In a paper published in Ecology in 1985, David Wilcove provided experimental evidence in support of the idea that nest predation could be the cause for the decline of migratory songbirds in small woodlots in North America. Using fresh quail eggs places in artificial nests, Wilcove showed that predation was higher in smaller than in larger woodlots, near suburban neighbourhoods than in isolated rural areas, on the ground than a metre or two above the ground and, finally, when the nest was an open cup, which most migratory songbirds use, than when it was a cavity. Thirty-one years after the paper was published, I spoke to David Wilcove about his interest in doing this study, memories of fieldwork, and what we have learnt since about the decline of songbirds.
Citation: Wilcove, D. S. (1985). Nest predation in forest tracts and the decline of migratory songbirds. Ecology, 66(4), 1211-1214.
Date of interview: 31 August 2016 (via Skype)
Hari Sridhar: I wanted to start by asking you about your motivation to do this piece of work. By looking at your publication profile, I came to know that this was a paper that came out of your PhD. What was your motivation to study the effects of fragmentation on migratory songbirds for your PhD?
David Wilcove: Well, so I’ve always been interested in birds ever since I was a little kid, and it was therefore logical that, going into graduate school, I was going to do a dissertation that involved some questions revolving around birds. Also, while I was in graduate school, around my second year, I made the decision that I very much wanted to do an applied thesis. The science of Conservation Biology was then really starting to emerge, and I really liked some of the papers I was seeing. It captured both my interest in ecology and my concerns about the loss of biodiversity. So that meant I was looking for a thesis that would focus on birds and have a strong conservation component. And so, in that context, when I became aware of the nascent literature on the impacts of forest fragmentation and what it was doing to songbirds in North America, I got really interested in the question, and I thought it would be a good way to combine those foundational interests of mine, in birds and in conservation. And from that I developed a dissertation research plan, and in fact, the artificial nest experiments that were the basis of the paper you’re talking about, they were really only a minor part of the initial plan for the dissertation. But they worked out surprisingly well, and so they became probably the most significant part of the thesis.
HS: Tell us a little more about how this particular piece of work fitted into your thesis as a whole, in relation to everything else that you did.
DW: Sure. So, if one goes back to the early 1980s when I was a graduate student, there was a small but growing interest in the declines in songbird populations that had been documented in a few of the parks and protected areas in eastern North America where there had been breeding bird censuses conducted over long periods of time. There wasn’t a huge number of those sorts of studies, but the ones that were out there seemed to show pretty consistent and dramatic declines in songbirds. And people had connected those declines to forest fragmentation, and there were a bunch of studies that, essentially, looked for correlations between various life-history attributes, or attributes of the forest fragments themselves, and the disappearance of the birds. But they were really correlational. And there had not been, at that point, a lot of work explicitly trying to get at some of the mechanisms that might be driving the decline. There had been a paper that came out, I think, shortly before mine, by Stanley temple and Margaret Birmingham that compiled some evidence that cowbirds were a factor in the songbird declines in eastern North America. The brown-headed cowbird lays its eggs in the nest of other songbirds and in so doing lowers the reproductive success of the host species. They had put together some evidence that cowbird numbers had increased, that they were parasitizing the nests of songbirds closer to the forest edges than in the interior of the forests. And that was about it. So, there was clearly a need to dive a bit more into the possible mechanisms. And it seemed to me a possibility that nest predation rates might be high given the fact that there were all sorts of scavenging mammals and nest-depredating birds that liked to hang around agricultural lands and suburban areas, and perhaps they were moving into the fragments and harming the forest songbirds. So I thought this would be a really interesting thing to look at. And a couple of years earlier, two researchers, who I believe then were at University of Illinois – Betty Loiselle and William Hoppes – had published a little note in one of the ornithological journals where they had put out eggs on Barro Colorado Island – which is an artificially created island in the Panama Canal region, created when the floodgates of the canal were closed and the water levels rose – they put eggs on Barro Colorado Island and on the adjacent Panamanian mainland, and found much higher rates of nest predation on Barro Colorado island, which she and others attributed to the very high populations of coatimundis and other nest-robbing animals that were on Barro Colorado compared with the mainland. And so, I realized that I could modify her technique by creating artificial nests with real quail eggs and then put them in forest fragments of different sizes and forest fragments in agricultural areas versus suburban areas.
