Donate
In this episode of K9 Conservationists, Kayla talks candidly about some mid-season musings.
Links Mentioned in the Episode: None
You can support the K9 Conservationists Podcast by joining our Patreon at patreon.com/k9conservationists.
K9 Conservationists Website | Merch | Support Our Work | Facebook | Instagram | TikTok
Transcript Thanks to Volunteer Katie Homan
Mid-Season Musings, transcript updated Katie Homan 10/12
Hello and welcome to the K9Conservationists podcast, where we’re positively obsessed with conservation detection dogs join us every week to discuss ecology, odor dynamics, dog behavior, and everything in between. I’m your host, Kayla Fratt. And I’m one of the cofounders of K9Conservationists, where we train dogs to detect data for land managers, researchers, agencies and NGOs. Today, I am sitting in my car at the wind farm on lightning stand down. So there’s been a couple lightning strikes, and we have to wait in our car for at least a half hour until the closest lightning strike is more than 30 miles away for safety.
I have been meaning to record a bit of a season update podcast, and I guess we’re gonna do it now. This episode may be a little bit wide ranging. I don’t have my professional mic with me. I’m just recording in the car. So if it’s too casual for you, just go ahead and skip until next week, but I’ve got a lot of thoughts, and I want to share them with you. One of the biggest things I’ve been thinking about this summer is how nice it is to be able to work with both of my dogs at the same time. Not that I’m running both dogs at the same time, like per turbine per search, but I will search one turbine with Barley and then one with Niffler and then one with Barley and one with Niffler. So it’s not that they have days that each dog works, they just get to trade off. It’s been really illuminating for me to see how much Niffler has grown and see how excellent he is as a search dog and also see some of his gaps compared to Barley, and actually see some gaps that Barley has compared to him.
So you know, I’ve been hesitant to talk about this too much because I don’t want to give the impression that my dogs are in any way not ready to work or in any way unprofessional. But they’re living beings, and they’re different, and they’re constantly learning and updating. So one of the biggest things is that handling Niffler and handling Barley feel really different. Despite being the same breed and obviously having the same trainer, me, they behave really differently in the field. Niffler is more likely to walk odor in the field, so he’s more likely to show a change of behavior, maybe turn up when and source the odor a little bit and then move on, and not source it fully or alert. He’s gotten much better about that throughout the season, but particularly if we’re having a day with really heavy casualties, and he’s already found a lot of casualties that given day, he starts to struggle a little bit more. And I think basically he gets hot and tired. So we’ve even had a couple days or a couple of turbines where it’s been really, really heavy. And after the first couple passes along the turbine, I will put Niffler away and bring Barley out and they will get traded out. And this is also one of the downsides of having dogs that are highly toy motivated.
Last season, Niffler worked well for food, and this season, he has been less consistent for food. So if he’s starting to get hot and tired, it’s really hard for me to find a good way to motivate and reward him. Overall, he just seems to deal with the heat less well than Barley, which is not what you would expect. Niffler is short haired and gray; Barley’s long haired and black. But Niffler is more likely, in the middle of a search, to want to go take a break in the shade than Barley, and then Barley is more likely to make false alerts.
So Barley, as I’ve described him in the past, is really desperate for his toy and really desperate for his reinforcers, and I think that comes into play, particularly if we have a plot with no casualties on it. Then he’ll start guessing or asking questions about other things. So, so far this season, he’s brought me to a dead tortoise, a couple of dead mice and a dead frog. In those cases, I just tell him, No, buddy, that’s not what we’re looking for. Thank you, let’s go, let’s keep searching. And it’s actually been getting better over the course of this season.
Yesterday, actually, we had two plots where he had nothing to find and he didn’t make any false alerts, which is a big deal for him. I think one of the things that’s really interesting about this is I think many of us find it more distressing to have our dogs walk odor. So, to have our dogs show a change of behavior, not fully source it, and then potentially, for whatever reason, you’re able to figure out that they did have something there. That feels worse, to me at least, than it is to have a dog make a false alert. I’m not quite sure why because both are technically errors. But I think, at least for this particular project, I don’t mind too much that Barley is a really liberal dog and will make false alerts because I can pretty much always confirm. And if I can’t confirm, it’s rare enough that I can’t confirm, then we’ve just got him on a variable reinforcement schedule. So, if he takes me to a tiny bit of bone or a tiny bit of hair, if I can’t confirm that it’s from a bat, I’ll just praise him and we move on, and that doesn’t bother him too much.
With scat, though, I think I would actually prefer to have Niffler. I would prefer to have a dog where…very rarely will Niffler make an alert that I cannot confirm. So that’s really nice for scat. Because sometimes with scat, a lot of times with scat, I cannot confirm one way or the other. Barley is a little bit more stressful, stressful to handle for scat projects because he is more likely to lead me astray and then end up running genetics lab fees up. Again, we’re working on that, I feel really good about upcoming scat projects and being able to handle Barley really well on those. But it’s really interesting how they have just different types of errors. And I haven’t actually sat down and done the formal signal detection theory calculations with them, but just anecdotally, I can definitely see it.
