Maine Farmcast Episode 06: Precision Livestock Management with Dr. Derek Bailey and Dr. Colin Tobin

On this episode of the Maine Farmcast, Dr. Colt W. Knight, Associate Extension Professor and State Livestock Specialist for the University of Maine Cooperative Extension, meets up with Dr. Derek Bailey, Professor Emeritus for New Mexico State University, and Dr. Colin Tobin, Research Animal Scientist at the Carrington Research Station for North Dakota State University, to discuss some of the new technology emerging in livestock behavior and precision grazing including sensors to pick up location and activity, virtual fencing, and cameras that measure and weigh livestock in near real time. I attended the Society for Range Management (SRM) National Meeting in Sparks, Nevada this January where Dr. Bailey received the SRM Sustained Lifetime Achievement Award to take part in the Precision Livestock Management and Virtual Fencing Symposia. I was able to corner Dr. Bailey and Dr. Tobin to discuss their research and visions for the future of livestock production.

Episode Resources


Transcript

Colt Knight: 00:25

Alright. Welcome to the Maine Farmcast. I’m your host, Dr. Colt Knight, Associate Extension Professor and State Livestock Specialist for the University of Maine. And I am in Sparks, Nevada at the Society for Range Management Conference. And I am joined with Dr. Derek Bailey, Professor Emeritus from New Mexico State University, and Dr. Colin Tobin of North Dakota State University. And these guys are personal friends of mine and colleagues that we’ve all been involved with tracking cattle with GPS collars, and using sensor technology, and I thought it would be good if we could all get together and discuss what’s happening in the industry these days. So Colin, why don’t you go ahead and introduce yourself first?

Colin Tobin: 01:12

Thanks, Colt. It’s great to be in Sparks out of the frozen tundra of North Dakota. So at North Dakota State University, I’m the research animal scientist, with the Carrington Research Extension Center. I run a feedlot operation with a drylot cow herd, and then we also finish animals. And then on the side as well, we do integrated crop livestock work where we run cattle on croplands through different parts of the year looking at soil, animal, and plant interactions.

Colt Knight: 01:47

Thanks, Colin. Derek? Why don’t you introduce yourself?

Derek Bailey: 01:50

Yes. I’m Derek Bailey. I’m a recently retired professor from New Mexico State University from our Animal and Marine Sciences Department where I did a lot of tracking livestock and working with sensors. I’ve been tracking cattle since early ’90’s, like 199- or late ’90’s, 1998. And so I’ve tracked a lot of cattle, and now I’m currently working as the Director of Research and Outreach from Deep Well Ranch in Prescott, Arizona where we’re continuing to do precision livestock technologies and evaluating new technologies and developing algorithms to detect animal well-being concerns as well as using – developing algorithms and techniques to evaluate performance and impacts to rangelands.

Colt Knight: 02:45

What would you say are the top tools that we use in precision livestock to measure the behavior of the animals? Like, GPS is definitely one.

Derek Bailey: 02:59

Right. So the exciting part for me, Colt, is that, back in ’98, I mean, I’ve – for the last – for a long time, for years, ranchers have came and asked me, said, “hey, can I use these new GPS collars you’ve been working with to find my cows?” And they never – I never could do that because we always did the all the data was stored on animals. You’d have to get some animals in, put the equipment on, put the collars on, turn them out, get the cattle back in, and then look at the data. So it was great for research, but no real value. The cool stuff that’s happening today is that we’re getting real time tracking and real time sensors. We can see it real time, or near real time. Maybe in absolute real time within 20, 30 minutes. In other cases, maybe every day, but it’s totally different than we had before. And we can use this information, and that is the coolest stuff that’s available. Real time tracking and real time sensors – it’s an evolving field, and it’s exciting.

Colt Knight: 03:59

I always like to tell producers, I think in the future, with all these sensors that we’ll have on our animals, we’ll be able to get text alerts. If they’re out of the fence, maybe they’re limping, maybe they’re getting sick. And I think in the future, we’ll have that ability. I don’t know how long it’ll take, but I think it’s coming down the pipeline.

Derek Bailey: 04:24

I absolutely agree. And I just say, there are companies that are providing that information, but it’s still developing. It’s a developing field. There’s a lot of work that can be done, and probably should be done, to develop better algorithms and new approaches to write those. But you’re absolutely right, Colt. It’s gonna happen. It’s already here, and it is – this whole field and the entrepreneurs working in this field, they’re getting after it, and it’s exciting. And I think producers are gonna benefit from this in the long run.

