Using ebird (1 Viewer)

I've just discovered ebird, and am currently dipping into it as I go to help find sites to check out on an ongoing Peru trip. Am I missing something or is ebird totally nuts? I've just been to the central plaza in urban Cusco, and am struggling to imagine the vast majority of the 77 species ebird tells me are possible at the location, particularly the rain forest and aquatic habitat obligate species.

Check out pretty much any ebird hotspot in Peru and you'll find this. Am I missing something, or is the idea of ebird to pinpoint and name a spot on a map, and rattle of a list of birds that you might have seen within 100km? I've so far ended up in a couple of fine eucalyptus plantations in the hope of seeing the reported specialities apparently living within! Even using the bar charts to thin out the nutty stuff isn't fail safe.

I can only imagine what a dangerous I'd tool ebird can potentially be too, with birder X being happy that they saw bird Z at site A, because birder Y initially misidentified one there and put it on ebird, so that's what birders P,Q and R also figured they must have seen.

Are we in for a crazy future, or am I missing something?

Sometimes people do play a bit fast and loose with the locations and include stuff recorded over a large area. I think the ID of species is supposedly checked, though this may be done more diligently in some areas than others.

Ebird is still extremely useful, although as with any reports you should use with a measure of caution.

Larry Sweetland said:

Are we in for a crazy future, or am I missing something?

Click to expand...

Hi Larry,

I've been using eBird for 9 years and on 6 continents. I have also used it in Peru, and found it to be overall quite accurate and reliable. However, you have to bear in mind several things when interpreting the data:

1. The quality of the data is a function of the quality of the submissions and the quality of the review process. For the review process, eBird relies upon local experts volunteering their time; and coordinating a global network of local volunteers is not an easy task. Both the submissions and the review process are likely to be of highest quality in the USA and Canada because of the number of birders in those locations and the popularity of eBird there. (I believe Australia is probably in the same class). But outside of those areas the quality of the process may be hit or miss.

2. In areas frequented by bird tourists from other countries, such as Peru, you have to be alert to people trying to enter data when they aren't completely sure where they were birding, because they were on an organized tour and weren't selecting the locations themselves. So mistakes can happen. Also, they may tend to enter for “Cuzco” all the birds they saw on a trip that started and ended in Cuzco, but actually traveled quite far outside the city. One way to see if this is happening is to drill down into the data and see what mileage and other information is given in the actual list. They shouldn't be doing this, and ideally, eBird reviewers would catch it. But I imagine Peru is one of those places where the eBird reviewers are always playing catch-up, because there's a lot more data than there are local reviewers to review. I have received comments from reviewers for South America a year or more after I have entered data.

Hope this helps

Jim M. said:

Hi Larry,

I've been using eBird for 9 years and on 6 continents. I have also used it in Peru, and found it to be overall quite accurate and reliable. However, you have to bear in mind several things when interpreting the data:

1. The quality of the data is a function of the quality of the submissions and the quality of the review process. For the review process, eBird relies upon local experts volunteering their time; and coordinating a global network of local volunteers is not an easy task. Both the submissions and the review process are likely to be of highest quality in the USA and Canada because of the number of birders in those locations and the popularity of eBird there. (I believe Australia is probably in the same class). But outside of those areas the quality of the process may be hit or miss.

2. In areas frequented by bird tourists from other countries, such as Peru, you have to be alert to people trying to enter data when they aren't completely sure where they were birding, because they were on an organized tour and weren't selecting the locations themselves. So mistakes can happen. Also, they may tend to enter for “Cuzco” all the birds they saw on a trip that started and ended in Cuzco, but actually traveled quite far outside the city. One way to see if this is happening is to drill down into the data and see what mileage and other information is given in the actual list. They shouldn't be doing this, and ideally, eBird reviewers would catch it. But I imagine Peru is one of those places where the eBird reviewers are always playing catch-up, because there's a lot more data than there are local reviewers to review. I have received comments from reviewers for South America a year or more after I have entered data.

Hope this helps

Click to expand...

I agree whole-heartedly with both of your points. I have found both to be true myself. I still use eBird a lot, though only in the States. It's an excellent tool for so many things.

I wouldn't have minded if I hadn't nearly gotten myself arrested, having concluded that the only place in the plaza the Taczknowski's Tinamous could feasibly be hiding, must be under the skirts of the women who were trying to get us to have our photos taken with their baby alpacas.

