Posts Tagged ‘Carolina Bays’

Permafrost as far South as Georgia during the Last Glacial Maximum

June 4, 2020

This is at least the 7th article I’ve written about Carolina Bays, but I keep coming across new and fascinating studies of these curious geological features. (See: ) These oval shaped depressions occur across the Carolinas and Georgia.  Their origins baffle scientists, but the commonly accepted explanation is they are topographical formations resulting from Ice Age wind and water erosion.  (Extraterrestrial explanations can be ruled out because Carolina Bays are of different ages, and there are 500,000 of them compared to just 250 known impact craters on the entire earth’s surface.)  I’ve long understood how wind and water erosion shaped the depressions, but I’ve never been satisfied with explanations for how the land initially subsided.  Some think wind simply blew unconsolidated sediment out of the pits, while I’ve suggested the land subsidence occurred due to peat fires (as occasionally occurs today).  In a new book Chris Swezey of the U.S. Geological Service proposed the initial subsidence of Carolina Bays was caused by discontinuous patches of permafrost that thawed during summers and collapsed.

Formerly, scientists thought permafrost (ground that stays frozen year round) extended as far south as northern Virginia during the Last Glacial Maximum, but Swezey believes there were patchy discontinuous areas of permafrost as far south as Georgia.  Carolina Bays resemble geological features found in southern Alaska today where permafrost is scattered and temporary.  The land swells and collapses and fills with water in oval depressions.  Northern Alaska hosts continuous permanent permafrost.

Millions of Arctic Methane Hotspots Detected by NASA – Global ...

Discontinuous patches of permafrost create lakes in southern Alaska that resemble Carolina Bays located in the upper coastal plain of Georgia and the Carolinas.

Average temperature and sea surface level through 35,000 years related to 1990 level

Temperature graph showing average temperatures and sea level fall during the Last Glacial Maximum. Note the dips at about 27,000 and 24,500 years BP.  This is when discontinuous permafrost could have developed on some Georgia and Carolina soils.

I believe this map is misleading.  It shows the southern extent of the boreal forest zone, but zonal forest types as we know them today didn’t exist then.  From the below referenced paper.

Georgia must have been much colder during Ice Ages than I thought.  Most Carolina Bays formed between 35,000 years BP-15,000 years BP when glaciers expanded to cover Canada and New England.  Some date to earlier stadials of the Wisconsinian Ice Age.  Patchy permafrost in the Carolinas and Georgia likely occurred during especially cold phases of the Ice Age that probably lasted for decades rather than centuries.

Landscapes in Georgia during the LGM must have been varied and interesting.  Wetlands on Carolina Bays likely attracted summer populations of ducks, geese, and swans.  Sand dunes from dried out riverbeds rolled over the land, smothering mixed woodlands of pine, spruce, and oak.  Arid conditions favored grasslands that fed horse, bison, and llama; in turn pursued by giant lions and dire wolves.  Strange as it may seem, caribou and stag-moose ranged into this latitude.  Zonal vegetation as we know it didn’t exist then.  Instead, habitats were patchy and species compositions were dissimilar to those of any existing types of forest.  Local microclimates might favor oak thickets, open spruce woodlands, mature pine forests, grassy meadows, small marshy wetlands, or bare soil.  Less than 100 miles east of the inner coastal plain the climate was markedly warmer.  Land extending into what today is the Atlantic Ocean  hosted more warm weather species of plants and animals because it was closer to warmer ocean currents that moderated coastal climates.  Inland, the boundary between cold and warm climates frequently fluctuated, contributing to the patchy unstable environments unlike those of today.


Swezey, Chris

“Quaternary Eolian Sand Dunes and Carolina Bays of the Atlantic Coastal Plain Province, USA”


Inland Dunes of North America

edited by Lancaster, Nicholas and Patrick Hesp

Springer Books 2020



Inner Coastal Plain Deserts of the Ice Ages

October 4, 2017

A new study reinforces evidence, indicating some regions of southeastern North America were harsh environments during climatic phases when the ice sheets that covered Canada were expanding.  The scientists who wrote this paper took cores of sediment from 2 Carolina Bays (Jones and Singletary Lakes) located in Bladen County, North Carolina. Carolina Bays are elliptical depressions found on the Atlantic Coastal Plain that were formed during Ice Ages.  They were created by a combination of peat fires, and wind and water erosion.  The peat fires lowered the elevation, wind blew out the dried unconsolidated sediment, and wind-driven water shaped them into elliptical formations.  Jones and Singletary Lakes were also studied in the early 1950s in 1 of the first paleoecological studies of late Pleistocene environments of the south.  The new study analyzed pollen composition, charcoal abundance, and biomass; and the authors compared their results to the earlier study.  The data was dated using radio-carbon dating.

