iNaturalist as iSpot

It is now nearly a month since we took the decision to migrate from iSpot to iNaturalist. We: the iSpot 5000 team (those iSpotters with more than 2000 observations and 1000 identifications and comments) and about 150 iSpotters who undertook the migration survey.

The biggest issue by far is getting used to iNaturalist. The philosophy is so different. iSpot was simple: it was about getting an ID and evaluating the accuracy of the ID. Everything centred around that. Yes there were interactions, and badges, and the taxonomy surfer, and the quiz and projects: but the philosophy was getting an accurate ID and evaluating it. To that end the reputation system with its expert and knowledgeable reputation was core. Outside experts regularly contributed to the IDs making the IDs really powerful. It was clearly designed for old eyes and computer novices – simple and logical.

iNat is quite different. It seems to centre more around the user and promoting the user. IDs are almost incidental: a world expert in a group is just another anonomous user with no acknowledgement or reputation of their status. There are tools to allow you to make dozens of IDs every hour, but your contribution to the ID is the same as the world expert and a green schoolkid: IDs by world experts to species level are drowned by well-meaning contributions like “plant”. And it is busy! There are observations and species and users and places and projects and they all contribute to the milieu and philosophy. The word “species” means at least four different things: any taxon, species only, species and subspecific taxa, and taxa that are research grade. Starting from a place and selecting a species gives different layout and results from starting from a species and selecting a place. iNat was clearly designed for young eyes and computer gurus: everything clicks, fonts start at almost-invisible serife 1 pt and anything you might want to do can be found on the page – somewhere, and not necessarily using the wording that you might expect! Click and discover!

It will take quite a while to make the transition. Clearly iNat has lots of features that we only dreamed about on iSpot (I still believe that the iNatters visited the iSpot forums and implemented most of our iSpot wish list). The four items we will miss most are the reputation (iSpot was leagues ahead), the interactions (should be easy for iNat to implement), society badges and keys. But iNat has features that are supercool. The ID tool is magic. The guides are really cool (and in time could include keys). The observation fields and projects work really well (once one gets used to them). And who knows: with a bit of training the machine ID might be able to identify southern African plants without giving us lists of irrelevant North American ones.
And we already have our plant dictionary half on. All the current names. Synonyms are coming soon with common names. And then our animal names ...

So if iNat is going to be home, what have we got? What are we starting with?

Let us look at the iNat stats for southern Africa as at 25 November. You can see these for yourself at https://www.inaturalist.org/observations?created_d2=2017-11-25&place_id=113055 (although this stats was for half way through the day):

39 374 Observations (= verifiable; vs 43 453 tota - I must find out what "verifiable" means)

The top observers out of 1002 observers are (over 500, ***= “iSpotter”) –

gawie 2,394 ***
tonyrebelo 2,046 ***
johnnybirder 1,664
calebcam 1,579
markuslilje 1,500
Joachim 1,460
zanskar 1,216
i_c_riddell 1,186 ***
florem 1,064
muir 830
snidge 822
csavy 771
tapaculo99 762
tonybenn 652
shauns 617 ***
graeme 581
colin25 514 ***

The top identifiers (our of 1086 identifiers) are (over 800) –

johnnybirder 18,247
jakob 11,748
calebcam 6,319
alexdreyer 3,162 ***
ldacosta 3,089
tonyrebelo 2,142 ***
wildnothos 1,746
wouterteunissen 1,656
john8 1,591
henrydelange 1,510
kokhuitan 1,301
martingrimm 1,197
colin25 1,101 ***
karoopixie 987 ***
alexanderr 985 ***
joachim 980
alanhorstmann 966 ***
shauns 931 ***
beetledude 862 ***

But the crux of the issue are the species records.

