Birders: iNaturalist and eBird

#41

Thanks for clarifying – that makes a lot of sense. Do you think it is practical to expect this kind of data outside of eBird? I imagine it would be difficult to quantify for groups like plants, insects, and so on. Not only because there is so little effort data collected, if at all, but because the circumstances are just too variable. For instance plants only emerging in certain years, insects being present but just not visible. You would get a lot of “false absence data”.

2 Likes
#42

Insects, fungus, annual plants, bats, rodents, etc. all present their own challenges, but so do birds. The issue of accounting for detectability and observation error is something has been a huge topic in among wildlife biologists for the past few decades. I get the sense (maybe @fogartyf can weigh in on this) that other disciplines in field biology (botanists, entomologists) have been less concerned about it. Here are a couple of interesting links from the USGS and NPS:

https://www.usgs.gov/centers/pwrc/science/occupancy?qt-science_center_objects=0#qt-science_center_objects

https://www.nps.gov/olym/learn/nature/upload/OccupancyModelFactSheet.pdf

@fogartyf is right that using presence only data presents some challenges, however it is quite common now to create species distribution models without effort or absence data – e.g. using herbarium records only. Some ecologists are skeptical of these models though.
Here is an interesting recent paper that contrasts distribution modeling based on “presence-only” data with a model built from quantitative abundance data:

https://www.nature.com/articles/s41598-017-18927-1

Here’s another new paper on sampling bias in herbaria (I think this is open access, let me know if you have trouble accessing):

https://nph.onlinelibrary.wiley.com/doi/full/10.1111/nph.14855

I’m actually pretty optimistic that we can create new statistical tools to account for some of the issues with iNaturalist and other citizen science data that have been mentioned above. What these data lack in detail we may be able to make up for in quantity of observations and the diversity of observers.

3 Likes
#43

I think it depends on the scale of data collection. There is plenty of rigorous data collection (e.g., with standardized protocols, effort information, study design, etc.) by researchers working for universities, NGOs, and agencies. The downside is that it’s expensive, so it’s difficult to achieve the sort of spatial or temporal coverage that you can achieve through volunteers and community science. There definitely is a trade-off between quantity and utility of data.

@andy71 Unfortunately, I work on those issues of detectability for birds, which as you mentioned are one of the groups most plagued by imperfect detection and movement when trying to model populations. My sense is that some other subsets of field ecology have been less concerned about these issues, some rightfully so (if you are studying oak trees I think you can reliably train observers to not miss them!) while others are probably ignoring a problematic source of bias (flying insects seem particularly problematic because of their mobility and variation in adult flight period).

#44

I have not been successful in adding clickable links in ebird. I think it is good as supplemental info, such as additional pictures of the bird or even a link to an iNat observation of an insect being eaten. However, the photos in eBird are used in illustrated checklists and that would be lost. I would not be inclined to go to a link. I like the image within eBird. Of course I am not thrilled with having to double post either.

1 Like
#45

Huh, weird. I don’t do it very often so I didn’t remember how to do it. I’ve just checked and I can still get it to work (I was wondering if the new checklist format messed it up), but I don’t see what the difference is between what you did and what I did…
I learned how to do it from this help article. In the “Sound and Video” section there is some example code, if you copy that does it work for you?

<a href="http://soundfileurl">Audio Recording</a>

1 Like
#46

Hey thanks. I copied/pasted that code and voila…it works. Thanks again!

1 Like
#47

It looks like the typewriter quotation marks got replaced by left and right quotation marks somehow.

#48

I use both ebird and iNat, although I have only just joined iNat and am in the process of uploading my data.

For birds, I think both sites have their pros and cons. Ebird, for sheer volume of data collection, cannot be beaten. I guess you could say I am a heavy user of ebird.
On ebird I can upload many checklists, sometimes twice a day, and ebird gets my data on species count - how many of that count are male/female - adult - juvenile. And then I can add a code for these sightings ie feeding chicks - in appropriate habitat. Those codes then inform the collection.

