How do you use iNaturalist data?

[Tony Iwane]

Hey folks,

There are a lot of you here on the Google Group who use iNaturalist data in your professional lives, and I'm interested in what kind of data you use, what you use it for, and what iNat can improve on in this area. I'm especially interested in folks who are using data in a governmental/bureaucratic sense, because (as someone with an education/outreach background) I personally have little direct knowledge of what is needed in those instances. For example, someone in another thread on capitalization said they trade data with others and were hoping that the common names were all in the same style; I would have thought that in most cases just scientific names would be needed, but I really don't know, so I'm curious.

Charlie and Tony Rebelo (and others), you have mentioned this kind of data being used to make policy, and I'd love it if you can share specific examples if possible, just to give me and others a more concrete idea of how you're utilizing iNat data.

Thanks!

Tony Iwane

[Charlie Hohn]

Here's how I use the data for a state agency:

I do tons of natural community mapping. Seeing what species of plant occur in an area can be very valuable. I am also very interested in the status of plants that occur in wetlands on a geographic level - their ranges and which habitats they grow in. So I look at the range maps a lot. For both of these purposes good geospatial data is really important which is why i care so much about being able to ignore or filter observations with poor locational accuracy. 

I use it in the field. This blurrs with recreational use at times but iNat is very good for quickly gathering data to help with mapping natural communties, mapping range, or creating a quick plant list in a wetland. I don't do this on most private land for a variety of reasons but might do more if there were 'secret' observations and more integrated ways of sharing that data with landowners.

I use it to help with plant ID. Sometime outright by asking, but rarely. I track plants I key out in the scope and try to get photos of small features... track what habitats they were in... and use that in the future to help me identify new ones.

I sometimes use it for outreach or communication - for instance in a nature preserve owned by a land trust i added all my data to iNat since they had a project. I've done the same with schools. 

I use it to track invasive plants whenever I can because it is quick and easy.  I will, sadly, be marking every case i see of the stupid Emerald Ash Borer which is now in Vermont :(

I am experimenting with the idea of using iNat and Ebird to trakc some of the hard to quantify aesthetic and recreational values of wetlands. I got this idea from a professor at UVM and I think it's really naet. For instance since iNat and Ebird are well used in Vermont I can see how many independent individuals have been to a given wetland and use that data to get an idea for how popular it is for nature observation. 

There isn't an official account for my job, because my job isn't pushing iNat per say (would be amazing if they did, but some things would have to change first, or time would have to pass, but i do use it to share things with co-workers and collaborators. For instance another state employee has a project for tracking invasive plants.

There are a few more things, if you want more detail let me know!

[Charlie Hohn]

oh yeah i do sometimes share rare plant observations with Natural Heritage Inventory, if they aren't collection prone and are obscured in location. I don't do this as much as I could because the obscuring isn't perfect.

[Ian Toal]

Hi Tony:

Although I don't use iNat data at a professional level (actually at any level besides for my own learning), I do see a couple of ways that it could prove to be useful. As an example, some of our major  cutworm species don't overwinter here, but migrate annually from the south. Theoretically, the northward movement of the moths could be tracked by agriculture departments and used to advise farmers when to start looking for damage.If the data were precise enough it may even be able to pinpoint which areas of the country are likely to experience outbreaks. I suspect the latter might need a more rigorous monitoring system than simply relying on random observations. And, as Charlie says, it could be used to track the spread of Emerald Ash Borer which is now in Winnipeg/Manitoba.

I have no idea whether any of our government agencies use iNaturalist data, but it would be a good resource for these kinds of things. The more people are looking and recording, the more complete the data sets will be. Maybe I should bring it up with some of them. 

Hope this is somewhat useful, 

Ian

[James Bailey]

I have not personally used it for hardcore professional cases but I have been in touch with people who do (or try to).

iNat data is most useful for occurrence data...that's pretty vague, but that is what makes it so useful. You can pull up a map of plant observations in an area, or even add your own. There is no special fee or secret degree requirement to contribute to iNat. That means that there is a lot more data available than traditional databases. Obviously there are errors too, but those are usually not a problem.

