How does a hotspot work on Ebird? (1 Viewer)

[THE_FERN]

can make the data more meaningful

...

it has limitations in terms

nope: less meaningful. When you round like this, you lose information. For example, any info about habitat preference (perhaps by time of day). There's simply no reason to do this.

Birds seen at hotspot X as opposed to birds seen at GPS location x.xxxxx y.yyyyy.

Nope: nothing stops there being similar system to the one they use when you submit a list. Before you submit, it does a database search for all species seen within (I think) a 20 km radius of the point and flags any species on your list which isn't in that set. So you could have the same functionality attached to a button called "what species are near me" [We know that it's doing a database search because it'll flag species even if it's a non-hotspot and a place you've never been to before.]

species data to plan for a visit

As everyone does... And in the old days we did the same with the "where to watch birds" books, the pagers, the hotlines, the online sightings pages, the... i.e. whatever preparation you do you should be aware of possible mis-identifications (even in the guide books) and of potentially new occurrences.

could be detrimental

Nope because they suppress these.

I'm happy to use it as it is, and be aware of these limitations

well so do I, but, please, please, please comply with personal information regs and good sense and let me download the GPS tracks I've submitted. You have them all in a postgis database after all...

 

[jmepler]

eBird

After having dipped on what now must be dozens, if not hundreds, of birds , solely due to the hotspot system of eBird, I am becoming quite irritated when people start defending it with made up arguments.

eBird is for scientific research, mainly phenology, population, and distribution trends. It is not a twitching tool, even though many people use it as such.

Even if eBird would track your exact GPS location when you enter a particular bird, that doesn't tell you where the bird is, just where the person was when they entered the bird.

... please comply with personal information regs and good sense and let me download the GPS tracks I've submitted. You have them all in a postgis database after all...

There is an extension for Chrome that will allow you to download a gpx file of your track. I have not tried it yet, but it sounds like it would do what you are asking for, just not within eBird.

 

[THE_FERN]

Even if eBird would track your exact GPS location when you enter a particular bird, that doesn't tell you where the bird is, just where the person was when they entered the bird.

of course. That's always true for all recording media, including cameras. We've already made this point...

There is an extension for Chrome that will allow you to download a gpx file of your track. I have not tried it yet, but it sounds like it would do what you are asking for, just not within eBird.

Interesting: thanks for that. It should be part of the main ebird site of course. And it's possible to get the data out of the phone if you know what you're doing. At some point I'll compare the 2 methods. Would want to check the extension doesn't lose anything (it's working off the data sent to google rather than the original).

I've also not tried creating "artificial" tracks and uploading them. One of my gripes with the app is that it's too easy to swipe up in Android and dismiss a session part-way through. This stops the recording of the duration and the GPS although of course you can continue to add species. On many occasions what was one session ("checklist") got broken into 2. I'd like to reassemble the parts to a whole.

 

[qwerty5]

I am completely lost here. Don't people who use eBird simply click the birds seen in the app as they go? On a smartphone, which nowadays is guaranteed to have a GPS in it? What stops eBird from retaining and showing this data, bird by bird? There would be zero additional effort from observers needed for that, just a change in the software.

Because it would require the user to input each bird right when and where they saw it, something I doubt hardly anyone does. I tend to input birds in groups, so I'm not walking around looking at my phone rather than the birds. This means it would be hardly more accurate than a hotspot.

As far as tracks go, I am pretty sure researchers can access them with the permission of eBird.

Any system is only going to be as good as the users who input the data. eBird wants to be used by as much of the public as possible, so the system is designed to be user friendly. A more complicated system will either result in less participation, or simply inaccurate data (and there's plenty of that already).

Now eBird isn't perfect of course, but I don't think adding to the knowledge and time required to use it will help.

How does having exact positions of birds "muddle the map"? If anything, hotspots do that, because with hotspots, there is no way to tell in which areas the birds are common and in which they are rare, because the icon for "one observation 50 years ago" and "20 observations per day" is the same - one hotspot. And if users really prefer to see birds bunched up in hotspots on the map, there is aboslutely nothing preventing eBird from having an option for this while displaying the map - the software behind the map could easily lump observations into hotspots on request.

If I'm looking at a species map for a local area, I don't want to have to click on a million different points to see the observations. And in my local area, a specific personal location doesn't really add anything to the accuracy, because there are accurate hotspots for pretty much everywhere. The problem comes when people don't use them, or don't use them accurately. This happens an astonishing amount, which I think goes to show that eBird doesn't need more complexity for people to mess up.

Finally, when looking at ranges, population trends, etc., I'm not sure what difference it makes rather the bird was reported exactly where it was seen, or with a checklist of other birds at a nearby hotspot.

I must say I don't like how the hotspots are set up exactly. I think having hotspots defined by areas would be great. I really like the way that eBird shows National Wildlife Refuges, with all the hotspots and personal locations brought together, allowing for an easy review of what birds are there at what time of year. I wish they would expand this to all places that have multiple hotspots within the same park, wildlife area etc.