Colin Tobin: 04:55

I think where we’re sitting right now, January 2024, we’re figuring out that there’s all these different sensors that we can use, and we can stick batteries on them, and we can put them on collars. We can put them on ear tags. But now we’re at the point of which ones do we actually need to use to look at behavior or welfare? So with Dr. Bailey, we’ve done – all three of us have done quite a bit of tracking of livestock with GPS. Some of the more in-depth sensors that they’ve come out with now, like with the accelerometers, which if you ever take your cell phone and want to change the direction of your screen when you flip it from straight up and down to on its side, that involves an accelerometer to change your screen angle. We can put those on cattle now and we’ve been doing them through research to look at changes in their behavior, not from their location standpoint or how fast they move in their velocity, but how intense their movements are. So between that, GPS, and then maybe some other types of sensors, such as pH rumen boluses that go inside of the animal to detect changes in core temperature, changes in pH of the rumen. So I think right now we’re in the phase of one, we’re trying to make things more durable because cow grade is much farther above military grade than anyone had imagined. But also kinda trimming all of the sensors that we do act- that we have access to, but which ones do we really need?

Colt Knight: 06:39

And right now, the technology is not available to ranchers and producers. They can buy the technology, but they don’t have any way to interpret the data. It’s really labor intensive, and it takes a lot of computing power. But we do have folks that are working on developing computer programs that will actually analyze the raw data from these sensors and just give us useful information. And with the machine learning and AI, I think that is going to exponentially increase our ability to use these technologies.

Derek Bailey: 07:20

Right. And so we – there are some technology that are already being used, and they do have the dashboards and ways to inform us. I get text messages all the time from some of these rumen boluses that Dr. Tobin was just talking about. But they were designed for operations that are very intensive where there’s all the things that are handy. Like, all the animals are close. They go by a certain reporting point. There’s probably Wi-Fi, like dairies. Some swine operations, even feedlots were all animal sensors. But on pasture, the challenge is much higher. And so those of us who like to work with pastures and challenges, we’re going to – we’re working on that. So I think, and even those companies are having their own thing, and so integrating the different company stuff, and different technologies, and preparatory developments are going to make that challenging. But we’re all working on it, and I think it’s a really bright and exciting future. There’s lots to be looking forward to. Producers can buy stuff now, but you worry whether it is gonna solve the needs that they think they’re going to solve. Because they’re working on it, and it’s changing all the time.

Colt Knight: 08:40

You mentioned the swine barn. I was in Des Moines for the National Extension Agent Conference this summer. And we visited a state of the art farrowing facility for pigs. And they were building it to be Prop 12 compliant. And for those of you that aren’t familiar with Prop 12, California passed a law that said if you want to sell pork in California, you have to meet our welfare and space requirements, which are two or three times that of normal swine operations. And I think the politicians that did that thought it would shut down the swine industry. I don’t think they did it to benefit the animals. I think they did it to try to stop the swine industry. But the swine industry in Iowa responded by building Prop 12 compliant farrowing houses, and they are integrating all this technology that we’re talking about now. They have got accelerometers built into those pigs so that they know their activity level, so they can tell where they’re – or if they’re going to get sick, or if they go off feed. There are contact sensors that lets them know how much feed they’re eating, kind of like a GrowSafe system for cattle, but they can measure how much. And their gestation crates, which are not traditional gestation crates, but it gives the sow a chance to go be by herself. She can walk in and out on her own. And the sensors will measure how long she’s in there and everything to keep it up. And one of the things that Colin’s working on now, they actually have overhead cameras where they’re starting to get weights and fat content on these pigs just by using those overhead sensors. And Colin, you’re trying to develop this for beef cattle right now.

Colin Tobin: 10:34

Yeah. We’re in the initial stage of setting up a project where we’re going to use red, blue, and green cameras as well as infrared style security cameras to see if a producer could establish these cameras in maybe a working slash weighing facility to where a fat – a fat steer, feedlot steer would come in, be weighed, spend some time on a scale or in a head shoot. Measurements would be taken via both those sets of cameras. And then the end goal of the project is to try to create an automatic sorting gate out the front so that way those animals that are prepared and ready to go to finish for slaughter, they can be sorted one way, whereas animals that aren’t quite ready and need to be on feed a little longer, they could be sorted back towards the pen. And I hope that producers that send these animals for harvest could receive more premiums that are afforded by some of these slaughterhouses by having very uniform truckloads that are coming in that we that are certified Angus beef, or any other type of alternative marketing as well as higher quality grades such as yield grade – yield quality grades or lower yield grades.