Any research on distribution, seasons and perhaps even morphology can be done through iNat. Except with listed taxon. I was contacted by someone recently who wanted to map a listed species on iNat and collect data, but he was quite upset that iNat reduced the accuracy to 10km for obscured species, which made it useless for his very professional purpose.

[Tim. Reichard]

Tony,

This thread's big question is an interesting one in itself, even if the carry-over from the common name capitalization thread is rather minimal.

Outside my day job, I'm a volunteer wildlife surveyor working on a state wildlife atlas (https://www.marylandbiodiversity.com/). I and a few like-minded colleagues work with local parks/refuges in my area to help us build out the species distribution data for that atlas.  The usual deal with the park officials is that in exchange for arranging special access for us to survey there (we do a lot of nocturnal surveying, when parks are closed), we provide them with a copy of the inventory we find at their park.  We use iNat mostly as a tool to collect and organize our data; the identifications provided by the experts in the community are a great bonus.  It is easy to get our records into the system, easy to look at each other's observations and confirm IDs, easy to update the records with identifications when experts are consulted, easy to confer on-site about particular observations, and it's fast to export our finished dataset.

We send the park two documents of interest: the exported observation data and the corresponding taxon list.  Each park system uses their data in different ways, and I'm usually not privy to the details. One park system has asked us to survey new pollinator meadows they created, so our data will help them learn about biodiversity in a newly created habitat.  The various parks tend to request both common and scientific names for taxa, and I haven't asked them why since it has been trivial to include both. They may each have different reasons and different levels of importance.

We send the atlas a reformatted copy of the exported observation data, and they include it in their DB and data views. They prefer scientific names, but common names are included as an aid to resolving instances where their and iNat's scientific names aren't the same.  Lately my group has focused on surveying moths for the atlas, an order that needs lots of study:

https://www.marylandbiodiversity.com/quadHeatMap.php?order=Lepidoptera&

We have no expert identifiers available to us outside of iNat or other internet identification sites like BugGuide, BAMONA, and OdonataCentral. The two useful advancements iNat could do for our purposes is to continue to woo/keep more experts who are willing to help with identification and to create a clearer way to know if there is expertise behind an identification.


Tim

[David K]

Hi Tony - I have a hopefully related comment for you as I represent a volunteer entomologist group that uses iNat data frequently.  I'm the admin for

https://www.inaturalist.org/projects/moths-of-ontario

https://www.inaturalist.org/projects/ontario-butterfly-atlas

and we use iNat data in provincial atlases for each group (links to the atlases are on the project page if you are interested).  Both atlases are curated data aggregation projects, and vetted iNat data is standardized and combined with other observations from individuals as well as GBIF, eButterfly, BAMONA, museums etc. 

 

As an organization, we encourage users to consider iNat as one of the options for sharing data, and in the case of moths, it is currently our single largest data source.  One of the reasons that we like using iNat is that is reasonably easy to use for observers, and for us when we download data.  However, there are friction points that, if you could eliminate them, would make life easier for observers and data consumers.

 

There are several issues that I see from an observer’s perspective, and I posted a response on Carrie’s thread. 

 

For us, the mechanics of getting data out of iNat and getting it suitably scrubbed/formatted is bit time consuming (it’s a volunteer group with no real IT specialists). 

 

Our projects are focused on lepidoptera and life stage information is vital for us.  But there are multiple ways that observers can supply that info, through the annotations and a staggering variety of duplicative fields.  I and others have asked to have annotations available in downloads as far back as last summer, and despite Ken-ichi putting it on the to-do list, we still can’t access that simple data (see his note on July 25 on this thread: https://groups.google.com/forum/#!topicsearchin/inaturalist/annotat;context-place=forum/inaturalist/inaturalist/V06_UTD1cTU). 

