Digital Transformation in Birding and Ornithology: An Ethnographic Look at eBird

Everett Kehew

Abstract

Citizen science has long been the driver of ornithology, but the Cornell Lab of Ornithology has created a technology that has accelerated birding to scales previously unheard of.  eBird is a platform for birders to upload their bird sightings to the Cornell Lab’s database which provides a wealth of publicly accessible maps and datasets for use both in casual birdwatching and in organized conservation.  Its benefits are many, but its networked nature and incredible size has created unintended consequences that have gone completely unexplored.  This paper aims to bring some of these issues to light and to investigate how eBird affects birders and conservationists.

Introduction

Data-sharing between citizen scientists and major scientific bodies is not a new phenomenon, especially in the field of ornithology.  But when digital technologies enter the picture – in what Italian theorist Bifo Berardi calls a “digital transformation” (Berardi 88) – relationships between producers and those they produce for take on new and uncertain forms.  This is exactly what has happened since 2005 with the advent of eBird, the Cornell Lab of Ornithology’s crowdsourced database of bird sightings.  Having arrived on the cusp of the proliferation of smartphones, eBird is a highly mobile method of birding.  Birders can share what they see and in turn gain access to the logs of every other user, compiled by eBird into easily navigable maps and datasets.  While digital transformations have been studied in other contexts, little research has been done on the effect networked databases have on the field sciences.  eBird opens up a wealth of questions concerning how human relationships with nonhuman ecologies transform as they become digitally networked and offers an opportunity to interrogate the role that digital networks have in shaping conservation efforts.

Before embarking on this project, I had no idea what eBird was.  I stumbled across it by pure chance and immediately saw that it was worthy of deeper inquiry.  Online fieldwork began immediately as I clicked and scrolled through the layers and layers of maps and data.  To find out firsthand how birders used the technology, I contacted several of the top-ranked eBirders in the Ottawa region.  I was able to conduct phone interviews with two of them which I recorded digitally and transcribed.  The first interviewee, James, is one of the checklist validators for the Ottawa region and a professional entomologist who is very familiar with field science.  The second, Martha, is the top-ranked female eBirder in Ottawa and one of the top in Ontario.  The fieldwork, the interviews, and a short literature review constituted the entirety of my research, but the project is supplemented by a lifetime of living with my dad, an avid birder, who has taken me birding countless times; I draw from these experiences when I make generalized claims about what birding and the birder-world is like.  This paper will begin by looking at how eBirders create data, and then move on to examinie how people use eBird’s data.  It will finish by analyzing some of the issues that arise from eBird’s structure and scale.

Data Creation and User Experience

Data is created on eBird in the form of checklists.  Based on previously recorded bird sightings in an area, eBird will provide birders with a list of all the species they are likely to see on an outing.  This checklist is accessible on their website and can be completed on the computer, but most birders download the mobile app Birdseye and do their checklists on that. To submit data to eBird, a birder will input the species and quantity of every bird they see, geotag their location, add ancillary information and photos, and upload the completed checklist to eBird.   

Checklists are nothing new to birding; they have been part and parcel of birders’ experiences in the field since the inception of modern ornithology.  Regional and national birding organizations produce checklists for birders that are sometimes handed out in person at conservation areas or that can be downloaded online.  eBird differs in its portability, ease of access, and immediate transfer to a large dataset.  Prior to online databases, conservation organizations would gather checklists from birders and painstakingly find patterns and construct models to get some idea of bird population data.  Later, in the early days of eBird, birders would transfer their data from their checklists or notebooks to the website’s lists.  James said that “before, you had to take notes, go home, transcribe them” which “got in the way of work”, but that since they introduced the mobile app he feels “liberated”: “Now, because it's a mobile app, it never gets in the way of anything.  So, if I’m going for a walk, I might as well record birds while I walk.”  Since every byte of eBird data is produced when networked technology, birds, and birders intersect, eBird tries to maximize the ease with which birders can turn every interaction with a bird into usable data.  They do this through incentivization (explained below) and by pursuing greater integration into the lives of birders.  It makes sense when James says that he “might as well” log birds while he goes for a walk because eBird makes it so easy.

The checklist quickly becomes a fixture of eBirders’ lives.  James related that “it becomes a bit of a driver of your activities”:

“Right now, they’ve got a new thing if you enter new data every day of the year you'll be entered in a contest and fun little things like that. Whereas last year I missed some days where I didn’t put in any data, this year I've been going outside and standing outside for 5 to 10 minutes every morning before I got to work and doing a checklist.  It's fun, it's comparative.”

