??????????
With some time to kill on Saturday and no desire to just sit around worrying about how I was going to get to Freeport on time, I headed to Prospect Park to round out my list of common winter land birds. It was a pleasant day, and I got everything I expected at the feeders (Red-breasted Nuthatch, White-throated Sparrow). But the real highlights were the cold and miserable Wood Duck sitting in the small patch of open water outside the Audubon boathouse, and a pair of Rusty Blackbirds on the bridle path. Rusty Blackbirds have become more and more of a challenge to locate as their population plays out a catastrophic tailspin that has been going on throughout my entire life. And with blackbirds being about as un-charismatic as you can get and still be a warm-blooded vertebrate – with most people, in fact, being utterly unable to tell a Rusty Blackbird from a Boat-tailed Grackle from a European Starling, and consequently perceiving them as common as dirt – I sadly suspect that that isn’t likely to change in the near future.
Gratified, I came home with a checklist and hand. I sat down in front of the computer, and pursuant to my New Year’s resolution to do more Science!, I logged into eBird. And then, as always, I faced a dilemma – one of the major factors in driving me away from eBird the first time out, in fact.
eBird, quite rightly from a data aggregation perspective, requires you to enter a count for each species that you report – the old standby “x” doesn’t cut it. Their FAQs kindly indicate an openness to estimates and even a degree of guesswork, but this count requirement still paralyzes me, because I am neurotic. In particular, I worry constantly about the fact that birds have wings and that I may encounter the same individual twice without realizing it.
Feeder-watching is particularly bad for this. I see thirty chickadees (to pick an egregiously kinetic example) over the course of a fairly short feeder watch, and yet only have three specific individuals in my field of vision at any one time. I know (or at least strongly suspect) that the birds are taking seeds elsewhere and then coming back, so the right number isn’t thirty, but I also have a fairly good idea, based on the rate of turnover and how long it takes a chickadee to open a seed, that it’s not just the same three birds over and over. So what the heck do I enter? Normally I end up just doing a rough mathematical split of the difference, and entering 16 or something, but I always feel absolutely terrible after doing so. I’ve searched the FAQs for guidance on this issue and found none.
What say other eBird users? Am I missing something obvious in the eBird documentation? How do you count?
:: 
:: 
:: 
:: 
:: 
:: 
:: 
:: 
:: 
:: 
:: 
January 20, 2010 at 6:52 pm
I always estimate for bird like Chickadees and Cardinals and other common species I usually note but don’t actively count. I figure it all comes out in the wash in the end.
January 21, 2010 at 3:12 pm
Yeah, I know deep down that it’s better to have the estimate than nothing at all, but it still drives me crazy.
January 20, 2010 at 8:23 pm
You could send an email to Brian Sullivan or Chris Wood and ask. They seem open to questions and revising documentation to make things like this more clear. They might even write a post on feeder watching.
I try to follow a loose version of the GBBC rules when I’m feeder watching. I base an estimate on how many birds I see at a given moment plus how they’re behaving. So if I only see 3 chickadees at any one time, but they are coming to the feeder from multiple directions, I’ll report a somewhat higher number that takes that into account. It may still be an underestimate but is probably closer to the true number.
January 21, 2010 at 3:13 pm
That seems like a good protocol. And it’s good to know that the folks at eBird are responsive. I should shoot them an email asking and suggesting a clarification.
January 20, 2010 at 8:55 pm
How to count birds may be the biggest stumbling block on eBird, with pin-pointing locations a close second. This may serve as a starting point for the counting dilemma, from an eBird “News and Features” post:
http://ebird.org/content/ebird/news/trouble_with_X
Personally, I follow the same protocol John describes, melding what I can observe at the moment with behavior (their behavior, I try and stay out of it!). I think Nate nails it regarding the wash: exact counts are nice, but estimates are the reality, and they adequately provide the information projects like these are designed to collect.
-Mike
January 21, 2010 at 3:39 pm
Hi Mike! Thanks for dropping by with your perspective. It’s very helpful.
January 21, 2010 at 3:23 am
I would think that the Great Backyard Bird count protocol, to count the greatest number of individuals you see at one time, would be a good way to do it.
Watching the feeder, you see 2, then 3, then 2, then 5, then 3, then 4, then 1, you would enter 5 as your count.