 

I also wonder if you can’t populate the life stage annotation with the computer vision output.  It clearly recognizes larva for many common lepidoptera, so if the suggested ID is Monarch (adult), Monarch (caterpillar) or Monarch (pupa), that may improve data quality for everyone.  As an aside, I am very impressed with the results of the suggested IDs across life stages.

 

Another issue with data output is suboptimal location descriptions that we receive.  The input box on the web has a search box and, effectively, an output box.  What is curious is how disjointed those two can be.  See the attached example where I entered a popular camping area in a major park, both of which are google locations, however, the resulting iNat location description is bordering on useless.  We prefer specific location descriptions, so when we run into these examples, we either discard them, or look up the location for significant observations.  Neither of which represent good outcomes.  Why can’t the observer’s search term be the default location description?  This is also linked to the concept of saved locations that I mentioned on the other thread.  I believe that saved locations would not only simply observer entry, but would lead to better quality output in terms of accurate location descriptions.

 

We prefer to have the real names of observers, so the proliferation of vague handles is a nuisance and we message people with significant observations to ask for their names.  I wonder why you can’t adopt an ebird style default where people provide their real names.

 

The download options are also puzzling.  When I am downloading from a project, I do not understand why any fields other than those used in that project are presented as options.  Conversely, some fields that are added by users to observations in the project, but not used by me, are not available to download.

 

Good luck with the retreat.  I know that these are not particularly flashy upgrades, but addressing them would make your site so much better for groups interested in using detailed data.

 

David 

[Charlie Hohn]

There are a lot of issues with requiring people to use 'real' names on social media which have been discussed at length elsewhere but i will say that another issue is that there are lots of accounts tagged to organizations, not names. I haven't seen the sort of abuses of anonymity on iNat that would require forced real names anyway... So I personally wouldn't support that kind of thing.

[Sam Kieschnick]

I'm pretty lucky as an Urban Wildlife Biologist to use the tool of iNaturalist -- and use it I do! 

I use the data collected from the tool to directly work with policy makers and public land managers in the forms of city councils and park boards.  I show them the diversity of organisms that exist in their park/county/region, and most important (to the city council folks), I show them the constituency of naturalists that go to observe these organisms.

A few concrete examples of this in Dallas/Fort Worth, Texas: Rockwall Nature Preserve in Rockwall, TX.  This land was donated to the city of Rockwall (north east of Dallas), and the park board and city council looked at the diversity of organisms observed at bioblitzes in Rockwall, and they've chosen to devote the land (23 acres) to a nature preserve.  They'll be using the iNat data to control invasive, create interpretive signs on areas of high biodiversity, and design trails to go to various areas around the preserve.

The city of Arlington (300K residents) had a proclamation of citizen science day after the city nature challenge -- again, the mayor and city council recognized the biodiversity in the urban area but also how many folks wanted to go and seek it out.

As I go to any park board or city council meeting, I've got the iNata data on hand -- always.  If there are policy and land management decisions made in public areas, I show them just how many species are found there and how many folks have gone to observe them.  So far, it's been pretty dang impactful.

Also, personally, I use iNat as a training tool for myself - I really want to know the local flora and fauna, so that when I'm out with those policy makers, I can tell them the animals and plants that we run into.  iNat has drastically helped me in this.

That's how I use it professionally! 

~Sam

On Friday, March 2, 2018 at 1:10:46 PM UTC-6, Tony Iwane wrote:

[jane.e...@gmail.com]

Hi Tony, 

I am a postdoc with the the Natural History Museum of Los Angeles County and the Nature Conservancy. My research project is to analyze biodiversity patterns in the highly urbanized Los Angeles County. I use iNaturalist dataset to understand what kind of organisms can be found in LA area and how different organisms utilize different urban habitats. I have had several meetings with multiple city and county planing departments and agencies. Based on their feedback, there are the things I think will be helpful for a broader governmental/bureaucratic audience:

1) making species establishment means (native /non-native) info more explicit.  Although there is a checklist on iNaturalist website listing speices establishment means in the region of California, a lot of users aren't aware of such lists/information. Also, it might still be hard for people with limited analytical skills to linked the establishment means information with species occurrence list of a particular area.