Martha confirmed this effect of eBirding.  Because eBird ranks birders by checklist count and by species count at local, regional, national, and global levels, she mentioned that she coordinates with other female birders to make every attempt to keep women visible in Ontario’s top 10 list.  Aside from recognition in rankings, eBirders can also benefit materially from their checklists by winning contests like the one James mentioned above.  These are usually sponsored by Zeiss (a binocular company) and give awards based on varying criteria; inputting a certain number of checklists at night and recording 5 bird calls a month are some examples James gave.  The incentivization of checklists through awards and competition for rankings, coupled with the immediate accessibility of a mobile app means that birders can – and do – log birds outside of dedicated birding outings.  James lamented that Birdseye had yet to add a function that would enable him to bird while driving.  

When eBirders do go on dedicated outings, they log everything on eBird.  It is worth considering, however, that this information was provided by two very dedicated birders who run in circles of dedicated birders; there are certainly casual, occasional eBirders as well.  Nevertheless, the structure of eBird’s data-entry technology and the incentivization and ranking system that accompanies it tends to draw the energy of birders into eBird data production at all times.  A lifetime of ethnographic observation of my birdwatching father leads me to believe that bird-aware people are always looking for birds anyways, but eBird differs in the productive aspect of checklists; no longer does it suffice to see and identify a bird, but it “might as well” be logged and uploaded. 

Most of these checklists are approved automatically, but there is a filter that catches anomalous sightings. If a rare or out of season bird is inputted, or if an unusual number of birds are counted, it will trip the filter.  A regional coordinator will then contact the birder and ask for clear descriptions of their sighting before the checklist is added to the dataset.  James is one of the regional coordinators for the Ottawa region and he was able to provide some insight about the reliability of this form of data collection.

“Most birders are pretty good at self-policing, but you get the occasional person.  And we all know about them; there's one or two of them in Ontario. […]  But, you know, we're talking one person out of an entire province of birders so it's not a big problem.  Misidentification is a more serious problem, but that's what we're doing as local validators of data.  Basically, we write to people, we ask for more details, we then decide whether we're going to accept a record or reject it.”

Additionally, James has access to the parameters of the filter.  If rare or out-of-season birds are reported and validated with enough frequency, he can make adjustments so that the filter automatically accepts, say, kingfishers in February (an example he used).  As a result, data collection can remain largely unimpeded by validators except in abnormal instances, and the flow of sightings can appear immediately on the website for other birders to see.

Mapping and Data Dissemination

Every data entry that passes validation is made accessible to all users almost instantaneously, and rare sightings even send notifications to those who opt in to receive them.The data is accessible in a great variety of formats on the eBird website, and the Cornell Lab even tries to keep some of their more esoteric bird science open to the public.What most birders will interact with, however, is the map function, which presents every verified sighting ever on its precise coordinates on a world map.Many birders will tag the same location; for example, at a local conservation area, a birder will create a node that subsequent birders might then add their data to.Expanding the node will open a list of every bird logged in that location and will rank birders based on the number of checklists submitted and birds seen there.The result is a heavy peppering of geotags, colour coded to show the “hotness” of a site (how many birds have been spotted there).

It is worth noting, however, that this maps uploads and not birds.  Every checklist logged on eBird requires a person to be there seeing the bird, so the map maps birders as much as their quarry.  It is not unsurprising then to see that areas where people would presumably go on dedicated birding outings anyways quickly become hotspots; conservation areas, city parks, and waterways are all high in observations, and consequently have rich and long-term datasets.  City streets are less common, but James recounted one story of a friend of his who wanted to change that:

“It was about three years ago he was like “man there's a lot of holes in our Ottawa data.  I'm going to eBird every single bird I see for a year including every house sparrow in the city.”  And what a herculean task, he did it. “

Nevertheless, this is the exception that proves the rule, since most birders want to see lots of birds and especially the cool ones.  Few people are concerned with seagulls and sparrows, and logging data on eBird while in the city may not be an apparent thing to do.  Consequently, the data that is visible to birders is not necessarily an accurate representation of the amount of birds in an area. 

