Friday, April 1, 2016

My 1st eBird Anniversary

The following blog post is a year-in-review (April 1, 2015 to March 31, 2016) of my experience as a new user of eBird. I have been a birder for more than 25 years, and subsequently have had a long history of record-keeping and data management challenges. Enjoy.

-------------------

April 1, 2015 marked yet another major shift in my history as a birder and biologist; it was the day I started using eBird, developed by The Cornell Lab of Ornithology, to keep track of my bird observations. Since becoming a birder, sometime between 1988 and 1989 (it's hard to know exactly when one truly becomes a birder), I've been through a broad range of methods of storing and tracking my bird observations. And even though eBird launched in 2002, it took me 13 years to adopt it as a tool for recording bird observations! Some might call me a late adopter, but the truth of the matter is, eBird had some growing pains and the system I was using at the time served my needs.

My record-keeping in the early phase of my birding history was slim to none; some records were documented with photographs and fewer still were written down. It wasn't until the fall of 1990 when a good friend and birder, Peter Sherrington, encouraged me to keep track of my observations. At the time, I was cutting my teeth on the birds of Bow Valley Provincial Park, Alberta, Canada. Peter advised that not much was known about the birds in that park, and that keeping track of my records could prove useful and at least provide a basic inventory of species; with enough information a good picture of species' chronologies (i.e. time of occurrence) and habitat use could be determined. Eight years later I published my first book, The Birds of Bow Valley Provincial Park - Ecology and Status, based on 6,195 records representing more than 50,000 birds of 189 species.

A labor of love, The Birds of Bow Valley Provincial Park was published in 1998

Records from Bow Valley Provincial Park were initially entered into a basic spreadsheet...some obscure program that worked in DOS. However, after a short time I started to use one of the first pieces of commercially-available birding software, Avisys. Within Avisys I kept all of my Bow Valley Provincial Park bird records, as well as any others birds that contributed to a 'complete' life list (i.e., any new species seen, including any new species seen at the state/province/country level). Unfortunately, almost all of those digital records were lost to a computer that ultimately crashed to an irrecoverable state. Fortunately, I did have the foresight to print a hardcopy of my Bow Valley Provincial Park records as part of writing my book.

Having a hardcopy back-up of bird records from Bow Valley Provincial Park proved critical

In 1995 I was encouraged by another birder to keep a permanent record of all of my observations, regardless of where and when they were seen. And so, in early February 1995, I commenced keeping a hardcopy record in hard-bound diaries. I kept records from every location and every date where I saw birds...this included parks, party's, and parking lots; university campus grounds; road trips; airports; and so on and so on. In some years, I generated more than 10,000 bird observations!

More than 20 years of bird records documented in hard-bound matching diaries.

When I returned to school in 1996 to commence my undergraduate studies in biology at the University of Calgary, I put my "birding" on hold (not entirely of course). Thus, for the next four years I made no attempt to keep track of my records digitally. In fact, I became much less interested in tracking my "life list" (although I enjoyed seeing new species), and much more interested in the ecology, behaviour, and conservation of birds. By the year 2000 (now married and slightly more intelligent), I was out of undergrad school and working steadily. My life list was getting little attention, but my 'diaries' were reaching epic proportions, making the daunting task of entering those records less inspiring with each passing day.

I do recall that sometime between 2002-2007 I quite literally 'glanced' at eBird. This glance probably occurred early in that range of years, and I distinctly remember creating a profile and entering a few records. I recall finding the process utterly cumbersome (using a PC), and the end result being underwhelming insofar that I couldn't be bothered to enter more records. Thus, at the time, eBird was not for me.

In July/August 2008, after visiting Panama in February, I was visiting my wife's family in the United Kingdom, and was asked, "So, how many different species have you seen?" The impetus of the question was based on my recent trip to Panama, and that I had been running around southern England trying to add more new species to my life list. The question was fair enough, and in 1995 I could have answered that question without blinking. Ironically, however, the best I could muster was a guess that was probably off by 20+ species. What I realized in that moment was that I had lost touch of something that I really enjoyed: counting species.

