Cursed sunfish observations in the catacombs of iNaturalist
Recently, I wrote a simple query to examine sunfish (Lepomis) observations sitting at Genus level. My intent was to see how many hybrid sunfish were still unable to reach Research Grade status, and ID them if I was able. Boy, was I not ready for the horrors I was about to witness.
Its a complete warzone. Infinite longears without consensus due to splitting, greengills galore ID'd as every sunfish known, mystery YOYs, and of course the obligatory high-reflection-low-angle-surface-shot special all too common for other fish observations on this site. I think I'll try and parse through them but it'll take time, patience, and other user assistance due to dissenting IDs...
The query for any brave souls: https://www.inaturalist.org/observations/identify?quality_grade=needs_id%2Cresearch%2Ccasual&order=asc&taxon_id=49592&hrank=genus&lrank=genus&photos=true&place_id=any
תגובות
Curious on your thoughts on future increased splitting. It would seem that the average person is losing their ability to have any sense of what a species is (along with their ability to Id anything to species). This wouldn't really matter but for the fact that we use phrases like "preservation of species" that seem intended to have meaning to the average person. What would it mean to preserve species if the average person can no longer distinguish between any of them?
I'll try to jump in and help out! I wonder if it would be possible to post this in the iNat fish working group project. Seems like what that project is about.
@stockslager we (humans) generally agree that the biological species concept previously used to describe the foundation of taxonomy is flawed, and therefore we are beginning to unravel it. There are many other species concepts but we seem to gravitate the most towards the evolutionary species concept for new taxonomic definitions. We use these conceptual models and taxonomic rankings to help us organize life…which is a fool’s errand. Life is incredibly complex and can’t be hierarchically classified; think of all the different “rule breakers” there are for species within these different concepts. We have self-reproducers, frequent hybridization, and geographically isolated but otherwise similar looking individuals. Taxonomy aside, I like to think of species as existing on a continuum; some fish are just more different than others because they’ve diverged longer ago. For example, a smallmouth bass and a lake trout are much different because they have a very far-removed common ancestor, whereas a smallmouth bass and a redeye bass have a much more recent common ancestor in comparison. Without a PhD in genetics or phylogeny, that’s how I interpret speciation and how I wrap my head around phenotypic variability.
With that in mind, my thoughts on splitting are relatively neutral (I could care less what they call a largemouth bass, it’s still the same fish it was a decade ago and 1,000 years ago) except when it complicates management or preservation like you mentioned. In fact, I had this exact concern regarding the cutthroat trout reorganization. There are some examples of taxonomic revisions complicating management, and a prime example would be Gila spp. in the west. There has been fighting over taxonomy for what seems like decades (1, 2, 3, 4), and inevitably this just reduces resources available for conservation and management. I think at some point we have to accept that the Linnaean taxonomic system is flawed but it might make more sense to follow given the complexity of life. I also agree that a lot of this splitting results in alienation of the public, who are mostly disinterested in jargon and minutia associated with this sort of thing. TLDR; it's important to recognize the diversity of life without being a pedant.
Keep in mind this is also written from a fish-y perspective…not all of this may be true for fungi fans or bird nerds, but I guarantee you at the very least all the species concepts fold in similar ways.
I really enjoy your user bio, by the way.
@mkvam nice! I'll see if I can add this to a discussion there. There are a lot of fish a Genus level that really shouldn't be.
@operculum_ben you are absolutely correct! If it doesn't work out I did message one of the admins of the project about it. I fear that all of this work will need to be repeated once the Caddoan and Redstripe Longears are formally recognized.
Oh my gosh! Thank you for such a thoughtful response. Makes me want to go back to college and take all the 101 courses instead of picking one thing!
In theory all the Longear mess should be handled with taxon splits from the curators, save everybody a bunch of hassle. I learned from Pantosteus that trying to do it one at a time is insanity. Things went much better with the Bowfin split.
As for the piles of hybrids, mystery YOYs, and distant low detail shots - that's a lot of work but definitely some interesting stuff to see in looking through them. Some of those mystery hybrids are worth the digging alone
One thing I do wish we did on the fish side of iNat (since subspeices are out of fashion) would be to use species complexes or subgenera or something similar for groups of closely related species. Especially the groups with wide introgression. Maybe that doesn't make sense for fish, I'm no taxonomist. I do know those intermediate categories are in wide use of the plant, fungi and insect parts of iNat where it does seem to be useful at times
@nf4t I've never thought too much about species complexes, except for maybe cisco species in the Great Lakes. What additional functionality do species complexes provide? In my head, subspecies and species complexes are just different ways to characterize the same thing (e.g., variability that's extreme enough to notice but not extreme enough to categorize at a higher level). But maybe I am unaware of their utility.
I understand there are some problems with the subspecies concepts (1, 2; I have not read these but I do see that some of the key points I've heard regarding fish are brought up) but I'm not sure if more species splitting is the correct solution.
