Emily Bray, PhD: Maternal Care in Dogs

Jessica Hekman: This week I’m talking with Emily Bray, an animal behavior postdoctoral researcher at the University of Arizona and Canine Companions for Independence. Emily’s area of interest is canine cognition, and she specifically studies the effects of different styles of maternal care on the adult personalities of dogs. In other words how does, how your mom treats you, affect who you will grow up to be. In this episode, Emily and I talk about what’s known about maternal care generally in animals and people, before diving into her studies of dogs, and what she learned from them. I loved nerding out with Emily about one of my favorite topics, and getting some behind the scenes looks at what it’s like doing these studies. So Emily, thank you so much for joining me on the podcast today. I’m really glad to have you here.

1:34

Emily Bray: Thank you for having me. I’m so excited to be here.

JH: So let’s see, we generally start out even before talking about peoples’ credentials, or any of that stuff, asking about the really important questions, the dogs that you live with. And I checked ahead of time to make sure you do have one. Thank God, I had a recent guest who only had cats. That was just traumatic for everybody. But tell us about your dog.

EB: Yes, so I grew up with Labrador Retrievers, but I currently live with an extremely happy and sassy Golden Retriever/Bernese Mountain Dog mix. And so she was actually my husband’s dog before we met, and he always jokes that she’s the reason I even talked to him in the first place. But she is an incredible dog. She has the world’s best temperament. But she is also somewhat of a medical miracle. So when she was five months old, she actually stopped being able to walk because she had severe hip dysplasia. And since she’s a large breed, my husband took her to three different veterinarians before one of them agreed to perform bilateral femoral head ostectomy rather than putting her down. And I am so glad that he persisted because she’s done really amazing. So she is super bouncy. People are always shocked when we tell them about her hips. But in January, she’s going to be 12 years old. So she’s done really well with it.

3:00

JH: Wow, that’s amazing. After this I’m going to have to ask you for a photograph, because I would love to see a Golden Retriever/Berner mix. That sounds adorable.

EB: Yes.

JH: Yeah, so I asked you on because you have a very cool research topic, which is basically maternal care and maternal effects in dogs. So I guess, trying to think whether I should ask first, how did you get into that? Or how did you get into research in general? So when you started your Ph. D. program did you know that this was what you wanted to do?

3:34

EB: Yes. But I did not know that I wanted to do this when I went to college, I guess. Right, so I think my interest was first… I knew I was interested in dogs. My parents are both veterinarians. Now my brother and his wife are veterinarians too, so I’ve sort of been surrounded by animals my whole life.

JH: But you escaped, you escaped having to go to veterinary school. Good for you. (Laughter)

EB: Yeah, well, it’s funny. My family makes fun of me, because I was always like, I don’t know, like, veterinary school, it’s like so much science. And now I do science all day. And my parents are like, we have no idea what you’re talking about. 

Um, so yeah, so I went to undergrad, and they had actually just started a canine cognition center. So I just happened to be in the right place at the right time. My friend was their first lab manager, and she was like, “I think you should volunteer. I think you would be really interested in this.” And you know, I missed my dog at home. So I was like, what better way to spend my time. And then it turned out that I could do research and so I did an undergraduate thesis. I studied inhibitory control in pet dogs. And my senior year, they started a project with Canine Companions for Independence, which is an assistance dog school out on the west coast, which we’ll probably talk more about later, because spoiler alert, I work there now. 

But so as an undergraduate, I was able to fly out and meet these service dogs. I was so impressed. When I first met them I’m like, “Oh my gosh, it’s The Stepford Wives,” you know, the dog world. But then it was really cool, because when we started working with them, there were so many differences, even in this, you know, highly selected population, in terms of their personality and their problem solving. So I was just totally hooked. And I applied to grad school. I knew that I wanted to continue in animal research and specifically, I would love it if I could continue researching dogs. So yeah, then I went to grad school at the University of Pennsylvania. 

And it was pretty cool. When I was at Duke, the amazing thing about the Duke Canine Cognition Center is that I got all of this great training in how to do cognitive tasks in dogs. So I got to work with Dr. Brian Hare, as well as Dr. Evan MacLean. And it was really fun to do these experiments in a lab setting. And then I went to grad school at the University of Pennsylvania and I worked with Dr. Robert Seyfarth and Dr. Dorothy Cheney. And that was amazing because they actually are not dog researchers. They studied vervet monkeys and baboons in Kenya for 30 years. So they knew a lot about observation and animal behavior. And so it was really great to get that viewpoint as well. And actually, my first summer they said, “If you’re going to study animal behavior, you have to go to Africa.” And so I got to go study spotted hyenas for the summer, which is actually where I first became interested in maternal care. The graduate student I was working with, we would get up early in the mornings and go out late at night to the dens and watch the mothers nurse their cubs. Hyenas normally have about two infants at a time. And there’s a dominant and a submissive nursing position. So we would record all of this information on hyena nursing. And you know, as I’m sitting there watching hours of hyena nursing, I’m like, I don’t think I know anything about this in dogs. And so, yeah.

7:23

JH: Wow! So the dominant and submissive nursing position is that, that the mom has different positions or the babies do?

EB: The babies do. So that the dominant cub would have the better nursing position. I guess I probably phrased that wrong. But yeah, there’s-

JH: No that makes sense that the dominant one gets the access to more milk probably. Right?

EB: Yes. Which is interesting though, because I don’t get the impression that that is the case in dogs, that there is a better position necessarily. And there’s also just so many more of them.

JH: Yeah, there’s a lot of nipples. I don’t know if it’s true in dogs? I think in cats that they have particular nipples, that each baby has their particular nipple. Is that true in dogs as well?

8:11

EB: No, no, it’s not. I talked to Julie Hecht about this because she is interested in it as well, because it’s a thing in cats, but I don’t think it’s a thing in dogs. At least not that I’ve seen and I’ve watched a lot of nursing videos.

JH: I tried to, well I successfully fostered a little several hour old kitten onto a litter of kittens that were like a week old. And he picked a nipple that was already taken. And he and the one week old kitten got into a fight over the nipple in which they were both—it was like a girl fight where they were both whacking each other on the face with their paws but did not have claws yet. And I’m sitting there. I was like, “you’re an hour old dude…” It was amazing!

