Research Collaboration: Method for supporting non-experts to label in 'unpopular' taxa

OK, so what do you envision as the application for this training? i.e., once your novices have been trained, what data sets they then be using their new skills to label? Who and/or what sorts of projects would benefit from having access to this crowdsourcing potential?

Since you say it wouldn’t necessarily be intended for iNat users specifically, would the purpose be something like that sketched by @cthawley above – as a streamlined way for scientists to train research assistants for a particular project? The idea being that iNat’s body of verified observations would serve as the basis for training, and once the training model has been developed, it could be supplied with a new photo set and adapted with little work to a different group of organisms?

I think in most cases this would require some initial input from the scientists about identification traits rather than expecting users to intuitively figure out the differences.

As an example, take oil beetles (Meloe). There’s a good, concise overview of British species and their differences here: http://johnwalters.co.uk/research/oil-beetles.php

These are large, distinctive beetles which tend to be somewhat “underlabelled” compared to many other beetle groups. There are only a handful of species in the UK and continental Europe and I suspect, given proper guidance and feedback about identification traits and a suitable set of reliably verified photos, it would be reasonably easy to train a novice to identify the local species.

However, if merely given a set of photos without being told what to look for, I suspect most people would end up frustrated and confused, because they would try to use an obvious trait, like color, to distinguish M. violacea and M. proscarabaeus and would find it difficult to understand why this doesn’t reliably work. Or they might think that the differences in the shape of the antennae mean different species, when in fact this is a sex-based trait.

Meloe also undergo substantial changes in their appearance during the course of their adult life – they expand to at least twice their original length through eating. A novice who has learned to recognize them in their engorged state would (quite understandably) be likely to assume that a freshly emerged adult is a completely different organism altogether. Without an explanation about why the appearance is different, they may resist accepting what the computer is telling them (“the computer must be wrong”).

These factors aren’t unique to Meloe – lots of organisms have more than one form (subadult/non-breeding/breeding plumage in birds, sexual dimorphism, etc.).

Another issue that complicates identification is that the required traits aren’t necessarily always visible in photos because the person taking the photos has to know what to photograph. A lot of the observations on iNat are less than ideal for this purpose, and I suspect that a lot of the observations that are identified may not be correctly identified (see: lack of IDers). So any iNat photos used for training people would probably need to be verified first.

I am not suggesting that the idea won’t work, but these are some factors I see that are important to consider from the outset.

5 Likes

Even credentialed experts can be bluffing, you know. I am very well credentialed as a biologist (PhD, Fellow of the Royal Society of Biology and more besides) but that doesn’t make me an expert at identifying a whole host of plants and animals outside the narrow field of my particular expertise … but, my, were I to wave the letters around people would believe in what I was only guessing at. On the other hand, credentialed scientists probably know better than many where to find the necessary data to make an ID and that is more than half the battle … credentialed or not. Lots of non-biologist types out there who really know their stuff and I defer to their expertise.

5 Likes

Even for those, I’ll use it, with caveats. Say were talking moths, for instance. If it suggests a species or a genus, I will look at the taxon pages for multiple species in that genus. If I think the suggestion is plausible based on that, I might go with the genus (not the species unless I find that it is very distinctive). Or, I might go with the family that contains that genus. If it suggests a family, I’ll probably go with that family. It may not be ideal, but given the shortage of identifiers, I find that leaving it at “Lepidoptera” is simply not a viable option.

2 Likes

unfortunately triggers CV to say Seen Nearby. Which prompts me to go to ‘that’ distribution map, and tidy up where I can.

And it is necessary to hash these out. One of the obstacles that we keep coming up against, no matter how many ways this is discussed is

and there seems to be resistance to changing that. Just about every field guide has a whole chapter at the beginning explaining all the terminology, because nobody knows how to explain identification without it. If I have to remember what a trochanter or an epiproct are, or exactly which vein is R4, before I can make any headway – well, I’m not going to make much headway. And I say that as someone who has a postsecondary-level background in biology and training in the use of keys; how do we expect “non-experts” to get a grasp on this?

Simpler, perhaps, but not viable in the long run, especially as the number of new observers grows faster that the number of expert identifiers. I think that was the reason behind this thread. The question is, can we bring this out of the ivory tower where “non-experts” can access it? When every field guide feels the need to make a case for “why you should learn the scientific names,” that’s a pretty strong indication that people outside the ivory tower, on the whole, find them intimidating.