Colt Knight: 11:59

And Dr. Bailey, you’ve been at this longer than both of us combined. What are the changes that you’ve seen in the precision livestock world since we started integrating all these technologies in the ’90’s?

Derek Bailey: 12:15

I think that – well, the technology keeps getting better. And one of the things – one of the very, very few things that COVID did for us, now most of COVID was bad, but there was one thing that had helped. Because of supply chain issues that occurred during the COVID industry, the technologies for tracking and things have gotten smaller in real time. This real time tracking of containers make it – that technology has made it cheaper, and more available, and more useful for livestock. It just didn’t – we didn’t get it. It was a huge boost because of that because everybody needed to track where containers were all the time, and it needed to be in real time. That technology, not because of livestock, but there was a huge demand in livestock, but not because of that, because a container’s a bigger thing. We’re going to be – livestock producers are going to be the beneficiary of that as well. And I think that’s the biggest change is that created the ability for real time tracking and real time monitoring. And that is gonna make it change this idea of all of these sensors and tracking you’ve been using for research for years. Great tools for research, but now, finally, the exciting part is we’ll be able to do it for producers because the producers need the information in real time. And the next big challenge is to be able to process and use that information. All this huge data, we’ll be able to get it, we’ll be able to monitor it, but we need some way to make it from just data, need to be able to make decision making information. And that’s gonna be a challenge. The next big thing to do that, artificial intelligence. AI is gonna do that. And I worry about what AI would do to our society and to, you know, academic learning and everything else. There’s lots of worries about it, but it also could be useful. It absolutely could be useful, and I think that’s gonna be the biggest change in the next future is to be able to integrate that information.

Colt Knight: 14:11

Yeah. And the frontiers of all the sensor is really exciting. But to talk about some of the things that you’ve done in the past, you’ve used these different sensors to actually find cattle that are better suited for certain terrains. So let’s say that you’ve got a 100,000 acre ranch out west, and half of it’s mountains, and then half of it is in the valley, or down by the water tank, or something. And then you noticed some cows are lazy, and they don’t climb the mountain. And if they don’t climb the mountain, they’re not eating the grass that’s on the mountain, and you’re effectively reducing the efficiency of the pasture and lowering your stocking rate because you’re not using all of the pasture. And so you tracked those animals, and you isolated the ones that that were not as lazy and would climb the mountains, and then you found the ones that were lazy, and they just stayed down in the lowlands. And then you integrated genetic markers to find if there was heritability in those so that we could select animals based on their terrain usage.

Derek Bailey: 15:17

Yes. And that’s an exciting field because out in the west, it’s a big deal, especially on public lands. It’s that it – livestock, on average, tend to hang around the streams and riparian areas and overuse it and degrade it, can mess up fisheries habitat. But there’s always some cows that are just willing to climb, like you said, mentioned, Colt. And it turns out that that, like all traits, includes both nature and nurture. But our first look at the heritability of it, it looks like it’s something around, like, what weaning weight is. It’s something maybe 30% heritable, which means you could – we’ve made big progress in changing weaning weights. And I think –

Colt Knight: 15:56

That’s a moderately heritable trait.

Derek Bailey: 15:58

Moderately heritable. It looks good. We found some genetic markers that’s definitely associated. Not only we did, there’s a study in New Zealand. Same marker we found, they found as well. So that’s an all positive thing that maybe there’s something to this. We just need some more research on that area. But the exciting part about these technologies is it’s gonna create new phenotypes, new traits that we previously could never measure. Things like where they go on land, but also on how well they deal with stress, both cold and hot. All kinds of different traits that are really important for producers, but were hard to measure. We’ve made great progress on measuring things that are easy to measure, like weight and gain. Even carcass quality, you can – there’s all the imagery and all the things in the plants. You can get the data from them. That data is easy, but there’s a lot of traits that are really hard to measure. And all this sensor technology and good GPS tracking is gonna make these new traits so we can start selecting animals that are more efficient and fitter operations that we never could tackle before.