HS: How did you choose to do this study in Maryland?
DW: There was a team of researchers who had done some of the pioneering studies documenting the decline of songbirds in forest fragments of various sizes. That included Robert Whitcomb who at the time was with the Department of Agriculture, and Chandler Robbins, who was at the US Fish and Wildlife Service. They’d done some of the early studies on this topic. And I got in touch with Bob Whitcomb, and he gave me maps showing where some of these fragments were, so that, essentially, I could work in forest fragments where there had already been some bird surveys. So I knew that they were missing songbirds. And then I added the Great Smoky Mountains in Tennessee as a control site because that was the largest and most intact forest ecosystem remaining in that part of the eastern United States.
HS: I noticed that there’s a personal observation from you for the bird surveys in the Smoky Mountains. Did you later on publish the results of the surveys separately?
DW: Yes, I did. That came out in a journal that was at that time called The Wilson Bulletin. It’s now called The Wilson Journal of Ornithology. My thesis really had three main components. The first was this nest predation study. The second was to see if there had been long-term declines in the forest songbirds in the Smoky Mountains. And the third was more of a literature review, and an attempt to create a framework for studying habitat fragmentation. And that appeared in the Soule volume on conservation biology, the edited volume by Michael Soule. So let me just go back then to just take you through the three parts of it. So, the first part – the one that we’re talking about – was to see if rates of nest predation change as a function of forest fragmentation and the nature of the surrounding matrix. The second part was to see if bird populations had changed in the Smoky Mountains. Now the reason the Smoky Mountains were important is there was, at that time, a pretty strident but naive debate about whether the declines of songbirds were due to changes on the breeding grounds or changes on the wintering grounds. In particular, there was this hypothesis that the loss of tropical forests was harming their populations. And it was thought by many to be the most likely explanation for the declines. Now, the problem was that the long-term survey data from forests in eastern North America all came from heavily fragmented and isolated parks, places that had been fragmented for decades. There were no data showing what was happening to bird populations in large intact forest systems. So you could imagine that if you went to a large and intact forest and the populations of migratory songbirds were stable, that would suggest that it was indeed a problem connected to the fragmentation of forests on the breeding grounds, because the populations had collapsed in the fragments and were stable in the large forest tracts. Conversely, if you went to even the largest forest tracts left in the eastern United States and the migratory songbird populations had collapsed but the non-migratory species had not, then it would be strong circumstantial evidence that, in fact, it was something having to do with the wintering grounds that was driving the declines. I had fortuitously come upon some bird census data from the late 1940s in the Smoky Mountains. I wanted to go back there to see if the bird populations had changed. But I also wanted to do the nest predation experiments, because that was the largest and most intact forest that one could find in the eastern United States.
HS: What did you find with regard to the change in the fauna in the Smoky Mountains?
DW: Songbird numbers, particularly the numbers of migratory songbirds, hadn’t changed very much. They were more or less the same as they were in the late 1940s, which suggested that, indeed, the biggest driver of the declines that we were seeing in the smaller forest fragments in the eastern United States was due to changes on the breeding grounds – fragmentation, in other words. That was not a popular conclusion at the time. But subsequently, of course, the nest predation work proved to be consistent with that hypothesis.
HS: Who was your PhD supervisor?
DW: He was John Terborgh, who subsequently moved to Duke University. And John is, I think, correctly recognized as one of the towering figures in tropical ecology. But he was also one of the real pioneers in the emerging field of Conservation Biology.
HS: If you don’t mind my asking, I’d like to know how come John Terborgh wasn’t an author on this paper. It’s somewhat unusual, especially today, to find a paper by a PhD student that does not have the supervisor on it.
