Migratory Journeys Mapped: The Magic of eBird
[Lisa Kopp] Hi, everybody. Welcome to today’s webinar. We are excited to have you all here. We’re going to get started in just a second. We are just a minute away from noon here in upstate New York where we are hosting from. Hope everybody is having a nice day so far. Here’s someone joining, or I see somebody is joining from Columbia. Welcome.
All right, we are at noon, again, in the Northeast. I hope everybody is doing well. I know that there’ll be plenty more people trickling in, and I’m glad that you all were able to sign on. We were having a little bit of a behind the scenes panic because, less than a half an hour ago, Zoom was having some outages across the board.
So I just want to give a disclaimer that if, for any reason, something happens, we will figure out what we can do to either reschedule or send a recording. But hopefully, everything from a technological standpoint continues to work.
So we’re going to get into the good stuff in just a minute, and I’ll welcome our panelists on in just a second. But today we’re going to be doing some really cool stuff with the eBird team. We’re going to be talking about an incredible feature, the status and trends maps, and some practical application of eBird data and status and trends maps around conservation efforts.
So as I mentioned, where we are broadcasting from is Ithaca, New York. That is where Cornell University is. And so I want to read a statement acknowledging the Indigenous people as the original inhabitants of this area.
Cornell University is located on the traditional homelands of the Gayogoho:no’, the Cayuga Nation. The Gayogoho:no’ are members of the Haudenosaunee Confederacy, an alliance of six sovereign nations with historic and contemporary presence on this land. The Confederacy precedes the establishment of Cornell University, New York State, and the United States of America.
We acknowledge the painful history of Gayogoho:no’ dispossession and honor the ongoing connection of Gayogoho:no’ people, past and present, to these lands and waters.
For those of you who aren’t familiar with the Lab of Ornithology, we are home to a community of researchers and supporters from all around the world who appreciate birds and biodiversity and the integral role that they play in our ecosystem.
And our mission is to advance leading edge research, education, and citizen science that help solve pressing conservation challenges. This work, including today’s webinar, is funded primarily by people like you who choose to become a member. So if you enjoy today’s webinar, I hope you’ll consider supporting the work that we do by visiting birds. cornell.edu.
And today’s exciting because this is the second event of our migration celebration, which is two weeks of virtual programming via Zoom and then one very large-scale family festival in Ithaca, New York. So I hope that you all will take some time to check out our website, that I will drop in the chat, which has a listing of all of the events that are coming up over the next couple of days. You can register for upcoming webinars and check out some really great resources as well.
My name is Lisa Kopp. I’m on the visitor center team at the Lab of Ornithology, and I get to facilitate today’s conversation and serve as the moderator between the panelists that will be joining in just a second and you all, the audience.
So last few things are a couple of tech things. So many of you, I know, are here via Zoom. And if you would like to have closed captioning, you can go to the bottom of your screen and hit the More button or the live transcript button and choose to show or hide subtitles.
We’re also going to be using the Q&A feature for today’s webinar. So the Q&A is where you should put all of your questions for our panelists. And I’ll do my best to look through those questions, make sure that the ones that are getting asked most often are getting answered.
The chat is actually only available for tech assistance. So if something’s going on with the sound or the video, that’s a really great way to let us know that we need to check our settings or for a little bit of help if we can take some time to offer it. For anyone joining on Facebook, hello. Welcome.
You also get to participate. We don’t want to leave you out. So you can use the comments section to put any questions that you’d like to share. And I have some wonderful behind the scenes help who will be relaying those questions to me and allowing you to participate as well.
We’ll be putting some important links in the Zoom comments or in the Facebook comments as well. But please, be sure that you’re not clicking on anything that isn’t from the Cornell Lab of Ornithology. So I want to make sure that no one gets any spam or bots, so only from our official Facebook page. Only click on those links.
So I think those are all of my official announcements. Tom and Orin, if you want to hop on and say hello. There’s Tom and Orin. Great. Welcome to you both. Thank you for joining today. We’re just going to start with some introductions. So Tom, do you mind going first? And then, Orin, tell us a little bit about who you are, what you do, what brought you to the Lab of Ornithology.
[Tom Auer] Yeah, so my name is Tom Auer. I’m the geospatial data science team lead with eBird in the Center for Avian Population Studies here at the Cornell Lab of Ornithology. I’ve been working here for almost seven years. I’ve been a birder for over 20 years now. eBirder for almost 20. So I’m excited to contribute and use the data and analyze it to have an impact.
[Lisa Kopp] Great. Orin?
[Orin Robinson] I’m Orin Robinson. I also work in the Center for Avian Population Studies. I work with Tom on the status and trends team. I also work with the Conservation Science group within CAPS here. And the overarching theme of my research is using eBird and the eBird products, the stats and trends products, to help on the ground conservation or management applications.
[Lisa Kopp] Great. OK, so let’s get started with some basics. So today we’re going to be talking about status and trends and using status and trends maps to make conservation decisions or using it for research purposes. But let’s start at eBird because I know we may have some people joining who aren’t familiar with eBird.
So Tom, would you mind sharing with us a little bit about what eBird is and how somebody might be able to use it?
[Tom Auer] Yeah, so eBird is a citizen science project. It’s been around since 2002. And the main goal is to capture information when people go birding. So people have been out birding in the field for hundreds of years. We want to capture that information about birds so that we can use it to help birds.
