The Monthly Fern: Water Clover & its odd spores

"It is worth clarifying that these plants are not clovers." (Photo by Bill Dodd).

For July, South Dakota's fern-of-the-month is the Hairy Water Clover, Marsilea vestita. Among our ferns it's quite the oddball—in habitat, behavior, leaves, and especially the spores. They suggest that one of its ancestors was the evolutionary start of seed plants!

Water Clovers grow nearly worldwide. On the order of 45 to 65 species are recognized (experts disagree on number), of which 5 are native to North America. M. vestita, the Hairy Water Clover, is the only species in South Dakota (so far). It's known from many sites across the state, in shallow water and on mud. Plants are rooted (not free-floating) and tolerate seasonably dry conditions, for example persisting after a pond has dried up. In fact, wet followed by dry can aid in reproduction and dispersal (Montana Field Guide).

Marsilea vestita is rhizomatous and colonial; in wet habitat, leaflets usually are horizontal (MWI).

M. vestita in a field; in drier habitat, leaflets often are ascending (Mary Ellen (Mel) Harte photo).

Hairy Water Clover can grow as tall as 20 cm on moist soil, or to 40 cm in water (in order to reach the surface). Leaves are dimorphic—sterile and fertile—but neither is fern-like. Sterile leaves have blades with four rounded triangular leaflets, and look a lot like four-leaf clovers! No other plant can be confused with Marsilea (1). 

Sterile leaves of Hairy Water Clover have divided blades on long slender stalks (MWI).

The sterile leaves also are unusual in behavior. Water Clovers are the only ferns known to be nyctinastic—moving with the onset of darkness. During the day leaf segments are nearly horizontal. Then as the sun sets they bend upward, forming a packet of sorts (Montana Field Guide has an account of this and other interesting features of Water Clover).

Fertile "leaves" are located near the base of sterile leaf stalks. In shape and size they resemble beans or peppercorns (source of another common name, Pepperwort). Being unusual they of course have a special name—sporocarp (= spore body); oldtimers like Linnaeus called them capsules.

Marsilea vestita. The hairy sporocarps contain 2 kinds of spores (lower right). Britton & Brown 1913.

Young sporocarps are greenish, hairy and slightly soft. With maturity they dry out, darken and become very hard. In this state they can survive for many years; the record is said to be 100. The Marsilea sporocarp is an effective unit of dispersal, often by way of waterfowl digestive tracts.

Like almost all ferns, Water Clovers have sori—clusters of sporangia which contain the spores. But the arrangement is quite different. In a typical fern, sori are located on the underside of leaves. In Marsilea, they're neatly arranged inside the sporocarp.

A typical fern; sori are clusters of sporangia, which contain dust-sized spores (2). USDA Forest Service.

Marsilea sporocarp with sori; below it, a sorus with sporangia, which release spores (no source given).

A cross-section through a Marsilea sporocarp (above) reveals an orderly but complicated interior. Inside the container-like sori are 2 kinds of sporangia. This is where things get exciting. They produce 2 kinds of spores—male and female!

Most ferns, 99% in fact, release a single type of spore—tiny, 1-celled, asexual. But not Water Clovers. They're among the 60 fern species (out of c. 10,500 total) with male microspores and female megaspores. These are heterosporous ferns; interestingly, all are aquatic (more here).

When a Marsilea microspore bursts open, many sperm (aka spermatozoids) swim off in search of an egg to fertilize—not unlike sperm of typical ferns. It's the megaspores that are so unusual. Not only are they 10 times the size of a typical fern spore, they're complex, with specialized parts.

Marsilea megaspore, c. 0.8 mm long, with 2 cells (no source given).

Shortly after leaving its sporangium, the megaspore divides to become two joined cells. The upper cell will produce an egg, which gives off chemicals to attract sperm. If a sperm successfully wriggles through the opening and reaches the egg, fertilization takes place, leading to development of an embryo and then a baby fern.

