THE FACTORS OF SWAMP FORMATION

 In 1807 Alexander von Humboldt published Essay on Plant Geography. Humboldt made major contributions to botany, zoology, meteorology, oceanography, and anthropology, but is in particular known at the “father of geography.” The Essay is particularly remembered for relating vegetation to climate and elevation (Humboldt studied vegetation on Chimborazo, a 6264 m mountain in Ecuador with tropical rainforest at the base and a snow-capped peak). However, Humboldt (and his collaborator Aimé Bonpland) also linked plant distributions to geology, soils, geophysical phenomena, and human impacts. Earth and ecological scientists ever since (and probably before) have used a “state factor” approach to explain the spatial variation of various phenomena and how and why the plants, soils, climate, or other phenomena in different places differ from each other. It’s a fundamentally geographic logic that says, in essence, if the factors that determine the type of, e.g., vegetation, soil, or climate in different places are the same, then the plants, soil, or climate should also be very similar. 

In the late 19^th and early 20^th century, soil scientists in Russia and a bit later in the U.S.A., formalized the concept, the best-known statement being the so-called “clorpt’ equation:

where S is soil or some soil property, cl is climate, o for organisms indicates biological effects (originally focussed almost entirely on vegetation, r (relief) indicates topography, p is parent material, and t is time, usually thought of as the time available for pedogenesis. The trailing dots indicate the possibility of other factors not encompassed in cl, o, r, p, t that influence soil formation in particular cases, but not necessarily in general (for example sea-level change or contaminants). 

Swamp soil along the Waccamaw River, SC

The state factor conceptual model is to this day the underlying framework for soil mapping and continues to be used, usually implicitly but sometimes explicitly, throughout the ecological and Earth sciences. State factor approaches have been subject to considerable debate in pedology, mainly centered on three general areas. One is the independence of the state factors—clearly they are not independent of each other or of the soil itself, but the debate concerns whether they can be treated as though they are in certain contexts, and how to disentangle their effects. Second is how to actually use factorial models for prediction, given that each state factor may itself have multiple variables to describe it. Third is the association of the state factor model and related ideas with practices in soil taxonomy, pedogenesis studies, and the general praxis of soil science (I don’t like that word because it seems so pretentious, but I didn’t want to use “practice” twice in the same sentence) that are thought to have inhibited other approaches. With respect to the latter, soil state factors have sometimes been presented in opposition to studies of soil processes. The two approaches are actually complementary, with the state factors providing the environmental context and boundary conditions within which processes operate, and processes explaining how the state factors work. 

With that out of the way, what are the factors of swamp formation? 

Let’s start with climate. Though small, isolated wetlands can occur even in deserts if there are local spots where water accumulates and wetness lingers, a full-on swamp requires a humid climate, wet enough to support trees, particularly hydrophytes. Climate is only one factor that influences water availability, illustrating right off the bat the interdependence of state factors. Temperature regimes will also influence swamps—for example, consider the distribution of swamp tupelo (Nyssa biflora) shown below. You don’t find the tree north of central Virginia, or west of the 100^th meridian (which corresponds with the straight line on the left, a general divider between humid and drier climates in the U.S.A.). The concentration on the Atlantic and Gulf of Mexico coastal plains hints at elevation and topographic-related factors—interdependence again. 

Distribution of Nyssa biflora (swamp tupelo) based on U.S. Forest Service Forest Inventory and Analysis data (https://www.fs.usda.gov/nrs/atlas/tree/694).

Another obvious candidate is topography. Swamps will occur where water can collect and persist; at lower elevations, gentler slopes, and in depressions. Closely related, but also linked to factors other than topography and climate is hydrology, including influences such as flow regimes, tides, and groundwater dynamics. 

Differences among swamps are also defined by vegetation. Plant communities are obviously related to climate, hydrology, topography, and soils, but also to other factors such as seed sources, dispersal mechanisms, disturbances (e.g., floods, storms, fires, pests, logging), and ecological interactions among plants and other biota. Perhaps animals could be lumped with plants as an organisms or biota factor, but one can also argue that the presence or absence of critters such as beavers, alligators, nutria, feral hogs, and others (not to mention Homo sapiens) justifies fauna as a separate state factor. 

