As a long-time AviSys user, the news of its recent closure presented an unexpected dilemma. I love eBird, but the prospect of migrating all of my records into eBird had seemed too daunting to attempt, and so I had adopted a hybrid approach to managing my records. I used AviSys to keep my world, ABA, and some state lists, and also for offline data input when necessary; I would then export those records to eBird and otherwise used eBird for more routine record entry. Although a bit cumbersome, this method worked, and allowed me to take advantage of each tool’s strengths.
With no ongoing support and the possibility of no future taxonomy updates, using AviSys is no longer feasible. I considered my options. One was to migrate to another installed software tool and continue with a hybrid approach. The other option was to bite the bullet and export all of my data into eBird.
After I finished crying in my beer, I decided to try exporting to eBird first. eBird has a handy “Manage Import” that allows users to delete an entire import. So, if the process went poorly, I could always delete the import and go forward with plan B. It turned out to be easier than I thought, and now I’m happily monogamous when it comes to my digital bird records. In going through the process, I learned a lot, so I’m hoping that sharing my experience here will benefit other AviSys users.
PreparationFirst, to quote the cover of a famous fictional book, “Don’t Panic!” This process requires a fair amount of time, but you don’t have to do it all at once; the process is such that, once you’ve completed the initial import into eBird, you in bits and pieces whenever you have the time. In my case, I had approximately 30,000 records of approximately 2,600 species stored in AviSys; it took me an entire afternoon and a little bit of work the next day to complete the migration.
Second, understand that eBird and AviSys are designed somewhat differently, so you will need to account for that. In particular, eBird is not designed to be a listing tool. Each record in eBird is intended to correspond to a specific sighting at a specific location and date. I started using AviSys long after I started birding, so I had a number of “list only” records that I had put into AviSys on a fictional date that were in there simply for purposes of having those species reflected on my AviSys list totals. I was pleased to learn that eBird offers a similar workaround. If you have these sorts of records in AviSys, the important thing is to make sure you can uniquely identify them once they are exported. AviSys attributes and keywords are NOT searchable or filterable on eBird, so you will need to be able to identify these records either by a unique date, location, or combination thereof. In my case, I dated them all either January 1, 1930 or January 1, 2029 (which, when I started using Avisys so many years ago, admittedly seemed a long way away!). More on this below. eBird prefers you to use January 1, 1900, but if you haven’t already used that date, it’s easier to change it in eBird after the fact than in AviSys.
Lastly, be very careful as you proceed. eBird does not have an “undo” feature. If you inadvertently click the wrong button and make a sweeping change or delete something, you won’t be able to bring it back and may be faced with redoing a large portion of your work.
Step 1 – Exporting Your Data from AviSysThe first step is to export ALL of your data into an Excel file to import to eBird:
- From AviSys, choose “List Records” and then “List.” The sequence doesn’t matter.
- From the list of sighting records, choose “Export” and then “eBird Export File.”
- You’ll get a popup warning you that “This will be an unstructured upload to eBird – RAW data.” Click ok.
- You’ll be presented with the eBird Export Manager screen. Choose a name for the file that is obvious; I used “AviSys Export.csv” but it doesn’t matter as long as you can remember the name and the file location.
- UNLESS you have been extremely meticulous about recording precise numbers of every species seen on every list input into AviSys, choose “No” for both “All Species Seen Were Recorded” and “All Records Reflect True Counts.”
- Click “Export.”
Step 2 – Importing Your Data into eBird
- Log into eBird and choose “Submit Observations.”
- Choose “Import Data.”
- Select your export file (see above) and choose “eBird Record Format (Extended).
eBird will take all of your data and create a unique “checklist” for each unique date and location combination. eBird will initially ask you to “Submit” all of the checklists that it deems “ready to be submitted.” This means that eBird recognized all of the taxa as currently valid and all of the locations as corresponding to existing eBird locations. Go ahead and submit these; this is the easy part.
Step 3 – Cleaning Up Your Imported ChecklistsAt this point, you will see something like this:
The remainder of the process involves fixing all of these issues. Please note two things that apply to both categories of fixes. First, do not accidentally click the “delete” link beside any entry unless you really mean it. There is no “Are you sure?” prompt, so you will not get a second chance if you delete one of these items. You will have no choice but to resubmit all of those entries if you need them.
Second, please note that, once you have choose the species or location fix, you MUST click “Apply Fixes” at the bottom of the screen. If you don’t, eBird may not retain those fixes and you will have to re-select them. If you have a large volume of fixes, I recommend periodically clicking on this button as you work through the fixes. This is essentially “saving” your work, and also the “fixed” species and locations will be removed from the list, so you’ll see visible signs of progress along the way as the list of fixes becomes smaller.
Step 3a – Fix SpeciesI suggest fixing species first, because it’s considerably easier and less time consuming. You’ll see something like this:
Like me, you may be puzzled by the “Rock Pigeon” matches “Rock Pigeon” line. I subsequently discovered that eBird has separate entries for the old world, native Rock Pigeons and the new world feral Rock Pigeons. However, this screen won’t readily allow you to sort this out, unless all of your records happen to relate to one or the other. In my case, I had a mixture of both, and unfortunately, I was unable to find a “global” way to fix it. I eventually had eBird list all of my records of Rock Pigeon and had to manually change all of the new world checklists to “Rock Pigeon (Feral Pigeon).” [Team eBird note: If you change these before upload, the process is much easier!]
For the rest, you just select “choose species” and select the appropriate taxa for your sighting. Often, the apparent mismatch is the result of a name change; unless the common name and the scientific name from AviSys both match the current eBird taxonomy, eBird will ask you to “fix” the species. For example, the scientific name for Sooty Shearwater is now Ardenna grisea, and eBird wasn’t able to make that connection. The vast majority of my species “fixes” fell into this category. [Team eBird note: If you don't include the Latin names in your upload, and the Common Names match, you won't be prompted to correct species like Sooty Shearwater. Just be sure that your Common Name is the right one according to the eBird Taxonomy!]