In comparison, the megaspore's basal cell is huge, and rich in carbohydrates and fats. These will sustain the developing fernling in its first days, before it can photosynthesize. In this way, a megaspore is like a seed, which supplies nutrients for its young seedling. Perhaps an ancestor of Marsilea was the evolutionary beginning of seed plants (more here).

Some readers may be wondering where the gametophytes are—those tiny independent plantlets that are the sexual stage of ferns. Good question! Water Clovers do have gametophytes, but they are minute and NOT independent (another similarity to seed plants). For more about the life cycle of heterosporous ferns, see (3) in Notes.

Simplified?

Now we finally arrive at the long-promised answer to the burning question, "How many spores would fit in a typical [empty] can of soda?"

Fern spores are truly tiny. A handful looks like a pile of dust. To show just how small they are, Robbin Moran (2021) calculated the number of average-sized spores that would fit in a typical can of soda, which has a capacity of 355 milliliters. For spore volume, he used 125,000 µm3, assuming for simplicity that a spore is a cube. What do you think? How about a ballpark estimate?

Hmmm ... 777,000?

Maybe 10 million?

According to Robbin's calculations the answer is 4,440,000,000 (4.4 billion). Yikes, that's a lot! Yes, spores are tiny indeed (4).

Notes

(1) While Water Clovers are easily recognized, distinguishing the Hairy Water Clover from others in the genus is not easy, requiring sporocarps. If you intend to document an occurrence, be sure to collect both types of leaves. See Marsilea in Flora of North America for a species key and descriptions.

(2) It seems sporangia are readily mistaken for spores, as a search for images of "fern spores" suggests. Many of the images actually are sporangia clustered in sori.

(3) In the previous Monthly Fern, I made a big deal out of the 2-stage life cycle of ferns, which involves separate tiny green sexual plantlets—gametophytes—that give birth to baby ferns. Heterosporous ferns have gametophytes, but they are minute and not independent. They develop inside the persistent spore wall, where they give rise to either sperm-producing antheridia or egg-producing archegonia (gametophytes of typical ferns usually have both). For the female gametophyte, developing inside the spore wall provides additional protection for the embryo, but there's no access to sunlight to photosynthesize food for the fernling. That's why the nutrient-rich basal cell is so important.

Life cycle of Marsilea, a heterosporous fern (labels added). See Milne Publishing for details.

(4) I couldn't find Robbin Moran's article online. If you'd like to read more about fern spores, and all of Robbin's calculations and conclusions (e.g. 4.44 billion spores taken together has a surface area nearly equal to 8.5 ping-pong tables), send me an email address and I'll send you a PDF file.

Sources (in addition to links in post)

Hooker, WJ, and Greville, RK. 1831. Figures and descriptions of ferns, principally of such as have been altogether unnoticed by botanists, or as have not yet been correctly figured. Vol. 2. BHL

Milne Publishing. Marsilea. Accessed June 2025.

Montana Field Guide. Montana Natural Heritage Program. Hairy Water Fern—Marsilea vestita.

Moran, RC. 2004. The Natural History of Ferns. Timber Press.

Moran, R. 2021. Fern Spores, Soda Cans, and Ping-Pong Tables. Fiddlehead Forum (May–Dec).

Pinson, J. About Ferns, American Fern Society.

PremaBotany (Prema Iswary). December 2018. Marsilea.

A Restless Region on the Colorado Plateau

Castle Valley, 20 miles east of Moab, UT; Round Mountain rises from the valley floor on the right.

The Colorado Plateau is a thick block of crust in the Four Corners area of the American Southwest, an immense stack of sedimentary rock. It's remarkably stable, remaining a region of geological calm even when severe deformation was underway right next to it—uplift and faulting of the Rocky Mountains to the east, and stretching and breaking of the Basin and Range Province to the west. This is why the Plateau's sedimentary rocks are largely horizontal, like the deposits they once were (source).