Seriously though, it seems a shame that with ebird being filled with more and more info, there are some people thoughtfully pinpointing their sightings by creating a wealth of new specific hotspots, only to have a plague of people feeling the need to cloud the useful data by picking one of these hotspots to spew out what's probably what their guide told them they'd seen on the day they were somewhere between that spot and the moon.

Larry Sweetland said:

I wouldn't have minded if I hadn't nearly gotten myself arrested, having concluded that the only place in the plaza the Taczknowski's Tinamous could feasibly be hiding, must be under the skirts of the women who were trying to get us to have our photos taken with their baby alpacas.

Seriously though, it seems a shame that with ebird being filled with more and more info, there are some people thoughtfully pinpointing their sightings by creating a wealth of new specific hotspots, only to have a plague of people feeling the need to cloud the useful data by picking one of these hotspots to spew out what's probably what their guide told them they'd seen on the day they were somewhere between that spot and the moon.

Click to expand...

One thing to be aware of Larry is that some ebirds sightings are simply not reliable due to observer inexperience and error.

I witnessed several times earlier this year in Costa Rica, where clearly inexperienced birders were very eager to post sightings which were simply wrongly identified. How to weed these out is another matter altogether.


A

ebird has the biggest bird-data set in the New World and it is fairly well moderated. So you cannot, and should not avoid using it.
But ebird works with hotspots. This means that data are clustered spatially around those spots. This also means the location info on birds is far from exact. I've had discussions about ebird being used for scientific papers, and all I can say is that it is not very good for studies connecting the data with habitat / spatial analysis, at least not on the habitat / small scale, but maybe yes for 100*100kms or more.ebird is fairly well moderated.

I happen to be administrator for observation / observado, and while the dataset is smaller, a lot of the sightings are, location-wise, exact. Ofcourse they can still be flawed (if the observer is mis-ID-ing, or if the observer puts the marker in the wrong place), but many people ID fairly well and they submit sightings through the app (obsmapp or iObs). This means that you can often just go to a spot using the website + google maps, lift binoculars and see the bird (more or less, of course!).

just the example about Taczanowskii's tinamou:
In ebird, there are quite a lot of sightings around Cuzco, but they are clustered. So you have to dive into the info for each cluster and cross-search with trip reports etc. or the internet to get to the place.
You have e.g. Tambochamay, reserva privada de Santa Maria, Quebrada Quesermayo,...

In observation, you have much less sightings, only 4:
https://observation.org/soort/view/...&os=0&hide_hidden=0&hide_hidden=1&show_zero=0

1 of them, by Kees De Jager is without an exact marker. He probably entered this from his desktop at home, and didn't remember where exactly he saw it.
https://observation.org/waarneming/view/103818464

From the other 3 observations it's clear (when you click in the info icon), that sightings were entered with the app. Arjan's sighting is entered with IObs with accuracy of 10 mtrs:
https://observation.org/waarneming/view/123357567
and Robert Ketelaar's sighting is entered with Winobs with accuracy of 25 mtrs (Mark Zekhuis sighting is a copy of Robert Ketelaars's):
https://observation.org/waarneming/view/144737327
You also have the exact time of sighting.

So all in all, observation has less data, but at least it has two sightings (one this month) with exact location info.
In ebird, you are not getting there with a search, you have more data but the location info is (deliberately, designed in the database!) vague. So for me, the sad thing is, that people keep entering vague data into bird instead of exact location data in other systems...

What would I do, in the end? If I am passing by one of the observation sightings, I would give it a go. If I want a location closer to Cuzco, I will spend some more research time into the ebird data and give one of those locations a go. So in the end, I use them both, but I use ebird reluctantly because of bad data (by design), and I often curse on observations because of the lack of data (by less observers).

This speaks for itself on how reliable I think eBird is...at least for local use.

andyadcock said:

One thing to be aware of Larry is that some ebirds sightings are simply not reliable due to observer inexperience and error.

I witnessed several times earlier this year in Costa Rica, where clearly inexperienced birders were very eager to post sightings which were simply wrongly identified. How to weed these out is another matter altogether.

Click to expand...

eBird has filters which flag any rare sightings or sightings unlikely for a place or time for review; that usually takes care of people who see every sparrowhawk as a gos, for example.

The bigger question is whether to encourage or discourage non-experts from submitting sightings; that is something eBird had to decide early in its existence. Either path has plusses and minuses; they decided to encourage.

temmie said:

I happen to be administrator for observation / observado, and while the dataset is smaller, a lot of the sightings are, location-wise, exact. Ofcourse they can still be flawed (if the observer is mis-ID-ing, or if the observer puts the marker in the wrong place), but many people ID fairly well and they submit sightings through the app (obsmapp or iObs). This means that you can often just go to a spot using the website + google maps, lift binoculars and see the bird (more or less, of course!).