And so eBird is a great project. There’s a mobile apps for both iOS and Android. There’s a desktop app. And you go out in the field and you start a checklist. And then, when you go birding, you enter the birds that you see, how many of them you saw, and a little bit of information about your birding event. How many people were with? How long were you out?
And all that information then goes into our big database, which people then can use to find birds. We can also use for science. Lisa, you’re muted.
[Lisa Kopp] Thank you. I almost never mute myself on these. So do you mind doing a little demo of– I just put in the chat the eBird website– but to show people how they could get started.
[Tom Auer] Yeah, so if you go to eBird.org, this is where you land if you don’t already have an account you’re not signed in. And it gives you some great prompts to get started. You need a Cornell Lab account so you can fill out your basic info there.
And then, we also have a lot of great tools. There’s some videos you can watch if you learn more. And you can read more about the different features, how to find more birds, how to share your sightings with people. We have a great community. And then, one of the valuable tools that we offer for people is your ability to track your lists and keep them up to date. And then you can start to dig into status and trends and science information if you want to.
But I also wanted to just show what an eBird checklist looks like here for a minute. So one of my friends sent me this one from this fall. We have that basic info about when and where you were, so the date and time, the location, here in Minnesota, how long you were traveling, how many people you were out with, and then counts and some notes about the birds.
You don’t have to enter notes. You can just enter counts if you want. But it’s a complete list of all the species that we’re seeing that morning. And that completeness is really helpful to us. So I hope everybody will dig in and check out eBird if you’re not familiar with it.
[Lisa Kopp] Yes, and we’re actually going to be hosting a whole additional webinar with a couple of other folks on the eBird team to just answer your questions on how to use eBird. So I will put a link in the chat with that registration link so you can sign up. That is on November 9. And it really is going to be an open Q&A. So you can submit questions as you register, or you can submit them just as you would in a live talk. And so we’ll be able to help with all of your questions about eBird with that.
So one of the things-I love talking about eBird because it’s one of those tools that is so helpful for individuals, right? You get to track your birds. You to get your life list really easily accessible. And I didn’t grow up as an avid birder. I’m much more of a naturalist.
But one of the things that I loved learning about was this competitiveness that exists among really great birders. And so eBird also gives you the ability to check where you are in terms of other people’s lists in your area or what other people are seeing.
But big picture, eBird drives a ton of science, right? There’s so much that eBird data does for the scientific community and, Tom, that’s my segue way for you to be able to talk a little bit about status and trends and share with us.
Actually, I’m going to backtrack really quick. I want to talk about the complete checklist first because you mentioned that in passing, Tom. And I want to give Orin a second to share with us what that phrase means because it’s a really important one when talking about eBird.
And again, as someone who didn’t come into eBird as a really skilled birder, I was worried about whether what I was going to be submitting or what I was going to be seeing was going to be right. Or if I hadn’t been able to identify everything, was that OK? So Orin, do you want to talk a little bit about what a complete checklist is and how that becomes part of the eBird database?
[Orin Robinson] Yeah, so complete checklists are really important in terms of the science we do. They essentially make your observations usable by us in our status and trends project here. So what a complete checklist is, it’s not just the birds you are able to identify by sight or sound, but so is the information on where you were, how long you are out there, how many people were with you, how far you traveled, things like that.
And the really important part of that is eBird will ask you when you submit your checklist is this a complete checklist of all the species you are able to identify? And if you say yes, this is every species I was able to identify, then we can also infer what species you did not detect. Doesn’t necessarily mean that they weren’t there. It means you did not detect them.
And in terms of the [AUDIO OUT]
[Lisa Kopp] Uh-oh, ope.
[Tom Auer] He’s back.
[Orin Robinson] All right, where was I?
[Lisa Kopp] Oh, you’re getting a little choppy, Orin.
[Orin Robinson] Yeah, so I guess what I was saying is the complete checklist lets us infer what is not there. And with the types of modeling that we do, it’s just as important to know what’s not there as it is to know what is. And that’s why those complete checklists are very important.
[Lisa Kopp] Great. So what happens if I swear I see– I’m going to go totally off the rails and say I see a flamingo in my yard in upstate New York. How does that– can I mess with the data that you are using for real science? How does that get vetted? Or how does that kind of thing get flagged?
[Orin Robinson] Yeah, so if you entered that into your cell phone, right– and you are able to do that a box will pop up that basically asks are you sure about that. It’ll tell you that this is a species that is not seen in your area and ask you to describe it, how you came to that observer– or how you came to that identification, how you know it’s a flamingo, things like that.
And if you do all of that, you can submit it, but an actual human will then– a reviewer– there are thousands and thousands of reviewers for eBird one of them will contact you and ask for pictures, the location so they may be able to go find the bird, things like that.
So there are many, many automated filters and then human filters that these observations have to go through before that flamingo in Ithaca can go through.
[Lisa Kopp] That’s very cool, another example of how much those thousands of eBird reviewers, how much community plays a role in all of this. The data is submitted by people, and then it’s being reviewed by people. But also, it speaks to how you really do need to have that second human layer because sometimes there are rare birds that show up in places. And then, pardon the pun, but like birders flock to see them and get to experience those rare birds that do show up.
So great. Thank you for explaining that. So Tom, now I’m going to get back to where I wanted to jump to earlier, which is how is all of this data used? If people are submitting their checklists, their complete checklists, and all of that’s going into this database– which is gigantic– how is that getting used?