DW: Yeah, it was a different era back then, especially at Princeton. At that time, it was not customary to put the dissertation advisor on the thesis unless the dissertation advisor had been quite extensively involved in the design and execution of the project. John strongly encouraged his graduate students to design and implement their own research plans. That was the reason. It certainly reflected no disagreement between John and me or anything like that. Today, professors tend to be coauthors of the work done by their graduate students.
HS: Can you give us a sense of what fieldwork was like in those times? What was your daily routine when you were doing this work?
DW: Very tedious. It was not what I imagined myself doing when I thought of myself as becoming an ornithologist. I had bought dozens and dozens of little woven bird nests– little baskets that Canary breeders provide to their pet canaries when they want the birds to breed. I lined the baskets with straw and I had shipments of quail eggs sent to me, which I kept refrigerated. So basically in the morning, I would go out carrying eggs, straw baskets and straw, and place them in transects in forest fragments, alternating nests on the ground with nests in shrubs, and basically imagining I was a songbird and trying to figure out where I would want to place my nest. And then I would return at some fixed-time interval after that and check on how many of the nests were still intact and how many had been depredated. I did a second set of experiments using artificial cavity nests, which I made out of hollowed out tree branches with the help of Truman Clarke, a very kind scientist at the Department of Agriculture. I hammered these cavity nests onto the trunks of trees and put in the eggs and also monitored them, to see how they fared relative to the open-cup nests, because I wanted to see if the type of nest a bird constructs – cavity nests versus open-cup nests — would factor into how likely that nest was to get robbed.
HS: Did you try different kinds of material for the artificial nest? Similarly, how did you decide on using quail eggs?
DW: I had decided that chicken eggs would be too big, and I wasn’t aware of an easy way to get, say, zebra finch eggs, which might have been closer to the size of a North American warbler. So the quail eggs seemed like a reasonable compromise and there were various hatcheries that would ship me boxes of fresh quail eggs. And in terms of the lining: I think I decided to use straw just to keep things standardized. And of course, over the years, since that work, various aspects of that design were either criticized or were redone by other people.
HS: In one place in the paper you say that the experiment nests were more conspicuous than the actual nests of migratory passerines. So I was wondering if people have, in subsequent work, used material that matches the actual nests better?
DW: Oh, yes. That study precipitated, probably, hundreds of similar studies in subsequent years, and people did all sorts of things, including collecting real bird nests after the breeding season and then re-using them in a subsequent experiment.
HS: And did you do most of this work on your own?
DW: Largely. I had a couple of people who would periodically volunteer to help me, who were curious to see what was going on. And so, there were times during my field seasons when I would have an assistant. But, largely, it was my own work.
HS: In the Acknowledgements you mention the names of two people who helped in the field – E. Lowry and T. Clarke. Were they also students in your university?
DW: No. Truman was the scientist at the Department of Agriculture who helped me make the cavity nests. In addition to being a superb scientist, he was also a skilled woodworker, which I was not. And Elaine was a research scientist affiliated with Bob Whitcomb’s lab, and she generously volunteered to help me because she was interested in the project.
HS: You also thank someone by the name see C. Handley for help in identifying mammal tracks. Can you tell us who that was?
DW: Sure. He was then the curator of mammals at the Smithsonian. I had made some effort to figure out which animals were depredating the artificial nests, mostly by putting out track boards and by putting a clay rim around a different set of artificial nests to capture bite marks from mammals. Most of the time, the prints were pretty straightforward to identify. But if there were ones that were a bit problematic, I’d photograph them, usually with a coin as a reference point for size, and then show the pictures to Handley. I remember getting one tooth-mark impression on the clay rim of a nest, which I took to him and which he was able to match to a red fox. Most predators were identified with track boards, however.
HS: And I’m guessing this was before camera traps were easily available?
DW: Oh yeah, yeah. There weren’t a lot of people using camera traps back then. And I was working on a shoestring budget for this project.
HS: How do you travel between the sites you worked in?
DW: I drove between them. The farthest apart my forest fragments were was roughly 20 miles. The Smoky Mountains is a ten-hour drive from Princeton; once there, I hiked to my field sites.
