A new field for sorting observations observed in human-influenced landscapes

Lately I've been tracking observations by natural community type, as defined in Vermont with the book Wetland, Woodland, Wildland. (other states and countries have different classification systems). For instance, see Northern Hardwood Forest, Vermont's most widely distributed natural community: http://www.inaturalist.org/observations?field:Natural%20Community=Northern%20Hardwood%20Forest

Note that it ranks the species by number observed. It's not systematic, as i don't record every species every time, but it gives you an idea of what species you will see in each natural community.

After the idea independently coming up in several different places the last few weeks, I've decided we need a field to track different human-influenced habitats as well. Urban nature is a whole other world than natural ecosystems... but every bit as interesting and often not as well studied. Both are worthy of attention, but the nature where people live is the easiest for most people to see from day to day.

To make sure things are standardized, since there's no book to anchor to, i've proposed a bunch of categories. Others are free to propose others. Once we start using them it becomes a bit of a pain to change them with a bulk edit, so I wanted to run them by a few other users first. Here's my current list:

road/railroad/trail right of way
utility right of way
mowed lawn (includes home lawns, sports fields, playgrounds, etc)
Human-created open field (cut every year or two, but not a mowed lawn. Like many fields around homes in VT)
Disturbed wetland (cut over every year or two like open field but is wetland
agricultural area (farms)
Dense Urban Center (parking lots, cracks in sidewalk, etc)
Abandoned Urban Center (vacant buildings etc)
landscaping or garden (where plants are planted, pollinators may appear, weeds, etc)
managed forest or plantation (this one can intergrade with wildland since most areas are managed in some way but thinking of high intensity managed areas)
Invasive Species Infestation
Storm Drain/Channelized Waterway
Artificial Pond
Stormwater pond/sump/catch basin
Indoors (spiders, etc that find their way indoors, birds in airport terminals, etc)

I don't want to create too many but nor do I want to leave too much out. I'm not sure if these would work globally, but they seem good in the northeastern US. Once i get this field set up we can start seeing what species are in each one. Any thoughts? Thanks to @erikamitchell @srall @jogarton for being part of this discussion so far. Not sure who else might be interested. @bouteloua @kueda @kpmcfarland @cullen @sambiology @silversea_starsong perhaps? Feel free to tag others. Also feel free to use the natural community field, but if you're in Vermont let's use Wetland Woodland Wildland.

Posted by charlie charlie, July 15, 2017 12:46 AM



This is a fascinating idea, Charlie!

When I started down your list for the first time, I thought...hmmm...lumping roads with railroads? They're so very different in terms of species that you are likely to find. Like grouping beech-maple forests with oak-hickory forests. But I also see your point. They're all human-made paths. And very different from agricultural fields or storm drains.

I think it would be very useful to have a short written description of each of these habitat types to make their definitions more clear. Is there any formal research on such habitat types that could be used as a basis for the definitions? Or are you creating these habitat types from scratch?

What research questions rely on the natural community types that you are currently using? That is, what questions do the natural communities answer, or bring up? How would these questions compare to the anthropogenic community types that you are listing and defining?

I find the storm drain/channelized waterway item quite interesting. How would that translate as a community type in various areas of North America? I'm thinking in particular of an abandoned canal I walked along in southern New Hampshire last week, 50' wide, water barely flowing, a major transportation route 150 years ago, and now forgotten in the woods, or the C&O canal trail near Washington DC, or irrigation canals in New Mexico--should all of these be in this community type? Why? What would we expect them to have in common? And how would this type be applied worldwide? Now I'm thinking of the falaj systems in the Arabian peninsula, ancient irrigation waterways through ancient agricultural fields. Should all of these be classified as a single type: artificial channelized waterway?

Anyway, it's a cool idea and a fascinating topic for research! Who IS researching such habitats?

Posted by erikamitchell 4 months ago (Flag)

We don't have to lump roads with railroads! But i'd like to have a dropdown list and so i am trying to eliminate splitting too finely. If we split off roads and railroads... what about rail trails? Every type of trail use? I bet they ARE different. But to some extent tags can also be used there, or a second field. I didn't make a dropdown for the Natural Community field because there are too many of them (also i colonized a field someone else made :) ). But for these it may work.