When I returned from the United Kingdom I took it upon myself to build a life list. I took a relatively simple approach at first, which was to compile a list of species primarily from memory, with some aid from my diaries (at this point I had 29 of them, each containing more than 5,000 records). I built the list in about two days using Microsoft Access, but shortly after I realized that my interests ran much, much deeper than a simple world list. I wanted to know how many species I had seen in each country, and in particular in each of two provinces (Alberta and British Columbia) where I had spent accumulating most of my bird records. That interest then expanded to wanting to know how many species I had seen in other provinces and states. For example, if I had seen two lifers in Montana, I wanted to know the total number of species I had seen in Montana. Both of these latter tasks were too great to take on at the time, and so I continued only to keep a master list for the world and for countries up to date. I did contemplate using commercially-available birding software again, but when I did my homework I was as underwhelmed as when I tested eBird several years earlier, and I didn't re-check eBird at this time to see if improvements had been made. At least after investing effort with Access I could answer the question, "How many birds have you seen?"

In 2014 I decided to add to my Access database the date of first observation to each of my life birds, as I was interested in looking at how my rate of accumulation of life birds varied over time. Assigning a first date was relatively easy for 90% of the species I had seen, as I was able to go through my diaries for all of the various countries/provinces/states I had visited, draw from my Bow Valley Provincial Park records, and glean from my early rare bird submissions which were published in various rare bird reports. But for common species, such as mallard or house sparrow, I have no idea when I truly added my first record. Thus, for a few species, there will be some true unknowns that will eventually be assigned the date in which they first occur in my sources of information (i.e., photos or notes).

On April 1, 2015 I went birding with two friends, Megan Willie and Jayme Brooks, in the Greater Victoria, British Columbia area. Both Megan and Jayme were recording our days' bird observations via eBird, using a mobile app (I should add, I was also a very late adopter of an app-based mobile device, having only got one in late 2014). I hadn't used eBird for years, and hadn't really kept abreast of eBird technological advances. Suffice it to say, that after birding with Megan and Jayme for two days, I was intrigued to investigate further by downloading the iOS version to my iPhone, and the companion BirdsEye app (for BirdsEye, I bought the annual World Membership to suit my current birding interests). That intrigue was fostered further by the fact that both Megan and Jayme 'shared' their lists with me, meaning I didn't have to [redundantly] enter the records. It took mere days thereafter to become hooked on the portability, utility, and expedience of the eBird experience (both mobile and laptop), and after logging a few hundred local checklists, sharing checklists with friends, and ultimately being exposed to the Explore Top 100 (essentially, Top 100 anything, by year or location), I was undeniably hooked. Moreover, the obvious security of having my observations stored on a secure server, and shared with the broader birding and scientific communities, were values that resonated with me. And as for Megan, Jayme, and myself, one of our favourite phrases each time we saw a new bird was to use eBird as a verb: "eBird that sh!t".

The eBird app interface; a simple yet powerful tool for logging bird observations

Now that eBird was entrenched in my birding life (and yes, I still hand write all of those records into diaries), this left me with a big task: How do I get my tens-of-thousands of records from notebooks and Access into eBird? The obvious answer: Enter them. The challenge: How to do it in the most efficient manner that provides the information I want to see first?

I began with entering my life list, which didn't rest with just importing my Access list; that would be too simple. eBird requires quite a bit more information (when you have it), such as not just providing the name of the country you're in, but rather the exact spot you're standing in, or the area you're travelling through or around. eBird also prefers some measure of effort invested in birding: What time of day did you start; For how long did you count birds; Over what distance were birds counted when travelling. But perhaps most important is how many individuals of each species did you see. Fortunately, I had most of this information in my diaries, and so the data entry process began. Where I didn't have this information, I used eBird's 'historical' data entry option, which allows users to enter records with incomplete or unknown information on effort. Subsequently, users can have all of their data on eBird, and more broadly the data contributes to the inventory and time of occurrence of species.