 

[opisska]

Again: how would it hurt, if the map allowed to show the observations both lumped into hotspots and spread out? This would be easily implemented with a simple toggle on the map - you would lose nothing, I would gain a lot. Because if I now want to see where a specific bird is common, I have to click on a hundred hotspots if the area is "hotspot dense". If there were individual observations, a quick look at the density of points would reveal the same information within a fraction of a second to me.

And the same with data entry: no barrier of entry, no added complexity. Just retain data that already exist but is willingly discarded. I don't care that some people add the birds less accurately and that some people don't add them on the fly. There are some that don't and having and showing this data is strictly an improvement, there is nothing lost.

 

[Kirk Roth]

From a twitcher standpoint, the great thing about hotspots is that they come with their own datasets - you can check a particular place for abundance of species over the course of a year, and check for "ease of finding" at that spot. Obviously one must still put in the "birding work" - they aren't just handed to us!

It seems that some are more interested in individual data than population data - in areas where twitching via eBird is common there are often notes provided for the specific targets for chasing, if the observer is so inclined. Comments on this forum from the past several years seem to indicate that this is more of a problem in Europe than the Americas, perhaps?

 

[pppppvvvv]

...

nope: less meaningful. When you round like this, you lose information. For example, any info about habitat preference (perhaps by time of day). There's simply no reason to do this.

Yes, I like the idea of this, but you'd need to improve the accuracy of the data input for it to have any value.
I can only see complete chaos if you tried to get this level of accuracy from the way the data has been gathered to date.
It would work if people signed up to specific projects within ebird, to collect quality data such as accurate times and locations of the species of interest in an area. "The ebird New Forest Crossbill survey" for example.

 

[jurek]

Yes, I like the idea of this, but you'd need to improve the accuracy of the data input for it to have any value.
I can only see complete chaos if you tried to get this level of accuracy from the way the data has been gathered to date.
It would work if people signed up to specific projects within ebird, to collect quality data such as accurate times and locations of the species of interest in an area. "The ebird New Forest Crossbill survey" for example.

Somebody already pointed that it works fine in European systems: observation.org and inaturalist.org.

Ebird is becoming more and more burdened by it's own technological legacies. The list system, the hotspot system, lack of privacy, lack of non-bird wildlife.

It will become even more bizarre with time as places on the ground will change. A forest hotspot will half become a housing estate, a road hotspot will be made 3 times longer etc.

For science purposes, too, hotspots of wildly different size and shape create lots of room for subjectivity. Suppose some bird lived in two hotspots: one 4 times smaller that another, and died out in one. Did it decrease by 50% or 20%?

 

[THE_FERN]

Somebody already pointed that it works fine in European systems: observation.org and inaturalist.org.

Ebird is becoming more and more burdened by it's own technological legacies. The list system, the hotspot system, lack of privacy, lack of non-bird wildlife.

It will become even more bizarre with time as places on the ground will change. A forest hotspot will half become a housing estate, a road hotspot will be made 3 times longer etc.

Already happened (of course). E.g. Barnes Reservoirs in London which only existed up until 2000: wetland centre thereafter. Very importantly, and clearly of geographical significance (!!!), is that the hotspots are several '00s of m apart so you can easily distinguish between them. Like all other hotspots, they bear no sensible, predictably relationship to the area they go with (they're not at the centroid, north western corner or whatever such that you might be able to guess where they would be given a map of the location)

 

[qwerty5]

Somebody already pointed that it works fine in European systems: observation.org and inaturalist.org.

I know very little about iNaturalist, but from what I have seen it looks like a total nightmare. In general anything except a list system is incredibly undesirable to me. Lists are how birders have been keeping track for decades, and it seems like the best system when looking at large-scale population and range trends (which is entirely the point of eBird).

For science purposes, too, hotspots of wildly different size and shape create lots of room for subjectivity. Suppose some bird lived in two hotspots: one 4 times smaller that another, and died out in one. Did it decrease by 50% or 20%?

Population is going to be based off of numbers reported, not how many hotspots it was reported at. But as I said above, I do like the idea of making hotspots area based. The problem with this in the eyes of eBird may be the large number of knowledgeable local reviewers to define and update the area boundaries.

One eBird feature I would like (and use) is the option to drop a pin for a specific species, as an easy way to show the exact location of rarities or other birds of interest, without losing the checklist style.

 

[Jim M.]

Somebody already pointed that it works fine in European systems: observation.org and inaturalist.org.

iNaturalist is not a "European system." It was started in and is based in California.

 

[THE_FERN]

I know very little about iNaturalist, but from what I have seen it looks like a total nightmare.

Suspect you're alone in that. It's great, especially for things like plant phenolog

Population is going to be based off of numbers reported, not how many hotspots it was reported at.

Maybe, but actually most ebird stuff I've seen has been about probability of encounter. That's only loosely related to numbers if the species flocks together

One eBird feature I would like (and use) is the option to drop a pin for a specific species, as an easy way to show the exact location of rarities or other birds of interest, without losing the checklist style.

I'd like it exactly the other way around: a set of individual observations which I can choose to tag together as one "list" or "trip" or "checklist".

 

[Jim M.]