There are 6032 species (= verifiable (I need to find out what that means) out of 6295 total): In descending order:

Elephant 649
Giraffe 388
Impala 359
Zebra 338
Kudu 249
Egyption Goose 243
Warthog 242
Lion 241
Hippo 196
Baboon 193
Wildebeest 189
White Rhino 165
African Wild Dog 162
Guineafowl 158
Waterbuck 152
Nile Crocodile 150
Buffalo, Vervet, Spotted Hyaena, Ostrich, Nyala, Leopard, African Fish-Eagle, Cheetah, Springbok, Hadada Ibis, Gemsbok, Lilac-breasted Roller, Rock Hyrax, Blacksmith Lapwing, African Penguin, Black-backed Jackal, Laughing Dove (101), Bushbuck, White-backed Vulture, Leopard Tortoise, Steenbok, Hartebeest, Southern Yellow-billed Hornbill, Grey Heron (84),Southern Rock Agama, Black Rhinoceros, Grey Go-away-bird, Hamerkop, Southern Masked Weaver, Common Bulbul, African Striped Skink, Black-necked Agama, Bontebok/Blesbok, Southern Fiscal, Red-winged Starling, Bateleur (70), Angulate Tortoise, Brown Fur Seal, Klipspringer, Long-tailed Cormorant, Brown-hooded Kingfisher, African Darter, Bush Duiker, Eland, Pied Kingfisher, African Jacana, Southern Ground-Hornbill, Kori Bustard, Pale Chanting-Goshawk, Painted Reed Frog (60), White-fronted Bee-eater, Southern Red-billed Hornbill, Smith's Bush Squirrel, Nile Monitor, Fork-tailed Drongo, Sacred Ibis, Ring-necked Dove, Kelp Gull, Cape Dwarf Gecko, Red-billed Oxpecker, Puff Adder, Woolly-necked Stork, White-faced Whistling-Duck, Rainbow Skink, Saddle-billed Stork, Pied Crow, Banded Mongoose, Cattle Egret, Tawny Eagle, Flap-necked Chameleon,

Baobab 51 – our First plant

  • Another 55 species
    Yellow Pansy 34 – first invertebrate

  • Another 2 vertebrates – second invert Redveined Dropwing, then Welwitschia Bug!!!!! Then another 4 vertebrates: then Citrus Swallowtail
  • Another 36 species (incl Honeybee & Blue Pansy, Painted Lady, Broad Scarlet, Plain Tiger)

Next plant #2 is Pillansii templemanii (28 observations)

& 40 more verts to next invertebrate.

(if you are interested in plants: #3 = Welwitschia, #4 Pelargonium cucullatum, #5 Black Wattle, #6 King Protea)

Conclusion: It seems that iNat southern Africa has been used mainly by big game and bird tourists up until now.

Of course, it is possible that all the plants are sitting there unidentified: So what are the totals?

Birds 11,709 obs, 793 spp, 427 Identifiers, 348 observers
Plants 8,882 obs, 2,685 spp, 272 Identifiers, 328 observers
Mammals 7,108 obs, 195 spp, 386 Identifiers, 492 observers
Insects 6,351 obs, 1,310 spp, 320 Identifiers, 332 observers
Reptiles 2,285 obs, 284 spp, 204 Identifiers, 301 observers

Chelicerates 649 obs, 143 spp, 100 Identifiers, 168 observers
Amphibians 602 obs, 82 spp, 81 Identifiers, 135 observers
Fish 408 obs, 210 spp, 53 Identifiers, 49 observers
Fungi 321 obs, 85 spp, 60 Identifiers, 59 observers
Molluscs 227 obs, 107 spp, 46 Identifiers, 55 observers
Protozoans 5 obs, 2 spp, 3 Identifiers, 4 observers
&
Unknown 431 obs, 2 spp, 9 Identifiers, 88 observers

So iNat currently has about as many plant species as the Cape Peninsula, out of 22 000 plant species for s Africa.
So that is our baseline. That is what we are starting with. What are we going to make of this site?

Our next step is to design our iNaturalist southern Africa user interface. If you would like to participate, or have some cool ideas, please contact me. (Just remember we are at the southern tip of darkest Africa. Fibre-optics stops somewhere in the Mediterranean. Fast lines drown in the Atlantic and Indian Oceans. And in remote areas, emails may be almost as slow as ordinary mail, which is an achievement only possible in Africa. Our landing interface needs to take this into account).
Marion has some ideas which I will be circulating soon. If you feel strongly, contact me now, before ideas start stablizing.

