February 17, 2024

On limiting notifications

Q. Do I really need eight (8) notifications that my completely unambigous Peregrine observation is really a Peregrine?

A. No. Somewhat regrettably, I'm going to try turning off notifications of agreeing IDs.

Posted on February 17, 2024 01:35 PM by matthew_wills matthew_wills | 2 comments | Leave a comment

October 22, 2023

End of Season Insect Notes

(Not that the season ever really ends, but it sure is reduced over winter…)

I didn't see a single Common Buckeye/Junonia coenia in NYC this year. Only six observers did: these six observations compare with 77 observations in 2022; 52 in 2021; 106 in 2020; 390 in 2019; 16 in 2018; and 34 in 2017. (2019!)
*

Why do Bald-faced Hornets/Dolichovespula maculata do so well in the city but Common Aerial Yellowjackets/Dolichovespula arenaria not so much? ID may itself be an issue: Bald-faced are one of the easiest wasps to identify; Common Aerial, being rather Vespula-like, are much harder to parse. But look at these numbers: 1,193 maculata versus 41 arenaria. Only two confirmed arenaria nests have been observed in the city. Once the leaves drop, it’ll look like a maculata nest on every block. This is hyperbole, but they do take well to the city for some reason(s), and not just in the parks.
*

I saw my first Asian Mud-dauber/Sceliphron curvatum this year, at the Narrows Botanical Garden in Brooklyn. I also observed what I think is one of their mud nests, in Prospect Park. So there are now a total of 19 observations in NYC (the Bronx, Manhattan, & Brooklyn) by 17 observers. The first was reported in the Bronx in June 2020. As there were no observations in 2021, 2023 was thus the biggest year for sightings so far. It will be interesting to observe the spread—or not. According to Bugguide, they were first noted in North America in 2013 in Quebec.
*

Where are Green-Wood’s Organ-pipe Mud-daubers/Trypoxylon politum? There are plenty of old mud nests, but I’ve never seen a live wasp there. And I haven’t seen any nests that look like they were built this year. (I have seen the wasps in past years in Prospect Park, and did see new-looking nests there this year.) Of the 95 observations of this species in NYC, only a hymenopterist’s dozen are of the insects themselves. The rest are of those distinctive nest structures.
*

Computer Vision is remarkable, but when it comes to arthropods it still leaves much to be desired. A lot of organisms are going to Research Grade because of the CV and a few serial identifiers, some of whom seem pathologically eager to build up ID stats at the expense of the iNat mission, when species level can only be determined by microscopic examination.
*

For an identifier, the iffy observations that flood iNaturalist from reluctant students in classes or bioblitz are pretty disheartening. We’re still getting tours of local zoos from a college-level Wild Flora and Fauna class. Such junk data-dumps rarely mar the arthropods, but this fall did witness a dozen observations of the same Hackberry Emperor/Asterocampa celtis in Brooklyn by bioblitzers. This is a rare species for Brooklyn; these repeats increased the total number of observations of this species in Kings Co. to 33. That’s a jump of a one third, so it gives a quite misleading impression.

Along these lines, it’s depressing to see a series of unclear photos of the same damselfly in somebody’s hand. (Ditto for other insects in a jar.) You’ve gone to all that trouble to capture and bother the animal, but for no real purpose. Even with good photos, the points of damselfly ID—color, eye spots, thorax pattern, abdomen pattern—are not shown because cell-phones just don’t cut it, and the way the creature is pinned against the hand. Holding by the wings also obscures the stigma details, which are key for a few species.
*

It’s been ten years since I’ve seen a feral Western Honeybee/Apis mellifera hive/nest in Green-Wood. It was in hollow of a mature oak. I saw it active in 2010 and 2013. (The one other feral hive I’ve seen in there was in a dead tree that was later cut down.) There are 3,596 verifiable honeybee observations in NYC, but the vast majority of these should probably be marked Casual. They’re livestock, unless you have evidence of feral hives.

Posted on October 22, 2023 12:32 PM by matthew_wills matthew_wills | 0 comments | Leave a comment

June 12, 2023

Frustrations of an Identifier

More than 2.6 million people have put up observations on iNaturalist. Some 311,424 have done identifications. Our community-consensus identification system really needs more identifiers.

Doing the work of identification makes my own observations better. Hence my sometimes-cranky tips for being a better iNaturalist user:

Prep: read the instructions and the FAQs. There is even something specific for teachers poised to unleash a deluge of junk observations from reluctant students.

Scale: what size is the life-form? It may well be obvious; it may not be. It will be obvious with some sense of scale in the image. I’ve used coins, my 2cm wide thumb, my 12” long boots, one of the rulers I usually have in my kit. Sometimes I just eyeball it and make a note of the approximation. Which brings me to:

Words: just use your words! Add a short explanation, description, some context. Especially for dark, unfocused, incoherent (a landscape of multiple species) photos. What are we supposed to be looking at? What did you see that can’t be represented by the photographic technology at hand?

Taxonomy: don’t submit observations as Unknown. Make your way down from the top: No idea what it is: mark it as Life. On the scales most of us are observing, we all probably know the difference between Plants and Animals. Mark it so. Keep working your way down the taxonomic ladder, if you can. Is there anything more cringe-inducing than someone in an advanced college bio class actually broadcasting their inability to call a bird a bird, a mammal a mammal? (It may, in fact, just be their inability to understand this tool, but is this any less cringe-inducing?) The appalling tendency of City Nature Challenge-racers to pile up hundreds of Unknowns when they clearly know the difference between Plants and Animals, for instance, is a dereliction-of-duty and an insult to identifiers. iNaturalist promotes you as a naturalist; get into taxonomy and prove ‘em right.