One of the other big things that I’ve really seen that’s been really exciting so far this season is I don’t have to give Niffler as many accommodations as last season. Part of this is because if we pull up on a plot, and there’s a bunch of cattle there, and I know it’s going to be hard for him to concentrate, I have Barley. So, last summer, I think one of the things I regret most about handling Niffler, when he was so young, and working him when he was so young, is that I didn’t have a backup dog to alleviate the pressure on him. That was because I gave Rachel Barley for the season, and I’m really glad I did. That obviously worked out well for all of us in the long run. But it’s been really nice this season to be able to ask Niffler these questions of Hey, buddy, do you feel ready to search here? Are you able to focus? Can you work with me here? And if he says no, I can’t, then I can just bring Barley out in search with Barley instead. And last season, my only option was to figure out how to get Niffler to search or to leave the turbine and come back later, which sometimes is really, really aversive and upsetting to me.
Primarily, if it’s a really, really hot day, there’s a lot of gates that I have to go through and it really upsets my schedule for the day. Or, if it’s the last turbine of the day, and it’s like, oh my god, like how long am I going to have to sit here until the cows move in order to get Niffler to a place where he can focus around the search at this turbine. So it’s been really nice that Niffler is really starting to turn on and show more of those mature search dog behaviors. He’s able to work with more intensity and more consistency this summer than last summer, and his toy play is also really starting to turn on. Last summer, I spent a lot of time wondering if his toy play was really where I wanted it. And with maturity, I mean, really even like week over week, this summer, I’m seeing more and more of that intensity and that drive that I really want to be seeing from him. And that’s typical, you know, he’s 22 months old now. Last summer, he was nine to 12 months old, our field season ended two weeks shy of his first birthday. Yeah, so those are all really good things.
And then circling back to cattle and some of the things that we’ve done, I’ll start with Niffler, as Niffler and I have been working a lot on some ready to work protocols. So I mentioned these in another podcast, I believe the episode I did with Taylor from Patreon, but basically what I’ll do is if there are cattle near a plot, if they’re really close, and I just know it’s gonna be a problem, I just get Barley out. I work with Barley, as long as it’s safe. I don’t ask Barley to work so close to bulls or mama cows with calves that it’s dangerous for him, but he can focus around the cattle at pretty much any distance.
But if they’re far enough away that it’s a bit of a question or maybe they’ve moved off a little bit, I’ve already done my search with Barley, and I want to do some training with Niffler, then what I’ll do is I’ll open the trunk of the car so Niffler can see the cattle from his crate. I will just watch to see if he’s like peering around me and staring at the cattle and his eyes are wide and maybe his ears are either all the way forward or swiveling back quite a bit, which is for him a big sign of conflict, then okay, you know, we’re too close. But I’ll wait, and if it looks good, then I’ll start feeding him. You know, can you eat, can you disengage in order to eat? Because like the first question is, even while you’re staring at the cows can you, will you eat? Yes. Okay, great. Now can I pause and see if he will disengage from the cattle, look at me, then I will feed him again, we do that a little bit, then I get him out of the vehicle. And we do that a couple more times.
So again, can you voluntarily disengage from the cows and reengage with me? Once he’s shown that he’s able to do that a couple times, then we start doing some cue testing. So I’ll ask him to sit, I’ll ask him to down, I’ll ask him to do a hand target. And I am looking for the latency, so how quickly is he able to comply with those cues, the enthusiasm and the focus with those cues. And if and when he starts showing that he is able to comply quickly, happily, with a lot of focus and enthusiasm, then I will ask him to search. And if he can’t, then I’ll either keep asking them questions for maybe a minute or two. And then we just put him away and we either do the search with Barley or we move on if it was just a training session. And I’m not going to beat a dead horse with it, where it’s like, if the cows are too close, and he just clearly keeps saying no, I can’t focus, I can’t think with them right here. Okay, cool. We’ll go search, we’ll go to the next turbine, I’m not going to stand there staring, like stepping on his leash just like waiting for him to disengage because staring is inherently reinforcing for Border Collies, in particular, and really, dogs. So I’m not going to like put him in a situation where he just gets to practice ignoring me and staring at cows because that is not a situation that is going to get better.