I am very familiar with how natural communities work. How those are usually classified? a bunch of data is collected that includes plant species, their life form (tree vs shrub mostly), and their percent cover. Then statistics such as cluster analysis and TWINSPAN are used to separate them and you can run correlations with abiotic data you can also collect (amount of litter layer, soil pH, disturbance history, whatever) and look for correlations. I'm not an expert with this stuff but have helped do it several times.

to my knowledge no one has done so with these human influenced areas, but i am not certain. They have been mostly ignored by science until recently. At work I am doing plots in wetlands and that will include very disturbed wetlands. I also know someone who is doing a study of the vegetation in stormwater basins which are kind of like the basins @srall mentions. @Jogarton is doing a study or survey of roadsides in Vermont which is one reason I started talking about this. My first response was 'did you talk to @erikamitchell ?" and she said she had. which is awesome :)

Presence or absense data isn't really enough to classify community types, or at least, it's harder to do. I wish iNat were easier with percent cover but it *could* be used that way. You don't even need a defined plot as long as you survey a set area with internally consistent community assemblages. We did a hybrid approach in California which combined making a map (too broad an area to do plots everywhere) with doing percent cover estimates in set polygons. Seemed to work great for classification. If everyone did percent cover for the different plants in their lawn, excluding planted grass, one could start learning about different lawns. I hate lawns for the most part and think they are way overused, but that doesn't mean we shouldn't understand them better.

Of course, all natural communities are defined by animals and fungi as well as plants. And in the case of disturbed or urban landscapes, i suspect the animals may tell you more than the plants, especially insects. After all they are able to use landscaping plants and such. There are insects and birds in places where there are essentially no plants growing on their own, not even invasives. I don't know of any work on defining natural communities by insect or fungi.

When I create the 'types' i will make written descriptions in a blog... or we can do it in some other format like google docs or even a wiki on iNat (i don't think they will care). Though, i will warn that I am super busy so that may take too long. I kinda want to get started with this so will either create something soon or start using tags maybe.

Old abandoned landscapes are a whole other issue too! Of course most of Vermont including many areas that now match as a natural community again (like Northern Hardwood Forest) were once farm fields or even villages (see little river state park). And certainly the human influence of the past DID affect the future trajectory of the forest. But with so few old growth stands to go by it's hard to say how. And as we move forward in the next few decades the oldest abandoned farms are moving back towards many old-growth characteristics!

So much to learn :)

Posted by charlie 4 months ago (Flag)

This is a nifty research idea, Charlie! If there really has been little to no research in this area yet, then we'd have to make it up as we go along. To me, that suggests that we probably won't get the choice of fields/categories exactly right on the first try. I'm thinking maybe we could try out some version this year, on a specifically trial basis, to see what is out there and what goes together and what doesn't. Then we could apply this information back into the category lists, and refine the lists next year by adjusting the fields according to what we have learned. For research purposes, that would mean creating a new list of categories next year, so that we can keep track of the meanings of the categories/groupings that were being used at the time they were applied. So maybe this year's list should have the word "-trial" in it, or "-17" or some other notation.

At some point, I would love to do some analyses on the different kinds of habitats I have walked this year. My gut feeling before doing the stats is that foot paths are different from railroad beds, and different from road margins. Paved road margins are different from dirt road margins. What can be lumped and how, I don't know. Where do hedgerows fit in? The strip of grass between sidewalk and curb in the city? Urban gutters?

And then...I bet most iNaturalist observations fall into one of these path categories--most people walk on trails when they go into the woods. Or on trails in parks. I'm guessing that there are probably few observations that are away from trails, roads, sidewalks. Thinking of my own walks, probably 95% of my observations are along a trail or road of some kind. Only maybe 1-2 times per month do I wander straight off into the woods, away from any trail.

Posted by erikamitchell 4 months ago (Flag)

Absolutely. For the trial version we could also not create drop down categories, but that would require a lot more coordination and might get confusing. Another option is to log all the choices we are using, here, and then create one with a pulldown later when we finalize it. What do you think?

Yeah, I do roam off trail a lot when i'm doing work stuff but I agree, most things are definitely along a trail. But for trail stuff I wouldn't mark as trail unless it's directly influenced. Like if I stand on a trail and see a paper birch tree in the woods 5 feet away i would not mark it as trail. Everything has edge effects but at that point it just seems too minor to be of note. But who knows.