In my particular case (the GBBC) when I use that system, I know it greatly underestimates the number of Ravens (my main bird), because they will be found in groups through out town, at the dump, and at dog teams. In that particular scenario, I station myself where I can see the dog teams and town, do the count on a sweep, and then move on to the dump adding those birds, less any arriving as I do. Closer but still under estimates the true population.
January 21, 2010 at 3:41 pm
Sounds like a good protocol.
And I know it’s all a matter of perspective, but I have to admit I think it would be cool to have Ravens be my main winter bird. At least for a couple of winters – then I’d probably be craving a nice trip to Florida and some ibises.
January 21, 2010 at 11:41 am
I would generally follow GBBC-type rules except that I tend to take long walks. So then I try to add sightings up in my head, but only if I think the sightings were far enough apart that they weren’t the same birds.
I still do use the dreaded “x” on occasion, like when I am in Central Park on my lunch break with limited time and walk past the reservoir and see the giant cloud of gulls. There are Ring-bills and GBBs and Herrings but damned if I am going to try to even estimate what the proportions are.
January 21, 2010 at 3:43 pm
Yeah, I’d be hard-pressed to come up with an even vague estimate for some of those gull swarms.
I wish I knew more about the typical everyday activities of some of these common birds, how far they typically travel and so on. Hopefully once I go back to school I’ll have access to more journals and can find out.
February 2, 2010 at 11:36 am
Different counting techniques have to be applied to different habitats and different ways of birding, so it is a complex matter and one that’s difficult to answer within the frame of a blog comment, and no, I am not an eBirder anyway, so my answer may not apply to your world of birding. I generally apply the following three techniques:
Open habitats (mudflats, lakes, ocean, meadows etc.): a continuous scan from one end to the other where you count all individuals of all species. In multiple-species assemblies, this may require multiple scans. I can simultaneously count three species “in my head” (I use commas, e.g. I count 2,3,6 – 2,4,6 – 3,4,6 – 3,4,7 etc) and have a counting device in each hand which means I will count 5 species in one scan. Birds flying through my field of view either get subtracted or added, depending on whether they fly against or with my scan direction.
Walks through more dense habitats (forests, parks, settlements etc.): I count the birds I see and then add up the totals. Of course a bird I have already counted might fly ahead of me and then be counted again, but the odds of that happening are roughly identical to a potential second bird perching along my path ahead (out of sight) and moving away before I get there, so it levels out. Of course if a generally scarce species obviously flies ahead of me I don’t recount it because e.g. 5 Northern Shrikes on a 100 metre stretch of hedgerow would be ridiculous.
In perfect honesty, I don’t ALWAYS really count. Sometimes, I just write down an estimate.
Stationary counts (e.g. backyard, feeder, waterhole etc.): If the count is embedded into a scientific project (like eBird or maybe a private, personal project to get seasonality tables of birds seen at your feeder) I write down the greatest number of birds seen at one time during the count period as this is the only really reliable figure I can get. Any other number would be guessing and thus not reliable. Of course, if it is just a once-in-a-lifetime count for my personal records, e.g. a day at a waterhole during a birding holiday far away abroad that will not be compared to other counts on that waterhole or used in any other way scientifically, I’ll just guess the total.
And if I encounter e.g. a large flock of gulls but don’t count it at species level, I’ll leave it out because the information alone that the species was present, without any quantitative idea, is not very valuable anyway.
Well, again: if it is just for my private, personal records, I might write down they were present “in high numbers” or so, but I would not enter them at all into any scientific data base.
Phew!
Long comment. I should have added a nonsense sentence somewhere to check if anyone has been reading it to the end…
February 4, 2010 at 10:00 am
“I should have added a nonsense sentence somewhere to check if anyone has been reading it to the end…
”
This implies that the rest made sense, a rather brave assumption.
February 4, 2010 at 4:38 pm
Huh?
Nah, I get what you’re saying, although ebird strongly discourages the dreaded “x” even for things like the gull count you mention. In fact I’m not even sure they let you do the x at all – it choked up an error last time I tried, although that may have been coincidence.
February 5, 2010 at 8:08 am
Oh, I never doubted you’d get what I was saying (I wouldn’t easily underestimate you), I just had one of my self-irony fits
The problem arises when you only use eBird to keep track of your observations. If you have two lists, e.g. an excel sheet and a scientific data base like eBird, you can use “X” on your private list and leave out those data sets completely in eBird. That is how I am keeping track of my observations.