And then for Barley, one of the things we’ve had to do a couple times over the summer, because Barley is able to work in much closer proximity to the cattle, we have had a couple of times where some curious young steers turn around and stop just minding their own business and are actually following Barley around. And they seem genuinely really curious. They’re just like young steers, just they don’t really know what’s going on. And in a couple cases, I have actually had the opportunity to use a lot of Barley’s fancy safety training last summer, which has been fun, nice, good practice. So we’ve done several times where I’ve done his emergency down at a distance, where say, he’s 20 meters away from me, and I will tell him to lie down and he will stop searching at that exact moment and just lie down. It’s great. He’s very good at it, it’s one of his best behaviors. And then I can turn around and move the cattle off if need be, and he will actually hold his down while I am shooing cattle away, which is great. Then a couple times, I’ve also had to call him to car wash, which is our cue for middle. So he’ll come and stand in between my knees. And that has been really great if the cattle are close enough that I actually want Barley with me. And then I can put him in between my knees, feed him some hot dog, and think about what we need to do, in order to move away from the cows or get them to move away. I’ve only had to do that once. I just misjudged how far away the cows were. And one of them was really, really curious and bothering us. And in that case, we actually ended up aborting the search and going back to the vehicle. The cows had moved off over a rise, and I thought it was fine to take him out to search and then they came back.
So you know, we make mistakes, it’s okay. And I think all of this circling back to having the two dogs. I’ve been keeping track of how many finds they make each day and each week. I can’t share those numbers with you because it’s confidential for the clients. But I can tell you, Barley does seem to find a couple more each week. Most weeks, it’s really, really close, though, which is really nice to see. It’s been nice to have the data to prove to me that Niffler is already almost as good as Barley at finding bats. Again, because Niffler is a little bit more likely to walk odor and Barley is a little bit more likely to make a false alert. It’s actually not surprising to me that Barley is finding maybe five or 10% more bats then Niffler but he also is making more errors in that inclusionary area.
As I’ve been reading our science highlights and digging into the research and teaching our class and all these things, I’m really noticing there’s not as much literature as I would like out there comparing dogs, between dogs and comparing their efficacy. So sometimes the studies will say for example, Dog A had a 73% detection rate and Dog B had a 97% detection rate, but I haven’t seen any studies that really specifically dig into this. I think this is also one of the things that sometimes worries us about volunteer dogs or novice dog handlers because we don’t really know. A lot of the literature out there is written about these the best of the best dogs. They’re the dogs that are trained by the professionals, have had maybe a couple of years of experience and that are then going out there and showing these really great numbers. But one of the things that I have seen a little bit less, and there are some papers out there on this, don’t get me wrong, but a little bit less about the variability in dogs and how to best determine that.
And you know, from project to project that might be fine. So for example, here on the wind farm, my understanding is that they’re constantly calculating our detection efficacy, our searcher efficiency is what they call it. So it’s okay, if myself and another dog team have different levels of searcher efficiency because they’re modulating that in their algorithms. And then they’re using that to create a custom algorithm, is my understanding, for this wind farm so that this wind farm can feather their turbines appropriately based on literally the conditions on the ground that we are discovering with the bats. So, you know, correlating our finds with our searcher efficiencies and with the weather and the time of year, and all that sort of stuff, to basically create a custom plan to reduce fatalities on this wind farm, which is really, really cool.
So one of the things that I’ve seen with Niffler, as far as walking odor, so again, that’s like, he seems to catch odor, but doesn’t necessarily source it all the way. I think for him, it mostly comes down to fitness, and heat tolerance. So he, as a 22 month old boy, tends to gas himself out pretty quickly. He searches hard, he runs fast. And his particular play style is also very physical and very intense. So he tends to wear himself out more quickly. And Barley is just in better shape. You know, this, one of my friends who’s very into fitness, and all this said to me when I was talking to him about this, he was like, Well, you know, there’s a reason you don’t see 22 year olds winning Ironman, it’s like 40 year olds, and Barley is just in better shape. Barley also has more experience; Barley has been doing this for significantly longer than Niffler has even been alive. And he’s more motivated, which I think comes down to a really interesting little infographic that I made a while ago now about problem solving and troubleshooting, what went wrong if your dog misses a hide. So the way that I know that this happens occasionally in the field, because obviously if my dog doesn’t alert, how do I even know that he’s missed a hide, and this has actually never happened in training.
Sometimes I’ll see enough of a change of behavior that I go over and investigate. I see that he showed a head snap, moved upwind a little bit and then moved on, and I can walk over and actually find a bat not far from where he was that he never alerted to. So when I was talking to one of our patrons, in a private mentoring call with her, we were talking about this, and I basically broke it down in a couple in a couple different ways, when you’re questioning like what went wrong in a search, and why did your dog not alert.
So if you’re trying to figure out why your dog missed a target, generally most of this, you’re only going to know if you actually placed the target out and you know where it is, so you’re in like a training situation. But as I said, there may be a case where I actually see a bat. And this happens very, very rarely, like a couple times a year that I find something that the dogs missed. So I also don’t want to make it sound like Niffler is doing this all the time. This is something that over the course of like dozens and dozens of finds, it’s happened a handful of times with Niffler, if that, but it’s actually I’ve never seen it happen with Barley. I’ve never seen Barley show a change of behavior, and then not alert to something that I later realized was a bat. Because we will also obviously see changes in behavior where the dogs source odor upwind, walk away, investigate and then choose not to alert and that’s something different. That’s what generally is going to be a correct dismissal.