EB: Yeah, that sounds more like a hyena than a dog interaction.

JH: Yeah, yeah. So okay, so then you got into maternal care in dogs. And so the way I first encountered you is through some very cool papers that you published and those were part of your PhD you said. 

EB: Mhm, yes. 

JH: And that was working with CCI Canine Companions for Independence as well, I believe.

9:24

EB: That was actually not, so it’s confusing. So I was working with Canine Companions at Duke. But then when I went to Penn, I actually started working with the Seeing Eye. So the same organization Dr. Eldin Leighton worked for. I believe he was a guest on your podcast recently.

JH: Yeah.

EB: Yeah, and so I actually worked with Dr. Seyfarth and Dr. Cheney but I also worked with Dr. James Serpell, who is at the Penn vet school and he had previously done research with The Seeing Eye with his C-BARQ, which again I believe has been covered on your podcast. And so he’d worked with The Seeing Eye as well as some other large, you know, assistance dog organizations. And so when I started graduate school, we drove out to The Seeing Eye and had a conversation with them that we were interested in studying maternal behavior, but also just, you know, their behavior in general and looking to see you know, kind of always the, the hot question in the working dog world of how we can predict success and what factors we can be looking for. 

So yeah, so it happens that The Seeing Eye campus, so they have their Morristown campus. And that’s where the dogs are trained, and the people go to be matched with the dogs. And then they also, about 30 minutes away, have a breeding station in Chester. And that’s where all of their litters are whelped, which makes for an amazing observational opportunity from a research perspective. So it made a lot of sense that we could do this work at The Seeing Eye because, you know, it’s a two hour drive. I should say at this point too, I always like to shout out to the undergraduates that helped me in all of my research. A lot of the research that I’m talking about today would not have been possible without them. There were 5 Penn undergraduate women, and one from Franklin and Marshall. And they would drive out with me, you know, so it’s like, close enough to be driving distance but still kind of far. And so we often would spend multiple days there because it didn’t make sense to do the drive back and forth. But yeah, so that was great.

11:42

JH: Yeah, the laboratory where I work has a similar relationship with Guiding Eyes for the Blind. So I know exactly what you mean about like well, it’s not that bad, but sometimes you spend the night.

EB: Yes.

JH: Yeah. Okay, so then you started studying maternal effects. But in order to put that into context, it might make some sense for us to take a step back before we talk about what you did and what you saw. And talk about some of the research that has gone before. And one of the things that’s very cool about Emily’s research is that she has done it in dogs, specifically. But there’s a lot of work that has been done on species that are a little bit easier to study, rats specifically. 

EB: Yes.

JH: So do you maybe want to give an overview of those rat studies and sort of like what, how you understood them to be like, what they were, how they were setting your foundations and what they were setting you up to expect to see in dogs? 

EB: Yes. There have been a lot of studies of early maternal care in rats, and mice, I believe, rodents broadly. And right, it’s great, because the generation time is much smaller than in dogs. And basically what they find is that when you observe these rats, they do this behavior called licking/grooming and arched back nursing. And so right, that’s sort of what it sounds like they’re licking their pups and they’re nursing them kind of standing over them. And it provides a lot of tactile-

13:16

JH: Giving them easy, easy milk bar access.

EB: Exactly. And it gives them a lot of tactile stimulation. And so you can sort of categorize these rats as either high licking/grooming, or low licking/grooming. And it turns out that based on the licking/grooming status of the mother, the rat offspring have certain outcomes. So basically, the mothers that are prone to high levels of licking/grooming and arch back nursing, tend to produce these offspring that have, what we would consider better outcomes. So they’re less fearful, less anxious. The thing that I’m really interested in is you actually even see cognitive differences in the rats. And so they’re better problem solvers, for example, you know, in like the little maze tasks that they’re giving them. 

Um, and so the other thing that has been done a lot in rats, is sort of the other side of that scale. So basically, if you separate a mother from her pups for a long amount of time, that is going to then be detrimental to all sorts of down the road outcomes. And we see this in primates as well. However, there has been some interesting work where if you sort of manipulate the length of the separation—so right, like if the, you know, it’s like 24 hours, whatever, that’s bad, because you know, the pups need their mother. But if it’s these shorter separations sometimes that actually leads to better outcomes, which is interesting. And I think one, there’s sort of different hypotheses for why that might be. One idea is that if you do these short separations, and then put them back together, it sort of induces the mother to then, you know, lick and groom, and do all of these behaviors that we know are positive for the pups. 

Um, but also sort of on the other end of this, and I think this has been looked at more in the primate studies, and also as, as the primates are a little bit older. You know, the rat studies, this is on a timescale of like the first week of life, right? So really early. Whereas in primates, a lot of the work has been sort of over the first year of life and, and what those separations might look like. And there’s been one group that finds these short separations, again, are actually beneficial to the offspring. In ways, you know, that if it had been, if you just completely separate them, like obviously that’s bad, right? So yeah, so that’s sort of a super brief overview of the work that has been done. And so in dogs, there’s really not that much research, and a lot of the research that has been done was done sort of earlier, and it didn’t then longitudinally follow any of the puppies. So like, you know, you can describe the behavior, that’s great. We know what it looks like, but we don’t know what effect that’s having. And so just recently, there’s been a handful of studies that have started to look at that. Both in terms of even just you know, eight week puppy behavior, and then all the way out to I think the longest is two years. But again there’s like, two studies that have looked at that.

16:48

JH: Yeah it’s hard to follow for that long for sure.

EB: Yeah. Well, and especially in dogs because it’s on the scale of years, unfortunately. Well, fortunately, yeah.

JH: And because they end up being in homes, right? Like it’s so much easier with the rodents, where they’re just there in the lab, and you can just sit there and watch them.

EB: Exactly, exactly.

JH: Yes. All right. So you had this, this backdrop, and then you went to The Seeing Eye and studied it. So the first question that you asked was, can you even see any differences in maternal care in different dogs? Are they all basically exactly, and you didn’t expect them to be all exactly the same. But the first thing that has to be done is to actually describe what are the differences that we see? And how wide ranging are they? And what breeds were these dogs? 

EB: Sure.

JH: So that we know how homogeneous the whole population was.