2) customizing "area of interest'. Right now, iNat users can search a certain location and the observations within it, however, some of these locations aren't always aligned with what users have in mind. It will be helpful to make it feasible for users to define their own "area of interest". Features such as, manually drawing a polygon or uploading a shapefile of boundaries will make a difference.

3) improving location accuracy. I understand the necessity of obscuring location info. However, as a research who analyzes species occurrence at a very downscaled level (US CENSUS block /block group level), I found that location obscuration can be a big hurdle to answering fundenmental questions like how different urban factors construct urban biodiversity locally... This rings true to a lot governmental planners as well, a lot of housing development, green infrastructure or city greening projects are site-specific, iNat information can be really useful for them to predict the impact of these projects on species distribution and richness. However, the 10 km buffer range of observations often make these information less reliable and accurate, hence less used.. It would be really useful to open some possibilities for researchers to get access to less obscured datasets. 

Hope this is somewhat helpful,

Jane

[tony rebelo]

I have not been on iNat long enough, and I dont have curatorship powers, so I am not providing iNat data to anyone at present.

Besides on iNat one can easily download data oneself (with the danger of obscured localities if one is unwary): this is one of the strongest plus features of iNat vs iSpot.

On iSpot downloading data is impossible, and most parties came through to community creator for data, obtained as backups once a month.

Almost all wanted an index of the accuracy of identification, and the iSpot summed reputation was routinely provided with any data, with scores in terms of "Expert Equivalents" where 1 EE was an identification by a national or international specialist in that taxonomical group: EE scores varied between 0.001 (an ID by a total novice) and up to 16 EE.  This system was explained to data users, and it was up to them to decide on a level of acceptable identification based on their needs.

The first project - and the one we used to lure our most productive entomological expert  (Beetledude: one of the curators of the national insect collection)- was the Harmonia Ladybeetle project - in which the data were used to map spread of this invasive species in the region in a global appraisal,published recently.

There were numerous requests by taxonomical researchers wanting distribution data for their taxa: I guess in iNat you wont see these as they just download them, without any records (it might be worthwhile for publicity to require down-loaders to log a brief "reason for use" each time they extract data).  Our leverage on iSpot was requesting them to review the IDs before we gave them the data (so that it would be more accurate), and requesting them to correct any IDs based on their work.  Unfortunately my computer and backups were stolen in 2016 so I dont have my records of what these projects were by taxon and group.  

There were also numerous requests from Environmental Impact Assessment practitioners (required by law for any developments over a certain size in threatened ecosystems, or that would impact threatened species.  These generally had to be vetted by SANBI, as sensitive data required an contract regarding displaying and dispensing these data.  Two provincial conservation agencies requested annual updates of all data on Red List and Sensitive Species, in part to vet and check EIAs, but also for their own planning and reporting

Data were fed to schemes such as the National Plan for Conservation for designing optimal land acquisition plans.  How valuable these data were and what impact they made on the schemes is not known.

iNat (and iSpot  previously) are used by the SANBI National Red List programme for evaluating Red List status of animals and plants.  For plants several hundred volunteers post observations onto iNat and enter data pertinent to the programme using the quick Redlist (https://www.inaturalist.org/projects/redlist-s-afr) and detailed CREW Site Sheet ( https://www.inaturalist.org/projects/crew-site-sheet-s-afr) projects.  On iSpot these worked OK, but on iNat the projects are superb and handling this information and data stream.  There are of course complaints about the inadequacy of the iNat accuracy of ID, and that Experts IDs - often especially recruited prior to extracting data for a taxonomical group) are often not reflected in the IDs obtained in the data downloads.   There has been discussion about discarding the  IDs from iNaturalist and using that of local specialists instead. 