So if your dog misses something in a search, or in training, the first question is if they encountered the target odor, and are you sure? Because if they didn’t, or if you’re not sure, then we’re actually looking at like a search strategy odor dynamics question. Was their target odor available? Did the dog pass upwind of it? Was the odor lofting or doing something else unusual and weird in a way where they were not able to encounter a target odor, or was the odor looping where it goes up and then drops down and then goes up again? So they could encounter target odor that’s quite a ways away from the source and then if they move away in the way that they normally would when they’re problem solving, they wouldn’t actually encounter more target odor for a while, because that odor is like up a couple of meters above them, or even dozens of meters above them. Or is the odor like caught in an eddy behind a wind turbine? So there’s a high concentration of target odor, but they would have to actually get out of that eddy and again move a ways upwind in order to find the target odor again. So, it could be that they didn’t encounter target odor, or it could be that they encountered target odor, but the odor is doing something weird. It’s not in like a typical odor cone that we always think of, that we always hope for, so they struggled to source it. Okay, that’s the first question.
That again generally is going to come down to handling on some level like if you didn’t handle the dog appropriately in odor well, and I don’t want to say appropriately even because we don’t always know like these things change so fast from minute to minute with weather and wind shifts, and there’s microclimates, and there’s all sorts of complicated stuff. So it’s not like it’s your fault. But, potentially, if you see your dog show a change of behavior, but they don’t source the odor, there’s some things that we can do as a handler to support them and help them source that odor appropriately. Or they just missed the target odor entirely. In those cases, in training, I’m not going to do anything to fix it because if they didn’t find the target odor, they didn’t encounter it at all. I’m just going to treat that as a blank, a blank search. Great, move on.
Next, does the dog really know their target odor? And are they sure that this particular sample falls into the category of their target odor? So, this may happen if your dog has only ever been trained on frozen samples, and this is their first time encountering a live, a real thing? I mean, not alive, but like a wild version of it. Or they’ve only been trained on zoo samples and this is their first time dealing with a wild sample, or they’ve only dealt with frozen, frozen or dehydrated or whatever. Or they’re they just don’t know their target odor very well yet at all. So then it’s a question of, okay, we need to go back and re deal with imprinting. Then we’ve got our question of confidence. Is the dog confident enough to source odor and alert in the situation? For example, when we talk to Search Dog Foundation, it’s easy to imagine a case where a dog performs well in a contrived search environment. But then we get them out into, you know, a disaster after a tornado or something or the dog is overwhelmed and doesn’t really know, they don’t have the confidence to alert in that situation. And I’m trying not to say they don’t know how, because then the next thing is skill. Is the dog ready to problem solve this puzzle? Are they prepared?
So that comes back to the example of eddying or looping odor that is partially a handler problem to figure out how to assist your dog in those situations, and also partially a dog problem. Does your dog have that combination of confidence and skill to say, Oh, I got my target odor, I know what I need to do. And I understand the process in order to try to find my target odor. And I think Niffler sometimes still lacks a little bit in this confidence and skill area. Sometimes, you know, he really hates being wrong. And sometimes if I see him catch target odor, and I don’t respond quickly enough, he will pull himself off of it and come with me instead. That’s a confidence issue. He’s getting much, much, much better with that. We’ve done a couple really brief training sessions to work on it and it’s working wonders. But you know, that’s a little bit of a factor for him. And then skill, you know, he has primarily, well exclusively, worked on wind farms. So he doesn’t encounter as much eddying or looping or fanning or like fumigation, even like all of these different odor dynamics, things that can be going on, because generally, we just have straight winds across a prairie.
And then lastly, we’ve got motivation, and that is, is the dog sufficiently motivated for the task at hand. So if your dog shows a change of behavior, but doesn’t source the odor, well, you know, again, they walk the odor. We may see this in a dog who has more drive to hunt than drive for the reward that you’re offering them. They’re actually finding it more fun to search than to play with you at the end. So that then comes down to a question of communication and training and finding the appropriate motivator. And working through that.
And then the last thing I don’t actually have in this little graphic is fitness, which is what I was talking about originally with Niffler. I’ve noticed this is much worse if he’s had a lot of finds in a day and if it’s a really, really hot day. So that seems to be a little bit of motivation, because he’s starting to have competing motivators, who would rather stay cool than source his odor.
Let’s see, what else do I have going on in my head? So one other thing that I’ve noticed this year, so last year, in my episode on fringing, fringe alerts with Stacy Barnett, we were talking about how Niffler sometimes had a habit of alerting before he had appropriately sourced his target. So he would alert three meters from a bat, six meters from that, like way further away than I would like him to. And what I’ve actually realized is, I trained this, of course, I was so determined to teach Niffler to search independently of me and not pay attention to me that I was actually intentionally ignoring his changes of behavior until he alerted. Therefore he learned that the only way to get me to pay attention to him was to alert. So he would actually alert when he encountered target odor, and then bring me closer and closer to the target odor. So I basically trained him to do a distance alert or proximity alert.
