Flower:

Branching clusters arising from leaf axils and at the tips of branching stems, with a few flowers at each node. Flowers are 1/8 inch across with 5 white to greenish or pinkish petals. In the center is a short green to whitish ovary topped with a feathery 3-parted style and surrounded by 8 white stamens longer than the petals, though the stamens are often sterile. Cupping the flower is a pale green to whitish calyx, tapering at the base to a slender stalk.

Leaves and stems:

Leaves are alternate, 2 to 10 inches (to 25+ cm) long, to 4½ inches wide, toothless, mostly broadly egg-shaped, pointed or abruptly narrowed to a pointed tip, straight across to heart-shaped at the base, on a stalk shorter than the blade. Surfaces are hairless; early leaves have minute, pointed hairs along major veins on the underside, though these hairs do not persist.

Stems are hairless, hollow and bamboo-like, light green to reddish or with reddish flecks, and usually clustered. Old stems may persist to the next season before disintegrating. Plants often form dense thickets from long, woody rhizomes.

Fruit:

Fruit is a dry seed 2 to 3.6 mm long, shiny black-brown at maturity, with 3 broad, whitish wings, but seed production is usually poor, the plants primarily spreading by rhizomes.

Notes:

Bohemian Knotweed, also known by synonym Fallopia × bohemica, is a hybrid between 2 non-natives, Giant Knotweed (Reynoutrias sachalinensis) and Japanese Knotweed (Reynoutrias japonica). Japanese Knotweed is known as one of the top invasive species world-wide and this hybrid is just as bad. Like its parent, its tough rhizomes can break through concrete and it is incredibly difficult to eradicate, since root fragments can resprout. Minnesota designated it a Prohibited Control Noxious weed in 2020.

Bohemian Knotweed is intermediate between its parents: Giant Knotweed is uncommon; there are currently no herbarium records for it and very few Minnesota records of it in other databases I've checked; it has quite large leaves (6 to 12+ inches long) with heart-shaped bases and young leaves have twisted, multicellular hairs along the veins. Japanese Knotweed is much more common, though it and Bohemian Knotweed are frequently mistaken for each other.

There are hundreds of records on iNaturalist and EDDMapS for both Japanese and Bohemian Knotweeds, though the true abundance and distribution of either is murky and the images posted on those sites don't always show characteristics that could result in a positive ID; I would consider many reports marked verified as suspect. Thus the county distribution maps are incomplete to say the least. Multiple references note, in early growth, a difference in the hairs on new leaves, but we don't have the evidence to back up that claim. When flowers are present it is much easier: Japanese Knotweed flowers have very short stamens and Bohemian has comparatively long stamens. Late season plants lacking flowers should be distinguished by the leaves: the largest leaves of Japanese Knotweed don't exceed 6 inches long and leaf bases are mostly straight across to broadly wedge-shaped, where Bohemian leaves are distinctly longer than 6 inches and leaf bases are straight across to somewhat heart-shaped. But the bottom line is: both are invasive and should be eradicated.

If you have an invasive Knotweed on your property: the Minnesota DNR has more information, including control measures, that may be helpful. The best alternative for your landscape? The native Spikenard, Aralia racemosa!

About the genus Reynoutria: According to the Centre for Agriculture and Bioscience International (CABI), "Fallopia japonica was independently classified as Reynoutria japonica by Houttuyn in 1777 and as Polygonum cuspidatum by Siebold in 1846. It was not until the early part of the twentieth century that these were discovered to be the same plant (Bailey, 1990), which is generally referred to as Polygonum cuspidatum by Japanese and American authors but, following Meissner's 1856 classification, as Fallopia japonica in Europe (Bailey, 1990). Galasso et al. (2009) proposed transfer back to Reynoutria based on rbcL plastidial sequence analysis and Reynoutria japonica Houtt. is cited as the preferred name in The Plant List (2013)." The trend in databases such as iNaturalist and EDDMapS has been to revert to Reynoutria, and we have followed suit.