As I did when compiling my life list in Access, I went through all the different countries I had visited first (eight, not including my home country). This task produced the largest majority of species on my life list, somewhere in the range of 80-85%. I then systematically entered records from Canadian provinces (except Alberta and British Columbia where more than 98% of my Canadian records are from). At this stage, including those records that I had already processed since starting to use the eBird app in April 2015, I had more than 95% of my life birds added to eBird. I then systematically went through the species I was missing, which were all either from Alberta or British Columbia. Most of the remaining birds were "speciality" birds, and generally were easily found in my diaries from pelagic trips, trips to the alpine, or one-off trips I had made to chase a particular species. By August 2015 I had my entire life list on eBird...a great feeling of satisfaction.

What I hadn't anticipated when entering my records was: 1) dealing with differences in taxonomy updates, and; 2) dealing with regional coordinators (reviewers). eBird uses the Clement's Checklist of World Birds, and periodically updates the list as changes occur (i.e., name changes; species are lumped or split). The records I had entered into Access fortunately followed the Clement's taxonomy, but I had used a version that was two years older that what eBird was using at the time I was adding my life list. This meant doing a bit of sleuthing, primarily of tropical species, to figure out which species I had actually seen among those that didn't match. It didn't take too long to fix these problems, and subsequent taxonomic updates by eBird are largely automated. It was particularly satisfying that when in August 2015 the eBird Taxonomy update occurred, I had nothing to do but look at my automatically revised list.

Last published in hardcopy in 2007, the Clement's taxonomy of world birds forms the backbone of eBird. The list is now updated annually, in digital form.

"Rare" birds, as eBird classifies them, are species that are not necessarily rare. Instead, they tend to be species that occur outside of their expected normal range, occur in part of their range at unexpected times, or represent an unusually high count of that species. Over a period of more than 25 years of birding, I've certainly found my share of rare birds. Yet in the digital eyes of eBird, all of these records require vetting by a regional coordinator (a human!), regardless of when they were originally seen. The process is excellent indeed, although when I was entering large numbers of historical records from multiple locations I certainly I had a fair number of simultaneous follow-up questions to respond to from multiple coordinators. My experiences with regional coordinators, whether they were from Arkansas, Arizona, Panama, Peru, Morocco, or anywhere else, have all been excellent. In most instances I was able to provide satisfactory notes, provide additional clarity on the location, correct an accidental entry (i.e., a species had been split and I assigned the wrong one under the current taxonomy), and in two instances remove records because of uncertainty. While I personally strive to provide accurate records, I maintain wholeheartedly that if I have uncertainty in a record, after being questioned about it, I will remove that record if I have no additional information to support it. For example, I recently removed what I thought was a record of a herring gull from a checklist I submitted for Souss-Massa National Park, Morocco. While I did have good field notes, I did not have enough information to reliably distinguish the bird from a yellow-legged gull, which to my surprise, occasionally can have pink legs! I thank the regional coordinator for working through the process with me; and I learned something too!

Now, a full year into eBirding (as many call it), I have amassed 1,849 checklists comprising 1,823 species. Of those checklists, 728 were for the period April 1, 2015 to March 31, 2016, and my processing of historical records to date resulted in 1,121 completed checklists. I'm currently processing my historical records from Bow Valley Provincial Park, mostly for sentimental reasons but also in consideration of updating my book. I have merely scratched the surface of entering all of my personal records, but after a year of using eBird I find it increasingly immersive on many fronts...currently I am using eBird to plan a 17-day birding trip to Australia, and I used it extensively to plan for Ecuador and Peru in 2015. I did not use the eBird mobile app when in Ecuador and Peru however, as most of the time I was not within cellular range. eBird does allow "offline" checklists, in which case you get an initial position from your phone's GPS and then verify the location later. I preferred to use a separate app (GPS Tour app for iPhone) to get Latitude, Longitude, and Elevation, which I then wrote into a waterproof field notebook with start and end times. I found that this wasn't overly cumbersome, and didn't detract from birding in any way, shape, or form. After returning from those trips, I entered the records online, which was better for reviewing and making any necessary changes...for example, the guides on both trips often used species names that were not represented in the current eBird taxonomy, which meant sorting that out in the field would have taken away from birding.

