lunes, 04 de marzo de 2024

Where orchids bloom

Every year we ask the same question - when and where should we go to see orchids in bloom?

You can of course go to the inaturalist website, filter for some orchid species and date range and then see where there is observations. It's not a perfect way, some observations are of orchids not in bloom (some of them have pretty leaves after all, like Goodyera), and some of those observations might not be on public land (but require for example a research permit). But in general inaturalist observations are a good indication as to when and where you can see a species.

I decided to use the inaturalist API to get data about the amount of different orchid species observed in each lat/lon square. Plotting the data for Europe (using matplotlib) looks like this:

We get a mosaic of colored squares (well, almost squares, Mercator projection means lat/lon are not exactly the same distance) and the color tells the amount of orchid species in bloom. In this case filtered for May. By using interpolation it can also be made to look a bit nicer:

There is no extra information - but it just looks much cooler now. Next I also opted to draw a blue ocean and confine the color to land, to make it look neat. And I added some pretty pictures (of the 8 most observed species in a month).

So, looking at the above map, I suppose I could go to a lot of places in May - Vienna has a nice purple circle and would be closest to me. And if I want to see 50 species in bloom at the same time, Southern France or Italy.

What does it look like a month earlier?

Quite a change. The only yellow color that early in the year is way south in Crete. Greece and Southern Italy definitely look colorful already, as do Southern France and Spain and Portugal. What if we look a bit later in the year, like July?

All gone! The good place to see orchids now appears to be the high Alps in Austria, Italy, Switzerland and France as well as the Pyrenees in France and Spain.

Here's a map for the US in June:

Florida is a good place to be for orchids year round - but in the summer the Northeast is just amazing with orchids everywhere!

As final step, I create a picture for each month and stitched them together into an animation.


Europe!


North America!


South Africa!


Australia!


The world!

Most tropical areas, like Southern Florida, have plenty of orchids year round so my animations are not very interesting there.

Anyway, back to the reason for doing this all - where should I go to see orchids in bloom? I still don't really know...

Publicado el lunes, 04 de marzo de 2024 a las 08:03 PM por elias105 elias105 | 2 comentarios | Deja un comentario

martes, 15 de noviembre de 2022

What should your state's state-flower be?

Recently I saw a map of what each state's bird should be by Sam Allen and thought I could use the same technique for state flowers.

This is what the current official state flower is for each state (from Wikipedia):

Some states more recently added a "state wildflower" in addition to the "state flower", for example Ohio's state flower has always been the carnation, which is not even native. So at some point they decided to have large white trillium as the "state wildflower" - a much better choice and what I used above.

The highlighted in pink states are the ones which do not have a unique flower. It's only 12 states, and the following flowers were picked by more than one: Coreopsis, goldenrod, mountain laurel, prairie rose, violet. Some states did not exactly pick a flower, for example Maine has a tree, white pine. Which is a vascular plant but not a flowering plant. Some chose the flowers of a fruit tree like peach or another introduced species like red clover or peony. Some have an actual species like Minnesota's showy lady slipper or New Hampshire's pink lady slipper which are rare orchids. Some just use an entire genus like Coreopsis for Mississippi and Florida.

I decided to only use inaturalist observations as my data source, and only research-grade observations of native flowering plant species (sorry white pine and red clover and peony). I also excluded plants with less than 100 observations, as well as a set of wind pollinated flowers (so no American beech or big bluestem even though they are commonly observed). And I excluded poison ivy and poison oak which are among the most observed flowers. It shows the general problem with using observation counts - just because something is observed a lot doesn't mean it's liked a lot. Also, using the "native" filter in inaturalist is problematic as not everything has its native distribution filled in yet, but my hope was that it should cover all those often observed species in the US at least and it would keep all the introduced species off the list. Either way, as a first pass I assigned the most observed flower to each state:

Four states (green background) are particularly interesting here as the most observed plant on inaturalist is in fact their official state flower: Arizona, California, Illinois, West Virginia. Really have to commend Illinois for picking milkweed as state wildflower in 2017. For another ten states (blueish background) the official state flower is in the top 10 most observed flowers. 31 states share their state flower with another state.

To avoid the duplicate assignments I added another pass and now for each state select the flower where the state has the most relative observations out of all states. For example in Mississippi the pale pitcher plant makes up 2.76% of all observations. This is the most compared to any other state - in Louisiana only 1.61% of observations are of the pale pitcher plant. Each state happens to have several flowers where it has the most observations - California in fact has a whole 624 such flowers, while Pennsylvania and Maryland have only 11 each.