PS: Would you like to help with the unknowns? You can:
https://www.inaturalist.org/observations?place_id=113055&subview=table&iconic_taxa=unknown

Lähettänyt tonyrebelo tonyrebelo, 27. marraskuuta 2017 16:05

Kommentit

"verifiable" means that your observation can be verified; other users know you observed it by the photos or sounds you have provided. If it does not have photos or sounds, it is given "casual", because there is no way of verifying that observation.

Lähettänyt calebcam noin 4 vuotta sitten (Lippu)

Oh: so there is no point in flagging these observations as:

Based on the evidence, can the Community ID still be confirmed or improved? : No, it's as good as it can be

Lähettänyt tonyrebelo noin 4 vuotta sitten (Lippu)

"It seems to centre more around the user and promoting the user. IDs are almost incidental." I disagree with this statement. I think getting correct IDs is still the central part of iNaturalist. It would be nice to have some kind of badge system that acknowledge professional experience but I still think experts are recognized and sought out by those looking for help with tough IDs. If you can get an observation down to family iNat shows you the top identifiers of that group which are often members who specialize in those organisms.

Lähettänyt vermfly noin 4 vuotta sitten (Lippu)

"If you can get an observation down to family iNat shows you the top identifiers"
Hmm: I would have expected experts to get far better than family level.
From iSpot most experts would be far under the radar: they are seldom among the top identifiers by numbers until one gets to tribe or genus level for insects or genera and species for plants. But because they are obvious on iSpot with society badges and reputation one can easily see this. Analysis of this on iNat would be quite difficult: one would have to check all the profiles, and sometimes it is not obvious whether users are experts (or not). As you say "if expert help is sought with tough IDs" then those experts will be very low end users. It would be nice to get some figures ...

Lähettänyt tonyrebelo noin 4 vuotta sitten (Lippu)

I think you misunderstand me. The experts are making IDs down to the species but iNaturalist will show you the top identifier of things in the family if you get your ID of an observation down to that level. For example if you look at a record of the mushroom Suillus pungens you will find the top identifiers in the lower right of the page. Two of the top 5 are authors of one of the most comprehensive ID guide to mushrooms of the Pacific Coast.

I think having society badges and possibly a way to post one's experience more visibly would be a very nice addition.

Lähettänyt vermfly noin 4 vuotta sitten (Lippu)

I think I do understand you.

Mushrooms are different though: there are relatively few identifiers. Often one only knows the family after one knows the genus, so one does not have lots of contributions to the higher levels and fewer as one gets to lower taxa - as with many plant and insect taxa.

(Interestingly, many of our mushrooms on iSpot were identified by Europeans (Americans were not common on the site): most of our macrofungi are "invasives" from north temperate regions (or more correctly, commensals of north temperate tree species planted for silviculture and parkscapes) - so we anticipate that Americans and Europeans will provide many of our fine-level IDs here as well).

Lähettänyt tonyrebelo noin 4 vuotta sitten (Lippu)

I see there are currently 6 comments posted, but how does one get to see the content, I'm assuming they're not personal.

Lähettänyt tuli noin 4 vuotta sitten (Lippu)

cancel that. I've found them

Lähettänyt tuli noin 4 vuotta sitten (Lippu)

I see it fine? and not personal.

Lähettänyt calebcam noin 4 vuotta sitten (Lippu)

sorry, didn't see your last comment!

Lähettänyt calebcam noin 4 vuotta sitten (Lippu)

on the ID:user issue
I think the format is unhelpful, not only is the potentially very long repetitive and bulky list of agreements a bit of an eyesore. The way I see it if one had a situation with a rather lengthy list of agreements possibly further inflated by a number of sheep going along with the flow and then an up-to-date expert comes along and sees they're all talking old hat and a recent revision has changed things his/her correct ID will be lost in the clutter and although significant will possibly remain unnoticed. Or am I as usual missing something?