So far, from what I’ve seen, almost everyone who attempts to address this issue wants to approach it as, “well, we’ll train them to think like us.” And maybe that’s the problem.

4 Likes

Then there are the gracious helpful comments
For example
https://www.inaturalist.org/observations/192091034

I think there are 2 intertwined issues here. The first is that there are a bunch of specific morphological bits and pieces that you may need to find on an organism and unless you are going to have a bunch of “does it look like this picture or that picture?” questions, those bits and pieces need names/labels. One of the reasons I like the keys in the Borror & DeLong (& whoever is currently updating it) insect book is that they have a lot of drawings and will reference them in the keys so you can often ignore the specialized vocabulary. However, it’s textbook-sized (and priced) and generally only gets you down to family/subfamily. And I suspect that producing an image-based key is going to be outside the time and money budget for a lot of taxonomists, especially if they can’t use existing images. I would be curious to know if the lack of mark-ups on field guide photos is due to a lack of resources or feeling that the photos will be less attractive if they have circles and text imposed over them.

To me, the iNat equivalent of this seems like it would be taxon photos that have been marked up to show the identifying characteristics. I’m aware of a few of these for plants where observers have taken the initiative to collect local relatives/look-alikes and mark the differences. And I have favorited a caterpillar observation that the observer put segment labels on so I can refer other people to it. But as far as I know, there isn’t a way for an expert to mark up someone else’s photo and get it back into iNat to use as a taxon photo.

The second issue is people assuming that other people are familiar with the terminology or being pedantic about it. For a while, I was using “abdominal legs” to describe prolegs, especially when talking about differences between caterpillar and sawfly larvae, because I associate that mix-up with less general insect knowledge. But I stopped after I got sick of comments from people (usually not the observers) pointing out that those were really prolegs.

1 Like

For grasses, I don’t think you can avoid using specialized terms like spikelet, glume, lemma, palea, ligule (unless you want to go with older terminology that referred to lemmas a fertile glumes) because the average person doesn’t see these (except sometimes spikelets), doesn’t think about them, doesn’t have words for them.

However, one can label them, if one has the room. This plate is from our Field Guide to the Grasses of Oregon and Washington, which gives one page to each species of grass. (The upper third of the page is text, using technical words as needed.) Our Field Guide is a BIG, heavy field guide for over 400 taxa. (Imagine providing this level of detail for a whole flora!) The book includes an identification key because we feel one is needed for identifying these often similar species.


Copyright Cindy Roche!

7 Likes

I value minimizing technical terms, and sometimes I can do it. For example, in most cases “hairy” can replace pubescent, villous, hispid, strigose, and the many other terms for plant hairs. However, sometimes one needs to distinguish between the hair types and then the technical terms matter.

We use technical terms to provide precision and brevity. Often one needs a whole phrase to replace a technical term. That impedes reading and makes books, website entries, etc., larger. Remember that until relatively recently, illustrating books and journal articles was really difficult and expensive, so biologists had to develop a language that means exactly what we intend it to mean. Technical terms can reduce confusion – when they’re understood. (The book “Plant Identification Terminology” by Harris & Harris illustrates each term with a drawing. This book belongs on the desk of every English-speaking person trying to ID plants.)

By the way, Flora of the Pacific Northwest, 2nd edition, consists of keys for over 5,500 taxa. It uses technical terms and supplements these with very small line drawings at each key lead. The usefulness of the drawings varies, but overall they’re very helpful.

To sum up: There are things we can do to reduce the dependence on technical terms and we should do them, but overall this vocabulary can’t be avoided.

6 Likes

Actually, the main thing I noticed in my grass identification course was that, after you have done all the dissecting scope work and keyed it out with all the technical bits, you step back and see that they actually look very different visually. Bromus hordaceus is visually very different from Bromus diandrus. Once you know them, you can tell them apart by sight. Different genera are even more visually different; once you know Melinis repens or Lolium perenne, you aren’t going to mix them up with either of those Bromus species. The problem isn’t that the technical details under the scope are necessary; it is that those are what taxonomists know best how to use.

I agree that after you know what to look for you can often avoid looking at the details. But how do you communicate that look, that gestalt? That’s hard. Often you need the details to divide the plants into the groups that you then learn.

