I went to Syngenta in an assistant breeder role in the hybrid wheat program. After some rearrangement in the department, I took the lead breeding role for the U.S. Central Plains, which was for both the inbred wheat that we were already commercializing as well as the hybrid wheat that we were gearing up to launch. It was really cool getting to be in more of a commercial role because I was seeing the things that were going out to farmers. I felt like “Okay, I'm building this trait into wheat that is going to be really beneficial for farmers.” For example, one of the things that we pushed really hard in my program was the standability of the wheat. The idea behind this is that as you're increasing yield you have to make sure that the straw is tough enough to withstand both the Kansas wind and the yield that you're putting on top of it. That was a neat trait that I worked on along with fertility restoration for hybrids, which could be pretty tricky.
Working with and for the farmers was really fun. I would think about how to do data analytics to make the most sense to the farmers. For example, how do you design trials to get data from all the environments that are important for the farmers? You don’t want to just make something that you know performs in a few areas or for a few farmers but does well for all kinds of farmers. That’s a pretty exciting job.
“If you’re just looking at data, and don’t have the bigger picture of how the plant grows in the field, you can get really misled. You always need to be thinking about what the farmer needs.”
J: I was really excited to hear that you come from a wheat background growing up - that part I didn't know before this interview. I think the fact that you come from a wheat farming family really deepens the relevance of what you've been doing and why, and I expect you have a great ability to understand what will make sense to the farmers and in an actual farming environment. For example, the critical aspect of timing is something that you probably understand in your bones.
S: In Kansas and Oklahoma, where I grew up, timing was all about water availability. In Oklahoma, farmers actually plant quite a bit earlier than they typically do in Kansas because they're trying to make forage for the cattle, and that’s a whole different kind of cropping issue.
J: You’ve been helping the wheat research community marry phenotypic evaluation skills with genomic data. Now we have even more tools becoming available, such as remote sensing. What do you think are the skills that students should have?
S: From what I've seen in the different panels I've been on, it’s important that researchers are at least conversational across the different areas of expertise. It's great to be specialized, but you really need to understand the other pieces of the puzzle. I think what it really boils down to is that you need to make sure that all the pieces are in place to have trialing excellence. Without well-done trials underlying the data, you can't go anywhere. You need a well-supported trialing staff with the right tools to get the types of phenotypes you need so that you can make breeding improvements. I think that's a really big piece and one I think people should have great respect for - It’s the quality of the trails that makes your data.
J: I think the quality of the trials is something that people often lose sight of. People get excited by all these gadgets and think that these new tools are going to solve all the problems. But it really comes down to this: If you don't have good plots, your data doesn't mean anything even if you use the best equipment to collect the data.
S: There's been a lot of technology that’s helped us have better plots, too. When I started, planting was done using a manual trip and drawing your own lines across the field to know where to trip (by looking at the lines). The technology advancement of GPS-trip has been a huge improvement, not only so that you have uniform plots, but also for using remote sensing tools. When you're using remote sensing it’s important that the plots are more reliably shaped and sized and that the position of the plots is easily identifiable. So, this is one way that technology has helped us develop better plots.
J: When you first step into a field trial, what do you think about?
S: When I go to a field there's a lot of things on my mind. Firstly, I think about how the trial is a representation of the breeding program and the institution. How does it look to others from the road? Are you taking care of your weeds and squaring off your plots? These are things that are very important to maintain the farmers’ trust. The farmer, who has ownership of that land, will want to see that you are being a good steward and also representing them well. For the neighboring farmers who drive by, the plots they see demonstrate your ability to produce a good variety following good agronomic practices. This is an important part of the relationship between the breeder and the farmer.
Secondly, I assess what the crop is experiencing. When I get out of the truck, I scratch my boots in the ground or pull out my knife to scratch in the ground and see how deep the moisture is. Then I look at the growth stage of the wheat, check for any weeds we need to clean up, determine if we need to spray fungicide, etc.
From there I start looking at the genotypes themselves. Do my checks appear to be where I expect them to be? Do my lines make sense to me? In Syngenta, we had commercial trials that were up in the front of our plots that were both commercial sales tools and were adjacent to the yield trial plots. I could check those to see how different varieties were doing – I could see that one variety is going to be the shortest and earliest and another is going to be really late. Checking these helps me know what to look for in the other genotypes. I also look at leaf health. Do I have a certain disease is there going to be something I can screen today? I might rate the quality of the plots – to make sure I capture if a plot is a garbage or not (i.e., can I get useful data?).