17:41

EB: So it was pretty homogeneous, but more heterogeneous than a lot of organizations in that they had German Shepherds. And then also Labrador Retrievers, Golden Retrievers, and then most of them were crosses between the Labs and Goldens. And all of the breeding happens at the station and the whelping also happens at the station. So sort of divided into two sections. So you have the breeding side, and then the whelping side. And then they live there with their mom.

JH: And so then, so how did it work? Did you set up a camera and video record?

EB: Yep (laughs). So we had, you know, tripods and camcorders. And basically I would, over the first three weeks of life, I would choose three days from each week, and then do four 10 minute recordings during those days. So basically, we would have about two hours of footage per week over those three weeks, and we were able to follow 21 litters, and all of the puppies from each of those litters. So it ended up being 138 puppies. Um, and yeah, we went and I think it was between, it was basically the daytime, so like, 7am to 6pm. Around there. Um, and yeah, we just had these video recordings. And then the fun part was we had, you know, a million hours of video recordings that we went through and watched and coded for the maternal behaviors. 

And so they were: time spent in proximity to the puppies. The way that we operationalized that was actually, a lot of the listeners are probably familiar with this, but when the puppies whelped, they whelped into one of those like plastic kiddie pools. And then they would stay there for the first three weeks of life. Which was nice because it meant that we had this proximity measure where basically if the mother was in the pool, she was in proximity, but there was also a larger pen that she could choose. You know, she wasn’t being forced to be in the pool. So it was about her choice of where she was spending the majority of her time. And so I think, again, you were talking about the variation of mothers. We for sure saw variation in mothers and I think, for example, on that particular measure, the mother that was the most in proximity was there about 90% of the videos, compared to the lowest was probably 20% of the videos. And I’d say that the medium was like 50, or 60%, 55% of time spent in the pool. 

JH: Wow!

EB: So you get a wide variation. Which is good for the purposes of, you know, looking at what this then predicts. 

And so, proximity was one of our measures. We looked at licking/grooming just like in the rodent studies, nursing, and again, I did differentiate between the different types of nursing. By that I mean the body position of the mother while nursing. So the most common would be when she was lying laterally, so kind of on her side. The milk bar was open as you would say. Then there’s ventrally, which is she’s lying on her stomach as well but you know, a little more on ventral. Then vertical nursing, which would be either nursing from a sitting or standing position. So past studies have not differentiated between that, but I did. Partly because of the rodent literature, right? Because you have that arched back nursing which is distinct from just normal nursing. And so again you see variations. I was interested in trying to capture that. 

And the other behaviors were contact, so how much was she physically touching the puppies. And then finally, this sort of vigilance behavior, where the way that the Seeing Eye breeding station is set up is very, I think from a welfare perspective it’s really nice. Basically, the dogs are kept together, but it’s this circular space. And rather than, you know, a traditional kennel where it would be sort of like down the line. And the benefit of that, especially with shepherds, which as you know, are pretty vocal dogs, is that whenever anyone enters the area, all the dogs can see them. So the hope is that the barking is not as much as maybe it would be in some of these more traditional setups. But anyway, what that also meant is that, when we were watching back these videos, there were also, you know, we were videoing throughout the day and we weren’t, the one requirement was if we were videoing we didn’t want a human in with the dog, you know, because that would be distracting her potentially from her mothering or lack thereof. But you know, people are in and out, just like their day to day lives. And so we would see that sometimes, and other dogs are in there as well. So sometimes the mother would sort of be looking out into the larger room. And again, we would code that and think of it as a sort of vigilance behavior.

23:14

So we end up with all of all of those measures that I just described, but it’s sort of like, okay, now how do we turn that into data, right? Into a number that we can then start to explore. So, basically we did principal component analysis, which is basically just fancy terminology that means we’re summarizing our data by giving each mother a single score. And then all of the maternal style behaviors, as it ends up, are highly correlated to one another. So they all load onto that one component. So in other words, mothers who are spending a lot of time with their puppies are also spending a lot of time nursing them, touching them, grooming them, displaying vigilance behaviors, and vice versa. Um, so I mean, that’s a little simplistic, but basically, that’s what that is. We can give them a single maternal behavior score. 

Once we have that score, we can sort of look at its properties and what does maternal style look like? So the first thing that we noticed was that maternal behavior was stable over time. So like I said, we looked at them over three weeks. We found that mothers behaved similarly to themselves across those three weeks. The other thing that we did with our measure- I should mention so there, like I said, there’s a handful of other studies that have recently done this same thing where they’ve observed the behavior, the maternal behavior, and then try to link it to later outcomes. And they, as well, were able to do principal component analysis and found a single principal component. So this is encouraging, right? This was in beagles. And another one was in German Shepherd military working dogs. 

So that’s encouraging, that it seems that this is a thing in dogs, right, that you have this maternal behavior that can be described this way. But what we also did is that we validated it by showing that it was related to other measures of maternal care. So specifically, we collected some salivary cortisol, which is often used to look at stress levels in dogs and other animals. And basically, our procedure was that each night a few hours after dinner, so they wouldn’t have the food residue in their mouth… although I regretted this because they, a lot of them just weren’t salivating very much cause they’d already had their food. They’re like, “we know we’re not getting more.” But anyway, we made it work. And so we would take a baseline measure, and then our manipulation was that we would remove half of their litter for five minutes. This was meant to be a slight stressor, nothing out of the ordinary. The puppies got weighed every day. So you know, they’re used to the puppies sometimes leaving but you know, potentially a maternally stressful experience, slightly. And then we would sample their saliva again, to look at cortisol after, 20 minutes after that event occurred. And so what we found first of all, this was sort of interesting, just the baseline cortisol before we even did the manipulation, we found an association where the dogs that were the most nurturing mothers let’s say, had the highest levels of cortisol just at baseline, which was sort of interesting. But then also what, after you remove their litter and put it back, that subsequent peak in cortisol, even when controlling for baseline, was still higher in these more motherly dams. So we thought that was a good validation of our measure. 