Image result for Bladen County, North Carolina

Location of Bladen County, North Carolina.  This is the site of the study areas discussed in this blog entry.

Image result for Singletary Lake, north Carolina

Photo of Singletary Lake, a Carolina Bay.  Scientists took a sediment core at the bottom of this lake and analyzed pollen, charcoal, and biomass abundance over the past 50,000 years.

Between ~60,000 years BP-~30,000 years BP climate fluctuated drastically between warm wet interstadials and cold arid stadials.  The glaciers covering Canada advanced then retreated then advanced again in fits and starts.  During glacial expansion more of earth’s atmospheric moisture became locked in glacial ice, causing prolonged droughts, but this moisture was released when glaciers were in a meltwater phase.  Oak and grass pollen increased during meltwater phases, and so did charcoal abundance.  An increase in vegetation meant there was more biomass to ignite and burn during electrical storms.  Oak and grass were fairly abundant from ~43,000 years BP-~32,000 years BP.  The environment mostly consisted of woodland and grassland during interstadials,  but about 30,000 years BP the situation deteriorated.

Ice sheets maintained a steady expansion from ~30,000 years BP-~21,000 years BP.  The initial drought that struck the region during this phase killed vegetation and caused a temporary spike of charcoal because the dead biomass was so flammable.  But after this initial spike, fire was rare to non-existent here.  Sand dunes rolled across the landscape because much of the region was sparsely vegetated.  I believe scrub oak thickets with thorny plants adapted to arid climates covered much of the landscape, but this type of environment doesn’t produce much pollen.  Thus, the amount of vegetation on the landscape then is understated in the pollen record.  For this reason I don’t believe the landscape was as bare as the authors of this study concluded when they wrote it was a “windswept sandy desert with riparian communities of pine and oak.”  Nevertheless, it was an harsh environment of thorny thickets interspersed with areas of bare soil and long distances between water and wetland environments where some trees and grass still grew.  Some tough species of mammals that could survive in this type of environment included horse, flat-headed peccary, helmeted musk-ox, and hog-nosed skunk.  Bison evolved into a smaller species more capable of living in a drier natural community. Overall, wildlife populations probably declined during this climatic phase.

About 21,000 years ago, the ice sheets began retreating and precipitation increased.  Oak and grass gradually increased in abundance, and eventually mesic species such as cypress, basswood, hemlock, and beech invaded the resulting wetter habitats.  ~12,000 years ago, man colonized the region and overhunted megafauna into extinction.  Human-set fires combined with an increase in biomass not being consumed by megaherbivores caused a great increase in fire frequency.

I’m skeptical of 1 claim made by this paper.  The authors estimated the average annual temperature and precipitation levels based on plant composition assumed from the pollen record.  During the Last Glacial Maximum they estimated the average January temperature at these sites was 20 degrees F, while the average July temperature was 68 degrees F.  However, they use 2 dubious assumptions.  They believe the pollen grains from northern species of pine can be distinguished from those of shortleaf pine, a southern species.  This is a doubtful assumption that I will examine more thoroughly in my next blog entry.  Moreover, the spruce pollen probably originated from an extinct species of temperate tree known as Critchfield’s spruce.  I don’t think they can estimate average annual temperatures based on pollen composition, unless the exact species are known with more certainty.

The outer coastal plain and the continental shelf, which was above sea level from ~80,000 years BP-~7,000 years BP, likely hosted richer environments than the inner coastal plain during stadials.  Sea breezes and weather fronts spawned in the Atlantic Ocean brought more moisture to the coast, allowing this region to maintain a mosaic of woodland, grassland, and wetland; while the inner coastal plain suffered greater aridity.  These fronts usually dissipated before they reached the inner coastal plain.  The coastal region likely served as a refuge for plants and animals that later re-colonized the inner coastal plain when climatic conditions improved.