[Tom Auer] Yeah, so if you’ve been to the eBird website and you go to the Explore page, you’ll see lots of ways to dig in and look at the data. So here we can see the raw data for a bald eagle right throughout the year. And one thing everybody might notice is that there are gaps, especially if you look at Alaska or high Arctic Canada.
eBirders can’t go everywhere, and not every eBirder is the same. So what we do with status and trends is to dig in and deal with filling those gaps and smoothing out that variation between people so that we can understand what the true– close to true, if we can’t get perfect truth, but we want to try to estimate how many birds there are across the entire landscape throughout the full year.
And so that’s what the real goal of the project is. And so some of you might be familiar with some of these animated maps. Here, I’m going to go ahead and click Play. And we can watch here a broad winged hawk as it completes its migration North in the spring breeding season throughout the Eastern US and Canada and then returns South.
And so, at each one of these pixels– I’m going to let it play a little bit further we can zoom in and get this relative abundance number, how many birds you would expect to see if you went out on a checklist on April 26. And this ordering of relative abundance is extremely valuable to science. And that’s the key piece that we’re looking to make with the eBird status and trends.
[Lisa Kopp] That’s amazing. And I can imagine that this is very helpful for people who are looking to understand migration. And so part of why we wanted to talk about this during migration season is to help understand what’s happening during these times of year where there’s a lot of movement, and then also how that might help people plan when and where they’d like to go look for birds. Would you mind talking a little bit more about that?
[Tom Auer] Yeah, so this tool certainly works for that. What I’m showing here is a new website for status and trends that finally puts that information in an interactive form. So if you’ve been to our website before, they’re still available. But maybe you’ve seen these static maps that you can download and play as animations.
But the exciting part is we now have these fully interactive maps. So if you’re trying to plan a trip somewhere, and maybe you want to see where the most Broad-winged Hawks are during this week, you might be able to zoom in and see, not surprisingly, that Pennsylvania is a big Avenue for Broad-winged Hawks right? That’s where Hawk Mountain is.
But maybe there’s a place near you, if you’re in Pittsburgh, you may not have as good a chance of seeing a Broad-winged Hawk fly over. But if you go a little bit east, you might find a ridge where, on the right conditions, right days, you can get Broad-winged Hawks moving through your neighborhood.
So it’s really important information for science, but it also is fun, especially with this new interactive form, to dig in and explore what’s happening with these species at the given week of the year in your neighborhood.
[Lisa Kopp] We have a good question. Brock is asking, I tend to keep paper logs of my observations. Is there some value to entering them even if they’re a few months after the fact? And why I wanted to ask that now is because I think it’s interesting to think about what the data is that you’re pulling from, how many years back is it looking at? Does it pull from current, or can you talk a little bit about that?
[Tom Auer] Yeah, so the data that goes into eBird status and trends, right now, it’s from a period of what you see on the website of 2006 through 2020. So we’re always trying to estimate through the last year. We do this in an annual cycle. So the last 15 years of data is most important because that’s where we have the most data in eBird.
So yeah, If you’re out there writing paper logs, as long as you’re keeping track of how long you were out and where you were out, you can go to the web portal and enter that data afterwards for sure.
There’s also some-if you check it out on our frequently asked questions in status and trends, there’s a good list of the things you should do if you want to contribute data, specifically in best practice for use in this scientific way.
[Lisa Kopp] Great. I also just added into the chat, someone mentioned in the Q&A that there’s a really great free course on Bird Academy, which is also a nice way to be able to go and learn about how to use eBird if you’re, again, trying to get started, which includes information on using these tools and products, not just on the beginning of using eBird.
Great. So there’s another-oh, I was going to ask. There’s a very techie question in here that, actually, I don’t even understand, but I want to make sure to acknowledge. Daniel’s asking how are you aggregating the data into pixels on the map? Do you want to talk a little bit about the incredible work you do from the technical perspective?
[Tom Auer] Yeah, so that’s a good question because it really highlights the amazing spatial detail, the resolution we have. So here, I think we’re seeing at about nine kilometer resolution. But if we zoom in– let the map load. There we go. You can see that we’re actually making these estimates predicting how many birds are at every pixel that’s three kilometers by three kilometers on the side, so nine square kilometers.
And that resolution comes from the NASA remote sensing data. So we use a lot of satellite imagery data from NASA to fill in those gaps, right? So we take the eBird checklists, and we use statistical machine learning methods to combine those with where the birds are on the ground with that information about the landscape and then predict all those locations.
So each square that you see when you zoom all the way in is a three kilometer by three kilometer pixel. And that’s where we’re making that estimate of how many birds are around. So here, on average, you would expect to go out, we’ve got an estimate of half a bird. That means they’re not super common. But if you intersected with the right habitat during that week, you have about a 50% chance of getting a [INAUDIBLE].
[Lisa Kopp] We have a question from Devin who has asked, how do you get to the status and trends from the Species Explorer? I’m trying to follow along, but my map looks totally different.
[Tom Auer] Yeah, so that’s a good question. There’s two parts to the website, the Explore page-and we try to keep them separate for a reason. If you go to Explore, everything you see there is about the raw eBird data, right? So this is the information that comes from– I’ll just jump to it here.
If you submit a checklist, within a few hours your observation is going to be on the map. And this is all of that data from every time period and whether it’s complete or not. So this is all the data, and there’s a lot for bald eagles, so it’s taking a moment. I’ll just zoom in here where you can see all the points. There we go.