For 21 states this makes no difference to using the flower with the most observations and Arizona, California and West Virginia still use their official state flower just like before. However this time in addition to those also Colorado and Ohio get their official state flower assigned, the beautiful blue columbine and the large white trillium, respectively. Illinois now gets switched away from its state flower - while common milkweed is the most observed flower in Illinois, the just as cool Eastern shooting star boasts the most relative observations compared to other states, even though it is only the 24th most observed species. But each of the more often observed 23 flowers in Illinois have more relative observations in one of the other states. 9 states would still have the official state flower in their top 10 picks with this algorithm - that is the state flower is indeed a flower which has the highest percentage of observations in that state, we just picked another one.

Personally I quite like this map, but when following Sam Allen's argument for his bird map - the assigned flowers are not truly unique to the state. For example Ohio gets its state flower large white trillium assigned which makes up 1.78% of observations in the state. But in Michigan large white trillium also makes up 1.67% of observations, only slightly less. (In fact when including non-research-grade observations Michigan would have more relative observations.) In this example the gap in observations between Ohio and Michigan is 1.78%-1.67%=0.11%. However looking at all the observed species in Ohio, purple cress only makes up 0.55% of observations but the gap to other states is 0.17%. So in a way purple cress observations are more unique to Ohio than large white trillium. What would the map look like if we pick the flower for each state where it has the largest such gap to any other state?

Arizona and West Virginia still use the official state flower as always. And interestingly, Texas now joins them. Maybe some Texas officials have come up with this same algorithm before? California, Colorado and Ohio switch away from their state flowers as they have other flowers more uniquely observed in them. California in fact picks Toyon which isn't found in any other state (just in Mexico). Colorado now picks the common starlily which is also a beautiful plant, but certainly not as iconic as the blue columbine. Worse with Ohio, instead of the trillium it now gets purple cress which is a pretty spring wildflower but grows everywhere. For 13 states the assigned flower is still the same as with using the most observed one. California, Colorado, Michigan, Minnesota, New Mexico, North Dakota and Ohio still would have their state flower in their top 10 picks.

What does all this tell us? I don't really know! But it's clear that Arizona, California, Colorado, Connecticut, Illinois, Kansas, Michigan, Minnesota, New Hampshire, New Mexico, North Dakota, Ohio, Oregon, South Dakota, Tennessee, Texas, Utah and West Virginia chose a flower that can be backed up by inaturalist observations. The absolutely uncontested state flowers must be Saguaro for Arizona and Great Rhododendron for West Virginia, every single measure confirms them.

You can see the actual data I used here.

Publicado el martes, 15 de noviembre de 2022 a las 01:50 AM por elias105 elias105 | 0 comentarios | Deja un comentario

lunes, 17 de octubre de 2022

How I found the Hairy Shadow Witch

I was starting to feel the onset of the usual depression setting in when the last orchids of the year disappear in the Midwest, sometime in October. So when a long weekend was coming up for me the second week of October I looked at orchid observations on iNaturalist and decided to drive south and treat myself to a new orchid or two and maybe lift my spirits a bit.

Scanning through October observations I found out about a really cool orchid which I had no idea even existed, Ponthieva racemosa (or "Hairy Shadow Witch").

BONAP distribution map for Ponthieva

The closest locations in Tennessee and Virginia both looked to be about 8 hour drives from Columbus, OH. And while it would mean a lot of driving it did not look impossible for a 3-day weekend. I decided I'd prefer the ones along the coast in Virginia since I've never been to the East Coast and I'd have a chance to also find Spiranthes odorata (or even Spiranthes bightensis) that way, other species I had never seen before.

The next step was to find as many pictures of Ponthieva as I could. iNaturalist observations proved not very useful for planning as the taxon is obscured everywhere, so the only information for any observation was the month, but not which day (or even week) of the month. But looking at pictures outside of iNaturalist I noticed that a lot of the pictures from Tennessee, Virginia and North Carolina were from mid to late September and I started worrying that I would drive that far just to find flowers already done the second week of October. After a while I noticed that a lot of the pictures I was looking at were taken by the same photographer, Jim Fowler. He found them in South Carolina over several consecutive years, and dates ranged from about first week to third week of October, with the first week pictures having a lot of buds at the top still and third week ones with bottom flowers done and only some flowers left at the top. So now I really wanted to go to that same spot in SC since I had a good chance of them being in full bloom in the second week of October.

Going to Charleston, SC however is an almost 10 hour drive from Columbus, OH - definitely not possible to go there and back when also considering gas and bathroom breaks. I had not taken any vacation yet this year and so asked if I could take the remaining 4 days off that week at work, even if at rather short notice - and I could. So now I would have plenty of time to get there and even stop at some other locations along the way.