In saying that, I have only been on ebird approx a year and I am noticing that in my country (Australia) the majority of users add no extra data to checklists. They just punch in species and number seen. Ebird is a powerful tool to inform conservation efforts and most people don’t use it for that. It’s being used to collate ‘lifers’ and personal bests. Like collecting Pokemon for the Pokelibrary. Which is fine. But it is also a real shame to see opportunities for broader data collection go to waste.
As mentioned above by a couple people, there is also a real lack of communication on ebird that is not good. Sightings removed off species list or hotspots and you wouldn’t know it unless you check up on your hotspots and see species count disrepancies. Which wouldn’t bother people who are just collecting bird counts, but if you are on ebird to collate your data to inform studies etc… it is a little confronting,

Still, despite those cons, ebird is the best for bulk upload of data.

iNat, well, I’m new, so I’m still learning my way. I have Vulnerable bird species I uploaded the other day and I wonder…if nobody from the community ends up ticking them off and making those Vulnerable bird sightings Research Grade, then do those sightings never get collected by researchers? That would be a shame. Because, just as with ebird, I use these two sites due to my desire to make my data collection of value to research.

#49

Oh…and ebird does not allow Camera Trap sightings. iNat does which is awesome.

#50

I’ve been using eBird for years and just started on iNaturalist. One feature that I use in eBird nearly every day that I’m not sure is available in iNaturalist is exploring sightings by area… county, state, or specific hotspot. In deciding where to go birding I might check out where the most interesting birds are being seen first.

#51

If I’m understanding correctly, you can use the ‘explore’ section on the website to do this. For example you can search for observation by a place (e.g. Roane County, Tennessee):
https://www.inaturalist.org/observations?place_id=310
by a place and a taxon (e.g. Roane County, Tennessee snakes):
https://www.inaturalist.org/observations?place_id=310&subview=grid&taxon_id=85553
or view these observations as a frequency by species grid:
https://www.inaturalist.org/observations?place_id=310&subview=grid&view=species
here’s a tutorial for the explore section in text form:
https://www.inaturalist.org/pages/getting+started#expweb
and video form:
https://vimeo.com/331279497

#52

This is one of eBird’s policies I don’t understand. They allow nocturnal flight calls that were recorded without you being present, as long as you make a separate account for them; why not camera trap records? Although I don’t think many people would object if you went and did that for one of your own camera traps anyway. Posting observations from a public feeder cam would be a different issue (does iNaturalist have a policy on that?).

#53

Yeah feeder cam type observations are quite a different beast aren’t they. In a roundabout way they are nearly a captive style of data. Unnatural and contrived. Data on natural behaviours in those settings are skewed towards Captive as opposed to Wild. You definitely wouldn’t want those obsrvations on iNat…and if you did perhaps they could indeed fall under the auspices of Captive and the member would need to tick the appropriate box.

Wild camera set ups, to capture species presence and behaviour in a wild setting, those are high value sightings. Not as high value as in person sightings, but by the sheer fact that the human is not present you can capture the species acting naturally. Naturally on its own or with others of same or different species. A lot of this you don’t get with an in person sighting. So, ebird misses out on such vital data, but iNaturalist benefits from allowing it.

It was so good to start adding some of my camera trap observations to iNat yesterday. A bit of a relief really. As I have been surveying buttonquails as a hobby, with particular interest in a Vulnerable one, and I had camera trap sightings of these vulnerable species and ebird didn’t want them.

1 Like
#54

To me it’s not so much having a camera on a feeder that’s an issue. There is not much difference from someone observing first-hand a bird at a feeder, which is still a wild bird.

I was thinking of the public feeder and nest webcams that for example Cornell has on its website, which can have hundreds of people watching from around the world. They may not know the accurate location, and many people could post observations of the same animal at once. I think eBird’s biggest concern might be that the camera doesn’t let you get a full experience of all the birds in the area.

#55

I’m pretty sure iNat doesn’t “allow” camera trap photos unless you are the “controller” of the trap… ie you set it, and you process the images. Otherwise you have no way of being sure the imprinted date/time and location are accurate

1 Like
#56

I’m fairly sure ebird have an option for netting surveys? Or perhaps netting for a specific project. I was very surprised at the time that ebird accommodates such invasive methods but does not allow sighting data from camera trapping.

#57

That’s good, kiwifergus. That would prevent observations like feeders and nest cams.

#58

unless you placed the camera, of course!

1 Like
#59

Of course. :-)

#60

Yep, there is the banding protocol which allows them to know when someone is doing that so they can exclude it from certain studies.