So yeah, all these red locations are ones that somebody have submitted in the last 30 days. These are the raw data. If you’re up at the top header here and you go to Science, you will find the status and trends here as the left item. And then that lets you search for species. And you can also explore all of the species that we have too, as well. So if you want to learn about snow goose, you can jump in and you’ll get that information where you can zoom in and interact with information about snow goose.
So we try to keep them separate because they’re different things. One’s an input. One’s an output. One’s raw and has a lot of noise from people. And one we’re trying to smooth that out and make it useful for science.
[Lisa Kopp] Great. Jon asks, how can relative abundance be used for planning where to go to try to see a specific species?
[Tom Auer] Sure, well, I’ll give an example because I’m going down to Florida next week. I’m flying into Jacksonville, and I want to see a Bachman’s sparrow because it’s been a long time. And I’m going to fly into the airport, and I want to go over and check out this area just south of the Okefenokee.
And so I can see here that there are some really high abundance pixels compared to these other ones. The darker the purple, the more likely the species is. And so there are some areas over here that, some of them aren’t too far from Jacksonville or the airport. I can know that I have a pretty good chance of going over there if I get in the right habitat and pay attention, Bachman’s sparrows are pretty quiet this time of year, so I might have to do some fishing. But it gives me a good idea of where the most birds are.
One thing to keep in mind with relative abundance is that it’s a combination of how frequently species occur, right, whether you see one at all or not and how many you would hope you would see. So Bachman’s sparrows, I would go out, and I might see one or two. But pelicans I might go out and see 100s. So this relative abundance number, if it’s above one, that means I got a pretty good shot, I think, I’ve seeing this bird.
[Lisa Kopp] Great. Do you want to shift gears and talk a little bit about trends?
[Tom Auer] Yeah, so let me switch over my screen share here. If I can get zoomed out. It’s always hard to get a PowerPoint started. Let’s see now. Here we go.
So I’m really excited to announce that this fall we’re going to release a lot of information about trends. And I think people have been waiting for this for a while. All that information that I showed about status and trends, relative abundance was a single year. It was an estimate for 2020. This year, we’re going to put out information about 2021.
But trends is that change in relative abundance through time. So it’s incorporating that information about occurrence and count. And it’s basically that this keeping tabs on how the species is doing. And what’s really, really exciting for us about this, aside from the fact that we’re doing it with eBird data, which very few people have ever done, is that we’re doing it at a really relatively fine resolution.
So what you see here on this map, every dot is a pixel that’s 27 kilometers on the side. It’s a little bit coarser. This is expensive to calculate. But within that, we’re estimating what that rate of change is, the percent change per year. And so, on this map, if you see a red, that means a decline. The darker the red is the stronger the decline is. If you see a blue, that’s an increase. And the darker the blue the stronger the increase.
The circles on this map are sized according to how many birds are there at the end of the year, so 2020 in this case, so the places where the species– in this case, wood thrush is really abundant. You see big circles.
The white circles can get a little confusing, but they basically mean that there’s either no trend or we don’t know what the trend is. We’ll tweak it just before we release it, but in statistical world that’s a little bit hard to suss out. But this is really exciting for us because this is a great opportunity to have eBirders realize how their sightings can be used to monitor and take care of the world’s birds.
So this fall, look for some time into November to see, hopefully, at least 500 of these maps for different species go on to the eBird status and trends website, along with an update of all the species that are already there.
[Lisa Kopp] That’s a big release. 500 is no small number.
[Tom Auer] No, we’re going to try to do our best with North America, trying to get as many North American species as we can.
[Lisa Kopp] Tom, I don’t have this in our script, but I’m curious if you would mind sharing a little bit about your background. Are you a statistician? Are you a computer scientist? It feels obviously, you’re a birder. So what’s the magic mix that a lot– this is– my background is in education. I could never wrap my head around how you create something like this. Would you mind just giving a little bit of background on how you are able to do this?
[Tom Auer] Yeah, sure. So I got to give a lot of credit to our whole team about this. I’m a bit of a hybrid person. I’m a half biologist, half geographer. My geography interests have been very information science focused, so familiar with stats and machine learning and geographic information science, GIS, which helps me work with our team to get these maps in these forms and understand the methods that are behind there. But Daniel Fink, who has been leading this trends effort, is just a brilliant statistician and machine learning expert.
He’s got a PhD, and he’s been working at the lab for over 15 years now, really focused on these kind of questions. And so to produce a product like this takes a lot of immersion in data science and statistics and understanding what’s involved when you’re trying to take noisy data and make reliable estimates out of those. And Orin’s familiar with a lot of those methods too.
So we’re always, as a team, trying to hone and refine and dig deeper to get the best methods we can to make the best estimates here we can.
[Lisa Kopp] Great. Thank you, and there are a lot of really good questions coming in, which we’re going to have time to answer more of, but I want to make sure that I give Orin plenty of time to talk about his work too. So thank you, Tom, for all that. We’re going to come back with quite a few more great questions in the Q&A.
So Orin, you are using this data in a different way from the researcher, the practical application side of things. Could you speak a little bit about that perspective, why these maps matter? Obviously, they’re fascinating to look at and to understand. But how are how are you using them for your work?
[Orin Robinson] Yeah, so these maps are awesome to look at. It it’s fun to be a part of getting them to that point. But once they are at that point, I could seriously spend hours just looking at the different species.
But the really important thing for conservation or management planning is the data that’s behind those maps. As Tom showed you the animated map of the Broad-winged Hawk moving, in each jump you saw was a week. So we know, based on this data for every week of the year, what those Broad-winged Hawks are doing. And we know that for thousands of species now from the status and trends project.