Lähettänyt tuli noin 4 vuotta sitten (Lippu)

No: you have summarized it perfectly. The idea on iNat is that then everyone must then agree with the expert (if they know that they are an expert - how?, and if they care, and if they are still around).

But we are finding this problem with our subspecies IDs. Most people dont have the fieldguides or papers to subspecies and therefore cannot agree, but want to contribute to the level of species that they are happy with. So the identifications stay at species level, when in fact they are adequate for subspecies level and knowledgeable people have done the work, but the system is just not coping with this. That is what I meant when I said that the system caters for the user rather than the ID. Some sort of reputation system is needed. But it needs to include reputation outside of iNat, otherwise it will not address this problem.

Lähettänyt tonyrebelo noin 4 vuotta sitten (Lippu)

A few comments & suggestions:

observations without an initial ID are unlikely to be found by users focussing on certain groups. So this batch of observations
www.inaturalist.org/observations?place_id=113055&subview=table&iconic_taxa=unknown
would immensely profit from having a coarse ID, even if something obvious as "dicots", "arthropods", "animals" is added. Ideally done by the user her/himself when uploading observations, but such IDs can be also added by any other user www.inaturalist.org/observations/identify?iconic_taxa=unknown&place_id=113055

observations from Southern Africa are indeed dominated by vertebrates, but that is to some extent due to a lack of expertise in other groups. For instance, those knowing the SA flora are invited to skim through plant observations, which is likely to shift the stats above quite a bit (apart from taxonomically more balanced new observations): www.inaturalist.org/observations/identify?place_id=113055&taxon_id=47126.

in the calculation of the Community ID, sequence matters, so a sequence of IDs from "plants" -> "dicots" -> "Crassulaceae" -> "Crassula younameit" is quickly refining the Community ID, and in the specific example, only 1 user would need to confirm the species ID to have the Community ID showing the latter.

I highly recommend thoroughly exploring iNat's functionality. Some stuff is quite a bit hidden. Create a test observation. Play around with it. Invite other @users to screw the ID up. Examine what is needed to bring it back on track in terms of IDs. Assess the functions of Data Quality Assessment, Annotations, Observation Fields. Loads of stuff, and many things certainly quite different from iSpot...

Hope you like it here!

Lähettänyt jakob noin 4 vuotta sitten (Lippu)

Welcome to the iSpot folks! iNat does take a bit of time to get used to, but as Jakob says, after playing around and practicing quite a bit, I hope you feel welcomed and encouraged to curate the regional data.

I have felt overwhelmed when I examine the entirety of the iNat database, but when I take off manageable bites (from a specific region — in my case, north central Texas), I find that a great group of knowledgeable ID’ers come along and help with the curation. Lots of mistakes are made in the process, and it can look a little messy, but as with all natural history collections (physical and digital), they’re eternally changing.

I think that’s one of the more important things that I’ve gained from iNaturalist: a sense of community. I published a few papers while in academia that literally no one read (I couldn’t even convince my parents!), but with iNaturalist and citizen science, I find that more folks are engaged and actively participating after using the tool. It’s got more of an impact on public policy and nature appreciation for the general public. I’ve found that to be pretty powerful. Just my two cents!

Again, welcome! :)

Lähettänyt sambiology noin 4 vuotta sitten (Lippu)

Thanks Jakob and sambiology.
Dont think we came here without testing! A team of over 30 of us tested the site before deciding on it as an option.
Cumulatively, I dont know all we did, but I have done 2300 observations, 2700 identifications (including reviewing all Proteaceae (from souuthern Africa, and our genera worldwide), Bruniaceae, Penaeacea, Stilbaceae). We have - with the iNat team - uploaded or confirmed 55 000 current plant taxa for southern Africa and 47 000 synonyms. We have already refined - with the help of iNatters - quite a few of our bird, mammal an herp ID's to subspecies.