HS: You thank someone by the name of S. Coleman for permission to work in the Great Smoky Mountains. Was he the park director?
DW: No, I don’t think he was the director. I believe he was the senior naturalist. I needed his permission both to put out artificial nests for predation experiments and to put up flagging tape to conduct bird surveys. He was also a fantastic source of information about the history of the park and the locations of good study sites.
HS: Do you remember roughly how long it took for you to write this paper, and where and when you did most of the writing?
DW: Let’s see, I remember finishing up the second field season towards the end of the summer. And we knew what the results were. Bob Whitcomb encouraged me to send the paper to Science. So, I submitted the paper, initially, to Science, and it was rejected. And after that, I rewrote it and sent it to Ecology, where it was eventually accepted. And I think I probably submitted the version to Science in the fall after my last field season, which I think would have been the autumn of my fourth year as a graduate student. So, I would have submitted to Science at the beginning of the fall of 1984. Shortly after it was rejected there, I rewrote it and submitted it to Ecology. And I think that it was in press by the summer after my fifth year. I don’t remember the publication month.
HS: Did you submit the papers before you submitted the thesis?
DW: Yes, I would have submitted the nest predation paper prior to writing and submitting my dissertation.
HS: Did you write the first draft when you were back in the university?
HS: Did it get reviewed at Science or was it rejected without review?
DW: You know, I don’t remember. I think it may have gone to review.
HS: Do you remember anything about the peer-review at Ecology?
DW: Well, the answer is yes; I went through a round of revisions with the journal. But what I remember is that you could submit your research as either an article or a short note. And I submitted it as a note. And then after the revisions, the editor told me that it was now long enough that I could either make it an article, which would delay its appearance by some months, or as a note, in which case it would come out more quickly. I opted for the article and the delay. And, in retrospect, I’m glad I did, because I think articles got more readership than notes.
HS: Could we quickly go over the Acknowledgements to get a sense of who these people were and how they helped?
DW: Sure, if I can remember them.
HS: C. Brown.
DW: That would be Charles Brown, who was a fellow graduate student, who has had an illustrious career studying cliff swallows.
HS: G. Cox.
DW: That was George Cox, who would have been the editor of the manuscript when it was at Ecology.
HS: A. Dobson.
DW: That’s Andy Dobson, who is now a colleague and close friend of mine in Princeton. At the time he was a postdoc.
HS: E. Greene.
DW: That’s Eric Greene. He’s now a professor at the University of Montana. He was a fellow graduate student of mine.
HS: H. Horn.
DW: Henry Horn, who was a faculty member on my committee.
HS: R. May.
DW: That’s Robert May. Perhaps the most renowned ecologist of his generation. He was a member of my dissertation committee.
HS: C. Mitchell.
DW: That’s Carol Mitchell. She was a graduate student.
HS: S Robinson.
DW: That’s Scott Robinson, now at the University of Florida. He was also a graduate student.
HS: J. Seger.
DW: That’s Jon Seger. He was a postdoc at the time. He’s now at University of Utah.
HS: How did you prepare the one figure that they’ve used in the paper?
DW: That was done pretty much by hand with rub-on letters. That was the pathetic way we did things back then.
HS: How was the paper received when it was published? Did it attract a lot of attention?
DW: It’s very interesting that you ask that. I was certainly pleased with the results and excited by them. My committee members said it was a good paper. By the time it came out, I had finished my graduate degree and I was working for an environmental organization. So, I was sort of out of the loop a bit and I didn’t fully realize just how interested people were in my paper. Sure, I knew it was a paper that had attracted some interest, but it came as a surprise, when, a number of years later, I was able to see the number of citations it had gotten. In June 1988, The New York Times published a big story about my work, focusing on the nest-predation study. But that was three years after it first appeared in Ecology.
HS: Do you have a sense of what the paper has been cited for mostly? Has it been cited appropriately most of the time?