So that’s kind of the Holy Grail of dynamic conservation, right, knowing when and where to provide on-the-ground conservation or management action for migrating species, right, because they’re not there all the time. You need to match up your limited resources with when you’re going to get the most bang for your buck, so to speak. And these maps really go a long way in helping us do that.
[Lisa Kopp] That’s great. Do you have any specific examples of either projects or species that have benefited from using these maps?
[Orin Robinson] Yeah, definitely. So one that comes to mind is in the central valley for shorebirds. You may have heard of the bird returns project where the rice farmer [AUDIO OUT]
[Lisa Kopp] Oh, no. Orin just froze up again. Let’s give them a second, see if we can get him to get back. If not, we’ll jump to some of these questions in the interim. Hopefully, he’ll sign off and sign back on. I’m going to blame Zoom’s earlier issues for this kind of thing right now.
OK, so Tom, are you OK to jump back into some of these Q&A features until we get Orin back to talk more about the bird returns project?
[Tom Auer] Yeah, totally. I’m scrolling through. There’s some good ones.
[Lisa Kopp] There are some. Oh, well, and Orin’s back. Hi, Orin. Welcome back.
[Orin Robinson] Yeah, hopefully, I stay for a little while this time.
[Lisa Kopp] OK.
[Orin Robinson] So I was talking about the bird returns project where rice farmers in California and the central valley are paid, essentially, to create temporary wetlands for when these birds are passing through. And not all shorebirds migrate at exactly the same time, so they can use these status and trend information to pinpoint the peaks of migration for species of concern or see when the most species will get the benefit.
What we’ve seen in that is that some species have definite peaks. They’re in and out very quickly. And some stick around for quite a while during the season. So knowing things like that can help them understand where to provide those temporary wetlands, for how long, when to put them there, and really help them spend a lot less money on management.
Also, doing work right now [AUDIO OUT]
[Lisa Kopp] Oh, no.
[Orin Robinson] Did I–
[Lisa Kopp] You’re still there. We heard doing work right now.
[Orin Robinson] -with waterfowl, and now I’m getting “your internet is unstable” message. Yeah. Shocker. Yes, so we’re doing similar work with various groups that manage waterfowl, different joint ventures. Ducks Unlimited is part of this work. So those are a couple of exciting uses of the status products.
[Lisa Kopp] Great. I already have seen questions about how these kinds of products could be used to understand climate change or how birds are being affected by, I think, calamities. I heard someone someone use that word. Do you have any ways to– or any examples of how these kinds of maps are being used to understand those changes or track them or respond to issues because of climate change?
[Orin Robinson] So in terms of climate change, I think our trends information would be better equipped to handle that versus the status because that’s going to be things you see that happen, potentially, gradually over time. And our trends information is really going to capture exactly that, the trends in the bird populations that could be related to effects from climate change.
We recently did a study that looked at the– it was couple of years ago the very cold anomaly that happened in late January, February. And we combine raw eBird data with some GPS tracking data from mallards. And we’re able to show that we could actually use eBird data to track waterfowl move– [AUDIO OUT]
[Lisa Kopp] Oh, all right.
[Orin Robinson] Am I still missing?
[Lisa Kopp] There we go. We heard tracking waterfowl, and then you froze again.
[Orin Robinson] Sorry, so using eBird to track waterfowl during this very short window of time was roughly right in line with waterfowl that had GPS trackers on them. So we were able to see across multiple species of waterfowl and the different flyways how they reacted to this incredibly unseasonably cold period. So we can use eBird to track things like that.
[Lisa Kopp] That’s great. Orin, do you know of-we’ve got a question in the Q&A from Niren asking, are organizations like American Bird Conservancy, ABC, using status and trends data to direct funding and identification of conservation projects?
[Orin Robinson] I know that various state agencies and some joint ventures are doing that. We know specifically about the American Bird Conservancy. But yes, that absolutely is happening, and we’re trying to help out as best we can.
[Lisa Kopp] Actually, Marcia just made– or maybe it’s Marcia– in the chat suggested, maybe if you turned your video off, we would still have your voice, but it might save your internet connection a little bit.
[Orin Robinson] Yeah, we could do that.
[Lisa Kopp] –again, we could try.
[Orin Robinson] Yeah.
[Lisa Kopp] Am I right in remembering that the bird returns project is with The Nature Conservancy? Is that–
[Orin Robinson] Yes.
[Lisa Kopp] Yeah, OK, so that’s another–
[Orin Robinson] Yeah.
[Lisa Kopp] Yeah, OK, great. Wonderful. So we have a lot of questions here in the Q&A that I want to make sure we’ve got some time to answer. So let’s see, where do we want to start? There are so many good ones.
So we had a couple of questions that I see Tom has already flagged as wanting to answer live. What are some of the shortcomings of eBird data that you hope you can overcome. And I’d love to hear from both of you on this because, Orin, from the practical side, I’m sure that if you could wave a magic wand, there are some things you’d love to be able to have right away. But Tom, do you want to start just from the data, the tech side?
[Tom Auer] Yeah, I do think about shortcomings a fair amount. Gosh, the one that just keeps popping into my head is just encouraging people to make an effort to count birds. A lot of birding, traditionally, you went out and just detected a bird. I saw a black and white warbler today.