This is not to say the landscapes are boring. In fact they're spectacular—sweeping vistas with colorful roughhewn features. For the last 10 million years the Plateau has been rising, invigorating streams and accelerating erosion. The result is a seemingly endless collection of buttes, arches, rimrock, fantastical spires, deep winding canyons, and more.

The Colorado Plateau covers c. 130,000 sq mi; note complex topography in adjacent areas (source unknown).

Valley of the Gods near Bluff, UT; vertical and horizontal erosional features are common on the Colorado Plateau.

Entrenched meander of the San Juan River, cut through horizontal strata; Goosenecks State Park.

Looking down Castle Valley. Is this a standard Colorado Plateau landscape? Bryant Olsen photo.

Castle Valley appears to be dominated by vertical and horizontal features, as is typical on the Colorado Plateau. Is it another example of recent uplift and erosion? Only partly. Its story is much more complicated—repeated flooding, prolonged deposition, unusual deformation, weird intrusions, and finally ... collapse.

About three hundred million years ago, not far east of today's Castle Valley, the great Uncompahgre Range was rising. At the same time the Paradox Basin was subsiding along the range's base, and filling with sediments eroded off the mountains. Critical to the Castle Valley story, seas repeatedly flooded the Basin (1). Then whenever sea level dropped, the saltwater left behind evaporated and deposited evaporites, including lots of salt (halite). At least 29 such cycles took place over a period of at least 8 million years. The result was extensive deposits of evaporites at least 4000 ft thick—the Paradox Formation (2).

Ancestral Rocky Mountains c. 300 million years ago; darker blobs are major ranges; Castle Valley location approximate (modified from Soreghan et al. 2009).

Being evanescent creatures, it's difficult for us to think of geologic structures as ephemeral. But the rock record clearly shows that they are. Even mountains have lifetimes. The great Uncompahgre Range is now gone, razed by erosion. The Paradox Basin also disappeared, filled to overflowing with sediments and then deeply buried. But in a sense both are still with us. Uncompahgre sediments are widely displayed in colorful rocks across the Colorado Plateau. And the Paradox Basin maintains a ghostly presence—dramatic, but difficult to explain.

After deposition of the last Paradox evaporites, the region was inundated by tropical seas—source of the impressive layers of limestone, sandstone, siltstone and shale that line Plateau drainages. Then about 200 million years ago, when the supercontinent Pangaea started to come apart, there was a shift to terrestrial deposits—dune sand, volcanic ash, and sediments from rivers, lakes, and inland seas. Under this immense "lithic layer cake" lay the Paradox salt, deeply interred but not dead (source).

Salt is a strange kind of sedimentary rock. Sediments such as sand and mud can be compressed to form dense rocks, but salt remains nearly unchanged under pressure. It's weaker and less dense than the rock around it, but also plastic—with properties of both solids and liquids. It can flow to escape from its "stressful surroundings", deforming any rocks in its way (source). Deformation can occur at the scale of landscapes, and the old Paradox Basin has many fine examples: fins and arches of the Fiery Furnace, closely-spaced meanders on the Colorado River, and at least seven parallel northwest-trending valleys, including Castle Valley.

Dotted elongate blobs are parallel northwest-trending valleys. Blue circle marks intriguing overlap of Castle and Spanish Valleys with La Sal Mountains (3). Modified from Doelling 1985.

Castle Valley begins at the base of the northern La Sal Mountains and extends northwest about 12 miles. It's a broad valley, to 2 miles wide. The southwest wall is capped by a nearly continuous outcrop of the erosion-resistant Wingate Sandstone, known as Porcupine Rim. On the other side of the valley, the wall is less continuous but equally dramatic—carved into mesas, buttes, rimrock and spires.

Porcupine Rim—Jurassic Wingate Sandstone caps southwest wall of Castle Valley.

Northeast side of Castle Valley; strata tilting away from valley center visible at arrow.