Also, easy though it is to separate Bromus hordeaceus from B. diandrus, other brome species complicate the picture. Bromus hordeaceus resembles B. secalinus and B. commutatus which are even harder to tell from each other than they are to tell from from B. hordeaceus. B. japonicus and B. squarrosus are really hard to tell apart (lemma margins help a lot). Bromus diandrus looks nothing like B. commutatus but only size distinguishes it from B. sterilis. (All the bromes mentioned in this paragraph are common annual weeds; you can’t argue that you only see one of a group of similar species.)

In a big complicated group like Poa (bluegrasses) there are some gestalt differences – Dr. Robert Soreng, world Poa expert, can often ID photos that my colleagues and I can’t tell apart without seeing the little details! But since we don’t know understand the gestalt differences, we are happy to find any non-overlapping difference in anther length, leaf sheath closure, hairiness of the lemma keels, or other technical difference that allows us to make the call. As we begin to learn gestalt differences, we remain happy to have these details to fall back on for individuals that don’t seem clearly one or another species.

I work on fine-leaved fescues a lot (Red Fescue and the dozens of species that have been called Sheep Fescue & some others). To ID them, not only do I measure lemmas (and this means I need to know what lemmas are) but I cut cross sections of their leaves to see the arrangement of fibers (sclerenchyma bundles). Do you think I do that because of some mad attachment to leaf cross sections? Or a wish to exclude others from fescue ID? No. It takes time and it’s barely possible to do in the field with sharp scissors and a hand lens. I look at leaf cross sections and other people do it too because that’s the easiest way, pretty much the only way, to distinguish some of the species. I wish it weren’t so, but it is. Yes, there are gestalt differences between some of the taxa, complicated by maturity of the flowering branches (= inflorescence) but if you see many of these species you come to prize the technical differences. Clarity! (Relief!)

Some groups of plants are easy. Some become kind of easy after you learn them but are hard to learn. And some are irredeemably difficult, technical. I wish all groups were easy and non-technical! But I don’t get what a want (again!). Sometimes one has to delve into the technical stuff or just leave the observation at “spider” or “slug” or whatever. Unsatisfying but real.

5 Likes

I don’t know why I keep disagreeing with you. We do agree that the language can and should be simplified. (Just try keying Reed Canary Grass in an older agrostology book – if you don’t know what it is, you misinterpret a basic point and can’t key it.) We agree pictures help. You seem to think that technical language can always be avoided and I disagree, strongly, but I think we agree that books and websites for plant ID fail to the extent that ID can’t happen. (attempted sort-of apology)

2 Likes

Today i was thinking about this, and it seems to me that grasses – maybe not all – are amenable to the Peterson System. Looking in my Peterson Guide to Pacific States Wildflowers, I have it open to a page where the heading says, “5 petals united as trumpets, bells, or flat wheels.” There are seven species illustrated on the page, in four families. I turn to the next page, and the heading says, “5 petals; many stamens; pinnate leaves with stipules at base: cinquefoils.” That page illustrates 9 species, all of which were Genus Potentilla at the time of publication.

What makes the Peterson System so effective is that it groups species morphologically rather than phylogenetically. Sometimes the two are one and the same (like the page of cinquefoils), but not always (like the page with representatives of four families). Phylogeny has its uses, but identification is not necessarily one of them; morphology is much more likely to be useful for identification purposes.

It occurred to me today that grasses could be grouped by Peterson-style morphological headings, which would then allow the illustrations to emphasize the differences between similar-looking species. So Cynodon, for example, could be on a page with a heading describing its appearance like the spokes of an umbrella, which would also show Digitaria and Eleusine which also look like spokes of an umbrella. The little arrows which are characteristic of the Peterson System would then be used to point out the key differences.

2 Likes

I agree that morphology is much more important than phylogeny for plant identification. I get so frustrated with people who use tiny little traits when there’s a big one almost slapping you in the face! I mean, the purpose of an identification key or a field guide is to get the user to a name, not to teach morphology.

I think that although you could use something like the Peterson system for grasses, and that might be a very good thing for many species, you’d have many, many pages equivalent to Fall Warblers or Female Ducks, but maybe worse. I say this as a person who spent part of today trying to distinguish one small, alpine, bunchgrass-type Bluegrass that lacks cobwebby hairs from about four other small, alpine, bunchgrass-type Bluegrasses that lack cobwebby hairs.

5 Likes

Reminds me of discussion at https://forum.inaturalist.org/t/should-curators-have-a-critical-point-of-view-when-changing-inat-taxonomy/46339/97

2 Likes