Swamp along the Black River, NC

Soils are another state factor. In the swamps I have worked and played in the Carolinas and Texas/Louisiana the main differences among soils are associated with the texture (particularly sandy vs. clayey vs. organic and combinations thereof), drainage class (obviously related to hydrology), geomorphic environment (e.g., floodplains, terraces, infilled oxbows, depressions), and presence or absence of specific soil features such as argillic horizons (clay-rich subsoils). Finer distinctions are related to mineralogy and pH. 

Independently of closely related topographic and hydrologic factors, I consider geomorphology a state factor. This includes the formation and modification of specific features such as natural levees, oxbows, and point bars, and the extent to which the setting is erosional, depositional, or both/either. In the classic form of the soil state factor equation, the type of swamp or its characteristics (S_w) is a function of climate (cl), topography (r for relief), hydrology (h), vegetation (v), fauna (a for animals), soils (s), and geomorphology (g):

This is a conceptual framework, not an equation to be numerically solved. For one thing, there are any number of specific variables or indicators that could be associated with each of the factors. For another, as we have seen, all the factors are interrelated and interdependent.

I will explore this framework further in future posts. 

CAVITY SEARCH

Tree cavities are important habitat for many types of wildlife, and the swamps of the southeastern U.S.A. are no exception. Furthermore, some of the trees in those swamps have some hellacious cavities. In the Carolinas these may be home to birds such as wood ducks and other ducks, promontory warblers, chimney swifts, and pileated and other woodpeckers. The apparently extinct ivory billed woodpecker called such cavities home. Cavities in swamp cypress and tupelo trees also host some important bat species such as the big-eared and the mouse-eared bat. River otters, black bear and alligators may also den in such cavities. I have read that beavers may also den in tree cavities, but I have observed no evidence of this, even in swamps with many beaver and many tree cavities. 

This post is basically a photo album of some of the best and most interesting cavity trees I've seen in the coastal plain swamps of North and South Carolina.

Grinnell Creek, NC

Great Pee Dee River, SC

Neuse River backwaters, NC

Black River, NC

Crabtree Swamp, SC

Enterprise Creek, SC

Holly Shelter Creek, NC

Northeast Cape Fear River, NC

Core Creek, NC

Pinetree Creek, NC

RAVINE SWAMPS IN THE MULTIVERSE

Any landscape, ecosystem, hydrological system, etc., in the grand scheme of things, is a historically contingent snapshot. Evolution, from genes and alleles to the entire Earth system, is ongoing. Change is constant, in climate, geology, ecology, and so on. And disturbances happen—hurricanes, floods, tornadoes, fires, earthquakes, human activities, and so forth. Each historically contingent environment represents a unique outcome of a single developmental pathway—but these are only one realization of what could or could have happened. All the possible pathways constitute an evolutionary potential space, which I call the geographical multiverse. The multiverse term plays off the idea of manifold timelines, familiar to quantum physicists, science fiction fans, and viewers of Rick and Morty. The “geographical” tag emphasizes the concern with environments that we experience on our planet, as opposed to the cosmological or subatomic scales typically associated with physicists and philosophers takes on the topic. 

I’ll leave the particulars of the arguments for other times and writings (but if you are interested, feel free to e-mail me). I’m a firm believer that anyone proposing theoretical notions, at least in the environmental and Earth sciences should be able to “walk the walk” with some real-world examples, and at last we arrive at the point of this post, which leans a bit toward the scientific side, though there is not too much jargon. An empirical example of the geographical multiverse notion is presented for ravine swamps along the Neuse River estuary in Craven County, North Carolina. The area features shoreline bluffs that stand about 10 m above mean water levels. These are the valley side slopes of the Neuse River drowned by Holocene sea-level rise. They were fluvially dissected during lower sea-level stands in the Pleistocene. The bluffs are therefore interspersed with steep-sided valleys containing hardwood swamps, typically perched atop clay- and organic-rich swamp soils approximately a meter above mean low water. These are ravine swamps, dominated by water tupelo (Nyssa aquatica) and bald cypress trees (Taxodium distichum)(Figure 1).

 

Figure 1.  Flanner Beach Swamp (A) and Tadpole Creek ravine swamp (B), photographed in winter. In (C) the steep slope between adjacent upland and Tadpole Creek is shown. The uprooted trees shown were blown over during Hurricane Florence in 2018. 

The ravine swamp example was chosen because it illustrates how the multiple pathway concepts apply even to small areas, and as a convenient illustration because several different landscape system states are evident within a small area. The ravine swamps are also affected by press disturbances (principally sea-level rise) and pulse disturbances, mainly tropical and extratropical cyclones. I have previously studied the impacts of Hurricane Florence on the ravine swamps and have regularly observed them for the past decade. 