Both the Frog and Reptile Red List assessments mad heavy use of iSpot data (and the ADU Citizen Science data)

Data are also used for planning Red Listing and monitoring field trips, and for compiling field guides for field teams: the iNat guides are especially exciting in this regard, potentially saving hundreds of hours of work. 

SANBI has run several national projects on iSpot, and all are continued onto iNat (even those officially closed).  These include SeaKeys where data were used to designation marine conservation areas: the Sea Fish and Sea Slug atlasses were particularly well subscribed, but severely hampered on iSpot by the lack of a cell phone app.  A Fishistory (fish history) project using iSpot was abandoned because of the lack of cell phone functionality on the site.

BioGaps is the latest SANBI initiative which aims to map the Nama Karoo to provide baseline data to determine the impacts on the fauna and flora of  proposed fracking in the area.  In addition to expert groups, citizen science groups  have been active in the process (see https://www.inaturalist.org/projects/biogaps-s-afr - but most of our data are still to be migrated from iSpot - which does not have efficient ID curation tools that iNat does).

Other projects for which we routinely provide data are Roadkill to the Endangered Wildlife Trust (https://www.inaturalist.org/projects/roadkill-s-afr), monitoring of the Western Leopard Toad (https://www.inaturalist.org/projects/wlt-monitoring - all data still awaiting migration: these data are used for planning activities each year, as well as annual updates to MoUs with roads, water (esp. clearing canals) and parks (esp. mowing parks and verges) about sensitive areas and times).

With alien plants, both the City of Cape Town and the City of Durban has "spotter networks" recording instances of aliens in their areas, and both use iNat(/iSpot) for ID validation and mapping of more difficult to ID aliens.  The Early Detection and Rapid Response units of both cities, as well as the national unit, use projects to management citizen contributions to the project (https://www.inaturalist.org/projects/alien-early-detection-rapid-response-s-afr).  Attempts to get a programme started in Mozambique faltered when iSpot was unable to provide promised APIs to link to cell phone apps (considered the best way to get georeferenced pictures in remote areas that autoupload where a cell phone signal can be found, that can then be externally verified to feed back to the field workers), unfortunately, this opportunity was lost and funds reallocated.

Various other organizations and projects use iSpot (elasmobranches, seahorses, cetaceans, caterpillar rearing), but as they were relatively autonomous it is hard to evaluate how much they benefited. Only a few have so far migrated to iNat (https://www.inaturalist.org/projects/dying-fynbos-s-afr,   , https://www.inaturalist.org/projects/reared-from-larval-stage-s-afr,https://www.inaturalist.org/projects/waterbeetles-s-afr), although two new ones are in the pipeline (https://www.inaturalist.org/projects/alien-biocontrol-abundance-survey - by two independent research groups; https://www.inaturalist.org/projects/avian-influenza-s-afr). No doubt there will be more. 

Attempts to set up Climate Change Watch type programmes could not be accommodated on iSpot.  We have not explored this on iNat yet, besides posting a project to gauge background interest (https://www.inaturalist.org/projects/first-last-records-for-season-s-afr)

We are aware that iSpot data were used by large-scale commercial plant poachers from Europe to plan and execute collecting trips in southern Africa for succulent plants.  This despite having relatively advanced (much more so than iNat) obscuring procedures in place.  However, iSpot was just one resource used among many.  Allegations surface from time to time about locaities put onto iSpot being pouched shortly afterwards, but these are anecdotal and the only two cases for which evidence was provided were possibly cases of use by baboons and porcupines: undoubtedly though some exploitation does occur, but whether it is by "informers" versus CS networks has not been ascertained.  Naturally care is taken with Rhinos, but accommodating paranoia where smaller reserves wish us to remove Rhino data from their reserves when they advertise rhinos in their brochures and websites has not been attempted.   

ta

Tony