Spencer, Jessica; et. al.

“Late Quaternary Records of Vegetation and Fire in Southeastern North Carolina from Jones Lake and Singletary Lake”

Quaternary Science Review 174 October 2017

McCullough Millpond, Burke County, Georgia

March 17, 2014

I planned to visit Big Dukes Pond in Jenkins County, Georgia, but after 3 days of rain, the access road to this Carolina Bay was impassably muddy.  Carolina Bays are interesting geological formations created by a combination of peat fires and wind and water erosion.  Carolina Bay formation was especially prevalent during the last Ice Age (See  As a consolation, I visited McCullough Mill Pond in Burke County where I’d spotted a few wood storks while I drove to Big Dukes Pond.  This body of water is near the Burke County/Jenkins County line and is not far away from Big Dukes Pond.  Both host the same species of birds.

Map of Georgia highlighting Burke County

Location of Burke County

Map of McCollough Mill Pond

I can’t find any information on the history of McCollough Mill Pond, a property of Burke County.  It’s a manmade body of water formed by the damming of a creek that flows into the Ogeechee River.  The stream doesn’t seem to be rapid enough to have ever been involved in the operation of a grist mill, but I must be wrong.

Wood storks are federally endangered.  They used to be common in Florida, but most of the larger colonies have been forced to move to Georgia and South Carolina due to suburban development.  Two major nesting colonies occur within a 40 minute drive from my house–Big Dukes Pond and the Silver Bluff Audubon Center in Jackson, South Carolina.

New Pictures 049

Photo I took of the wood storks (Mycteria americana) at McCollough Mill Pond.  Click to enlarge. They are in the center way back.

Fort DeSoto County Park, Pinellas County, Florida, USA - Nov 15, 2006 © William Hull

I found this much better photo of wood storks on google images.

New Pictures 050

Cypress trees, willows, and red maples grow in and around McCollough Mill Pond.

Wood storks require large areas of shallow water where they can use their big ugly bills to forage for small fish, tadpoles, frogs, and crayfish.  Mark Catesby, an early 18th century naturalist, saw wood storks foraging on open pine savannahs after heavy rains caused flooding in this type of environment.  Vast agricultural fields have replaced much of this habitat.  Farmers dig drainage ditches, so their fields of cotton, peanuts, and corn will drain in time for spring planting.  This is not good for wood storks, though crayfish and amphibians can still survive in the deeper ditches.

I always see many birds when I drive through Jenkins County.  In fact, the name of the road on the south side of Big Dukes Pond is “Birdsville Road.”  Last summer, I even saw a rare swallow tailed kite flying high in the sky in this county.  On this occasion I counted 18 species between Big Dukes Pond and McCollough Mill Pond including wood storks, great blue heron, common egrets, turkey vulture, black vultures, a kestrel, broad-winged hawk, red-shouldered hawk, eastern kingbird, mockingbird, cardinal, crow, mourning dove, robins, and at least 3 small species that wouldn’t stay still long enough to allow me to identify them.  The agricultural fields that form the most common type of environment in this area are ideal hunting grounds for hawks.  The leftover grain attracts rodents, but the open grounds don’t afford much cover.  Still, it’s a poor substitute for the original stands of longleaf pine savannah and extensive wetland sand cypress swamps.

New Pictures 052

The pond on the other side of Highway 56.

New Pictures 053

I think the small tree in the foreground is some type of bay tree, a characteristic species of Carolina Bays.

Bay trees, the characteristic tree of Carolina Bays are in big trouble.   The invasive ambrosia beetle (Xleborus glabratus) is killing these trees wherever this insect spreads.  It’s probably only a matter of time before it reaches this area.