This does not, however, point to a deficiency in the data that the Cornell Lab produces; they have found ways to combine eBird data with other data in such a way that they can correct for inconsistencies and errors, and even predict how many birds may be in an unbirded area (eBird).  The data produced by eBird is high quality very well-regarded, and the models and analyses drawn from it are made available on their website.  While the data may be excellent, the manner in which users access it still plays an enormous role in the effect it produces, and for both interviewees and their friends this is done almost exclusively through the map. 

Feedback Loops and Conservation

Both James and Martha brought up an incident occurring several years ago when a group of great grey owls was spotted in the Ottawa’s east end.  According to James, “there were thousands and thousands of people went to see them.  It was on CBC”.  He was thrilled to see the number of people who came out to see them, but was also disappointed in some of the side effects of the birds’ popularity:

“Every time I went there were families with kids, you'd set up the scope, it was awesome; lots of people getting turned on to birding.  But my old crony birding friends hated it, hated it.  Its like the paparazzis of birding, right?  And people were baiting the owls and feeding them, and people were really just pissed off.  I thought it was great.  Now how many new birders were created by that situation?”

This points to an unintended consequence of eBird’s mapping system: hotspots and rare bird sightings are apt to create feedback loops.  When a spot becomes popular, birders flock to the location because they know there are birds there and, since they see birds there, they contribute checklists to the area.  That area may not necessarily have more birds than another spot, but it will certainly have more birders.  At places like conservation areas this is not abnormal, since, in all likelihood, they were hotspots prior to eBird, but any emergent spot can quickly spiral into overcrowding.  Additionally, if a charismatic species returns to the same spot every year, everyone who wants to see it will know precisely where to go, as the map is searchable by species and season.  This phenomenon occurs over time at hotspots and seasonally in certain locations, but flash-in-the-pan birder massing occurs within days if rare birds are spotted, as happened with the great grey owls.  James seemed to see two sides of this issue:

“The more this stuff is in databases, I think the better it is for conservation.  Not everyone agrees with me on that when it comes to rare species but that's my own feeling.  The problem is that there is a lot of photographer pressure.  For example, there's a bunch of short eared owls in Quebec right now near Luskville that are known, that have been eBirded, and I talked to a guy yesterday who was there last week, and he said that there were 72 photographers on the road. Hah hah.  So, there's no doubt that eBird can be used in a negative way.”

James points here to a growing tension between photographers and birders that has been mounting for some time (Spears).  Photographs of rare birds are highly sought after by professionals and amateurs alike, and many birders are concerned about the impact on bird communities caused by the presence of so many people, especially considering the frequency with which controversial baiting tactics are used. 

            Aside from concerns about respecting birds, both interviewees also mentioned a qualitative decline in birding experience that results from the upsurge in eBird-enabled birders.  James aired some of his displeasure when he explained why he rarely uses eBird in Ottawa anymore:

“Now, I can't go out birding in Ottawa without seeing other birders.  And as a social thing that's great, but from a birding perspective, one of the great things about birding is discovery.  So, you go out and you find something that's not supposed to be here, something unusual.  Well the chances of doing that now are much less, because there's like a hundred-plus retirees out there beating the bushes every day.  And if you're not retired, you know, you're not going to be able to find something novel. […]  I find it less exciting to bird locally because I rarely find something really surprising […] Even in my own backyard, you know it's like there's other birders combing my own backyard!”

Large birding organizations helping people find birds is not a new phenomenon, as attested to by field guides that no birder would be caught without.  Since these field guides have long been based-on data produced by citizen science initiatives, they also evince the reciprocity between database and user that is fundamental to most contemporary networks.  Like most digital networks, however, eBird differs from its predecessors in scale, accessibility, and timeliness.  While field guides might give a region’s birders a general sense about what to expect around a certain time of year, no technology before eBird was able to notify every single birder in an area of the exact location of a rare bird just instants after it is spotted. 

            Despite these kinks, James is still very enthusiastic about eBird’s mission.  For him, more data translates to more conservation and healthier ecosystems. 

