Fatamorgana, Weakley's key lists six Calystegia species for North America. Those species are divided into subspecies, when appropriate, in order to more thoroughly describe population characteristics. For instance, C. sepium ssp. angulata and ssp. limnophila have white flowers. The other three C. sepium ssp. have pink flowers. Each subspecies name represents features that separate it from the others (type of pubescence, shape of leaves, etc.). So all five of these subspecies are C. sepium, and the ssp. designation just adds more information about that population's appearance.

Unfortunately we cannot see the detail of the first white flower. Berteroa incana may be a possibility but without a close up of the flower ..... https://extension.umass.edu/landscape/weeds/berteroa-incana

We call it Swamp Sunflower out here. Grows in the ditches and wild places. Continues to grow from the roots as well as drop seeds. Hardy. Perennial. Comes back faithfully. Can be cut back so that it stays lower to the ground if you only want flowers 18" high for example. Single stems easy to pull out if you want it thinned or have stragglers you don't want. Flowers are long blooming and look nice. Repeat bloomer especially if you deadhead. Heidi

I use google lens for suggestions sometimes, and it can give me good leads?(sometimes), but other times can be light years off, with absolutely ridiculous results. Does anyone know if Seek is all AI, or do they have people at I naturalist that give the answers for seek? I know a couple people who are credited with thousands of ID'S at I nat . One of them used to be a regular contributor at this forum. I'm thinking Google lens learns from all my internet searches and thus over time gets smarter. I asked someone who was praising a plant app, if they would still remember the plant species, if they ran into it a year later, and were without their app. I don't think much plant knowledge is retained if you rely on an app. I don't understand why a lot of people boast about their apps. You're basically stating that you let a machine do the thinking for you, so nothing about that is something I would be proud of. I respect and admire those who come up with correct ID'S from their own minds and stored knowledge. I totally agree with Ken about apps replacing human discussions, ( teaching, learning), and taking all the fun and enjoyment out of plant Identification.