Saturday, March 16, 2013

Oh, e-, please tell me (At what level do I record birds for eBird?)

I am up to date for my eBird records for this year! Well, at least I think so ... I do have lots of 3x5 cards on my desk and may find more reports.

Tonight I faced a familiar dilemma while entering data for a local place.At what level do I  want to enter data?

When I first started entering eBird data, I tried to be as specific as possible. For large areas with diverse habitats, I knew there was a huge difference between Nature Centers with bird feeders and the ducks on the lakes in the same park. Hey, even as a new birder, I could figure this out ... :)

I initially thought about how the data might be used, and as a person who previously ran biological surveys (albeit in the marine environment), I tried to go with the more specific.I made lots of personal sites.

As I learned more about the workings of eBird, I saw many more  "Hotspots" start emerging. I wanted my data to be included. I was pleased that a few of my suggestions as Hotspots were accepted for eBird (Thanks eBird - I do not suggest hotspots lightly...). And so it (i.e., the family of Hotspots) grows. ;)

Anyway, I started recording data more into Hotspots than the more specific personal locations I had previously used. Heck, I am not doing surveys, and get my birds into the right area. I am positive that on the national level of eBird, it matters little. I learned how to merge my past location data (yes, I still need to do more merges) into Hotspots.

It matters not at all from a "listing" point of view. My birds are recorded. I can see my records anytime I want! And, it does not matter from a county review, or the "top one hundred", or basically any eBird slice of data. I feel great that I am making a valuable contribution to national (and now international) bird distributions! Cool!!! Most of my records are birds I see on my deck each morning over coffee, and even those records are valuable. After a couple years of recording data, I now really get to see when birds appear and leave my deck. It has been fun recently to see the winter/spring border of arrivals here and comparing them with past records - and I can see my reports over time without going back to my journals!!!  Cool!

YET .. something is missing!

By "taking the easy route" for most sites, I have lost data (no - not really, most of the data are still in my journals - which BTW probably no one will ever see... not that it matters on a grand scale...)

Tonight I entered all my deck birds for the last two weeks (up to date!), then went on to record birds Judy and I saw at Belle Isle (BI), Wayne County, Michigan  last week and other places. I discovered a Hotspot that I think was not there when I first started recordng BI birds. I had a personal site called Belle Isle Nature Center. I found an eBird Hotspot at the same place called Belle Isle Nature Zoo (actually a more correct name because the Nature Center (NC) there is really an extension of the Detroit Zoo - they have Steve - a horny special deer - and his gals in an outside enclosure ... except when Steve is too frisky...). I merged my site with the Hotspot.

OK. But I had given up thru lazieness separating my bird data between the Nature Zoo and the remainder of Belle Isle (main Hotspot). It is so much easier to record Belle Isle for everything! It bugged me because of the huge difference between we see there at the feeders and what we see on the rest of the island. The difference is especially huge in winter!

Again I doubt it means little to national eBird data. Yet I know a great birder, Allen Chartier, has been studying birds on Belle Isle for several years and has extremely refined data. Allens's recent BI checklist on eBird can be seen at:

And I also think about the the fact the Nature Zoo folks probably would love to have data to justify their existance (and probably also the cost of bird seed) and the eBird records would be invaluable for this.

Far beyond the present example, I have always been bothered that the local Metro Parks show up in eBird as Hotspots; for example, Stoney Creek Metro Park - Oakland County or Stoney Creek Metro Park - Macomb County. I guess at Stoney Creek, I can see the smaller chunk of Oakland County as one Hotspot. Fine. You are either looking at the lake itself or walking the trails at West Branch, and the areas are very close and you can guess at where the bird (aquatic vs. forest) was seen.

The problem remains with the much larger Macomb County part of Stoney Creek that includes the Nature Center with feeders. I do not know if anyone is doing surveys there, but if I worked at the Nature Center, I would certainly love to have records of birds seen in my immediate area of responsibility!

Similarly, Kensington Metro Park. Most of the park is in Oakland County. The smaller piece in Livingston County is basically watching birds from the boardwalk or taking a hike around the smaller lake. OK.  Fine. Yet, the Nature Center is in Oakland County - an area with quite diverse habitats!  The Nature Center with its trails (where people can place seeds in their hands and have birds feed - WOW!)  is a very special place indeed!

Oh! Dilemma!

Why should I care? Heck, the same thing exists at Streamwood Estates (where I live). We have a wonderful fairly long stretch along the Clinton River, but especially in winter, a walk along the river yields few birds except near areas where neighbors have feeders. Makes sense to me... I would much rather hang out near a fast food place in winter! I record my deck (feeder) birds separately from my river walks. And my site (combined = eBird "yard") is not anywhere as large as the afore-mentioned parks. Well, I guess this area is my survey site and it matters.

But the question I propose (Hey did you think this was just a "ramble" without a question?) is:

At what level of detail do we want to record eBird sightings? For the referenced cases, I guess I would like to think someone cares about the difference between nature center feeders and what birds are sighted "in the wilds". One solution is to have hotspots for NC areas, then label the remainder of the larger (eg., park/county) area as - for example, "Belle Isle except Nature Zoo"; "Stoney Creek Macomb County except Nature Center", and so on. Using eBird, it is easy to create a "patch" or "yard" that includes multiple sites, so if you wanted to see all  Belle Isle Birds, or all Stoney Creek Metro Park birds, you just select and include them as part of the yard or patch!