If you have instances where the records may actually relate to more than one eBird taxon, then you will have to pick one or the other here and then go fix them checklist by checklist. Obviously, you should “default” to the one that is likely to appear on more checklists.
If you are unsure what to do, simply click “choose species” and start typing the name into the search field. For example, I knew that Black-mandibled Toucan had been lumped, but was not sure if the combined species name. I typed “Black-mandibled” into the search field and immediately saw “Yellow-throated Toucan (Black-mandibled) - Ramphastos ambiguus ambiguus/abbreviatus.”
Remember, when you are done, you must click “Apply fixes” at the bottom of the screen to save your work.
Step 3b – Fix LocationsFor me, this was the part the required the most time and effort. eBird will present you with a view sorted according to the number of affected checklists. For my records, location-fixing fell into two general categories: specific locations and general locations. I had a number of records that were identifed only at the city, county, state, or, in a few cases, country level. Those are the easiest to fix. When you click on “choose location” you’ll be taken to a screen asking you to match your Avisys location name to an eBird location name. For general locations, choose “Find it by city, county, or state” and then choose the appropriate level of specificity.
The specific locations will likely require more work, both because the level of effort to associate them with an eBird location is greater, and because there tend to be a lot more of them. If you have an existing location in your list of “My Locations” that you know corresponds, then you simply select that location. Similarly, if you happen to have precise GPS coordinates, you can search by that method and identify an appropriate hotspot or create a personal location.
If neither of those options is available, then you have to use “Find it on a map” and attempt to search for the right location. Note that the “search” feature on the map screen only searches Google maps location; it does not search eBird hotspots. Map searching can be a fairly tedious exercise for this reason; there might be an eBird hotspot location with the name you need, but if you don’t know where precisely to look and it doesn’t correspond to something searchable via Google maps, then it may take some time to find it. On several occasions, I found myself doing Google searches in an effort to find more information about a birding location to inform my search. For example, the town of Afluente is a well-known birding stop in northern Peru near Abra Patricia. Finding it on the map was quite a challenge, and only by working my way out from other locations that I knew to be near it was I able to find the right hotspot (which is called “Bosque de Protección Alto Mayo--Llanteria/Afluente vicinity (km 389-391)).
Another technique that might help is to go back to the “Find it by city, county, or state” tool, enter as much information as you know, and make sure the “Show me Birding Hotspots in the area” option is checked. Then, on the next screen, you can scroll through a list of hotspots in that area in the hopes that one may look right. I also found a number of instances where the eBird hotspot used a different name than the one by which I knew the location; sometimes this was due to language differences (e.g. Spanish language names in South America), and sometimes this was due to alternate names for the location. For example, there is no eBird hotspot for “Rudong,” a well-known birding location in China, but there is a hotspot for “Yangkou,” which appears to be the tidal flat area in the Rudong area that is most well-known to birders.
Note that some map searches on eBird only show hotspots for the area you search. If your search was for a particular county, in some cases you may only see hotspots in that county. If you move around the map to another county and don’t see the right hotspot, it may not be because there isn’t one. Try a different search focused on that neighboring county.
Step 3c – Cleaning Up Your “List Only” RecordsIf you have some “list only” records, then your next step is to find them, change all of their dates to January 1, 1900, then hide them from eBird output (because they are not technically “valid” records). If you adhered to my earlier advice to ensure that you can uniquely identify those records either by date, location, or a unique combination thereof, then you need to locate and edit each one of those checklists. As previously noted, eBird will create a separate “checklist” for each unique location/date combination. If, like me, all of your “list only” records were dated January 1, 1930, then all you have to do is go to “Manage My Checklists” and then sort by “Date” so that your oldest checklists appear first. Then, edit each checklist to change the date to January 1, 1900 and to check the “Hide from eBird Output” option. That option automatically “invalidates” the records such that they do not show up on maps, bar charts, etc. However, they DO still appear on your lists and your list totals.
Step 3d – Taxonomy CleanupThe last cleanup step is to review all of the major taxonomy updates since the last time you updated your taxonomy in AviSys. Probably the easiest resource for this is the annual eBird taxonomy updates. Go back as far as you need, and then skim through the relevant updates, looking for places where you might have lumps or splits that might need to be tidied up. Unlike when you input a checklist directly, when you import a checklist into eBird, eBird does not alert you to records that are rare or unusual; either you or an eBird reviewer will have to find them.
If you are not sure if an update might have impacted you, go to your “My eBird” screen and click on “Your Life List.” You can then do a text search to see if that species appears on your eBird list. If it does, you can click on that species to see all of your records for that species, and evaluate from there whether you need to make a change. If you are unsure, one good starting point is to check the relevant checklist to see if that species is flagged as “rare.” If it is, and in particular if you see the split taxa also listed as not rare, you probably need to update your list. If you miss a few, then likely the eBird reviewers will catch it, but it may take time. In my case, some reviewers contacted me almost instantly.
Optional Step – Removing Duplicate ListsIf you have using both AviSys and eBird together, then you will likely have a number of duplicated checklists as a result of this global export. It is relatively easy to go to “Manage My Checklists,” list all of your checklists, and scroll through them looking for duplicates. Typically, if you follow eBird protocols, your “global export” checklist will be listed as an “incidental” observation and your original list will be listed by some other protocol, so it is usually easy to identify the duplicate and delete it. I listed this step as optional, because I have not done it yet and have thus far seen no problems from it.
Hopefully, at the end of this process, you’ll find that you can finally put AviSys to rest and use eBird solely for data entry going forward.