And then the other way that we did it was a behavioral experiment where basically, I would wait for the mother to be nursing each week, one day, and I would kind of follow the set script where I would go in and I would say, “Hello,” kind of just stand there for one minute. And then I would sit there for one minute looking at them. And basically, they could choose to interact with me or not. But the idea is like, oh, here’s this potentially inviting distraction. Like, are you going to, you know, stay and mother your puppies, or are you going to be like, “Oh, my gosh, human! Drop everything!” And again, we find this nice association where the moms who in real life are like the most motherly as we know by our component, are also the ones during this little experimental manipulation that are less likely to leave their puppies, and want to spend the time with them. So that was interesting. 

The other things that we looked at, people always ask me about this, and I think because it’s interesting, is: so what are these factors that might be playing a role in maternal care? So for example, everyone always wants to know about breed. So obviously, I can only speak to the breeds at The Seeing Eye in this particular instance, but we did find that Labrador Retrievers had higher maternal behavior scores than the German Shepherds.

28:17

EB: I’m not able to say anything about the Goldens, mainly because [of] sample size issues. We didn’t have as many straight Golden Retrievers in our study. Um, but yeah, so that was interesting. And then we also found an effect of litter size on maternal style, which is perhaps not surprising, whereby the mothers of the larger litters display a lower quantity of maternal style. And this mirrors what is found in mice as well. And, you know, it’s physically difficult for a mother of 10 to interact with each individual puppy at the same rate as a mother of two. And I would think that tending to larger litters is likely more demanding and exhausting, right? So that might potentially lead the mother to spend less time with them. But we do see that association. 

And then finally, the other thing that I think is really interesting, and people are often interested in, is this (unclear). And by that, I mean, you know, the number of litters that the mother has had before. And there have been a few studies on this. I think our study would agree with this as well, that the first time, dams look different than dams that have had multiple litters. Regardless of how many that might be. And so, basically, we found that the first time moms showed higher levels of maternal behavior overall. The other study that I was looking at that has looked at this found kind of similar, but basically what they found was that the mothers that have been mothers before showed a lot, you know, their behavior was very stable across time. Whereas they found in their first time mothers, it was increasing over the weeks and then was like highest in that third week. So, yeah, so anyway that’s what we found in terms of the demographic features. Although I should say they’re not simply redundant with the maternal behavior component. So there’s more than just these factors that are determining how a mother cares for her puppies, and, you know, individual differences and all of that. So, yeah…

30:29

JH: Yeah, for sure. It’s easy to overemphasize and say that our findings predict everything, but they never do. I’m curious, did you check differences in cortisol levels between the first time moms and the ones who’ve been moms before?

EB: Oh, interesting. Um…

JH: Hard to remember, right? I’ve just got to think that the first time moms were super stressed out by the whole experience.

EB: Yeah, yeah, right. Well it’s always so funny when you see a first time mom whelp, and she’s like, “what just happened?” (Laughter) 

JH: What just happened? (Laughter)

EB: You’re like, what must be going through their heads? No, that is a great question. I could go back and look, I don’t think we looked at that specific breakdown. But I mean, let me think about this, though… if we see that the, I think probably yes, because just going by the fact that we know that the first time moms, based on what I just said, are the ones that are showing the highest levels of maternal behavior. And we also know the moms that have the highest level maternal behavior have the highest levels of cortisol. Yeah, but I guess I wonder if that’s what’s driving it?

31:36

JH: Yeah, that’s a good question. Right? Interesting. Yeah, you would need larger sample sizes, even to untangle all of that. And I am sympathetic with how much dog TV you had to watch. I did that for my masters. I watched a lot of dog TV to score behaviors, and I’m very sympathetic with how it is mind numbing after a while. You start out being like, “I get to watch dogs on TV!”  (Laugher) After a while you’re like, “Oh my God. No more, no more dogs sitting, licking puppies.” 

I noticed and it’s really nice, I noticed how careful you are not to put value judgments on the different types of maternal care. It’s really easy for us to say that the, what you’re referring to as the more motherly dogs, or the higher levels of maternal care, it’s easy for us to refer to that as good moms and the other ones as bad moms. And I know that a lot of people in these fields of maternal effects emphasize that we shouldn’t think of it as good or bad mothering. I don’t know if there’s anything more you would want to have to say to that or how the different effects can be appropriate in different environments maybe?

32:39

EB: Yeah, no, I think I agree with what you’re saying that it’s hard to assign, and I don’t think we should assign value, because I think part of it even might depend on your end goal, right? So like, a “good” mother, for a certain working dog population might be different even for another working dog population, versus for a pet. And it’s like, what the goal is for the puppies, right? Because I think too, we can’t really assign value. Sometimes I feel like, one way that I do hear people assign value, and this is valid, is they refer to a mother as a good mother if it’s not a lot of work for them. Right? Like for the humans involved.

JH: Yes, that makes sense. 

EB: Because obviously, all of these dogs, the humans are pretty heavily involved. I mean, you know, on a sliding scale, I’m sure, but that’s just the nature of it. And so I’m always interested, because I feel like we could definitely find that actually, these moms were like, “you’re the one that’s having to go in and clean up all the poop and whatever.” But actually, then the puppies turned out great. So it’s like, sorry, you just gonna have to do more work. (Laugher) But, you know, that’s one way of thinking about it. But I think that’s the point, right? There’s multiple different outcomes you could want. And based on that, mothering you might think would be good or bad. Does that make sense? 

34:05

JH: Would push you in one direction or the other, right? So it might be preferable for certain outcomes and less preferable for other outcomes. But that doesn’t mean it’s good or bad.

EB: Yes. And you might, right it might be different among those outcomes, or most likely is. So I guess that’s why I am careful about it. 

JH: Yeah, no, as you should be. I just wanted to call it out because it was clearly something that you had previously thought about, but I wanted to make sure that listeners who hadn’t thought through that had the chance to understand why you were using that language. Which also sort of sets us up for then the next question was what affects the different styles of mothering?

EB: Totally, right. And I think honestly once we talk about this, it might also become more clear why I’m not assigning value. So we had collected all of this maternal style data, and then we were able to follow the puppies longitudinally until they had an outcome in the guide dog program. And we actually do find effects of maternal style, all the way into adulthood. And so in some ways that just in and of itself is pretty awesome, right? Because on one hand, we know that early experiences are so important. But on the other hand, the pups are with their mothers for such a short period of their lives. And yet here we are seeing these effects up to two years later. So what are the effects? 