Colt Knight: 17:12

And what that’s gonna mean to the farmers, ranchers, and just the general public is we are going to create more efficient animals that use less inputs to produce more outputs.

Derek Bailey: 17:26

And with less of an environmental impact as well.

Colt Knight: 17:30

And, you know, the animal welfare aspect is always a big concern these days. And with these technologies, we can monitor welfare and make sure that nothing is being mistreated.

Derek Bailey: 17:41

So Dr. Tobin and I, especially him, we’ve done work to show that these things have real potential for welfare issues.

Colin Tobin: 17:49

Yeah. Down in Arizona, we use accelerometers and GPS units to track water deprivation. We did a study in the middle of June, Central Arizona, on one of the hottest days of the summer, I think. Or at least it felt that way. But we looked at how those animals change in their behavior when they come to water, when they want to drink, and then as they stand in water – at the water tank, but there’s nothing available. And we were looking at GPS units, and they did typically. They stood around the water tank. They didn’t go anywhere. They didn’t go back to brush and lay down like they normally would when they get a drink and wanna go loaf the rest of the afternoon. But one thing with the accelerometers that we found was the converse. They didn’t become less active. They actually became more active in that time. And, actually, with the activity, they stirred around around the water tank quite a bit, and they became more aggressive with each other and their calves, trying to hook on them, trying to get right up close to the water tank as close as they could just in case a little bit of water would squirt out of that hydrant. They didn’t want to miss out. So that’s one big one on just managerial. Maybe there’s a way that, in the future, here again, we see this comp- or this collaboration between GPS units and accelerometers to see they’re not leaving watering areas, and they’re becoming more restless. And then on the opposite side, we can use these devices to look at health. We worked on a study with some colleagues out of Australia where we looked at animals that were infected with this bovine ephemeral fever, which is very similar to West Nile for the U.S. producers. But as those animals that contracted bovine, ephemeral fever, a three day sickness, some of the symptoms are just motionless. They sit there, stand there, all stoved up. They get a little bit of a fever. They drool, but they just don’t move. But when they do have to move, it’s huge bouts because they’ve got not so much a limp, but it just hurts to move. So instead of moving all the time, it’s just in one giant move. And so that’s one thing that we did find. We looked at accelerometer data over time, and we saw this kinda change from constantly – consistently variable movement between grazing, and drinking, and loafing, to just a flat out drop in activity when those animals became sick. So I think in the future, going back to your point earlier about that text message that you’re going to get, I think those are very applicable ideas that are going to happen. Maybe in 2024, but probably by, you know, by 2030, we’ll be walking around, and we’ll be getting texts left and right whether we’re calving, or we’ve got a sick animal, or there’s something even worse going on out there. Right. Just to kinda follow-up, it’s exciting is that, we – based on Colin’s work, we’re carrying on with that and found work with computer science faculty and using artificial intelligence, deep learning, machine learning, try to develop good algorithms to see these changes behavior that you can detect with the sensors. And so you can do as soon as possible detect if there’s a problem without false positives. I mean, you don’t want to be crying wolf that there’s a problem when there isn’t one because there’s nothing more frustrating in the world to say there’s a problem when there really isn’t one. So you gotta get good information, detect the problem without false positives, and AI is the answer to do that. And with Colin’s data, we followed up with some other data. We’re able to do really good detection. And once that’s detected, the idea would be the sensors would send the information from the animal to the cloud, go through detection algorithms, and then send something – useful information about the animal that the producer could go react to, check it out, and hopefully treat them sooner, and not have many false positives. We just don’t – that is just something we can’t want. So the whole idea of detection and AI is just exciting, cool stuff that’s going to happen, I think, in the near future. And it’s fun to be kind od on the ground for this.

Colin Tobin: 22:35

I think one additional thing that’s really going to help out everybody is the fact that we don’t have to see those animals to decipher whether they’re sick or not. Now if you’ve ever – if you’ve been in the feedlot, usually the feed truck drivers can decipher sick animals a little bit faster than most pen riders. You know, those animals that don’t wanna come to feed, those feed feed truck drivers can decipher them real fast, whereas animals are – domesticated livestock are animals of prey. I mean, they don’t want to be looking like the sickest animal out there because they’re going to get picked off in their natural order of things. So when the that pen rider goes out there, those animals are gonna get rid of that sickness or anything that’s ailing them, and we’re not going to be able to see them as efficiently as we would if we didn’t – if we weren’t out there intervening in what they’re doing.