 

Figure 2 below shows the topography of the study area, with the fluvially dissected valley sides, and the tributary drainages truncated by drowning of the Neuse valley during Holocene and contemporary sea-level rise and by shoreline retreat along the Neuse estuary. There are three general types of ravine swamps. The smallest, less that 0.1 km² in drainage area, are generally seasonally and occasionally flooded, with an intermittent discharge to the estuary. Larger ravine swamps such as those shown within the box on Fig. 2, are permanently flooded, with at least some standing water even during droughts, and at least some drainage to the estuary except during droughts. Larger ravine swamps such as Otter and Dam Creeks shown on the map, maintain permanent connections to the Neuse estuary. At their mouths fluvial inflow to or backwater flooding from the Neuse may occur. Larger tributaries (generally with drainage areas >100 km²) also occur, but these are not considered ravine swamps. The focus is on two ravine swamps the middle category, Tadpole Creek and Flanner’s Beach Swamp (these are not formally or officially named; names used by locals are applied)(Figure 3). 

Figure 2. Shaded relief map of the study area derived from 10 m horizontal resolution digital elevation model data (3DEP from the U.S. Geological Survey). The boxed area is shown in Figure 3.

Figure 3. Google Earth^TM image taken in February, 2024 showing the Tadpole Creek and Flanners Beach ravine swamps. 

The ravine swamps differ in five key respects, summarized in Table 1. With respect to hydrology, they may be constantly inundated or occasionally have little or no standing water. They may be regularly flowing or predominantly impounded, with only slow (except during wet periods) or intermittent outflow. The connection to the estuary may be a single dominant channel, or a dripline, where multiple small outflow points occur. Salinity also differs—though none of the ravine swamps are regularly brackish, some experience occasional low salinity due to high water levels in or storm overwash from the estuary. The swamps also differ in the state of valley infilling, though all are infilling over Holocene time scales. Most are dominated by organic matter and fine grained (silt and clay) swamp substrates, but some (along with segments of others) are dominated by sandy substrate due to storm overwash into the swamps and/or sapping erosion of the steep valley sides. The ravine swamps may have standing water right up to the edge of the upland valley sides or may have aprons of encroaching sediment around the edges, indicating some reduction of surface area. Vegetation is primarily water tupelo (or swamp tupelo, Nyssa biflora) and bald cypress, but other hydrophytes (e.g. Salix nigra, Phragmites australis; black willow and common reed). Cypress has greater salinity tolerance than tupelo, and so tends to be more dominant where salinity incursions occur. The occasionally-seasonally flooded swamps also feature other trees such as Acer rubrum (red maple) and wetness-tolerant oaks (Quercus spp.).

 

Table 1. Summary of the major factors determining the state of ravine swamps.

A key feature of the cypress and tupelo trees is that they have quite specific hydrogeomorphic conditions required for establishment. The species are hydrochores (that is, seeds are dispersed by water), so newly colonized sites must be accessible to seed transport and deposition by water. Both species are shade-intolerant, require moist conditions, and are most competitive vis-à-vis other trees where soil saturation or high water tables are frequent. Once established, tupelo and cypress can persist and thrive under conditions of constant inundation. However, they cannot germinate underwater, and submergence of the tops of seedlings will kill them. In the ravine swamps, sediment deposits from storm overwash often provide good conditions for establishment (Figure 4A). Because germination cannot occur underwater, the pioneer trees must have become established at lower sea-levels (with lower water tables), during severe droughts, or following storms, which along with sediment deposits may supply large amounts of woody debris which could serve as germination sites. Germination on downed trees since Hurricane Florence in September, 2018 by cypress and tupelo has not been observed, however, though black willow has become established. 

 

The Neuse estuary is part of the Pamlico-Albemarle Sound estuary. Due to relatively few small inlet connections to the Atlantic Ocean, astronomical tides are minimal and water level changes are dominated by wind, as shown in Figures 4B, 4C. SW winds pile up water on the east side of Pamlico Sound against the Outer Banks, lowering water levels. NE winds have the opposite effects on water levels, and given the exposure of the study area, wave attack on the shoreline also occurs. 

Figure 4. Recently deposited sand on the margin of Tadpole Creek ravine swamp with young bald cypress trees (A). Figures (B) and (C) show the cypress headland of Tadpole Creek, from upriver during strong southwest wind (B) and from downriver during strong northeast wind (C).