Plants: sorry, app-magic, many plants can’t be readily identified with a single picture. The idea that you point and shoot and app your way to botanical knowledge is ridiculous. (Botanists must often be enraged.) Check out this new monograph on Violaceae: that’s botany, a life’s work connecting to others’ life works. Respect it. Meanwhile, there are hundreds if not thousands of shitty pictures of distant tree trunks on this site—no canopy, no leaves, no flowers/fruit, just bole. On the one hand, ok, in the vast energy-draining nexus of this database, another life form! On the other hand, come on!

Annotations: the problem with the app version is that it’s so limited. You do it all on your phone, and you’re missing out on the richness of Annotations, Projects, Tags, and Observation Fields to which you have access to on the web. (Basically, the system creates two classes of users, once again piling more work on the minority of users who do identification).

Identification: to reiterate, doing the work of identification has made me a better observer. I assume it works this way for everyone. But as it stands, lots of observers do very little identification. (Once again, using the app alone cuts you out of this part of iNat.) Check out profile pages to see the disparity: here’s one with over 5400 observations but only 100 identifications. This is a problem. There are simply not enough identifiers. New to the game, ok, but you know some things pretty well, or you soon will, right? Spotted Lantern Flies, say. Start with what you know. Don’t know anything? That’s a beginning you’ll soon outdistance.

Wild: iNaturalist is intended for wild organisms, not pets, zoo/aquaria inmates, farm animals, houseplants, garden plants, botanical garden specimens, or any plant intentionally planted by humans. All captive/cultivated organisms should be marked as casual observations.

Humility: community science databases are full of biases and errors. Plenty of things make Research Grade when they shouldn’t. (One big problem: RGs can lead to more false IDs because of the AI algorithm). Many species cannot be identified by photographs alone. Lots of insects, for example are identified to species by the shape of their genitalia, requiring specimens, higher-powered optics, specialized keys, and hard-earned knowledge. I have, for example, 729 observations of Hymenoptera that need ID, and I suspect a good portion of them will never be identified to species level. A bummer, but then how amazing is biodiversity?!

Give Back: consider contributing to iNat.

Posted on June 12, 2023 01:38 PM by matthew_wills matthew_wills | 2 comments | Leave a comment

May 3, 2023

Strategies for winning iNat during City Nature Challenge

To get the most observations, dump hundreds of barely helpful pictures without any taxonomic label, so that you pile up hundreds of “Unknowns.” With less than 10% of users doing IDs, this junk-data dump is a thumb in the eye to them; for NYC alone, over 8000 observations need ID after the four days of the CNC—many will never be identified because the single photo is useless.

Meanwhile, to get the coveted Top Identifier, you can second the original observer’s “Plants,” and add your name to already Research Graded-ed observations. For extra credit, blithely Research Grade zoo animals and botanical garden specimens.

Get those stats higher, higher!

Isn’t it fascinating how making everything a challenge, or, in other words, little subsets of the same monstrous capitalism that is devouring the natural world, invariably leads to corruption? In this case, junk data and cooked hierarchies of achievement.

Posted on May 3, 2023 11:13 AM by matthew_wills matthew_wills | 2 comments | Leave a comment

February 27, 2021

Observing Trees

I see a fair number of tree observations that are a single image, sometimes little more than a tree in silhouette. These are often impossible to identify. From my own experience, which is still on-going, I'll offer these suggestions for making better tree observations.

All of the following can be helpful in identifying a tree. As many of these as possible should go into an observation:

entire tree, or best approximation thereof
detail of leaf/leaflets, including needles for conifers
buds, bud scars, leaf scars
flowers and/or fruit/cones/pods, including any fallen to ground
bark

Don't forget scale. How big is the thing in your photograph?

Obviously, there have some seasonally dependent items here. But that's good; go back to that tree, watch it through the year. Mature trees may hold a lot of these things above eye-level, so scout the ground below.

In New York City, where I live, many of the trees we come across on the street, and in yards, parks, and cemeteries, are planted specimens raised by the horticulture industry. This means that, for iNaturalist purposes, they are captive/cultivated or "casual" observations.

There is an existing street tree map https://tree-map.nycgovparks.org. It isn't error-free; when it first came out I noticed that the two common hackberries I can see from my window were classified as hawthorns. (There's a way to submit corrections.)

So the actual urban tree may not be worth an iNaturalist observation as such -- good luck on the trademarked varieties sold by the industry! -- but because the tree is habitat it's worthy noting what kind of tree it is and pairing it with your observations of what's living on the tree. Trees are home to algae, fungi, lichens, mosses, and liverworts. Invertebrates are all over them, from the obvious pollinators to the ants crawling up and down them all day long. And that ticking sound in the fall? A katydid!.

Don't forget the reptiles, birds, mammals....

Look for galls; nests/drays; webs; leaf damage from herbivores and diseases; pupae; ambush predators; eggs laid on leaves; beetle excavations and frass. E.O. Wilson has said you could spend an entire lifetime of exploration around a single large tree.

Lately, I've been observing the types of trees Yellow-bellied Sapsuckers drill into for sap.

Posted on February 27, 2021 05:04 PM by matthew_wills matthew_wills | 7 comments | Leave a comment

Archives