First we looked at behavior when the dogs enter training. They would have been around a year and a half, around that. And they participate in what we call our cognitive and temperament testing. So it’s basically a series of standardized games that we play with each dog. And it’s supposed to give us a window into how they think, how they problem solve, how they react and recover, all those sorts of things. And so we did find some associations. One, we found that the highly mothered dogs showed higher activity levels in our isolation task and a quicker latency to vocalize during our novel object task. Both of those are actually potential signs of stress and anxiety. And they were also worse at one of our cognitive problem solving tasks. So they took longer to solve it and they perseverated at an unproductive solution even when it wasn’t yielding the results that they wanted. So again, back to this valence discussion. And this is maybe surprising, right? Because you’re seeing, you think of mothering, maybe normally as a good thing. And yet here are these dogs, showing outcomes that at least for a guide dog are kind of negative. And then actually, when we look at overall outcome, we find a strong association where the puppies of the less involved mothers are actually more likely to go on and graduate as a successful guide dog.

37:07

JH: Yeah, and it’s really surprising particularly because it’s exactly the opposite of what those rat studies found. In terms of like, confidence and cognitive abilities. 

EB: Yeah.

JH: So I actually talk about your studies in talks that I give about socialization in dogs, and about maternal effects and stuff. And so I have my hypothesis for why this is, but do you have a hypothesis? Probably a little more relevant than mine.

EB: I have multiple, I’d love to hear yours, if you want to say. Or I can give mine. I’m curious.

JH: I don’t remember if I got mine from you, or from the paper. I’m sure it did not arrive out of my own brilliance, right? I’m sure I got it from somewhere. And it was probably either from talking to you, because by the way, Emily and I have known each other for a while, or from reading the paper. But basically, my favorite hypothesis is that you have this population of dogs that are so highly selected to be guide dogs, that it’s possible that one of the things that you’re selecting for is very high mothering. And so you have dogs that are, so many of them are such intense mothers that it’s sort of too much. That there’s actually a sweet spot, and that in the rats it was easier to see the range. But in these dogs it’s maybe harder to see the range, because they’re all on one side of it, and the rats are maybe more spread out. And so maybe going too far in one direction is not good. Instead of being at 100%, you want to be at 80%. And these dogs are maybe going from 70% to 100% instead of 0% to 100%. So what looks like low mothering scores in these dogs compared to the rest of the world is maybe still high mothering scores. And what looks like high in these dogs is too high, if that makes sense. Although what you said at the beginning of our talk today was that you saw quite a range. So maybe that shoots that in the foot? I don’t know. So what do you think?

39:11

EB: Yeah, I don’t know. Well, I mean, that was very well said. I like that. And I agree with that. I think you’re absolutely right, that we’re only seeing— like they’re just a little snapshot of this bigger range. Or I like to think of this in a lot of my work, as this inverted U shaped function, right? Where again, there’s a sweet spot. And I think too in these rodent studies, again speaking of ranges, but you truly have the range and there’s no human intervention, right? They’re just letting them do their thing. In the guide dogs, they’re not gonna let them get to that, you know, the neglect/abuse part of this scale, the really poor scale. And even for example, if the mother isn’t spending enough time and the puppies aren’t gaining weight, they’ll supplement the puppies, right? So like, they’re not ever going to let the puppies fall off on that end. So basically how I like to describe it is that you can think of all of the mothers as adequate mothers. Or, I know we’re not using valence, but good mothers, right. And so within good mothers, you can go too far, just like you said, or…

JH: They’re helicopter mothers! 

EB: They’re helicopter mothers. (Laughter)

JH: (Laughing) Helicopter mothers. 

EB: Exactly. But it’s like, you can’t go too far the other way because The Seeing Eye isn’t gonna let that happen. You know what I mean? Um, so yeah, I think that’s a great way to think about it. Again, I feel like we’re just missing the other part of the scale because they’re not letting it be there. 

I was talking earlier about these primate studies and that group, I think they are in squirrel monkeys, they basically have the same. Or they have a slightly different hypothesis, but this idea that the lack of mothering for the dogs that are having the less motherly mothers, that that might be a little bit stressful for them. And maybe that’s actually not a bad thing, especially for a working dog. So maybe that’s almost like fostering some independence, they’re not having everything handed to them. So again in the primate work, the analogy would be these short term separations early on, where it’s basically like giving an animal a challenge and letting them prevail over it. And that’s gonna, that’s going to be good for them down the line. Even if in the short term, it looks like “Oh, you’re not getting as much mothering.” So there could be a little bit of that, as well. 

I think the other thing, or there’s some other hypotheses that are not necessarily mutually exclusive. So one could be that we know from the cortisol studies that these high mothering mothers have higher stress levels. And so obviously, the less mothering mothers, the mothers have lower stress levels. And so maybe having a more relaxed mom is better. Right? Or I know I just said stress was good, but maybe the mom’s stress is bad. Like, I don’t know.

42:42

JH: Yeah, no, I would love to pull out, I would love to see you redo the study with only experienced mothers, right? 

EB: Uh huh, okay.

JH: And just get rid of that whole confounder of the nulliparous mothers, the first time mothers. It’d be interesting.  

EB: To see if that would- yeah. 

JH: Yeah. It’d also be interesting to look at how much cortisol came out in the milk. Right?

EB: Mmm yes! Right. Because they do have a direct line. Right?

JH: Right. Right. So they’re more stressed. Maybe it’s not even the mothering, maybe that’s having this effect. Maybe it’s the amount of cortisol that’s coming through in the milk.

EB: Mm hmm. 

JH: Yeah. When two researchers talk about a study. (Laughter)

43:28

EB: Yeah, I know, that has always fascinated me that I have not yet tapped into looking at the actual milk, but I feel like that needs to be done, you know?

JH: Mm hmm.

EB: Because it’s so important.

JH: Yeah, that would be really interesting. Oh, which actually brings me… so when we’re talking about trying to piece apart the different bits of the picture, we had a question from Mikaela on Patreon, asking about whether dogs with a particular mothering style tended to produce puppies who went on to have the same mothering style. And I feel like that’s trying to pick apart genetics versus environment, in this case, environment being what the person who raised you acted like. And so we know that has been looked at in the rat studies. So I thought maybe you could talk about that and whether you were able to look at it in the dogs.