DW: Yes, I think so. There were a couple of parts to that paper that, ultimately I think, caught on with researchers. So, the general topic of habitat fragmentation became a very, very hot topic in conservation biology in the 80s; probably the hottest topic of that time. Also, the case of the songbirds was of tremendous interest to a lot of people, because it related to how many of the National Forests in the United States were being managed. The prevailing wisdom, in terms of forest wildlife management at the time, was that you wanted to open up large forest tracts; you wanted to put clear-cuts and openings in them, to attract more wildlife. So, to those members of the environmental community who were very unhappy about this logging, my research provided a piece of ammunition to use in fighting those who wanted to do more logging in the National Forests. In that way, it caught on with the environmental community; that much I was aware of. But, of course, as it caught on with the environmental community, the Forest Service and the land managers and the academic community really wanted to know more about the forest fragmentation phenomenon, which created funding and other opportunities for research. So, there was a lot of new research that was initiated on this topic. The second thing that was of interest to people, with respect to the paper, was that I had argued that the reason these suburban fragments, in particular, had such high predation rates was because of all of the meso-predators that occurred in suburban habitats. Now, I didn’t use the word “meso-predator” in my Ecology publication; that term was coined some years later by Michael Soule. But I was one of the early authors writing about trophic cascades. There certainly had been some before me, and there would be many after me, but it was nonetheless somewhat novel to think about the role of trophic cascades in terrestrial ecosystems. Finally, the approach I took, for all of its weaknesses, was an experimental approach. And at that time in ecology, there was strong, and I think, justified criticism of a lot of ecological research because it wasn’t experimental enough. Studies weren’t well designed or they were purely correlational. I had the advantage of doing an experimental study in conservation biology at a time when such studies were relatively rare.
HS: What kind of impact did this paper have on your career one and on the future course of your research?
DW: It was very beneficial to my career, in terms of getting my job at the Wilderness Society, which was the second job I held. I got that position in large part because the Wilderness Society was increasingly concerned about the amount of logging that was going on in the National Forests, and they wanted to have a scientist on board who could help them understand where logging was appropriate and where it was inappropriate. My background, as someone who studied force fragmentation, was very appealing to them. And that job was a wonderful start to my career. It got me involved in a whole bunch of land management issues that were interesting and high-profile at the time. My work, I should add, included the infamous case of the Northern spotted owl and protection of the old-growth forests in the Pacific Northwest of the United States. And my work on the spotted owl controversy, in turn, brought me in touch with a lot of people and led me to make contacts with scientists at other environmental groups, which, in turn, led to my next position at the Environmental Defense Fund. In a way, that Ecology paper had a trophic-cascade effect of its own, and it benefited my career quite a bit. I think, for some scientists, if they remember my name, they remember it for that Ecology paper. I hope that’s not true for all of my colleagues in science, in academia and environmental movement, because I’ve done a lot of research on different topics since then, but it may well be the case for some.
HS: What about in terms of the trajectory of your research? A lot of your research has been on migration. Was the work you did in your PhD, in any way influential in, your decision to work on migration?
DW: I think it certainly piqued my interest in the special challenges associated with protecting migratory species. And it was also my introduction to the issue of land-use change and how that affects biodiversity, which has been a major part of my research interests ever since. But I haven’t done a lot of work specifically on fragmentation issues since my graduate school days. That field, fortunately, grew more crowded, and there were all sorts of great people doing interesting studies, so I looked in other directions.
HS: Today, 31 years later, would you say that the main conclusions of this paper still hold true, more-or-less?
DW: I think so. The idea that fragmentation can exacerbate threats to species is valid. The way in which the matrix may affect what happens inside a particular patch of habitat, or I should say, the impact of the surrounding matrix on what happens inside a forest or other ecosystem, is clearly critically important. But, you know, the really interesting thing is, three decades later, we’re still–and by ‘we’, I mean the scientific community in general – we’re still not sure of the degree to which all these different factors are contributing to the observed changes in songbird populations. So, the hundreds of studies that have been done since that paper of mine, most of which were vastly more sophisticated than mine, have made it clear that the question of, to what degree are songbirds declining and why are they declining, is a really difficult question to answer. I think the factors that I identified in my paper are valid in many cases. But I don’t think they’re valid in all cases. And even where it is valid, I couldn’t tell you, with certainty, how it compares to other possible threats. It’s a complicated story.