When birding in remote areas, I still prefer hardcopy notes. I eBird the records later.

Like many things that I've become obsessed with in my lifetime, I tend to engage rather intensively. Thus, I have already reached out to eBird with suggested improvements, which to my delight were received openly from eBird developers (FYI eBird, I will be sending more suggestions!). I also decided to celebrate my 1-year anniversary of eBirding by becoming an eBird Partner, which financially supports development of perhaps the largest citizen-science project in the world, but also gives me a chance to beta-test new elements of eBird and provide constructive feedback.

I don't regret having taken so long to adopt the use of eBird, because in doing so would serve no purpose. But if you find yourself in a similar position as I did, in having a seemingly insurmountable volume of information to process, perhaps some comfort can be taken from my experiences above. I do have a lot of work left to do, but getting my life list and country lists onto eBird was relatively quick and painless, and maybe amounted to no more than 100-150 hours of effort. And there is no doubt that the mobile app (available on iOS and Android) is a breeze to use, and great for running trip lists while on the go.

Until next time (perhaps a year from now), happy (e)birding wherever you may be.


7 comments:

  1. I entered my first two ebird lists earlier today, and am inspired by your story.

    ReplyDelete
    Replies
    1. That's great. You're making a great contribution!

      Delete
  2. Hi Michael,

    I came across this post indirectly because of your eBird entries in Banff county. Not that I'm competitive, but I'm 17 species behind you :-)

    More seriously, I'm interested in your BVPP book - original or second edition. And also whether BVPP had the same boundaries when you were compiling the book, or if it's changed in the last 25-30 years? Would you, for instance, have included Seebe Dam or Hay Meadows in your 'survey area'?

    Thanks for your thoughtful comments on eBird. As with anything it's not an unmitigated good, but overall I believe it's more than good enough for every birder with an interest in conservation (or listing) to participate.
    Neil

    ReplyDelete
    Replies
    1. Thanks Neil. I think I can send you a pdf of the book...I'll look into it. I did include Seebe Dam in the species accounts, and if I recall correctly, Hay Meadows is to the south, near Chilver Lake? If so, that was included too. The book has a map showing the boundaries.

      I haven't quite yet finished entering my historical records for Alberta, including Bow Valley PP, but I have the majority of them in. If I can squeak out just two more species I can temporarily hold 1st place for Banff county, until you pass me :)

      Delete
    2. That would be great, Michael. And if you do update your book, be sure to let me know. My son and I would be happy to help with any 'feet on the ground' stuff.

      Hay Meadows is just north of Kananaskis village, close to Lorette Ponds. I'd never have thought of that area as BVPP until I looked at this map: www.albertaparks.ca/media/767879/2008_bvpp_map.pdf.

      To contact me privately, email info birdboy.ca

      Thanks,Neil

      Delete
  3. Michael,
    I came across your blog via a Manu post. We just recently returned from a 20 day trip, covering much if not all of the same ground. I thoroughly enjoyed your blog detailing this trip. Thanks for sharing!

    Comments re ebird and the phone app. Kudos to you taking the time to go back and enter all your records. I know it is important to the data base to have such info in its system. I was one of the original beta testers for the phone app. It is amazing what they created. I love using it in the field.

    Frances
    Lodi, CA

    ReplyDelete
  4. Thanks for your comments Frances. I still have another 2 posts for Peru to come, and should be posted in the next 1-3 weeks.

    I've really enjoyed using e-bird, both as a database and as an app. I'm heading to Australia in September and have been using it extensively for planning (along with traditional resources such as bird-finding guides). I recently became a 'partner' of ebird, as I can't wait to beta test their new products...the next one being a profile page for your records...looks interesting.

    If you're on Twitter, follow me @avispec and you'll know when I put new posts on my blog. Cheers, Mike.

    ReplyDelete