A more complete analysis of effects of the 2018 Hurricane is given in Phillips (2022). One key result was a transition of Tadpole Creek from a single dominant channel connection with the estuary to a dripline outlet. In 2024 Beaver moved into the site, damming the swamp at the outer edge and raising the water level about 1 m higher than before. At Flanners Beach Swamp, a dripline connection pre-Florence was converted to a single-channel outlet across deposited sand. Reduced inundation area due to marginal sedimentation was minimal at Tadpole Creek. Large areas of the Flanners Beach swamp were converted by Florence (Figure 5), and much smaller areas at Tadpole Creek. 

Figure 5. Outer edge of Flanners Beach Swamp before Hurricane Florence (A), following deposition of about 60 cm of sand during Florence (B), and in 2021 (C), looking toward the interior. The woody debris wrack line was deposited by a midlatitude cyclonic storm (nor’easter) in 2021. Note the pines in the background, which colonized a formerly inundated area covered by deposited sand.

During the most recent lower stand of sea-level in the region, the Neuse tributaries were able to cut channels and valleys, essentially dividing the upland landscape into incised channels and valleys, unincised channels at the upper, low-order portions of the channel network, and undissected areas. The incised areas experienced truncation during sea-level rise as the Neuse River valley was inundated (Figure 6). Some of these areas became ravine swamps. 

 

Figure 6. Possible evolutionary pathways in the study area. Larger gray arrows indicate multiple pathways leading to other (non-ravine swamp) states.

The watershed area of the tributary streams is determined by the original (pre-SLR) drainage area, modified by possible reorganization during SLR, when rising base levels differentially affect the streams. The lower reaches of larger tributaries are essentially drowned embayments of the Neuse estuary. These have drainage areas >100 km² (Slocum Creek, the largest tributary nearest the study area, has a watershed area of 130 km²). Smaller tributaries generally have drainage areas <10 km² (Otter and Dam Creeks have areas of 7.3 and 1.8 km²; see Figure 2). Tadpole Creek and Flanners Beach Swamp have drainage areas of 0.414 and 0.162 km², and the seasonal ravine swamps have drainage areas less than 0.1 km².

 

Those in the ravine swamp pathway shown in Fig. 6 then vary—and as outlined above, change—according to the aspects described above and shown at the bottom of the figure. Each of those criteria are determined and modified by multiple factors, the most important of which are shown in Figure 7. 

Figure 7. Major factors influencing five key characteristics of ravine swamps. SLR indicates sea-level rise. 

Five defining characteristics with two possible conditions each produces 32 possible states, in this case, with many potential transitions among them. Tadpole Creek ravine swamp at the moment is a constantly inundated swamp fed mainly by watershed runoff, secondarily by local water table fluctuations, and rarely subject to storm surge inputs, with a dripline connection to the estuary. Occasional salinity effects occur, favoring vegetation with some salinity tolerance (but not completely excluding species with little or no tolerance). Valley infilling is dominated by organic input from on-site vegetation and storm deposits of sand, with minor inputs from valley wall sapping. There is limited area reduction at present. The vegetation is cypress-tupelo swamp, with a transition in recent years from sawgrass (Cladium jamaicense) to Phragmites australis at the outer edge. A beaver dam is currently maintaining higher water levels and reducing flow velocities, allowing the floating aquatic plant duckweed (Lemna perpusilla) to proliferate.  Flanners Beach swamp, though only about 500 m away, differs in four of the five key characteristics.

 

This unique combination of characteristics is one of many possible states that could have (or could) develop. These were recently influenced by Hurricane Florence, whose impacts were strongly influenced by specific characteristics of the local and regional environmental setting, and specific characteristics of the size, track, and speed of movement of the storm (Phillips, 2022). Another large storm, departure of the beavers, a prolonged severe drought, or fire (for example), could quickly change the system state. 

 

The bottom line is that even when we consider a single type of relatively small feature, within a small area, multiple evolutionary trajectories and outcomes* occur. Ravine swamps are at home in the geographical multiverse. 

 

-----------------------------------------------------

 

*”Outcomes” does not imply any sort of final state; just the situation observed at present. 

 

Phillips, J.D. 2022. Geomorphic impacts of Hurricane Florence on the lower Neuse River: Portents and particulars. Geomorphology 397, 108026.