To give some sense of scale for the amount of hybrids:
https://www.inaturalist.org/observations/14654365
https://www.inaturalist.org/observations/17683453
https://www.inaturalist.org/observations/17755137
https://www.inaturalist.org/observations/18176457
https://www.inaturalist.org/observations/18527868
https://www.inaturalist.org/observations/19087665
https://www.inaturalist.org/observations/19087787
https://www.inaturalist.org/observations/19087944
https://www.inaturalist.org/observations/19088229
https://www.inaturalist.org/observations/20036057
https://www.inaturalist.org/observations/21012340
This is just a snapshot (like, first 200-250) unreviewed observations in that query.
@operculum_ben - basically those are all taxa groups that sit between genus and species to represent a clade of species within the genus. Functionally it just provides an extra layer of granularity without using the species/subspecies method. Really it is all the same thing to me, like you said this is trying to put dynamic systems into discrete boxes (which is an impossible task).
For a fish example - Shorthead Redhorse, Pealip Redhorse and Smallmouth Redhorse are all really closely related to each other (there is probably another on the Atlantic coast but that's for the future to tell). Way back Pealip and Smallmouth were described as unique species (like today). Then they were lumped into Shorthead as subspecies. As we've gone away from subspecies they've been elevated back to species but they are still very closely related, within the zones of introgression things are messy because there is a lot of gene flow and identification becomes dodgy. From an iNat perspective you can handle this a couple different ways - you could have a hybrid taxon and put all introgressed populations in there (not everybody likes that option), you could ID them to whichever they look most like which may or may not be meaningful, or you could ID to genus where they sit alongside truly unidentifiable past genus images.
None of those are bad, however since we know what clade they belong to a subgenus or complex construct could put them at an ID finer than genus but more coarse than species. A similar thing could be helpful with the various flavors or Longears, Ciscoes, etc were things collapse a bit in certain spaces to a group of introgrades. Of course to keep that accurate from a taxonomy standpoint they'd have to be actual monophyletic clades and not just visually similar lumping. Even then it probably wouldn't make sense unless COF implemented that structure and it would probably create more management work when somebody does more DNA work and finds that two species previously lumped as a single aren't even closely related to each other. It also doesn't really describe the weird things that happen with Ciscoes (and I imagine they may not be alone in that) https://www.researchgate.net/publication/280287350_Morphological_and_genetic_variation_in_Cisco_Coregonus_artedi_and_Shortjaw_Cisco_C_zenithicus_multiple_origins_of_Shortjaw_Cisco_in_inland_lakes_require_a_lake-specific_conservation_approach
TL;DR - evolution is messy, all taxonomy solutions destined to be bad. Changing structure from the way COF handles things probably isn't worth the hassle. And yet the drive to make more boxes until things fit better is still there - even though I know it to be a fool's errand
@nf4t well written! I think I understand species complexes a bit better now. When you bring up the three options we could use to ID fish (using shorthead/peamouth redhorse as an example), I think I tend to lean towards Option 2; if we are certain what group/complex it belongs to, then it should be assigned to one of the individuals in that group. After all, if a line on the map is all that differentiates them, it can easily be corrected or amended in the future if needed (for example, see all the blacknose dace observations I incorrectly ID'd based on range). I think most people on this site would agree that Option 3 is bad, because you're essentially saying "yup, no idea where this fish belongs" even if you're certain. I have done Option 1, mostly with longear sunfish x northern sunfish...its not my favorite thing to do but I think that is technically accurate if you follow taxonomy as it currently exists.
I'm not sure if species complexes will ever be fully recognized in fish taxonomy. There would have to be some really big changes in the way fish-people as a whole think, and like you state I'm not sure if that'd solve our woes.
I also just realized how I sometimes switch between all three of these ID options you described. Maybe I'm species-ist. Perhaps I should try and be more consistent.
@nf4t @operculum_ben In theory the species complex function exists - see the spectabile complex. I believe fish taxonomists are trending towards the use of complexes as 'species' like the sand shiner are now considered to be a complex of species. I like the complex approach for some things like that, but with the long ears they are identifiable to species generally. Perhaps if the ranges are understood well enough we could use hybrid or some introgressed taxon for introgression areas.
I generally think you should ID as low as possible - things can always be lumped later.
As an update to this, I've started tagging all hybrid fish with this field: https://www.inaturalist.org/observation_fields/7615. I figure it might be useful since some of the hybrids won't come off genus either due to uncertainty or too many dissenting IDs.
That’s a really interesting idea. If I understand u correctly, it kinda leaves a trail of breadcrumbs for the knowing.
@stockslager yes, that's right! I see it as an additional way for people to find and characterize potential hybrid observations on this site. Since this was created by another user, there were already plenty of observations with this field; that adds to the appeal for me.
הוספת תגובה