Click to expand...

eBird also has an app that can be used in the exact same way to provide exact locations, and has for several years. You don't have to use hotspots; it's up to the user. But exact locations also have a drawback, e.g. in parts of Asia where bird trappers operate.

Gladiator96 said:

This speaks for itself on how reliable I think eBird is...at least for local use.

View attachment 644716

Click to expand...

Actually, it doesn't speak for itself. Is it a confirmed sighting? In any event, Malta is one of the areas I spoke of where local experts are needed. But if it is confirmed, contact eBird and they should take steps to remove it. eBird users can improve data quality just as reviewers can.

[EDIT: I know the bird photographed is not a Ferruginous Duck but a domestic mallard-type.]

Jim M. said:

eBird also has an app that can be used in the exact same way to provide exact locations, and has for several years. You don't have to use hotspots; it's up to the user. But exact locations also have a drawback, e.g. in parts of Asia where bird trappers operate.

Click to expand...

That's a really good point - most people submitting records in Britain would be rightly nervous about pinpointing the locations of rare breeding birds, and recorders caution against submitting sensitive breeding records on public sites. Travelling birders need to be aware of local sensitivities too.

That said, in my admittedly limited experience of eBird, I think the 'hotspot' system does encourage the use of very broadly defined sites, when more specific data would be more useful for analytical purposes - I'm thinking non-breeding wader records from large estuaries for example. I know eBird encourage users to submit multiple geographically distinct lists, but the way hotspots are used when actually entering the data doesn't facilitate this - probably I just need to get better at using the app though!

Jim M. said:

Actually, it doesn't speak for itself. Is it a confirmed sighting? In any event, Malta is one of the areas I spoke of where local experts are needed. But if it is confirmed, contact eBird and they should take steps to remove it. eBird users can improve data quality just as reviewers can.

Click to expand...

Jim

It is a domestic Mallard type. The checklist is here and it needs to be reported to correct the error:-

http://ebird.org/ebird/view/checklist/S38423647

Gladiator96 said:

This speaks for itself on how reliable I think eBird is...at least for local use.

View attachment 644716

Click to expand...

Gladiator

The dataset is clearly very small. How correct generally is it:-

http://ebird.org/ebird/country/MT?yr=all

All the best

I started using eBird in December 2015 having been exposed to its use through Noah Strycker's Big Year. No one in a Big Year since or currently has endeavoured to record such a complete dataset save for some of the ABA competitors. Of course, doing that, there will be errors and there will be generalisations on locations. Complete datasets are its value and I've tried to use it in that way since.

Before that, personally, I have an incomplete mainly mental roll call of birding highlights. Others have notebooks on their shelves which can never really be interrogated. I wish that I had had something like it available when I started birdwatching in the early 80's. I can now see what I have seen and where for my entire year to date and indeed interrogate my sightings and photos similarly.

I use broad areas to record my local bird recording and if you are endeavouring to record more than 'highlights', I simply do not see what else is practical. It throws up anomalies and it would be crazy as a result for someone to try and see Great White Egret in Clevedon despite my seven records so far.

Temmie/Larry - do either of you attempt full bird recording and if you do, presumably you resort to species lists for broader locations??

I am a Regional Reviewer for 5 Regions - though have so far failed to do much if anything proactive. It has filters which will keep most anomalous records to a lower level.

I think that if you want to target a species, then you should look at the map of sightings and then look at the pattern of sightings within each record. You should bear in mind that the percentage of sightings is pretty irrelevant if the numbers of checklists are low. If the number of checklists is high, then appreciate why a percentage is low. I still think that an eBird dataset is still leagues further forward than say a Regional Bird Report from thirty years ago - a fair few of those on my shelves - but plainly not as good as a focussed trip report. It is an indication of the likely presence of the species in an area and then you'll need to build on that knowledge depending on habitat or similar.

As for reinforcing misidentifications, what has changed? It has happened historically through trip reports and eBird at least provides some mechanism whereby photos and similar material or descriptions may suggest misidentifications and can be peer reviewed and corrected?

All the best

In unrecorded areas it can be hit and miss. Sometimes it can be hard to gauge how common species are as a fair few of people seem to record with an X rather than a number so can be challenging to work out sometimes as a low estimate is in my mind gives a better picture of what your likely to see but again I thinks thats because different people use in different ways for example some use just to record their life list, but overall a very useful resource.