My markers on all the dirt roads criss-crossig Francis Marion

Over the next days I further studied all the inat observations of Ponthieva in SC to see where people might have found them. No exact locations of course, but everything pointed to Francis Marion National Forest, which is also where Jim Fowler had photographed his. In his blog posts he mentioned that they would grow along the dirt roads throughout the swamp forest. In fact the forest service he said, way back when building those roads through the swamp, used sand which contained a lot of crushed sea shells, which in turn created the ideal habitat for orchids. So they would actually grow right on the embankments of those forest roads.

I worried a bit about not being able to drive the dirt roads, hurricane Ian had just swept through the area and I only have a normal car, no truck or 4WD. Also, while those blog posts had some location description ("dirt road off Halfway Creek Road, close to the Santee river") it was by no means an exact location. And he mentioned that they can be hard to see (which, having see them now I would disagree with). He also said that there would be 1000ds of flowers along those roads. I'm quite good on my feet and not afraid to walk a few miles, so I felt if I couldn't drive the roads I would walk them, and in that case I would not likely miss the flowers. And if need be I could now spend two entire days just searching for them, walking a bit along every single forest road if I had to!


My car and a Ponthieva

I took the above picture right after I arrived in the evening - I took a short detour along Halfway Creek Road (a big paved road through Francis Marion) on the way to my motel for the night and just wanted to get an idea of the condition of the dirt roads, so I could make a battle plan for the next two days. The first dirt road I turned into was in perfect condition (much better than any in Ohio) and, not really having expected to already find one the day of my arrival, the orchids were growing right on the road and not hard to see at all. In fact they were much taller and showier than I had expected.


Most of the orchids were not right on the road but under the ferns on the embankments

That is not to say that I would have been likely to just find them without knowing where to look from the earlier research. The one right on the flat surface of the road was the exception - a lot of them liked to hide under ferns on the side of the road instead. When I explored more over the next two days I also found them trailing off into the swampy parts of the forest at times, away from the road. As is often the case with rare flowers, they can be locally abundant - this one definitely seems to like Francis Marion forest a lot. I found them on four different dirt roads, separated by several miles and each about a quarter of a mile long, and each comprising 100ds of plants.


Unlike with many orchid flowers, the lip is at the top with Ponthieva


Om nom nom (not sure this is an actual pollinator for it, but it certainly liked the nectar)

Jim Fowler's blog also mentioned finding Epidendrum conopseum, the Greenfly Orchid, growing on an oak tree in a 1700s churchyard. I googled for 1700s churches in South Carolina and realized there is only 3 or 4, one of them in the middle of Francis Marion forest and only reachable via a long dirt road. That had to be it and I could just put it in Google maps! So while not equipped for taking pictures of something high up in a tree, I got a picture that's at least identifiable, and the only epiphytic orchid I have ever seen.

Epiphytic orchids grow on tree bark

One thing to mention is, you should not be like me and blindly trust your GPS - instead double check the route first. It took me over 90 minutes to reach the church from the spot I was at, blindly following Google maps navigation, but only 18 minutes back. Turns out for some reason Google had me go along 8 miles or so of a dirt road, just to loop around at the very end and then go almost 7 miles back on the same road before turning right. Apparently it though you can't make left turns off this (one lane!) dirt road. Luckily I was not on a tight schedule and could still enjoy driving through the forest (twice!) and I even spotted a bald eagle flying in front of my car on the way to the church.

The biggest adventure of the trip was a Habenaria repens I spotted in the distance the next day, in the shallow water of a forest pond. There was a strange log swimming in the deeper water close to it though, and after staring at it for several minutes it moved a bit. And I realized it was an alligator! I come from Ohio where there is no dangerous animals in the forest (except for humans and ticks), so that was a bit of a shock and sent me running away. After much contemplation I decided against walking close to the orchid, even though I'm still disappointed that I was so close to it but then couldn't get a good picture.


I think this was a Waterspider orchid in the distance, but unfortunately it was being guarded

I also saw countless other lifers (176 according to the inat calendar). Two favorites are a sundew species I had never seen before, Drosera capilaris (Pink Sundew) and a plant I had no idea even existed, Burmannia biflora (Northern Bluethread). The Burmannia genus used to be in an order called "Orchidales" before DNA testing revealed it's not really an orchid at all.


Two favorite lifers, Drosera capilaris and Burmannia biflora

Lastly, on the way to Francis Marion I also stopped in Congaree National Park, and was extremely lucky to find 1000ds of Spiranthes odorata (Fragrant Ladies Tresses) in full bloom there. Hard to compare or find even higher superlatives when seeing so many incredible things, but this might actually be the biggest highlight of my entire trip, considering Spiranthes are my favorite flowers and this species was also a lifer for me.

Spiranthes odorata - despite its name I didn't notice any smell

Publicado el lunes, 17 de octubre de 2022 a las 02:11 AM por elias105 elias105 | 2 comentarios | Deja un comentario

Archivos