And now that I know that’s what’s going on, I’m fine with it. I’m realizing that for Niffler, I need to be a slightly more supportive partner in the field versus Barley, who is much more of a like, Screw you, I’ve got this sort of dog. And that’s again, that’s experience and they’re different dogs. Our dogs are not programmable machines that are always going to turn out the same way they are different individual with with different learning histories, different experience levels, different motivation levels, and we see all of that playing out. So it’s been really fun to like realize basically what happened with mufflers fringe alerts, the fact that I taught him that takes the load off of him. And now we’ve been working on, if I notice him turning upwind, I will rotate and pause and wait for him and let him do his search and let him source it as need be.So yeah, that’s been really nice to see.
And one thing to keep in mind when we’re thinking about, you know, the fact that Barley is more likely to make these false positives and Barley and Niffler is more likely to walk odor and or miss something, make a false negative. And again, like both my dogs, in training and testing perform really, really well, so we’re really talking about the margins here. But one of the things that’s important here is, in a lot of our cases, for conservation dog stuff, as I alluded to, when we’re talking about a dog making a jump from Frozen, or dehydrated to real or from zoo to wild, is we want our dogs to be able to be liberal and to generalize in a lot of cases.
So for example, here on the wind farm, my dogs were originally trained on two species of bat that had both been frozen. And then they get out here and we are expecting them to find birds, they had never been trained to find bird carcasses. We’re expecting them to find all different sorts of decomposition, a variety of different species. So we average here, probably five or six different species are like 95% of what we find. But still, that’s three or four species that my dogs were never trained on. And they’re not just finding whole bats. They’re finding wing fragments, bits of hair, a bit of dehydrated bone. So we do really want our dogs to be able to generalize and a lot of ways and there was some really great conversations from the team over at Latrobe University at the Australasian Conservation Dog Network Conference, like all talking about different ways to get the dogs to generalize the way that we want while remaining specific. So I think it’s just important to remember that generalization or being liberal is not always a bad thing. Although again, like everything here is a trade off. So just something to keep in mind there.
But yeah, so in some cases, again, here, it might not matter so much that we get some different searcher efficiencies from team to team. But if you’re really working on an invasive species project that may matter to you more, you may really want a dog who’s finding everything, and a dog who only has 75% detection rate might be a problem. So I think, overall, just handling two different dogs and seeing the difference. At this point, I’m relatively confident that that difference is real. I don’t think it’s just that, you know, there are definitely some weeks where Barley just has one plot where it’s like, okay, well, there was a bunch of bats on that plot, or Niffler has a plot with a bunch of bats on that plot and like that can throw things off, because we may get like a group of bats that all fly into the same turbine in the same night. And also, most days, I’m searching seven turbines, so on any given day one dog is searching four and the other dog is searching three. But I’m pretty sure right now that we are actually seeing a difference in success rate for my two dogs. And I think again, it’s just really important to be thinking about that and be aware of that and test that so that we can use our dogs more effectively, create good models and sign the right dog up for the right project. If I had a project where I knew that it was going to be borderline impossible to confirm my dog’s finds on a given search, and I only needed one dog, I would rather handle Niffler for that project, versus if I had a project say, again, an invasive species project where I’d rather have a false positive than have a false negative, I would rather have Barley for that. So anyway, it’s it’s just really interesting to think about and I don’t know if there’s enough literature out there on that.
Kayla Fratt
Patreon book club is in full swing. We just finished up Detector Dogs and Scent Movement by Tom Austercamp and we’re about to start Canine Ergonomics: The Science of Working Dogs. To join our book club for three bucks a month head on over to patreon.com/k9conservationists. We also offer monthly group coaching sessions for aspiring handlers, puppy raisers and pros, as well as a monthly rotation of free webinars, workshops and roundtables with experts. Again, three bucks a month, up to 25 bucks a month, depending on what level of support you want to give and receive. Check that out at patreon.com/k9onservationists. I hope to see you join us there soon.
Kayla Fratt
So one of the other things I’ve been thinking about is rewarding for re-finds and how, you know, whenever we get a question on this podcast, my answer is almost always it depends. And this is a really great example of this. So basically, a re-find is if I’m searching with the dogs, and they find a bat up in the northwest corner of a plot. And we’re searching north to south and the wind is blowing north to south. I hope that this is coming across visually, I know it’s a little hard to describe some of this sort of stuff. So they find something up in the northwest corner. So then when we’re doing our next transect, that’s a little bit further south, the dogs may encounter that odor cone again, and source back up to a bat that they have already found. In most cases, I have not rewarded for that. And basically, what I’m hoping that they’re able to learn is that, hey, if there’s a bat with a pin flag next to it, you don’t get paid for that, if there’s a bat without a pin flag next to it, you get paid for it, so basically using the pin flags as a signal that they will not be rewarded.