EB: Yes. Yes. So, unfortunately no, we’ve not looked at it in the dogs. Mainly because we haven’t been studying it long enough. Or at least I haven’t been studying the same population long enough. So but I mean, I think that would be fascinating, right? Because basically, you have to study the dog and then wait at least two years for one of her offspring to hopefully be picked as a breeder which, you know, most dogs are not picked as breeders.

JH: They don’t let you control that? (Laughs)

EB: They don’t let me control that, I know, I know. Someday… (Laughter) It’s probably for the best that I’m not in control of that, just for the record. Um, but yeah, so we can look at research and other species. And it seems likely that this is the case. That the moms with the highest mothering style are going to produce mothers with a high mothering style. But like you say, “Why?” Is it the genetics? Is it the environment of it, that they experienced that, or obviously, could be both. And so, there are some studies that have looked at this. So in the primate literature, one study found that over 50% of macaques who received abusive parenting went on to display abusive parenting toward their own offspring. Luckily, in dogs, I don’t think abusive parenting is even a term that we would use.

45:54

JH: No, but some bitches do eat their puppies. So…

EB: I have heard of that, actually.

JH: Yes. That one’s… that one’s a little harder to study because they don’t go on.

EB: (Laughs) The outcome was poor.

JH: But you can probably look at some of the surviving puppies and see how they did. So anyways, go on.

EB: So this is a happier one, the levels of mother/infant contact in vervet monkeys, was best predicted by the mother’s level of mother/infant contact as an infant, right? So those were highly correlated. And then similarly, in rodents, like we were talking about, this level of licking/grooming that females demonstrate toward infants is highly correlated to the level of licking/grooming that they received from their own mother. 

And so interestingly, in rodents, this actually does seem to be mostly environmentally mediated. And they can tease that apart through cross fostering studies, which is when the mom gives birth to her litter, and then they take half of her pups, and they put them on a different mom. And so you can know that okay, this is a high licking/grooming mom. So I’m going to take her pups and I’m going to put them on a low licking/grooming mom. And so then the question is: okay, so they genetically come from high licking/grooming, but environmentally experienced low licking/grooming, what are they going to look like as a mom when they grow up? And it turns out that they look like low licking/grooming, they look like the experience that they had. But again, we have not done that study in dogs. So, open question.

47:30

JH: No. In the rats, they um, just fascinating where they were able to see changes in part of the brain epigenetically. And they understand it super, super well, exactly what’s going on. And it’s so hard to do that in dogs, because people… like, cutting open puppy brains… I would not cut open a puppy brain, but I’m so desperate to know, just to see. I mean, it’s probably very similar. They’ve seen similar things in humans, looking at humans who die of like car accidents versus humans who die by suicide. And they’re able to look and see the differences in their brains in that same area of the brain, the hippocampus and the same epigenetic changes. So it’s probably true in dogs as well. If it’s true in mice and humans, it’s probably true in dogs. So it’s so cool. But I’m still waiting for the Star Trek scanner that will scan a dog’s brain and be able to say, what is the methylation status of, you know, of the glucocorticoid receptor gene in the hippocampus right here and have it tell me without me having to cut the dog’s brain open, but until then, we just guess.

EB: We just guess and study rodents. 

JH: Yeah. And so while we’re on Mikaela’s question, she also asked actually if having a C-section affected mothering at all. Were you able to look at that? 

EB: Oh, yes. Um, so no, I was not able to look at that. I think, it’s just a guess, fortunately it’s rare. Oh wait, no, sorry. I did look at that. But it’s super, super anecdotal. I mean, I didn’t look at it in the paper, but I went and looked at the data. And so let’s see, the sample size is low, as I said, because I’m just basically considering this an anecdote. But of the 21 mothers that we studied in the maternal style study at The Seeing Eye, three of them delivered by cesarean section. And so I think what you might hypothesize, or what you might worry about with the C-section, is that maybe the maternal instincts aren’t kicking in, because like, you know, the natural birthing process wasn’t allowed to occur, and when the mom wakes up, it’s sort of like what just happened? Even more so than when she’s naturally giving birth. But so when I look at those three mothers, they were all either average or higher on all of our maternal behavior dimensions that we measured. So, you know, anecdotal, but it doesn’t appear to be that they’re like, oh, they just quit, mothering never kicked in. And I’m sure you know, like everything, there’s probably individual variation, etc. But that is a super interesting question that unfortunately, just our sample size didn’t let us tease apart as much as…

50:20

JH: Yeah, you need a bigger sample size, man. Go back and watch more dog TV. Well, that’s cool then I’m glad that she asked that question, because I’m guessing because it’s anecdotal it didn’t make it into the paper. So we wouldn’t have known here on the podcast and nowhere else. (Laughs)

EB: That’s true.

JH: So I mean, I keep hoping that more people will do these studies. You did a little bit more analysis, you told me, I think this paper is not published yet? Where you were looking at some of the behavioral testing that is done.

50:59

EB: Yes. So that paper is published in Animal Cognition. That was, the maternal care did not enter into it. But we basically, like I said, when the dogs entered training, we did a series of what we always referred to as cognitive and behavioral tasks. And in our minds, we categorize them that way. Right? So if it’s a problem solving task, we’re like, “this is a cognitive task.” Whereas if it’s, you know, you opened the umbrella, and how does the dog react? That’s a temperament task. So what we were basically asking is, is that the case? Or are they kind of more intertwined than that? And what we find is that they are!

So if you try to, again, using the sort of principal component or factor analysis technique, you can basically either force the data, for lack of a better word, into the categories that you think it should be then you can look at sort of, “Okay, how does my model fit?” What’s the validity, like, right? Are these components then corresponding to things that we think they should be? Or you can have this sort of more bottom up method, where you’re basically like, here’s the data, how do the chips fall using this statistical model? When you do it that way, the results are actually much stronger. And what you see is that different tasks are clumping together, even though we might think of one as cognitive, and one as temperament. And I think intuitively, if you think about it, it makes a lot of sense. So for example, if you give a dog a problem solving task, but they’re kind of a very cautious dog, and they’re like, they touch it and it scares them, they’re probably not going to do very well on it. Not necessarily because they couldn’t have figured it out, but it’s like their fear is holding them back. So those things kind of go together in interesting ways. And I think the point of that paper is just that when we study dogs, it’s important to include both types of tasks, if we really want to get the whole picture, and also to continue to look to see how these things are interacting.