We have already given two courses in using iNat to potential users, and hope to roll out a few more regionwide (at this stage mainly to CREW in South Africa, but I have a commitment that if anyone can get 15 people I will come and give a course there).

And we are negotiating to migrate our 400 000 iSpot records here. There is no better way of learning a system - and figuring out the strengths and weaknesses of both - than trying to fit in data from one web site into another.

And in case you missed it: anyone can help identify our unclassified observations: they are here

Totally unknown:
https://www.inaturalist.org/observations?place_id=113055&subview=table&iconic_taxa=unknown
(or if you want to use the Identify tool to speed up the process then use this: https://www.inaturalist.org/observations/identify?verifiable=true&place_id=113055&iconic_taxa%5B%5D&iconic_taxa=unknown)
At present there are 330 of these.

These ones need support or finer IDs
https://www.inaturalist.org/observations?lrank=genus&place_id=113055&quality_grade=needs_id&subview=table&verifiable=any
or if you prefer ID mode:
https://www.inaturalist.org/observations/identify?verifiable=any&lrank=genus&place_id=113055&quality_grade=needs_id
Beware though: there are 8700 of these (a fifth of the observations for the region) - but so use the group filter to narrow it down to your group).

Lähettänyt tonyrebelo noin 4 vuotta sitten (Lippu)

I‘m sorry - I didn‘t make it clear that my comments are meant as an encouragement for newcomers who might be reading this. There has been obviously a lot of testing by many former iSpot users over the last weeks!

Excellent to have vastly enlarged expertise on iNat together with often stunning observations from a global biodiversity hotspot.

Lähettänyt jakob noin 4 vuotta sitten (Lippu)

Great work Tony and the Inat team. All the best for the inverts.

Lähettänyt fubr noin 4 vuotta sitten (Lippu)

Much of what you say was also true for iSpot. The IDs were dynamic, experts input is essential (but given emphatic recognition on iSpot) and making wrong IDs elicited rapid and useful responses. I guess it is true on all systems where interaction drives identifications.

At this stage we have largely uploaded the southern African plant dictionary, barring common names. There are still many issues with synonyms - our southern African synonyms are now sitting alongside identical current names (in our opinion due to outdated global taxonomy, but due process needs to be followed). Next will be the animal dictionaries.
And then setting up our own network: that will be done in parallel with the dictionaries, but probably take much longer for the MoUs to be completed.
And then we want to migrate all our iSpot data onto iNat: that is under negotiation with the Open University, although negotiation appears at this stage appears to be the OU stalling by hiding behind an EU restriction on providing personal data which they used in Sept 2014 to withhold email addresses for southern African users on iSpot from our data dump. If they have not resolved this over 3 years they are unlikely to do so imminently. We need that data to train the AI system: with over 20 000 species of plants it wont be as simple as for New Zealand. But already some reasonable IDs are appearing among the hundreds of North American names the AI proposes for our plants.
From our point of view the most amazing thing about iNat is that they listen and interact and do things!

Lähettänyt tonyrebelo noin 4 vuotta sitten (Lippu)

And 2.5 months later is has grown by 1/3. Quite impressive

Lähettänyt alexanderr melkein 4 vuotta sitten (Lippu)

25 November: 39 374 Observations (= verifiable; vs 43 453 total)
12 February: 66 497 Observations (veri, vs 72 152 total)

Leaderboard:
Elephant 866
Impala 566
Giraffe 538
Zebra 478
Egyption Goose 422 (rising)
Kudu 356
Warthog 355
Lion 327
Baboon 285 (rising)
Hippo 276
Wildebeest 270
Waterbuck 248 ("upset" below here)
Buffalo 238
Guineafowl 229
Vervet Monkey 222

Plants & Inverts:
Redvein Dropwing 85 (first invertebrate)
King Protea 67 (first plant)
Baobab 65 (position 127 - was still plant #1 only 4 days ago ...)
Honeybee 64
Christmas Butterfly 64
Quiver Tree 62
Julia Skimmer 53
Common Sugarbush 44
Common Pagoda 41
Black Wattle 39