I also think of Pte Mouillee (again with diverse habitats in a major area) as a candidate for more specificity (and I do see more hotspots there now than before...).

What do y'all think?

Thanks for reading my "ramble"!!! And - do please use eBird! It makes it easy to record your sightings and see your "life list" and offers an amazing variety of ways to "cut your data".

- "Dr. Bob"


Charles Swift said...

Dr. Bob - Forgive me if I'm misunderstanding your question but location specificity is important and useful to eBird. This allows for finer grained spatial analysis and association with habitat information. In fact Hotspots are primarily for birders and probably used little in eBird analysis. I would hope that as eBird evolves the hotspot concept will evolve to allow more specific locations under and umbrella hotspot (but it is perfectly fine to use a personal location if it is more specific than the closest hotspot). As a general rule stationary or short travelling counts at specific locations will be most useful to eBird. But of course birders should do what that can that fits with their birding style and preferences. Charles.

pauljhurtado said...

I always go with as fine a spatial resolution as possible. If a hotspot exists (or appears) where I'd otherwise put (or already have) a persona location, I use it. Otherwise, I stick with a personal location. At the other extreme, some data is better than no data, so I try and submit incidentals for potentially valuable observations with the best effort and location info I can include -- even if it's only ballpark. If I'm guesstimating a start time or location, I'm always sure to make that abundantly clear in the species comments (remember: nobody can see the checklist comments but you! Species comments, however are open for all to read).

Hotspots are great, but I always try and be specific about location info and err on the side of being too specific when it comes to using a personal location vs. a personal location. That said, I do try and follow this naming convention for personal location near hotspots:

Lastly, I try and also include exact (ideal; not always practical) or estimated counts for each species, reserving "x" (the eBird symbol for "present/observed") for cases where I was too lazy to even pay attention to numbers, or was otherwise unable to come up with a reasonable estimate. Again, I try and note "estimate" or "approximate count" in the species comment, and if my estimate is questionable I'll elaborate on (1) how the estimate was arrived at and (2) I'll try and give a range of possibilities.

PS: Probably not all that appropriate for the eBird Tech Talk email list, but definitel share on the eBird Facebook Discussion group page! ;-)

pauljhurtado said...

... ignore those typos ;-) That should read "using a hotspot vs. a personal location"

"Dr. Bob" said...

Thanks for comments! All info helps me figure out what to do. I find asking a question the hardest part - " i.e., what do I want to do with the data?" Charles - I am looking at eBird as a "birder" and not as a scientist. Right now I am working on a list of birds for a local park. Fortunately it is designated as a Hotspot, so if everyone lists their birds on the Hotspot, I assume eBird will allow all their records to show in the total species list. If, however, people are using a personal site in the park for records, I do not think I the compilation will catch their records. Yes, this is a different problem from the larger Metro Park situation I discussed. In the Metro Park example, think that I might want a total of all birds seen in the Nature Center, but folks were recording data into the larger county Hotspot.I would not be able to see the more localized data for "my slice". Nor would eBird have refined habitat data separating feeder birds from the much larger area. I am not sure of the answer, but at whatever level we record data, it at least allows good county distribution data.

C. Michael Stinson said...

Bob - as both a reviewer (for VA) and a lister, I say use the most specific level of detail you can. I face similar issues here in VA and I make lots of individual (personal) locations. Occasionally I use Hotspots as well, but often they are too general for my taste. And as a reviewer I can tell you that checklists with vague or excessively general location info are frustrating...

"Dr. Bob" said...

Thanks, Michael! And I agree - especially with the example! I do think that is what I was addressing! Now I have a different dilemma (that I will address in another blog...). In looking for data for a specific small city park, I learned that several birders had established personal locations for the park. It will be fine for eBird folks, but as a birder, right now I just want to get a total for all birds seen at the park, and the personal locations do not "roll up"!

"Dr. Bob" said...

The city park is an eBird Hotspot ...

Robert Gerald Porter said...

Personally, I think being as specific as possible (and as specific as you are willing to get) is perfectly reasonable. If it is on public land though I'd say share that specific location so others can pool data with you, and interested eBirders can generate locations reports better.

For example, in a conservation area local to me that is hundreds of acres in size, there was just one hotspot a few months ago. Since then I've added many hotspots (and they've been approved) so that myself and others can have the option of being specific, whilst the larger original hotspot remains for those who might have just drove around the area quickly and only are willing to make a single report. It's important that this larger "meta-hotspot" remains so that contribution is still quick an easy for people with less time/energy/experience, but its equally important I feel that more granular hotspot options remain available.

I would say that as long as hotspots ultimately can be as small as the reasonable viewable distance from a standing location -- for example, as you say a locations with feeders could be one, and a couple hundred feet away could be another if perhaps there's a body of water not viewable from the feeders, that would have completely different birds (waterfowl).

Personal locations I tend to keep for one-off incidentals (highway sightings, residential areas, etc) and for private property. Anything else I put in a hotspot so it can be used to generate reports (the bar charts) at a level more specific than county.

"Dr. Bob" said...

Robert Porter - I think you hit it! All advice points to being as specific as possible, and that is great given time and desire, but quite often it is a quicker "pass thru" and it is far better to get the info into a larger Hotspot than have no data at all. Thanks all for comments!!!