Colt Knight: 23:44

So we can’t have a discussion about sensors and livestock technology without bringing up virtual fence. And I don’t think the technology is quite there yet. But we are starting to test virtual fences. There are several companies here in the United States that are working with virtual fences. And the way that a virtual fence works is you would take a computer, a tablet, your cell phone, and then you would look at, like, a Google Earth image of your your ranch, and you would draw a fence line. And then the cow is fitted with a collar that has both audible and electric shock sensors. So as the cow approaches the fence, she gets an audible sensor that alerts her that she needs to turn back. And if she gets too close, the electric shock will prod her away from the fence. And so the idea here would be that they learn, in short order, when it beeps, go the other way, and then they don’t have to get shocked. But what is going on with that in the real world? So there’s a lot of folks looking at that now, and a lot of people trying it. And we – Colt, you already mentioned there’s, like, industrial grade, military grade, and cow grade. That is still an issue. I mean, it’s just difficult to make something strong enough to withstand the, essentially, the abuse of equipment that the cow will apply to it. So that’s an issue. Just make it strong enough. Secondly, there are issues in there. We actually did a study once where you have to worry about that. It’s very important to do the training because if an animal just gets shocked several times in the same area, it will tend to avoid it. We actually did a study where we shocked animals with and without clues. If they had a clue where the shock was and the clue left, a visual clue, no problem. If they got shocked and didn’t know and didn’t associate with that auditory, they would stay away. Just like if the animal gets shocked in a spot, they may not go back. So there’s a lot of potential issues behaviorally that could be there, and, plus there’s, like, welfare concerns. Do we want to have automatic shocking to our livestock and cause that pain? There’s a lot of information. Our colleagues are even looking at, like, cortisol levels about how what that does to them. I think it’s an evolving field, but the downside is the equipment’s got to get better, and you gotta know whether you really need it. And there are certainly some situations that I think it really fits. Situations where the boundary might change. It might be irregular. You need to move it. Maybe for targeting animals to eat certain things, like maybe even having sheep or goats stay in the spot to target a patch of weeds, or a patch of brush we don’t want. Those things look really good. But it may not fit everything because it’s not going to be cheap. It’s an intensive technology. I think you need to find its place, and I think it will have a place, but I don’t think there is really a truly silver bullet for all grazing management. And I don’t think it’s gonna be that. But I could be wrong, but in my opinion, I don’t think there’s gonna be a silver bullet.

Colt Knight: 27:24

What I’ve seen with virtual fence is the companies, when they release their promotional videos, they’re almost always, like, tame, mature cows and this idyllic pasture situation. And I think it probably does work pretty good there because there’s no line of sight issues. The cows are real gentle and tame, so they don’t want to go anywhere anyway. But what happens when we introduce calves or bulls? And I don’t know that anyone has tested that on virtual fence yet.

Colin Tobin: 27:57

There was some work, and I believe it was Oregon State, where they looked at creating a firebreak with dry cows, and then the second cohort with cow calf pairs. Dry cows, they were – they respected the virtual fence quite a bit up to a point. But the cow calf pairs, it seemed like when the calves would escape the boundaries because they weren’t collared, the mama cows would take off after them. And so that would be one drawback to it is mama cows are gonna disregard the fence, the auditorium and electrical shocks because we got to get to our calves. Compared to, you know, dry cows, or maybe potentially yearlings or something along those lines. I think one of the additional issues with virtual fence in a lot of scenarios is we try to graze pretty heavily. Animals wanna be selective. They want to eat the ice cream plants out there. They don’t wanna eat sticks and stems if they don’t have to. So a lot of times, I’m afraid that we’re trying to use virtual fences and enclosure device to graze things fairly heavily. And that’s where the failure comes into play where the animals just want to eat something nutritious. And regardless of the the beep or the shock, they’re going to go find something else to eat, and then they’ll come back and get some water potentially. But, you know, one – I think one value of virtual fence maybe rather than enclosing animals into an area. I think that’s gonna be more useful as an exclosure. Going back to it, you know, if we set up this paddock within a pasture, 20, 30, 40, however many acres to graze down to utilize, sooner or later we get to 30, 40% utilization, those animals are going to want to move. Because most of the leaf material is gone at that point. We’re back down to stems where most of the weight is. But if we can set up small exclosures within the larger enclosure, within the larger hard fenced enclosure. You know, maybe around riparian areas or key areas that we want to conserve at, you know, certain times of the grazing season or that setup. I think that’s more of a value to where those animals don’t want to be shocked, but there’s still plenty to eat where they’re at. So they don’t have to get into that loafing area because there’s going to be a beep and a shock that’s gonna deter them out of it.