HS: If you were to redo this study today, would you do anything differently, given the advances in technology, statistics, and all the other studies that have been done since then?
DW: Absolutely. There is no way that that one would repeat today my 1985 paper precisely as I did not back then . You might monitor actual nests, preferably using cameras. And in places where you couldn’t find the bird nests, you would have to use vastly more realistic-looking nests and eggs. You would have to hide them better than I did. And you would, again, probably watch them with cameras. And you should do the whole project on a much larger scale than I did.
HS: Do people still use artificial nests to study nest predation?
DW: Yes, you still see papers coming out that used this technique. But I think that the bar to getting such a paper in a scientific journal is higher now, as it should be, given all of the complicating factors that have been studied since. I did that work in the early 80s. People worry much more about the size of the eggs, the scent of the eggs, the impact of having people check the nests, placement of the nests, as well they should.
HS: What has happened to these particular sites where you work for this study? Have you changed a lot? Have you been back to these sites since the study?
DW: That’s a really thought-provoking question. No, I haven’t been back. I would be surprised if most of those fragments are still there. My guess would be that the ones in the suburbs have probably been developed and the ones in the agricultural areas have been suburbanized, if I can use that word.
HS: And the Great Smoky Mountain site itself is a protected area, so I’m guessing that’s unlikely to have changed much?
DW: It hasn’t changed much in terms of its protected status, but a former postdoc of mine, Morgan Tingley, about 5 years ago, redid my bird surveys – not the nest experiments – but my bird surveys in the Smoky Mountains. And looking at his data, it’s clear that things have changed quite a bit since the 1980s. I’m waiting for Morgan to analyze and write up the data, but it’s a different place in terms of its birth communities now than it was back in the early 80s.
HS: That’s something I wanted to ask you about. In the last sentence of the paper, you say, “it is possible that forest tracts of 900 or more hectares may eventually experience declines in breeding populations of migratory songbirds”. Is that something that’s happened? Have people been monitoring this?
DW: I haven’t kept up with the studies well enough to tell you that. Certainly, in the case of the Smoky Mountains, it looks like there have been changes in the avifauna, but some of those may have been induced by the loss of important tree species due to non-native pathogens. The hemlocks, for example, are dying across the park, and they’re very important for some of the birds. Whether the overall populations have dropped –I think we’ll have to wait until Morgan has analyzed the data.
HS: When was the last time you visited these sites?
DW: I haven’t been back to the fragments since shortly after graduate school. I did go to the Smoky Mountains with Morgan a few years ago for a short visit.
HS: Have you had ever read this paper after it was published?
DW: Yes, I did, but I don’t think I’ve read it in its entirety for at least 20 years. My suspicion is that I would want to rewrite parts of it and use a more sophisticated statistical analysis.
HS: Do you remember in what context you read it the last time?
DW: No, I’m afraid I don’t.
HS: If you compare this paper to papers you write today, do you find any striking differences in the way you write?
DW: No, I don’t think so. I hope I’m a better writer today, but I don’t think that’s a badly written paper.
HS: Would you count this as one of the favorites among all the papers you’ve written?
DW: Oh, it’s certainly it’s one of the papers I’m most proud of.
HS: Any particular reasons for that?
DW: Well, I think I think it had a real impact on the field of conservation science, and a strong indirect impact on the way a lot of forests in the United States are managed. And I’m proud of both those things.
HS: You cite a personal communication of data from a D. Bystrak. Could you tell us who that was?
DW: Danny Bystrak. He worked with Chandler Robbins at the Fish and Wildlife Service, back when I was doing that research.
HS: What would you say to a student who is about to read this paper today? What should he or she take away from this paper written 31 years ago?
DW: Probably two things.First, ask yourself what you would do differently to tackle the same question today, given the way ecology has matured as a field over the past three decades. And second, ask yourself what parts of the experimental design and especially the discussion did I get right.