And that works really well, particularly for Barley, Barley will actually source or source and then once he gets up to, you know, within a couple of feet of the target, if he sees a pin flag, he’ll actually just choose to walk away, which is great. That’s exactly what I wanted to see. But Niffler on the other hand, as I said, he’s a little bit more likely to walk odor, and therefore I would rather not. I have chosen to reward him in a couple of cases, because I’ve noticed there were a couple times where he like sourced odor halfway, didn’t go all the way up to the pin flag, and then chose to walk it. And that is what I want, but because I’m not always sure, sometimes you do have two bats that are very, very close to each other, I would rather have the dogs source it, and then make that decision very close to the bat.
Again, this is getting really granular, but I just wasn’t quite liking what I saw there. So I have started intermittently reinforcing Niffler for re-finds. Not every time. And then the funny thing is here, I know that because I’m doing intermittent schedule of reinforcement for refinds, and it is a leaner schedule of reinforcement compared to his intermittent schedule of reinforcement that he gets for actual first finds because I don’t reward them for every single bat, that’s important, we’ve talked about the other cases on this. I may be making his refind behavior more resistant to extinction. And I think that’s fine with me.
But it’s just really interesting. I am constantly updating and changing my training schedule, or my training plan and my reward plan in the field, because we’re not training when we’re in the field; we’re working. But of course, I have to decide whether or not to reward, and that fundamentally leads to learning. I’m constantly taking data on these sorts of things, and then modulating my responses. I’m not literally standing out there with a spreadsheet filling in numbers at this point, partially because we’re not, you know, I don’t have time. I’m being paid hourly, and my employers are not paying me to do that and I don’t want to take their money for a personal project. And partially, we’re constantly fighting heat out here and I don’t really have the time for that if I want to be able to get the dogs home before we hit 100 degrees. But I am constantly paying close attention to these things and modulating my responses based off of the trend lines that I believe I’m seeing. And if I’m starting to see a trendline that I’m concerned about, or starting to feel a trendline that I’m concerned about, feel a pattern that I’m concerned about, then I may start taking data in a more concrete way.
So for example, when I was a little worried about Niffler finding fewer bats than Barley, that was what started me on this route of counting the bats to actually see how different they are. And they’re not that different, not different enough that I’m worried about it. So yeah, and then some of the other things I’ve been experimenting with is generally I started at the upwind corner of our plots for the wind turbines. So if the wind is coming again from the north to the south, I will start at the northern corner, a northern corner, work from east to west, then move down to to south a little bit, then move, then search west east move down south a little bit more, east to west, you get the picture. So zigzagging across. I’ve been experimenting a little bit with starting at the downwind edge of my plot instead, just to see how that changes how the dogs work, how that changes, what we’re finding how that changes us dealing with refinds. The big thing, I’ve really noticed, this isn’t all that surprising, I walk more if we start on the downhill edge. Because if we start at the downwind edge, so if I start at the downwind edge, there’s a chance that my dogs are able to catch odor from a bat that is 100 meters away, upwind. Alright, so again, so say I’m starting at the southern corner, there could be a bat at the northern edge, that those outer cones will travel 100 meters, walk all the way up there, put it in my GPS, reward the dog put in the pin fly, blah, blah, blah, blah, and then walk back down to the southern edge to continue our search. Versus if we started the upwind edge, I have to do a little bit less of that. However, it does feel faster, particularly if we have very few bats on a plot like only one, two, maybe three. It seems faster to start at the southern edge of the plot. So just interesting stuff to be playing around with on that end.
So I think the last thing I’ve really got on my mind in the field right now is mental health in the field. I have struggled with it a little bit this season, I’m definitely really overwhelmed with a lot of grant writing in the evenings and just other projects, this podcast included, although this podcast is fun for me. You know, it’s a lot to handle. At some point soon, I would like to try to find someone to come on to the show to talk about mental health and self care in the field because we often are really sacrificing our minds and our bodies for this work. And in a way that is unsustainable.
And for me personally, a lot of this has stemmed from imposter syndrome. I think I’ve talked about this in the past on the show. But I am acutely aware of the fact that I only have four years of experience in this field, which is, you know, it’s not nothing, but it’s also not as much as I would like. The only way to get more experienced is to just keep doing it. I can’t rush that along. But I am acutely aware of being less experienced than a lot of people, yet still being a voice of this field in a way and that’s hard for me. I get a lot of requests for mentorship and for help that I don’t always feel ready for, but I’m also starting to recognize that that is a need and a void in this field.