And so it’s a bit of a shift to ask people to count, and I think we’re still in a bit of a transition like with eBird. There’s some great resources on the website about counting. But from my analytical perspective, I can still see that there’s a big chunk of eBirders who aren’t making the strongest effort to count. And so, yeah, that’s one place I’d love to see some effort put in.
[Lisa Kopp] Orin, any thoughts for you, either about shortcomings or wishes for the future of eBird data or products?
[Orin Robinson] Yeah, one of the things that I would wish, or maybe selfishly advise people to be cognizant of, is making shorter, more frequent checklists during their trips. That helps us with location accuracy in these models. It also helps us with some of the noise issues. If all of your checklists are roughly an hour or you traveled roughly a kilometer and a half on all your checklists and then you start a new one, that’s one of the things that also helps quiet the noise in that data.
[Lisa Kopp] That’s really interesting. Yeah, somebody had a question earlier about– and Tom might have typed in an answer– about going on longer bike rides, like if they’re going for a 20-mile bike ride, but they’re birding during that time, is that helpful data? And it sounds like– you guys can correct me if I’m wrong.
It sounds like data is data, and if it’s a complete checklist and they’re submitting it, that’s helpful. But maybe the shorter, smaller checklists would be helpful from what you were just saying, Orin? Tom, anything that you want to add to that, or is it–
[Tom Auer] Yeah, it’s just important to keep what you’ve seen with the particular configuration of habitat you’re in. If you travel long distance, that configuration changes, and it’s hard to know why the birds are where they are.
[Lisa Kopp] That makes a lot of sense. Another great question in the Q&A is how do you take into account reports with just an indication of presence, like the x, as opposed to what you were just talking about with counts, right?
[Tom Auer] Yeah, so x are part of the challenging part of the data. We really love people to try to count, even if it’s just an order of magnitude, 1, 10, 100. We are actively working on some research to use those x’s to tell us whether species was there or not. And then we would throw it out for the next phase, which is to understand how many birds were there. So trying to pull more of that in because it definitely does help with that first part of was a species there or not? That helps. So we’re trying to work on that. That’s just a little bit tricky.
[Lisa Kopp] Jackie had a really great question. If there’s a large increase in eBirders, can it bias the data to make it appear that there’s a greater abundance of a bird species just because there are more people out there reporting that bird?
[Tom Auer] Yeah, it’s a really good question about our trends analysis. And it goes into a lot of detail. It’s really hard to explain, but the method that’s being used, that Daniel’s worked on, is called machine learning. The first part is to model out that issue. So he’s feeding information to the model about that change in the number of sightings, and then that can go into the estimate of the trend so that we’re not seeing that.
And in a great example that I always use– I don’t have the map readily available, but if you look at the map, and it’s sad for great blue heron there’s a lot of decline. And great blue heron is one of the first, and probably there’s a bias to how much report people make about great blue herons, beginning birders and people who are new want to go out and record their great blue heron.
So if there was an issue with that, we would expect to see more and more great blue herons being reported, and we would expect to see an increase in the trend. But we’re actually seeing the opposite. So I feel pretty good about how we’re doing that.
[Lisa Kopp] That’s interesting. Deborah asked, you both used the word noise when you’re talking about data. Could you, maybe Orin, you could talk a bit about what you mean by noise when you’re talking about data.
[Orin Robinson] Yeah, so that’s just the variation in the things that impact what you saw on your checklist. So for example, if I have two wood ducks on a checklist and Lisa has two wood ducks on a checklist, she may have only been out for 15 minutes and been stationary. Whereas, if I was out for three hours and walked five miles to see my two wood ducks, those two checklists aren’t the same even though there are two wood ducks on them.
So statistically, we have to determine how those types of things impact the observation of species. Does that help?
[Lisa Kopp] Yes, David asked the follow up question, which is, so if I’m birding with a group, is it better for us to submit one checklist mentioning the number of birders in the group, or for each of us to submit on our own? Tom, do you want to–
[Tom Auer] So ideally, you’re sharing checklists with people, but there’s no worry about duplication. The way our modeling methods work is that we’re looking at the average count. We’re not summing up all the numbers. And we’re doing that a lot. We’re sampling and repeating, and sampling and repeating. So we’re grabbing random checklists.
And so it actually sometimes can help if you spread it out a little bit and we know that this person saw a few less species or this person saw a few more species. That’s helpful information sometimes. So no need to worry about duplication.
[Lisa Kopp] Great. We’re getting questions, both on Facebook and in the Q&A, about the interaction between Merlin and eBird. And we’re going to put a pin in that and have you all save those questions for our eBird Q&A webinar in November because answering those questions actually opens a whole can of worms about how you can connect things in your own account. And so I want to make sure that we’re not going down that rabbit hole. But I didn’t want to ignore all of you who are curious about that content.
So Fabian asks, is there any data about trends available for areas outside the US, or plans for that?
[Tom Auer] Yeah, so we’ve tested that in the past year, and I’m excited to say that, this fall when we release our maps, we’re going to have some from outside North America. So we’ve tested some and seen reliable results from South America, especially Argentina, Chile, Central America. Costa Rica has a lot of data. And then we’ll see about India and New Zealand. Those are both places where we have a lot of eBird data, and we’ve seen some good results. So I’m really hoping to get a good handful of each of those regions up on the website this fall.
[Lisa Kopp] That’s very cool. We have a question about privacy. So if you’re submitting data from your backyard, are there ways to protect people’s information, or is the data anonymized, or any tips on that?