Lähettänyt tonyrebelo melkein 4 vuotta sitten (Lippu)

The fact that the observations has increased dramatically for Southern Africa is encouraging, indicating (I believe) that the "old Ispot users" have/are settling in the new platform and taking hands with the "old INat" users to build the S.A community.
But, we still believe that the amount of observations is driven by Identifications. This trend we believe has shown it's self over and over again and most probably a common human trait - people photograph something because they want to know what it is. From this it is also evident that if knowledgeable users are active one tends to get an increase in specific taxa observations thus expanding the visual (pictorial) knowledge base of the site and ultimately spacial distribution data (a key input from Citizen Science).

From this our question is rather how many new species were added per month in Southern Africa apposed to how many observations of a specific species.

Could one make a monthly project/list for new species added???? Damn, maybe we need to attend one of your courses to figure out what this platform can do.

Lähettänyt fubr melkein 4 vuotta sitten (Lippu)

Yes: looking at 17 to 21st March for courses in Gauteng area. Watch the Facebook page for latest details: https://www.facebook.com/iSpotNatureZA

Lähettänyt tonyrebelo melkein 4 vuotta sitten (Lippu)

With the export and some pivot table/excel skills it is quite easy.

For the Project: Herps of southern Africa:

year no of species obs no of unique species added
2011 70 70
2012 173 202
2013 62 228
2014 176 310
2015 122 343
2016 109 355
2017 343 468
2018 210 495

Lähettänyt alexanderr melkein 4 vuotta sitten (Lippu)

So in 2018 in 1.5 months we have already the second highest number and 2/3 the 2017 number of observations

(I am assuming a "species obs" is an observation of a herp (- the only alternative is that it is an observation ID'd to species level).

Lähettänyt tonyrebelo melkein 4 vuotta sitten (Lippu)

species obs is the total species observed for that year only (irregardless of what was previously added)
unique species added is the total species for all years (so 27 new species for iNat in 2018)

sorry, a bit confusing

Lähettänyt alexanderr melkein 4 vuotta sitten (Lippu)

Amazing data, thanks for clearing up the confusion. what is semi unbelievable though is the 2011 data that tends to indicate that all observations were of different species. Can that be ?????????

Lähettänyt fubr melkein 4 vuotta sitten (Lippu)

No - both figures are species counts. The term "obs" is confusing and should be ignored.
So the first figure (after the year) is the number of species for the year, and the second figure is the cumulative number of species over all the years.

Lähettänyt tonyrebelo melkein 4 vuotta sitten (Lippu)

Please confirm below that you were notified of this comment.

Lähettänyt tonyrebelo melkein 4 vuotta sitten (Lippu)

Yes it works well. Notification Reads "tonyrebelo added a comment to a post ("iNaturalist as iSpot")"

Lähettänyt fubr melkein 4 vuotta sitten (Lippu)

Hi Tony,
Re the change to Dimorpotheca, do you have the citation details for the change, please?

Lähettänyt johnsteel yli 3 vuotta sitten (Lippu)

I presume that you mean Dimorphotheca fruticosa, or all of them? Does not matter: same reference. Please see the appropriate flag. The justification and references are there under comments.

Lähettänyt tonyrebelo yli 3 vuotta sitten (Lippu)

Sorry to be a nuisance Tony, but I still can't find it (not too good on computer stuff! - what is "appropriate flag"?)

Lähettänyt johnsteel yli 3 vuotta sitten (Lippu)

Sorry: it was not a flag but on the taxon swap: here - https://www.inaturalist.org/taxon_swaps/34986#comment-1926988
copied it here to be more accessible:
https://www.inaturalist.org/taxa/76661-Dimorphotheca-fruticosa/flags
(oh: if you go to Osteospermum fruticosum, inactive taxon the taxonomic swap is shown - click it to view)

Lähettänyt tonyrebelo yli 3 vuotta sitten (Lippu)

Lisää kommentti

Kirjaudu sisään tai Rekisteröidy lisätäksesi kommentteja