I know very little about iNaturalist, but from what I have seen it looks like a total nightmare. In general anything except a list system is incredibly undesirable to me. Lists are how birders have been keeping track for decades, and it seems like the best system when looking at large-scale population and range trends (which is entirely the point of eBird).

iNaturalist is great and has collected tons of data on many taxa where there was previously no or little public data. But eBird is better for birds because, with complete lists, it shows data about the absence of birds--not just their presence.

 

[qwerty5]

Suspect you're alone in that. It's great, especially for things like plant phenolog

I haven't ever tried to use it, so it might be better than I think. It seems to me it's heavily photo-reliant, and I don't own a camera. I can also obviously see the appeal for those who want to list more than birds.

I'd like it exactly the other way around: a set of individual observations which I can choose to tag together as one "list" or "trip" or "checklist"

I can't imagine using anything but a checklist during a morning in migration with 80+ species and hundreds of individuals. Unless of course you are satisfied to not record every bird (which I'm not).

 

[Jim M.]

I am completely lost here. Don't people who use eBird simply click the birds seen in the app as they go? On a smartphone, which nowadays is guaranteed to have a GPS in it? What stops eBird from retaining and showing this data, bird by bird? There would be zero additional effort from observers needed for that, just a change in the software.

How does having exact positions of birds "muddle the map"? If anything, hotspots do that, because with hotspots, there is no way to tell in which areas the birds are common and in which they are rare, because the icon for "one observation 50 years ago" and "20 observations per day" is the same - one hotspot. And if users really prefer to see birds bunched up in hotspots on the map, there is aboslutely nothing preventing eBird from having an option for this while displaying the map - the software behind the map could easily lump observations into hotspots on request.

After having dipped on what now must be dozens, if not hundreds, of birds , solely due to the hotspot system of eBird, I am becoming quite irritated when people start defending it with made up arguments.

You misunderstand the data that is in eBird. First, the app is only a recent development in eBird in the last few years--there is tons of data entered before the app became prevalent--and even now birders won't necessarily use the app and won't necessarily use it immediately when they see the bird. Second, the whole idea behind eBird is not to record exact locations of birds (which the app isn't designed to do), but to record survey data for areas--surveys that indicate the birds present in the area and the birds not present in the area. The information about absence is equally if not more valuable to researchers and can only be recorded if you do a complete checklist for an area--which eBird has always encouraged birders to record.

 

[qwerty5]

Maybe, but actually most ebird stuff I've seen has been about probability of encounter. That's only loosely related to numbers if the species flocks together

Most of the products (like bar charts and the species maps) are based on probability since their purpose is primarily for people to find birds.
However, their scientific products (like abundance and trends maps) are also based on numbers.

 

[Paul Chapman]

I remember the days when you would visit a country based off 4 photocopied trip reports if you were lucky and no one would dream of reporting the run of the mill to bird news services..... Maybe a Where to Watch guide would help or some conversations with mates.

Systems have advantages and disadvantages. Be thankful for gen & allow people to use them how best they work for their own recordings.

I put my Royal Cinclodes today under a Hotspot. Three covered an area of several hundred metres. How would people want these reported....? The previous records from there had the GPS co-ordinates from someone far more diligent but you would still need to search and find your birds.

https://ebird.org/checklist/S147812934

https://ebird.org/checklist/S145527749

That said - someone has created some comedy hotspots on my local patch. No idea who and that is notwithstanding the fact that I am a local reviewer.

Ho hum.

All the best

Paul

 

[THE_FERN]

Unless of course you are satisfied to not record every bird (which I'm not).

Unclear the one thing follows from the other. I do try to record all individuals but in big flocks I estimate numbers, same as everyone else. There's no reason why the location of that sighting should be subsumed within a longer track. And if I don't record the bird at the time I'm liable to forget it.

The information about absence is equally if not more valuable to researchers and can only be recorded if you do a complete checklist for an area--which eBird has always encouraged birders to record.

At the risk of stating the obvious, the difference here is a measure of effort. In ebird the gps track and length of time of the checklist serve as that. There's no good reason not to separate the 2 logically distinct things: a sighting of a bird (save where, when that happened) and the effort (path you walked) in order to see it.

Actually these days many use maximum entropy methods for estimating species distributions. One of their benefits is that they work without "known" absences---just use presence information. (Absence is rather difficult to estimate and under-recorded---even if we leave aside birders' tendencies)

 

[THE_FERN]

Most of the products (like bar charts and the species maps) are based on probability since their purpose is primarily for people to find birds.
However, their scientific products (like abundance and trends maps) are also based on numbers.

Numbers are difficult to estimate from ebird for many reasons. One is that lots of birders including some very good ones just record presence ("X"). Even for well-watched areas you have to prune and massage the data a lot. I'd want to be persuaded this doesn't materially affect the outcome...

 

[opisska]

I would like to note that I started this line of discussion in this thread in response to a post where someone describes how eBird reviewers create hotspots just to hide individual locations of rarities. So this has nothing to do with surveys, bird frequency data etc... It's just done because someone doesn't like having too many points on a map.