[Tom Auer] Yeah, so we don’t track people. We track birds. It is helpful to know who has seen a bird and when and where, but we don’t flip it around and look at all the places you’ve been and where you are. We don’t do that. And if anything like that happens, it’s a machine seeing it and doing an analysis and not a human looking at it.
All the data, the raw eBird data that gets output from eBird in our data deliveries does not have your name in it. It just has a user ID. So your name is not in all these analysis tools. And so there isn’t really a way on the website either to see all the places you’ve been. So it would take somebody a lot of work to try to track somebody, so to speak. And we don’t really track people. It’s not something that we need to do.
[Lisa Kopp] Yeah, OK, great. Thank you. All right, there’s a great list of all sorts of questions that you’ve marked that you want to answer, Tom. Do you want to start at the top of the list? How do you test the reliability of predictions?
[Tom Auer] Yeah, so with the trends, it’s really interesting. We actually it’s kind of a cool method, and I’ll talk about that one specifically is we actually simulate eBird data. So we have computer simulations of eBird data where we take data that looks like regular eBird data, and then we add a trend to it. So we may take 20 years of random eBird data, simulate similar looking eBird data, and then add, say, a 3% per year trend to that.
And so, then when we take that simulated data with a trend added to it and run it through our model, we really hope that we recover that signal. Can we see that 3% per year increase? We do that in a bunch of different ways. There’s tons of different permutations of that, increases, decreases, strong, weak.
And that lets us say, OK, for the species, we simulated a bunch of data. And can we see those added trends coming through? And if we do, then we know that the estimates themselves are reliable for that model. So we have a really good handle with the trends data on whether the estimates are good or not.
[Lisa Kopp] That’s so cool. Orin, I see you turned your video off, I think, in an effort to save your internet bandwidth. But I do have a question here, if you’re able to still hear. Do you have any recommendations for people looking to learn more about some of those conservation stories that you shared?
[Orin Robinson] Yeah, so a lot of, as those happen, they will be published in scientific journals. But also I know the lab has, in the past, through various media outlets, promoted those stories, like the tricolored blackbird a few years ago. And either last year or the year before, bald eagles, when it all came out that the numbers had exploded and that. So that was done using status and trends data.
So the lab seems to stay on top of putting those things out when they are of great interest. And just to be brutally honest with you guys, a lot of what I do with those groups is the boring stuff, the number crunching, the model building, things like that. So that’s not really stuff you’d want to read about anyway, at least most folks.
But when everything is complete and conservation or management recommendations are made, the lab usually– [AUDIO OUT]
[Lisa Kopp] Yeah, and Orin, I don’t know if you will have the answers to this, but someone asks if there’s any information about where people could volunteer or if there’s ways to contribute to this work. I would say a quick answer is, if you’re submitting data, you are contributing to the work. The data is built on people submitting their checklists. But do you have any suggestions for anyone looking to get involved in maybe their local areas?
[Orin Robinson] I really don’t other than what you were saying. If you are contributing eBird checklists, you are contributing to this data and to the status and trends. And down the line– [AUDIO OUT]
[Lisa Kopp] I think, even with video off, we just lost Orin a little bit.
[Orin Robinson] I would just say keep submitting checklists.
[Lisa Kopp] Great, thank you. OK, Tom, somebody has asked a question about the wood thrush population. The abundance between 2007 to 2020, it looks like there’s a geographical shift. What could be the reason for that?
[Tom Auer] Yeah, so it does look kind of like there’s an east, west shift going on here with wood thrush, and wood thrush is one of the most well studied species in the eastern forest ecosystem. So fortunately, we have some other evidence to tell what’s going on.
There’s been a fair amount of work in the Appalachian Mountains here, where you see a lot of increase related to forest management and practices. Wood thrush also like a transitional forest. So if there’s been logging that’s starting to transition to young forests, that’s going to be beneficial for wood thrush.
Further east, you’ll really see a lot of decline in the urban corridor. It’s very likely that that’s related to habitat degradation, loss of habitat as that area continues the urban intensification in that area continues to happen.
Overall though, one thing we can’t really answer here is that we also know that all these different populations have different wintering grounds. So species in Maryland or Pennsylvania are going to go to a different places than those in Wisconsin and Minnesota. And the ones that are in Mississippi probably go to a different part of Central America.
So that one we can’t quite untangle. It’s very possible that those eastern birds are going to a part of Central America that is undergoing aridification, that’s starting to experience drought, and the birds aren’t doing as well on the winter. Can’t answer that one conclusively, but it’s something to investigate.
[Lisa Kopp] Great. One of the related question, Peter asks, does climate change seem to be impacting the ability to predict species locations and density? Which I think is pretty much speaking to is it going to be harder and harder to do these trends maps over time? Or is the data going to become wonkier, for lack of scientific word?
[Tom Auer] Yeah, no, wonky is the right word there. I use that word all the time.
[LAUGHTER]
I don’t think climate change is going to make it harder to make these trends. I think we do see a fair amount of climate change signal. We can see elevational signal in the west. We can see some latitudinal signal in the east. I don’t think it’s going make it harder to make the trends.
I think it will make it harder to untangle the true reason that a species is declining. Is it because they can’t find enough food on the wintering ground? Is it because the temperatures are too high during the breeding season? I think that’s going to be challenging.
[Lisa Kopp] Yeah, I see here Andrew asked the question of can your data show certain signals like fire or smoke events that are affecting birds in status and trends or migration patterns? We talked a little bit about this with Kevin on Monday during our Ask an Ornithologist webinar. But do you have any answers to anything like that?