53:17

JH: Yeah, we as humans really like to categorize things. And I think as we, as science progresses, and we learn more and more, I think we’re starting to realize that we should not be categorizing so much. I like your phrase of letting the chips fall where they may. A lot of times in science, it’s important for us to just be like, let’s just see what the analysis says and not put things into groups.

EB: Yeah, something that I kind of like about this work with assistance dogs is when we, you know, often we do these tasks, and we’re like, yeah, we’re studying memory, we’re studying this. And it’s like, are we? Hopefully, maybe, I don’t know. But at least with this work, we can still, you know, it’s still a task that can be repeated. And if it has interesting findings we can go and see like, okay, dig deeper, what is it we were actually studying? But regardless, it’s like, oh, whatever it is, it’s important. And we know how to look at it now.

JH: Yeah, someday we’ll fully understand how the brain works. And then we’ll be able to understand exactly what’s going on with things. (Laughs) But until, yeah, until then. So what are you doing now? So you finished your PhD doing that stuff with The Seeing Eye? And I’ve hinted that you had ended up at Canine Companions for Independence?

54:34

EB: Yes. So I am a postdoc, postdoctoral researcher, whatever you call it. And so basically, I am half at the University of Arizona, at the Arizona Canine Cognition center with Dr. Evan MacLean, and half at Canine Companions for Independence, with their research team. And what that looks like practically is that I live, I like to say I live at the field site. So I live in northern California and I go to work every day at Canine Companions, because that’s where all of our study subjects are. And so these are not guide dogs, they are assistance dogs. So service dogs, as well as hearing dogs, and PTSD service dogs they have just started placing, and facility dogs. And those are dogs that they place with able bodied people, but then they take them to their place of work. So whether that’s a courthouse, or a school room, or like a therapy setting, and then they use their dogs in that way. 

So the exciting thing is that we have started a new maternal style study. Or I have started, it’s been two years now and it’s supposed to be finished, quite honestly. (Laughs) But, um, but it’s, it’s exciting. So in January of 2018, we received funding from the AKC Canine Health Foundation to study the effects of maternal style on future puppy behavior. Which, if that sounds familiar, it’s because honestly it’s a replication study in a lot of ways. Which is exciting, because we don’t have enough of those, you know, and especially in dogs, it can be difficult. So it’s a replication, but I’m also hoping it’s an improvement, we’re looking at even more. So we have a bigger sample, and some methodological additions that I’m excited about. 

So we are recruiting—so at The Seeing Eye we had 23. Here, we’re recruiting 60 dams and 240 puppies. So we ended up, at The Seeing Eye, every puppy that was born of a study dam, was included. Here, we’re only doing four from each litter. And that’s because we did a power analysis and figured out that gives you more bang for your buck. So basically, randomly choose four puppies from each litter, so we’ll end up with 240 total. And so the questions that we’re asking are very similar. First of all, what do we find in this different working dog population? So yes, they’re all working dogs, but technically the job that they’re doing is different. And theoretically, it’s going to require slightly different temperament, maybe different problem solving skills. So basically, I would predict that, you know, you would think that the behavior should be similarly affected by maternal style, regardless of what job you’re being groomed for. However, because there’s not this perfect overlap of characteristics found in the best guide dogs versus the best service dogs, it could very well be, again back to this valence thing, that the ideal mothering phenotype is going to look different between these two populations. So we’ll find out, hopefully soon. (Laughs)

57:54

JH: Is it more breeds as well?

EB: Less breeds, so it’s everything but the German Shepherd. So it’s Labs, Goldens, and mainly crosses? 

JH: I was hoping with uh, because sometimes the hearing dogs are, like, little.

EB: Yeah, they used to be, whatever, like corgis or something back in the day? But now they’re also the same. Sorry. Yeah.

JH: All right. Well, I’ll save up my money and send you some so you can analyze the cortisol in the milk before you’re done. (Laughs)

EB: Ok. Yeah, yeah, we did not milk the dogs. But so the other thing that I wanted to do with this study was to look more in depth at individual differences. So I didn’t mention this but in our first study, the maternal style score for the mother was assigned to each of her puppies, like we were not able to individually distinguish the puppies. This time around, we tried, we worked really hard, and we can. So they each have, it’s very high tech, they each have these velcro collars, that are different colors. It’s actually based on, Canine Companions does this normally, it’s based on their birth order, right? So puppy number one is red, puppy number two is blue. And then I have meticulously in Sharpie drawn unique patterns on each of them, right, so it’s like red stripes, blue chevron. And then that way, what it means is even an infrared light, we can tell the puppies apart. So now not only are we looking just during the daytime hours, but basically 24 hours, nighttime as well. Cause, you know, that could make a difference. The other thing, which, we didn’t actually even talk about this in the results of the other study, but maybe we can talk about it now. So we found that, like I said, maternal style affected outcome. But interestingly, if you looked at just vertical nursing, like if you pulled that out, you actually saw the opposite effect, where mothers that nurse the most from a sitting or standing position, their puppies were really successful. 

Um, you know, sort of in contrast to- and that also plays along with this idea of, if you think about it for a puppy, the vertical nursing is more challenging, right? It’s not just being given to them, they sort of have to like scramble and get to it. And so maybe, again, that’s like a little bit of a challenge for them and that’s beneficial down the road. But, huge caveat, the vertical nursing is a very rare behavior. So we did not sample it the way that you would want to, and most likely missed a lot of it. So this time around, I’ve specifically built in, we’re basically doing the same thing. These like 10 minute, focal samples throughout the first three weeks, but I’ve built in a day every week where we just do a 24 hour recording. And then we go back and only code that for vertical nursing, which is how you would want to do it for a rare behavior. So I’m hoping that will give us more insight into that behavior as well. We can see if we can replicate that finding or, or what that looks like.