A Young Carolina Bay in North Carolina

April 30, 2012

Thousands of oval-shaped depressions, known as Carolina Bays, pockmark the coastal plain from Virginia to Georgia.  Originally, there were about 500,000 Carolina Bays on the southeastern coastal plain, but farmers fully or partially drained most of them.  They range in size from a small fraction of an acre to tens of thousands of acres.  Some stay wet year round, while others may only hold water for a few weeks of the year.  Small bays devoid of egg and tadpole-eating fish are ideal breeding grounds for amphibians.  The wetland habitats, even if seasonal, provide critical environments for aquatic species, including rare flora, such as pitcher plants.

Aerial photograph of some Carolina Bays in South Carolina.  Most geologists believe they are formed from a combination of wind and water erosion as well as peat fires which lower the elevation of the depressions. 

A minority of scientists believe meteor or comet impacts created Carolina Bays, but convincing evidence debunks this theory.  There is a complete absence of the kinds of rock that are associated with extraterrestrial impact.  Moreover, the ages of Carolina Bays vary greatly.  Not all have been dated, but some are tens of thousands of years old, while others may be as young as 6500 years.  Instead, Carolina Bays were likely formed from a combination of wind and water erosion and peat fires.

Stilted trees growing in a dried out Carolina Bay.  A fire just burned the peat causing the elevation to decrease about 4 feet.  The tree roots appear as stilts.

All Carolina Bays are oval in shape and are oriented perpendicular to the prevailing winds of the Pleistocene.  Much of the loess, or sand constituting dunes in the southeast, may have even originated from Carolina Bays.  The two geological anomalies are closely interrelated.  During dry climate phases, peat swamps became dessicated.  Lightning strikes ignited fires that burned off all the peat, thus lowering the elevation by as much as 4-5 feet.  Wind blew the exposed soil to the northeast, explaining why a sandy lip can be found on the northeast side of a Carolina Bay.  Later, after the rains returned, the water table rose and ponded water driven by wind also eroded land along the same axis.

Most Carolina Bays formed during stadials of the Pleistocene, especially the Last Glacial Maximum (~28,000 BP-~15,000 BP), but Lake Mattamuskeet, an enormous Carolina Bay near the North Carolina coast just a few miles from Pamlico Sound, is a young one with origins dating to about 6500 BP.  Scientists knew the lake had to be younger than 80,000 years old because a high stand of the Atlantic ocean inundated this area until then.  When they took cores down to that aged level, they found a layer of marine sediment, including saltwater species of clam and snail shells.  Above this layer they found peat.  Charcoal and estuarine silt (also known as loess) was mixed in with the peat.  The former is evidence of peat and forest fires; the latter was windblown from what is now Pamlico Sound, but was then high and dry land because during the Ice Age the Atlantic Ocean receded many miles to the east, leaving exposed marine-derived soil.

The scientists used ground penetrating radar, vibracores (, and radiocarbon dating to determine that the present day site of Lake Mattamuskeet was a heavily vegetated forest from about 12,000 BP-8500 BP.  Peat swamps grew in 3 areas within this site.  Periodically, they would burn, and silt from dry land or what is now Pamlico Sound would be deposited here via wind.  About 7,000 BP the ocean level rose and filled in Pamlico Sound.  Lake Mattamuskeet began forming about 6500 years BP during dry spells when peat fires left depressions of dry earth further scooped out by wind.  These peat fires still occasionally occur as the above photo shows and they can last for over a year as they slowly smolder.  An Indian legend even supports the scientific study of how Lake Mattamuskeet formed.  Supposedly, the Indian legend claims a great fire burned here for 13 moons, creating a depression that later filled with water.  According to scientists, the 3 bays stopped forming ~5,000 years ago, and the water table rose, joining the 3 lakes into 1.

Lake Mattamuskeet in North Carolina is a Carolina Bay within a few miles of the sea shore.  It’s a shallow freshwater lake though a manmade canal was constructed in 1850 in a failed attempt to drain the water into the saltwater Pamlico Sound.  In this photo it looks as deep as a reservoir, but the depth is only 2-3 feet deep.  Reportedly, there is good fishing, crabbing, and duck hunting here.  Scientists studied the geological history of this Lake and determined it’s just ~6500 years old. 


Rodriguez, Antonio; Matthew Waters, Milas Pehler

“Burning Peat and Reworking Loess Contributes to the Formation and Evolution of a Large Carolina Bay Basin”

Quaternary Research 77 (1) Jan. 2012