Posted by charlie 4 months ago (Flag)

Just a few thoughts:

-I agree the various right of ways are very different!
-I'd be curious to hear about your differences between "lawn," "field," and "fallow field" in VT. Here I would have fallow field its own category and fields would either get lumped into either lawn or fallow field.
-"Disturbed wetlands" - this could be a degraded natural area or a stormwater area, an abandoned agricultural field, etc. It seems to be more a quality than a locational or human use category
-"invasive species infestation" is another quality, rather than locational or human use category, and could occur in any of your listed locations.
-the same sidewalk weeds will grow on a rural town's main street as a big city, so "dense urban center" probably isn't the best phrasing. Rural town and urban city are very different classifications for animals, though....
-"brownfield" is not so different from "abandoned urban center." maybe include abandoned urban center inside of brownfield?
-many "artificial ponds" are also "stormwater ponds"

Here's another take after a very quick search: "anthromes" or "anthropogenic biomes" http://ecotope.org/anthromes/v1/guide/ (not nearly as specific as your list, though it is more global in scope) I'm sure others have tried classifying these, would take some paper reading!

Posted by bouteloua 4 months ago (Flag)

ooh thanks! Very neat. I don't have a ton of time but a few things:

Lawns in my mind are mowed every week or two. Field is a place that is mowed either in summer or fall (and that does matter which but won't always know). A field is full of goldenrod and milkweed and such, a lawn is either a pesticide wasteland or has some little clovers and violets and such..

I need a way to track disturbed wetlands for a variety of reasons... stormwater pond would be separate for sure but otherwise would be looking at suburban seeps or wetland fields and lawns, etc. They are really different and I want to be able to parse that out. One could have a separate field for wetland but i feel like the iNat field system isn't robust enough to make that easy to do.

Invasive species infestation : i want to track the other things that survive in say, a dense patch of buckthorn or knotweed. Yesterday I found a buckthorn patch with maybe 8 native herbs beneath and even an orchid. Other times they are dead zones underneath. Why? would love to find out.

Not sure how best to delve out the city and town types.

I was thinking of brownfield as having some sort of toxic waste or pollution issue but maybe that doesn't matter.

Stormwater ponds are super different from landscape ponds people build. But how different I don't know. Maybe they could be lumped.

Posted by charlie 4 months ago (Flag)

@bouteloua since i was just talking to you, maybe you are interested in this.

i haven't forgotten about it, just busy.

Posted by charlie 4 months ago (Flag)

Hi all,
I'm late to the conversation here but would like to finally weigh in and hear more of your collective thoughts. I am getting pretty specific about how to classify roadside communities as part of a project studying the health and longevity of roadside forests. A lot of town road crews in Vermont want to clear the right-of-ways of their rural roads to make the roads easier to maintain. Thinking about @erikamitchell 's question, "What questions do the natural communities answer, or bring up?", for me, it's about future management recommendations. Unlike natural communities, these unnatural communities are constantly disturbed and shaped by human action. That could be anything from repeated clearing of roadside to disturbance from wind and dust caused by passing cars.

So, I'm wondering about classifying roadside communities based on what key species in key areas, and how these key areas would respond to recommended management. Any thoughts on the scale to which you would make divisions in categories of unnatural communities? For example, is it worth distinguishing between "edge species of a northern hardwood forest" and "early-successional northern hardwood forest"? "Wet ditch", "streamside community", or just "wet land"?

I'm also wondering about the merits of distinguishing between the unnatural community in the road right-of-way (usually about 25 feet from the road centerline) and the neighboring ecological community. One can be managed by the town, the other is managed by the landowner. But of course, the line between the two is blurred as far as nature is concerned.

It's a work in progress!

Posted by jogarton 3 months ago (Flag)

Hmm. I think if we get that detailed it will be hard to standardize between different people. But maybe that's ok. I'm traveling now and will have more to say soon. Haven't had time to look at this.

Posted by charlie 3 months ago (Flag)

Add a comment

Sign in or Sign up to add comments

Is this inappropriate, spam, or offensive? Add a flag