Geologists find Castle Valley intriguing. The floor is much broader than would be expected for its little streams, and quite flat. Rock layers on both sides of the valley tilt down away from the valley center (more easily seen on the northeast side, photo above). Most exciting is what lies beneath the surface. Wells drilled in the center of the valley revealed a long steep-sided bed of salt to 1000 feet thick! Castle Valley must be a salt anticline, an elongate convex uplifted fold cored by salt. On this geologists agree. But as to how it formed and what happened to it ... that's another matter.

Castle Valley (Google Earth). But where's the anticline?! Valley walls hint at what happened.

The diagram below shows a common explanation for salt anticlines. In the top panel, flowing salt accumulates to form a convex fold, pushing up overlying rock layers. Castle Valley salt is thought to have flowed and formed an anticline 300 to 200 million years ago (Ornduff 2006, Trudgill 2011).

Salt anticline in cross section; at the time of the top panel, Castle "Valley" would have been a long ridge.

Now the Castle Valley anticline is mostly gone. The second and third panels in the diagram show a possible demise, but first, a major disturbance very close by needs to be considered—uplift of the La Sal Mountains about 28–25 million years ago (3).

The La Sals are not a mountain range but rather clustered peaks. They're similar to volcanoes except that magma never reached the surface. Pioneering geologist AC Peale called them "eruptive mountains of a peculiar type ... igneous and yet non-volcanic". Recent studies indicate that magma stopped just 1–3 miles below the surface, making them shallow intrusions, specifically laccoliths.

La Sal Mountains rise 8000+ feet above the Colorado Plateau—a major disturbance! (source)

La Sal high country: La Sal Peak (right) is intruded trachyte; Castle Mountain (left) is still capped with sedimentary rock (Ross 1998).

Now we're faced with another question. If magma never reached the surface, why are the La Sal "intrusions" visible? Instead of being 1–3 miles below the surface, their tops stand over a mile above the Plateau. The likely answer is the recent uplift and erosion of the Colorado Plateau mentioned at the beginning of this post.

Starting about ten million years ago, both the Colorado Plateau and the Basin and Range Province (to the west) have been rising. But while the latter was stretched and faulted, forming its eponymous basins and ranges, the Plateau remained a single block. Eventually it rose about kilometer higher than the Basin and Range. Why? That's a puzzle not yet solved (source). In any case, streams were steepened and invigorated, and erosion sped up enough to reveal the La Sal laccoliths.

Recent uplift and erosion probably explain the demise of the Castle Valley anticline as well. Erosion and/or faulting of overlying rock would have exposed the salt to water. Being salt, it of course dissolved. When enough was removed, rock layers at the crest fractured and collapsed, creating a breached anticline. But others think differently. Regional extension may have been the cause, perhaps related to ongoing extension in the Basin and Range Province. Or as Naqi et al. (2016) safely concluded, "formation of the salt valleys might be attributed to multiple factors (i.e., extensional forces, salt dissolution, and internal salt flow) rather than a single mechanism."

A salt anticline's demise may start with salt dissolution, followed by collapse of rock layers at the crest.

Breached anticline; dashed line shows former continuity across crest (Grabau 1920, A Textbook of Geology).

Let's visit!

Castle Valley is a great destination for geotrippers. Enough remains of the anticline to see and appreciate what happened. The tilted rock layers of the flanks are now the valley walls. Imagine them reaching higher and arching across the broad floor. Consider the depth of the valley below the now-imaginary crest and think about how much salt and rock must have been removed! Then look toward the head of the valley, at the dark hill rising from the floor. That's Round Mountain—a little relative of the La Sal intrusions. It was exposed when the anticline was breached and deeply eroded.

Castle Creek Road (paved) runs the length of the valley, providing easy access. Spires, buttes and mesas on the northeast side can be reached from several parking areas. Round Mountain is a short distance south of Castle Valley Road via a rough eroded 2-track; I parked just off the paved road and walked. Tour the southwest side via the Porcupine Rim trail—highly recommended, though maybe not on weekends and holidays.

Castleton Tower is a short hike from Castle Valley Road.