Colt Knight: 30:39

And Dr. Bailey, you mentioned this, or touched on this a little bit when you were speaking about artificial intelligence. It’s like, what is it going to do to intellectual on humans? It’s like, are we just going to rely on AI? But with all these sensors and technologies that are coming in livestock, what about stopmanship?

Derek Bailey: 31:01

Yeah. I think that – I mean, it’s really – the thing is you don’t want to lose our relationship to our livestock. I mean, they’ll be able to – like, in stockmanship, we can use techniques developed by Bud Williams to handle our animals. We can spend time with them, and we can herd them places. We can – and we did some work where we actually used stockmanship to keep animals out of riparian areas. We’d go back in the middle of the day and move them out, and we really made a big huge difference by just going out and riding, moving them. And effectively, easily moving them away and putting them where there was other other feed, not by the water where where we didn’t want them, where they grazed plenty. And we’ve also used it to focus in areas, to graze areas for even fuel control and other things, to graze areas they wouldn’t normally graze. And we did that, and we were there. So we got – I think we had to develop that, and I think that’s an important tool. You just can’t wait, you can’t just rely totally on technology. I don’t think. It’s part of the great part, and some of the fun part, and rewarding part of having livestock is to interact with them. And you just can’t really use a technological thing to change that, I don’t think. I don’t think it is useful. It may come to a time, but I think it’ll be – I don’t think it’d be as effective. It won’t be as humane and more chances for problems. So I just think it’s important to keep that idea. Like I said, just to use stock – don’t use that as an excuse not to apply stockmanship, and take care of your animals, and spend some time with your animals. I think that’s part of – that’s part of – that’s part of our obligation to raising livestock. It’s certainly some of the most fun and rewarding part of owning your own animals.

Colt Knight: 32:47

So nothing is going to take the place of having experienced, knowledgeable ranchers and farmers raising livestock. These are just tools to help us and not replace us.

Derek Bailey: 32:58

That’s right. And I think that probably the big thing is then with these sensors and tracking, especially in real time, it may help us solve some of our labor issues because we may be able to use our sensors to help us check and train new employees that are not as well, and young people that are not as well, that they may not have the experience about how to do that. So you can use that. The rancher with experience can use the data from the sensors and the report from your employees to check those out and make sure there’s a balance. Like, did they really go out and check – did somebody really go out and check the water? Did someone really go out and look at those cows? This type of technology will help us decide that. And we may be able to use those sorts of things to help us with that, and also just make our time more efficient. It doesn’t – we’re not tucking it out, but we’re all getting busier. We all have other things. We all have – sometimes if we have our slow operations, we might have another job. And this will help us observe our animals with times we can’t. It’s not changing our relationship. It’s just making us more efficient, and giving it a second eyes and giving us a new set of data we’ve never had before.

Colt Knight: 34:13

Dr. Bailey, Dr. Tobin, it was great sitting here talking with you, and I think we’ve got to get to dinner tonight. So it’s good having you.

Colin Tobin: 34:24

Appreciate it, Colt.

Derek Bailey: 34:25

Appreciate it, Colt. It’s been fun.


In complying with the letter and spirit of applicable laws and pursuing its own goals of diversity, the University of Maine System does not discriminate on the grounds of race, color, religion, sex, sexual orientation, transgender status, gender, gender identity or expression, ethnicity, national origin, citizenship status, familial status, ancestry, age, disability physical or mental, genetic information, or veterans or military status in employment, education, and all other programs and activities. The University provides reasonable accommodations to qualified individuals with disabilities upon request. The following person has been designated to handle inquiries regarding non-discrimination policies: Director of Equal Opportunity, 5713 Chadbourne Hall, Room 412, University of Maine, Orono, ME 04469-5713, 207.581.1226, TTY 711 (Maine Relay System).