[Tom Auer] Yeah, and maybe Orin can speak to it a little more. We have a project coming up this fall, post-doc Andrew Stillman is working specifically on the fire question and how it relates to the management of fire-dependent species. And we’re hoping to start to incorporate that information about fires, fire severity, extent, duration. We’re going to start putting those variables from remotely sensed data into the models and trying to understand those relationships.
It’s harder, much harder to do during the migration periods. We don’t have great data about airborne smoke or things like that. But on the ground fire is an exciting topic that our team is digging into.
[Lisa Kopp] That’s great. I feel like, so often with these weather events or climate– things that we know are being caused by climate change broadly, we still just don’t have a lot of answers. And so it’s great to hear that there’s work being done to try to get some of those answers made or find out some of those things. Another great question is, do these mapping software techniques display declining species? So that was a question I had with the trends map that you were showing, will there be an overall– if there’s–what’s the balance of the blue versus the red?
[Tom Auer] Right, yeah, so maybe I’ll pull up the great blue heron map just since I brought it up. It’s a bummer, but yes. You know what I will say, the great blue heron is not doing well. We’re seeing a dramatic decline across the entire range. So I don’t see any blue on this map. There might be some in the intermountain west, which is an area we are seeing increases in, I think, likely due to changes in moisture patterns throughout the west.
But I will say we analyzed a lot of these species. And almost all of them have some area of increase. It’s not just a blanket bad story. But yes, our model has captured both increasing and decreasing species. And the great part about this method is that it does it through space.
So if they’re increasing in the intermountain west, we can see that. We got that detail and resolution. If we see that they’re declining in the Mississippi Valley, we’ve got that information. Hopefully, we can put this into the hands of people who can then act in the right places.
[Lisa Kopp] Yeah, hopefully. I think sometimes it’s really hard to hear all of the stories of declining species and all that needs to be done from a conservation perspective. But like you just said, there are lots of people who are doing that work, and the tools that you’re helping to create are really just going to put so much more information at their fingertips.
All right, we have three minutes left. So I want to see if there’s anything that you really want to answer, Tom, from this list of questions here. And Orin, if there’s any last minute thoughts you want to share before we get into our final announcements?
[Tom Auer] Maybe I’ll just answer broadly a few different questions that are about species that aren’t on the website. So status and trends requires a lot of computation, and not every species has enough data for us to do it. So we’re working through the most abundant ones. We’re hoping to put 2,500 species of status information up by the end of the year, we hope. So we’re growing every year on that front.
I’m also seeing some questions related to eBird growth overall. We are getting more and more checklists every year. We’re growing 20% globally every year, so it’s amazing how much more data we have. And specifically, in a place like Columbia, I didn’t specifically say that it wasn’t a place we couldn’t do trends.
I’ve actually seen some reliable trends from Colombia. The tricky part is there is that many species in Colombia also have ranges further east in Venezuela or in Brazil, which have lower data densities. And that drives down the reliability of our models. So maybe some of the Colombian endemics, we’ll throw a couple in and see what happens.
[Lisa Kopp] It’s great. Orin, any parting thoughts you want to share before I get into our announcements?
[Orin Robinson] Yeah, one question that I get asked a lot and that I did see in the Q&A was about hiding information for sensitive species. We do. There is a list of sensitive species that their locations are hidden. So when you go to those purple maps, like Tom showed, from the raw eBird data and you start zooming in, you’ll get to a point where those locations disappear. So that is a thing that is absolutely done in eBird to help protect those more sensitive species.
[Lisa Kopp] Thanks for answering that. That’s helpful and really important.
Well, Tom and Orin, thank you both so much. As is always the case when I do these webinars with the eBird team, my mind is blown, and I’m so grateful for you sharing this amazing work that you’re doing. Thank you to the audience for all your great questions and for being patient with us as we dealt with tech issues and jumped back and forth between the Q&A and the presentations.
So for those of you who signed up over Zoom, we are going to be sending out a recording of this video in the next couple of weeks. We’re going to wait till migration celebration is done, and we’ll send one massive email with all the recordings. But you can always check the Bird Academy’s website where all of our recorded lectures are.
There’s actually one last year from Tom and Orin which is interesting because there’s lots that’s been happening in the past year, lots of great information about the upcoming webinars. And as I said, we’ve got a great one coming up with other members of eBird who you can ask all sorts of questions about when it comes to your specific checklists and using it with Merlin and things like that.
So I hope to see you all at our next webinar, either this week, next week, or next month. And thank you all again for joining, and thanks, Tom and Orin, for sharing so much great information with us all. Have a great day, everybody. Bye.
End of transcripteBird’s massive citizen-science database allows both scientists and birdwatchers to better understand where birds go during their migrations. Watch the recorded webinar and learn the method behind eBird’s magic, explore eBird’s breathtakingly detailed Status and Trends maps, and find out how to use these tools to know when a particular species is most likely to show up near you this migration season.
- Get started with eBird by taking Bird Academy’s free self-paced eBird Essentials Course.
- Learn more about your favorite species locations and abundance with the eBird Status and Trends maps.
- Read about how eBird data is helping conservation action in California’s Central Valley.
This webinar is part of our annual Migration Celebration. Join us for two weeks of online events, family-friendly programs, and ideas and resources for your own migration activities. Visit the Migration Celebration website for the full schedule of events and recorded webinars.