This field is incredibly difficult to break into. Frankly, it can be really unwelcoming and very gatekeepy. I think that’s a huge problem in this field right now is that well intentioned people who are really concerned about the integrity of this method are making this field incredibly unwelcoming and incredibly difficult to break into. And I have spoken to literally, in the past 48 hours, three or four different people who are not yet members of Patreon, and they’re not in our course, they’re just trying to figure out how to get help, who have spoken to a lot of other professionals in this field and been rebuffed, or they’ve sent multiple emails and not gotten a response at all. I know we’re all really busy and not all of us want to be mentors, but that’s really tough, and it sucks and conservation needs us. And we we have a really powerful tool. And there is space for everyone here. And there is space for dogs who have a 75% detection rate. And for less experienced handlers, there’s a way to bring people into this umbrella and lift them up and lift up this method without excluding everyone. This is something that’s really important to me and I also feel really weird about it. I took a big mental health hit earlier this year when I was basically scolded by someone I really respected in this field for being irresponsible for launching our online course.
And I see what they’re saying, you know, we can’t teach all of this online. Absolutely not. But this field is so difficult to break into and the need is so dire in the conservation world. That it really sucks that we are mantling over this, this field and you know, trying to keep people out instead of trying to bring people in and lift them up. And yeah, my mental health has really suffered for that. I feel like a fraud a lot of the time. I feel like I don’t know what I’m doing. I feel like I’m not ready to be mentoring these people. But the reality is I am I know a lot more than everyone that I mentoring and I’m able to help them and when I’m not we’re able to ask questions and investigate and learn together and make this into a welcoming and exciting place, the way that I want it to be in the way that I wish it had been when I was new in this field.
You know, I’ve talked in the past about how hard it was for me to break into this field and how many, you know, just flat out rejections or how many times I was just plain ignored trying to break into this field. And this is not universally true. There are a lot of other amazing mentors and people who are really welcoming in this field. We’ve had a lot of them on this show, but it is definitely a gap. You know, I’ve been talking to people as far flung from, you know, I’m home based in Missoula, I’ve been talking to people from France and Poland and Finland and Kenya and Argentina and Costa Rica. There’s a huge lack of accessibility for this field. This is an episode that we’re going to have coming up in the future, but it really has tied into my personal mental health. And I think, you know, it’s just something that’s on my mind right now, this is mid season musings. So we’ll use that as a teaser for a lot of other upcoming episodes. And, yeah, let’s get to a Patreon question or two.
And then we have a question from, we have a couple of questions from Patreon that I wanted to go over. So Bec asks, considering the fact that many conservation detection dogs are trained on multiple odors and the gig economy of taking work where you can find it, if there was an occasion where you had trained your dog on an odor, which turned out to be incompatible with with a future job, could there be a case for retraining the dog to ignore that odor? And if so, how would you go about this? Is this something that can be done? And then she goes on to say, for example, you originally taught your dog to indicate on a species that is present in the environment of your new species, which may be a cryptic species, so it is vital that the dog only alerts to the presence of your new target species.
Yes, so this has been done. I have not personally done it yet. I believe Amy Hurt did this with Wicket at some point. I just remember overhearing a conversation or participating in a conversation when I was at Working Dogs for Conservation. And they were talking about how they were like, Kashi, I don’t think I’d do that again because it was really hard. Because you always are going to have that reinforcement history in there somewhere. And that’s, you know, fundamentally, the problem with eliminating any behavior from a dog’s repertoire is you’re always going to have that reinforcement history back there. So there’s a bunch of different threads we can pull on for this particular question. And I love this question. I love almost all of our Patreon questions. So one thing to consider is if the targets are visually distinguishable, so say for example, you’ve got a dog trying to find moose poop, which looks like Milk Duds, and you are about to go embark on a project where you are trying to find wolf poop, which basically looks like really, really gross hair filled bone filled, dog poop.
And moose and wolves obviously co-occur in a lot of places. I would say, Hey, it’s okay. You know, maybe give your dog a bit of cheese for the moose poop, you move on, the dog gets a nice reward, they get to continue searching. And then when you find the wolf poop, then you throw your big party with the Frisbee and the praise and the water and the break and all that good stuff. That’s probably what I would do.
And that’s like, I know, Keith Smith from Rogue Detection Teams talks about this a lot as like a huge benefit actually, like cross train the dogs, you know, and maybe reach out to like, your local herbivore group, or like the hunter conservation group in the area and see if maybe they want that information as well, maybe you can figure out a way to actually just study both species and improve, you know, maximize the efficacy of the dogs by bringing in multiple stakeholders. So that’s one way to do it. and that would be my preferred way.