We also had an across-company initiative to record data using a consistent rating scale and to try to rate all the plots (not just the ones that look good). Breeders are busy people and there's probably a lot of miles in between your fields, too, so you can't always take a rating on everything. Sometimes you have to take a rating on the best or just mark them with stars or something else like that. But as much as I could, I tried to rate on a full 1 to 9 scale. One of the things that was important for me to learn early on is the importance of variation of the line ratings within the scale. If you’re rating everything a 3, 4, or 5, that’s not as useful. You need to be seeing ratings across the whole range for the data to be useful. You might have to come back at a different time to rate the trait so that the scale is differentiated enough. Sometimes I took notes that were in my breeder shorthand to mark down things like one is tall, another is late, short, early, etc. I had a lot of little shorthand things that I could write out. I would use this information when I went to the crossing blocks. I would also go through some of the earlier generations of yield trials to identifying some favorites that I would cycle back into the crossing block. Most of the time those early favorites don’t really go anywhere (unless it's a whole family that you saw looking good, then you probably had something), but at least to cycle back into the crossing block and say, “I like this, and I'm going to move it back into the crossing block and I'm going to trust myself.” Confidence in your own notes is a big deal. That's what I'm saying about variation in the ratings – you need to be confident that something is good and something is bad, and get that full range of good to bad, not just mark everything in the middle.
Dr. McNeil’s first view of hybrid vigor -- F1 in CIMMYT, Ciudad Obregon, Sonora, MX.
J: What would you say if somebody said to you, “Why should I go spend time in the field when I can just look at the data and do breeding on the computer?” Why should students spend time in the field?
S: One of the things that we always said about wheat is that it's kind of like a cat - it's got nine lives. There are multiple ways to get to yield in wheat. How many tillers there are, how many heads? How big are the heads and how big are the kernels? There are also biomass kinds of traits. Is it a big plot? Did that big plot really fill as many heads as we want? If you were just looking at the NDVI or greenness values or other data on the computer like that - you could pick an all-around later type favored by most of the numbers. Being “later”, though, is not necessarily the best thing for a farmer. If you're just looking at data, and don’t have the bigger picture of how the plant grows in the field, you can get really misled. You might start selecting things that are taller and later – but taller, bigger, and later can burn you, especially in the Central Plains where we get late-season heat and some of the late materials will not fill as well. You always need to be thinking about what the farmer needs. How are you going to spread the maturities over their whole farm? What is important in their risk assessment? How does the plant earliness or lateness fit into their workload? Things like that.
“It’s so important that you’ve seen it and done it yourself in the first place. I think jumping in and being willing to help and go train in the field and learn those field things – is important in a breeder.”
J: If you were hiring somebody to be a new wheat breeder, what are the things you'd be looking for?
S: Firstly, being a team player. I want somebody with the soft skills to get along with the team and communicate within the team. I want somebody who will do what needs to be done and pitch in when and where it’s needed. Some days that means that the breeder with a Ph.D. is organizing seed packets because that's what needs to be done. I want someone that's not going to have any issue with that and will be the person with a good attitude and do what needs to be done.
Regarding the technical skills and those kinds of things – it’s important to realize that you don't know everything when you go into a job. You don't even know what you're going to be doing in two years. When I went into my job my predecessor and mentor was like “Just get in here and you'll figure out what you’re doing in about a year.” At first, that was really unsettling to me. I have a Type A personality. I want to know what I'm doing and what's going on. But he was right, it’s important to come in and respect the way a program works when you start and then see what you bring to the table. You don’t want to be too soft-spoken, and not what you think needs to happen or what is right, but someone who can respect what's happened to get the team to the point that they’re at when you join the group.
I also look for somebody that generally knows their way around the plants, the plant breeding cycle, and the bigger picture. What is our goal? What is the importance of my task to that bigger goal? As long as you can talk me through those basic concepts and get along, that’s what’s important for me. I think a lot of people that try can do really well at this. It means applying yourself to a lot of areas that you might not have seen yourself doing before.
I think it's important that somebody has a general concept of what their field is looking like and where their data comes from. It’s important to respect and understand how much hard work is required to make agriculture work and get good trial data. I don't think you can get that without actually being involved in those activities and having done it yourself. If you're leading a program and you're expecting all the data to drop into your computer, but you don't know how much work is required to get good data, it’s a major knowledge gap.
It’s important to have a “jump-in” attitude. I might be dumping packets on the planter today or I might have to help with this or that. It's so important that you've seen it and done it yourself in the first place. If you apply yourself, you can go talk to people and say “Hey, can I help with this?” - That'll open so many doors for you. I think jumping in and being willing - willing to help and willing to go train in the field and learn those field things – is important in a breeder. When we're looking for candidates, we want to know that somebody knows their way around the field and is also conversational in data analytics, molecular technologies, and the breeding cycle, etc. You don't have to know everything impeccably, but you should be conversational in a lot of areas, and that includes having hands-on field time. You should know that “this is more or less what my plant looks like that I'd be selecting for”, “this is when the busy-time is during the season,” etc., - those are all important things to think about as you're considering a career in agriculture and crop breeding.
You can reach Dr. McNeil at sbatten@ksu.edu, or her LinkedIn profile here.
“Agriculture is something you pass on through family.” Dr. McNeil investigating seeding depth with son.