I have found ebird excellent for preparing for trips. Yes, there are anomalous records, caused by dubious identification or lazy inputting, but if I spend enough time working through the lists and bar charts I tend to pick up that certain names show up more often on species that are otherwise unrecorded (or rarely so) at a location and know to be more wary of that observer's reports.

As a tool to providing a general list for a given area it is priceless - my trip preparations involve lots of lovely lists with species number ranked according to probability. As an example: during this year's Peru trip I had to decide whether to try for Henna-hooded Foliage-gleaner whilst at Limon. Local man Lino said it was tricky and my lists bore this out. I also knew from my lists that it was much more likely at Abra Porculla. I saved (potentially lots of) time by not trying at Limon and got it easily at Porculla.

Cross referencing with trip reports is, of course, also essential.
And using Google Earth (directly through the "directions" option) can be helpful in seeing whether turning up exactly at a hotspot, as marked, will put one in suitable habitat.

Yes it has flaws but with a bit of time and effort these are generally easy enough to iron out given a bit of time and effort.

I'm a big fan of eBird having successfully migrated all of my data there, out of BirdTrack. I prefer the layout, the ability to link photos and sound recordings to individual records, and the more open nature of the data I input. Anyone can access my data and use it. However, I still feel that BirdTrack, and Trecktellen are excellent resources, and I'd encourage all birders to use one of them. There are two main reasons for this:

Many of the complaints that birders have about these systems can be mitigated by putting more good data in. A misidentified ferruginous duck in Malta might skew the picture at a very local scale. However, if everyone contributed their observations, the data would clearly show the areas where ferruginous ducks regularly occur across Europe, and if more Maltese data was in there, it would show that ferruginous duck is very rare in Malta.

Second - these things are about more than holiday planning and keeping your own records. The data will be useful to conservation. At the moment in the UK we are still struggling to get past the old county records system where data are used to produce an annual report. These annual reports are a nice read but are of limited conservation value. Having each counties data in different formats, with no related effort makes these datasets difficult to do anything worthwhile with.

So, in a world where governments are less and less likely to pay for surveillance and monitoring, it falls to us to step up and make sure our data counts for something...

Fat Paul Scholes said:

I'm a big fan of eBird having successfully migrated all of my data there, out of BirdTrack. I prefer the layout, the ability to link photos and sound recordings to individual records, and the more open nature of the data I input. Anyone can access my data and use it. However, I still feel that BirdTrack, and Trecktellen are excellent resources, and I'd encourage all birders to use one of them. There are two main reasons for this:

Many of the complaints that birders have about these systems can be mitigated by putting more good data in. A misidentified ferruginous duck in Malta might skew the picture at a very local scale. However, if everyone contributed their observations, the data would clearly show the areas where ferruginous ducks regularly occur across Europe, and if more Maltese data was in there, it would show that ferruginous duck is very rare in Malta.

Second - these things are about more than holiday planning and keeping your own records. The data will be useful to conservation. At the moment in the UK we are still struggling to get past the old county records system where data are used to produce an annual report. These annual reports are a nice read but are of limited conservation value. Having each counties data in different formats, with no related effort makes these datasets difficult to do anything worthwhile with.

So, in a world where governments are less and less likely to pay for surveillance and monitoring, it falls to us to step up and make sure our data counts for something...

Click to expand...

Do you know if county bird recorders use Ebird records in the UK, Mark? I know this happens with BirdTrack, but am not sure if it happens with Ebird. One of the advantages of using BirdTrack in the UK is that my records are automatically available to county recorders so I don't have to send reports in again.

Andrew Whitehouse said:

Do you know if county bird recorders use Ebird records in the UK, Mark? I know this happens with BirdTrack, but am not sure if it happens with Ebird. One of the advantages of using BirdTrack in the UK is that my records are automatically available to county recorders so I don't have to send reports in again.

Click to expand...

As far as know, they dont. What I'll do is export all of my local data from eBird into a spreadsheet, and email that to the county recorder.

ebird is a fantastically useful tool, but like anything else, it isn't perfect and requires a bit of judgement.

With respect to the comment about dodgy Maltese Mallards: if you are a knowledgeable local birder, instead of poking fun at the record, why not contact ebird admin, and offer to become a reviewer? They would probably welcome you with open arms. As a reviewer, you can decide which species in your area should have records "quarantined" until accepted by a reviewer, either because they would be unusual or sensitive breeding species.