1:01:08

JH: And I apologize for this. But now I’m also going to give you another thing that I wish that you had done. You should have put accelerometers on the puppies so that you could see if they moved more. Because I’m thinking for vertical nursing they have to start recruiting their muscles sooner to like, reach to get the nipples right? So I’d put accelerometers on them so if they got a couple days more of exploring outside the nest, because they were moving earlier, that could be part of…

EB: We track when their eyes open, which is not the same thing, but like… Yeah, yeah, we did not. We actually for a while had Whistles on the mothers. And so we have a lot of data on that. But then the last, I don’t know, 10 or 15 moms did not, because Whistle stops recording. Yeah, whatever. But anyway, so yeah.

JH: For people who don’t know, by the way, Whistle is a GPS collar, but it’s also an accelerometer so it can track motion.

EB: It’s like a Fitbit for dogs.

JH: Yes, it is a Fitbit for dogs.

EB: Um, and then so the other, or one of the other things we’re doing with this study is a more in depth look at how maternal style is affecting the offspring. So in the first study, we looked at, you know, behavior and cognition when they turned into training, and outcome in the program. And we’ll certainly look at those things again. But this time, we’re also looking at stress responses and hormone levels, because we know from the rodent literature that that is definitely affected. So basically, what that looks like is me collecting fecal samples at six weeks. And then we’re also able to get oxytocin at eight weeks, from a blood draw, because Canine Companions collects blood from the puppies for DNA purposes. So we can also, you know, process it in a way that we can get oxytocin. And then we cognition test the puppies at eight weeks, which has been really fun and frustrating at times (laughs). But, same sort of thing, both cognition and temperament tests all rolled together.

And then obviously, like I said, the outcome in the program, and it’s a different program this time, so it’ll be interesting to see.

JH: Oh, that’s great. I’m gonna look forward to that. So what’s your status now? You guys are, I’m guessing, coding all those videos.

1:03:42

EB: Yeah. Yeah. If anyone wants to, hey, all the listeners, if you want to watch puppy TV, now that Jessica has downplayed it, it’s actually quite fun. But yeah, so we are definitely in the process of coding, wading through. I mean, even gosh, just like getting, you know, compressing the videos, getting them in the format. Like, I feel like I’ve spent years of my life already. But yeah, so we’re doing that. And actually, the COVID, fires, blah blah blah… we, knock on wood, have recruited all 60 of the moms. Three of them are pregnant, but are about, you know, whelp, in December. And then if all goes well, then we’ll be done with data collection in February, hopefully. So, um, yeah. And then and then we’ll be looking at it. 

I should say the one other piece of this study that, I don’t know if this is gonna work or not, but I was so interested because part of what would be really helpful if we can identify a phenotype that is going to produce the kind of puppies that we would want in these different contexts would be well, how do we know what kind of mom dog is going to be? Can we figure that out before she’s even had puppies? Which maybe not, but what we have been doing is collecting some hormone measures on the dams pre pregnancy, as well as, they also go through this cognitive and temperament testing. So we can look to see if there’s any sort of associations, you know, personality wise, or her hormone profile that might predict the type of mom that she would be. So stay tuned for that as well.

1:05:33

JH: Yeah, yeah, I wish we could do it genetically. But we’re going to take 1000s of these moms to find the relevant genes.

EB: I know. We need to find a way to get the phenotype in a way that’s not as labor intensive.

JH: Right, well, so I have this fantasy that with the Functional Dog Collaborative, eventually, it will, so start out by providing lots of support to breeders, but then be able to sort of loop that in to research in a useful way. So I have this vision of the citizen science approach, where around the country, maybe we have a protocol where we explain to people how to video record it and then they could upload that to YouTube, and then maybe we could train people to look at it. So I feel like we could start getting some decent samples sizes that way. 

EB: Yeah. 

JH: And more breeds, more breeds.

EB: And more breeds, yeah.

JH: So that could be a really fun thing. And you could organize it! Cause I don’t have the time to organize it… (Laughter) So what’s your plan going forward? You’re a postdoc, which is a temporary position. Which direction are you going after this? Are you going to stay at CCI? Are you gonna go be a full time researcher?

EB: Ah, great question. I would also like to know! Yeah, no, I think my goal is definitely to continue in academia. And I plan to continue conducting research as long as I can. So in addition to the maternal side of research, we have a lot of exciting projects in the pipeline. So one thing that I’m excited about, a lot of my research, you know, as we’ve covered today, focuses a lot on this early developmental period, in terms of behavior and cognition. But we’ve recently begun collaborating with the Dog Aging Project, which is based out of the University of Washington.

1:07:21

JH: Wow! My laboratory is collaborating, same thing. We’re collaborating with them as well.

EB: Yes, exactly. Which is super fun. And, and basically, what this gives us the opportunity to do is look at how cognition changes over time and what it looks like on the other end of the lifespan, which will be nice to see. But um, yeah, so I don’t know what my plan is, but I hope it involves lots of research. Cause that’s what I love to do.  

JH: Yeah, I hope so too. There’s so much amazing dog research going on right now.

 EB: Yes, yes. It’s exciting.

JH: So if, if people wanted to get in touch with you, or is there a place you would want to send people to learn more about all of this stuff? I totally get it if sharing your email address on a podcast is not something, you know, some people are like, “Oh, I’m trying to sell my services so I’m happy to do that.” Or you could tell people where they could go to learn more about CCI. But if people wanted to dig more into some of this stuff, what would your recommendations be?

1:08:21

EB: Sure! So I have a website. And you can probably, actually I know you can find my email there. But it is emilyebray.com, just my name. CCI also has a website, I believe it is cci.org. Um, and they do have a section on their website about our science and research. I think that’s one of the things that I love about Canine Companions is that we have a research department. That is not common for a service dog organization. And I’m also on Twitter, I believe my handle is Dr. Emily Bray. But it’s like Dr. if that makes sense.

JH: I will verify that with you and I will put it in the show notes so that people can easily find you.

EB: Okay. Oh, perfect.

JH: Yes. All right. Well, this has been fascinating. So thank you so much. I really appreciate your coming on the podcast.

EB: I really appreciate you having me.

Thanks so much for listening. The Functional Breeding Podcast is a product of the Functional Dog Collaborative and was produced by Sarah Espinosa Socal. Come join us at the Functional Breeding Facebook group to talk about this episode or about responsible breeding practices in general. To learn more about the Functional Dog Collaborative, check out www.functionalbreeding.org. Enjoy your dogs.

Copyright © 2022 Functional Dog Collaborative. All rights reserved.