“Short eared owls: everybody keeps them secret; they don't put them on eBird, which is a shame because it's a species at risk.  And what happens is there's a population near the airport that's not getting recorded on eBird. That means that when a consultant is hired to go and study that area for development expansion they don't have that critical species that's actually at risk recorded.  So, they go into the databases, they look, they go 'oh man', they go on their one trip in the middle of winter […] And then the place gets developed.”

eBird is constantly and reliably updated all over the world, providing information on migratory patterns and population that health helps conservationists address large-scale issues.  The fact that it is crowdsourced and publicly accessible also means that the data can be extremely effective at smaller scales, as local conservationists can use eBird data to coordinate projects that directly target places with rare or sensitive bird populations.  James is also thrilled that it gets more people outside and excited about birds; he hopes that this will lead to greater community involvement in protecting habitats.  The multi-level utility of eBird may certainly be effective for conservationists, but, as both interviewees pointed out, not everyone on eBird is a conservationist.  Because of the decline in quality of experience produced by crowds, and the pressure that eBird logging can have on rare species, many birders choose to withhold their data from the eBird database, preferring instead to keep the information within close groups – James calls them “cliques”.  There is an in-crowd in the Ottawa birding scene that spreads information using old methods of bird-news dissemination like telephone trees, in which one person will call ten people, who will then each call ten people etc.  While they still post frequently on eBird, they will make context-based decisions on when to not log something.

Conclusion: Immanent Databases

            Concerning data-sharing, birders and ornithologists have had a reciprocal relationship: birders provide the data, and conservation organizations analyze and disseminate it.  While birders are motivated by competition, collecting, and the undeniable the fun of birding, they also enjoy being part of bigger scientific undertakings and conservation efforts.  For birders like James, it is even felt to be a duty.  Were it so simple eBird would seem to be a win-win for birders, but the technology comes with many drawbacks. 

First of all, it makes the database immanent to the eBirder because it is embedded in mobile phones.  Whereas older databases did not produce the ever-present imperative to share data, eBird turns users into competitive, constant data sources.  This shift evokes Deleuze’s observation that emergent forms of control function by eliminating the organic structure of society and doing away with distinctions between the “different internments or spaces of enclosure through which the individual passes” (Deleuze 4) that characterized previous control methods.  eBird functions through “inseparable variations, forming a system of variable geometry” (4) that does come and go from the birder but captures labour at every encounter they have with a bird. 

The way birders act when using eBird also echoes what Bifo Berardi noted as one of the main processes of the digital transformation: “the capture of work inside the network, that is to say the coordination of different labour fragments in a unique flow of information and production made possible by digital infrastructures” (Berardi 88).  Workers in these digital infrastructures are highly mobile, “But at every moment and place they are reachable and can be called back to perform a productive function that will be reinserted into the global cycle of production” (90).  Birders might not be producing ‘semiocapital’ like Berardi’s cognitariat, but their leisure time has analogously been yoked to the production of capital.  It may seem silly to speak about birding using ominous Marxist language, but it should be remembered that eBird does function like a business; the full version of the Birdseye map costs $3.99 a month for their North American customers, and they receive funding based on the effectiveness of their data.  They are cornering the market on bird data (what a funny market to corner) and doing this thanks to the uncompensated birding labour of over a million talented birders.  On the one hand, it seems ridiculous to hold that against a scientific body, but on the other hand, it would be naïve to ignore the enormous power they have garnered as gatekeepers of the world’s biggest repository of ornithological data. 

And should ‘the world’s biggest repository of ornithological data’ be immediately accessible to every single user at all times?  As the interviews with James and Martha showed, this aspect of eBird creates feedback loops that crowd hotspots, disturbing both birds and birders.  Some birders attempt to counteract this effect by withholding data but do so at the risk of jeopardizing conservation efforts.  In the end, the overwhelming tendency of eBird users is towards data-sharing, and while elite birders may take satisfaction in protecting their finds from hordes of photographers they make little impact on the feedback cycles that seem inevitable with such a universally accessible, enormous, and precise dataset.  Birding has undergone its digital transformation, and as the problems inherent in such technologies become apparent, birders and ornithologists might benefit from a look to other digitally transformations as they seek solutions.

Bibliography

Berardi, Bifo.  2009.  The Soul at Work.  Los Angeles: Semiotext(e)

Deleuze, Gilles.  1990. “Postscript on the Societies of Control”.  L’autre journal 1: 3-7.

eBird. "eBird Modeling." Accessed April 15, 2018. https://ebird.org/science/modeling.

Spears, Tom. "The Purists vs. the Baiters: Fowl Play in Ottawa's Birding Country." Ottawa Citizen. February 09, 2017. Accessed April 14, 2018. http://ottawacitizen.com/news/local-news/the-purists-vs-the-baiters-fowl-play-in-ottawas-birding-country.