But say, for example, the species are not visually distinguishable. So, the example I often pick on for this is red fox and Bobcat. Very difficult to tell them apart, especially if the scats are not like pristine textbook scats, and they’re not really fresh. There, I probably would, instead of trying to untrain a dog, I would just find a different dog. Um, so ideally, you know, this is part of why I’ve got two dogs, you know, within K9 Conservationists, if say both of my dogs had already been trained to find red fox, you know, okay, maybe Heather and Ellie get to work on that project. I know that’s not necessarily realistic if you are like, for lack of a better word, a lone wolf operator and you only have one dog, but it is definitely something to consider and the reason that visual distinguish, distinguishability is that a word?, matters is because, so in that moose wolf example, I can always tell that my dog is alerting to moose. I just say oh, good job. Yeah, you did find that. Here’s a bit of cheese, here’s a bit of turkey, here’s a little bit of moose steak. And then when we find the wolf I can really throw the party, and I can choose not to collect the moose scat and therefore not run up our genetics lab fees or waste a bunch of sampling bags or bog down my backpack with a bunch of heavy poop, because I can tell the difference. Red fox and bobcat, that’s not going to be the case in a lot of cases. So you may accidentally reward the dog for red fox when you actually really want bobcat, and this can happen when the dogs are not properly trained anyway. For example, when I was talking earlier about Barley, this is something I would really worry about with Barley, and it would be something that we’d have to work really hard on with him.
Another solution, there’s a bunch of different potential solutions here, you could just go ahead, in our bobcat, red fox example. The big downside of that particular one is that red foxes are going to be more ubiquitous than bobcat, so if your target is bobcat, you’re going to end up with a lot of extra crap, literally. And this would be less problematic if say you had a dog who was trained on wolverine scat and bobcat scat and say you only wanted to find bobcat. I actually, I’m not sure, I think you can probably tell the difference between wolverine and bobcat but say for argument’s sake, you can’t tell the difference between wolverine and bobcat. Your dog is trained on wolverine, but you want to go do bobcat. There are so few Wolverines out there that odds are, even if you’re in an area where their territories do overlap, you are going to be able to just collect a couple of wolverine scats. It won’t be that big of a deal because there are just so few wolverines versus again, red fox, there’s gonna be a lot more red foxes. So then you’re gonna end up in a situation where you’ve got like 60, or 70, or 80%, red fox scat, and like 20% bobcat scat, and obviously, your project partners are probably not going to be super happy about that because, again, genetics are expensive.
So say, for whatever reason, you had to do it, and you can’t tell them apart, and you’re gonna end up with a bunch of false positives, and you’re gonna end up collecting them all, then potentially what you could do is go the route of training the dogs a really specific match to sample sort of protocol. I’ve been speaking to both Simon Gadbois and Ken Ramirez about a similar project for this. I have not done it yet. But my thinking is, so say, okay, we know we’ve got a big cooler full of a bunch of samples. They’re all separated, obviously, and we know a bunch of them are probably red fox. We don’t want to bother paying to send those in for genetics lab fees. Okay, let’s teach our dogs to match to sample and then I can put a sample out, and then ask the dog to match Hey, is this red fox, and is this bobcat?
Again, I’ve not done this yet, but I’m really excited about the idea. In speaking to a bunch of really good trainers, again Ken Ramirez and Simon Gadbois among them, I think there’s a way to do this with the match to sample protocol. And then what Dr. Gadbois and I have talked about is the idea of if your dog is trained on bobcat, but has been inadvertently collecting red fox, you may be able to basically just put out a lineup and the dog will, in a lineup situation, be better at making that distinguishment. I have not done this yet either. It’s a little bit different from Beck’s question, because it’s more about the collecting the wrong sample versus something that the dog was previously reinforced for. That probably muddies the water or answers the question, but it’s a fascinating question.
The last option, yes, would be going through an extinction protocol. So we did this with Maddie and Percy while we were in Kenya. Both Maddie and Percy were finding caracal and leopard scat when they were supposed to be just finding cheetahs scat. They had never been intentionally trained to find those other cat scats, but at some point, they clearly had it bad because they were alerting very consistently on them. So, we hinted at this in our Kenya wrap up episode with Heather and Rachel. So feel free to circle back and listen to that a little bit more, we may do a full episode on this. We’re still working on crunching the numbers from this. And we will hopefully be publishing on it as a protocol and an example of how to work through this. So stay tuned.
And I think that’s going to be all for now. We’ve got a couple of other kinds of Patreon questions, but I think they are going to be better suited to YouTube videos actually. So stay tuned for that if you have not subscribed to us yet on YouTube or at K9 Conservationists on YouTube. The big thing we’ve got up there right now is every single training session that puppy Niffler and I ever did. They are unedited, so you may have to scroll around a little bit to, you know, there’s a lot of blank video of a search area as I’m like putting Niffler away and then resetting hides or something. But you know, it’s all there. And for anything else that you may need, we have coaching groups and book clubs and all sorts of great stuff over on Patreon. That’s patreon.com/k9conservationists. You can buy stickers and shirts and mugs and all sorts of great stuff with really, really cute designs on them at k9conservationists.org and we will be in your earbuds again next week. Thanks so much!