La Sals on left, Round Mountain on right, rabbitbrush in foreground.

En route to Round Mountain, Porcupine Rim beyond.

Looking up Castle Valley from Porcupine Rim, La Sal Mountains on horizon; redrockrubi.

Refreshing shade, courtesy uplift and erosion of the Colorado Plateau.

Notes in addition to links in post

(1) Cyclicity of Paradox deposition is well documented, but the cause is debated. Glaciation seems to be most popular, specifically sea level change with alternating glacial and interglacial periods. Other possibilities include rise of the Uncompahgre Range and climate change. Trudgill (2011) concluded that glaciation-driven changes in sea level was the main cause; tectonics and/or climate change may have made lesser contributions.

(2) In my reading, I found a range of estimates for Paradox deposition: 29 or 33 cycles over 8 to 15 million years, producing evaporites 4000, 6000 or 8000 ft thick (Ornduff et al. 2006, Trudgill 2011, USGS).

(3) Geologists have long wondered whether the La Sal Mountain intrusions and salt anticlines such as Castle Valley are related. Thomas Harrison, who surveyed the Paradox Basin area in 1926, discussed the possibility in his report (1927):

"It is interesting to know that igneous intrusive rocks of very considerable importance are closely associated with the saline anticlines. The laccolithic La Salle Mountains occupy an area on and between two parallel anticlines ... A small isolated igneous stock [Round Mountain] surrounded by gypsum occurs in the Castle Valley salt [anticline]."

Harrison noted that salt anticlines were thought to be associated with lines of weakness dating from Precambrian time. Their uplift was followed by subsidence, and massive accumulation of sediments. Perhaps this "heavy load" generated heat that "liquefied rocks within the zone of fracture, resulting in the [magma] which formed the laccolith, and the Castle Valley stock" [Round Mountain].

Today's geologists may chuckle at the idea of a "heavy load" of sediments melting igneous rock below. In contrast, "lines of weakness dating from Precambrian time" are taken seriously. Many northwest-trending faults cut basement rocks in the Paradox Basin area. Ross (1998) concluded that "the locations of the La Sal Mountains intrusive centers along the trend of subsurface faults ... suggest that the faults were avenues of weakness for the ascent of magma in the upper crust. This is especially true for the northern and southern clusters of peaks [which coincide with salt anticlines]".

Sources

Doelling, HH. 1985. Geology of Arches National Park. Utah Geological Survey, to accompany Map 74. PDF

Harrison, TS. 1927. Colorado–Utah Salt Domes. Am. Assoc. Petroleum Geologists 11:111–133.

Ornduff, RL, Wieder, RW, Futey, DG. 2006. Geology Underfoot in Southern Utah. Mountain Press Publishing. For salt anticlines see Vignette 29, A Sea of Fins; for La Sal Mountains see Vignette 32, Intruders in a Sedimentary Domain.

Ross, ML. 1998. Geology of the Tertiary intrusive centers of the La Sal Mountains, Utah; influence of preexisting structural features on emplacement and morphology, in Laccolith complexes of southeastern Utah; time of emplacement and tectonic setting. USGS Bull. 2158: 61-83. PDF

Snyder, NP. 1996. Recharge area and water quality of the valley-fill aquifer; Castle Valley, Grand County, Utah. Report of Investigation 229. Utah Geological Survey. PDF

Soreghan, GS, et al. 2009. Hot fan or cold outwash? Hypothesized proglacial deposition in the upper Paleozoic Cutler Formation, western tropical Pangea. J. Sed. Res. 79:495-522.

Trudgill, BD. 2011. Evolution of salt structures in the northern Paradox Basin: controls on evaporite deposition, salt wall growth and supra-salt stratigraphic architecture. Basin Research 23:208–238. https://onlinelibrary.wiley.com/doi/10.1111/j.1365-2117.2010.00478.x

US Geologic Survey. Mid 2000s. Geologic Provinces of the United States: Colorado Plateau Province. Internet Archive WayBackMachine.