The role of fungus-culturing termites in turning Kalahari sand into an ecosystem of megaherbivores in the Okavango

(writing in progress)

See and

My references are McCarthy et al. (1991,

What I have learned from these studies:

The Okavango is not a delta, but an alluvial fan

The channels keep changing, as the accumulation of organic matter blocks them

Peat fires deplete this organkc matter

Most (96%) of the water evaporates, instead of sinking into the earth

Thus there is a pan-like effect, in which a considerable quantity of salt builds up in the Okavango

Islands are formed partly by termite mounds, via a mechanism of precipitation of solutes

Evaporation and precipitation are aided by the transpiration of trees

The centres of islands tend to be so saline that they are bare of vegetation

I infer that:

  • the accumulation of peat presumably acts as a bank for iodine, over the century of accumulation of peat in each channel, and
  • the concentration of calcium carbonate on islands, particularly based on termite mounds, would also act as a bank for iodine, in the form of iodate.

Iodine flows in, in extreme dilution, in the water of the Okavango River. It is somewhat concentrated by evaporation, and then greatly concentrated by evaporation from islands, aided by

  • transpiration from trees,
  • precipitation of calcium carbonate, and
  • binding to peat.

Iodine is thus stored in non-volatile forms, over time.
The peat fires that deplete the organic matter are mild, with combustion so slow that minimal smoke is produced. Thisnpossibly minimkses volatilisation, conserving much of the iodine.

Because the watershed of the Okavango River is flat and sandy, silt is scarce in the Okavango. Thus, remarkably, the main land-building material for the construction of islands is actually precipitated solutes, e.g. calcium.

The groundwater under the islands is actually saline.

I infer that the islands of the Okavango act almost like the pans of the semi-arid Kalahari, farther afield.

In both cases,

  • fairly fresh water surrounds the evaporational surface, which concentrates salt and calcium from the water,
  • ungulates are likely to seek nutrients by geophagy, perhaps at the periphery of the saline centre of the evaporative surface,
  • groundwater under the evaporative surface is likely to be saline, in contrast to the surrounding water at the same horizontal level.

Basically, both pans in the Kalahari and islands in the Okavango can concentrate iodine from the surrounding fresh waters, and store it in calcium-rich material at tne surface, where iodate is likely to be stabilised. In the case of the islands of the Okavango, leaching of IO3 by rainfall is likely to be prevented on termite mounds.

Peat layers aggrade by about 5 cm per year, to a cumulative depth of about 5 m, before the channel blocks and shifts elsewhere. This takes about one century.

"Water in the Okavango swamps is of the carbonate-rich variety typical of continental regiions" (McCarthy and Metcalfe, 1990).

Islands have precipitation of calcium carbonate from evaporation of the waters of the Okavango River. This precipitate is concentrated by termites.

In the capillary fringe area of islands, calcium precipitates out first (edges of islands), whereas sodium reaches the centre of the island in solution, precipitating out there to form a trona crust (

Island centres:
"During the rainy season, surface crusts are dissolved and leached down into the soil...the rainfall rarely contributes to the water table, and the rainfall and its dissolved salts are merely held in the profile...During the dry season, the salts are returned to the surface" (McCarthy and Metcalfe, 1990).

McCarthy and Metcalfe (1990): the observed concentrations of calcium in the Okavango are calculated to have required 22,000 years, to the nearest order of magnitude. Thksnis likely to err on the conservative side, because rates of evaporation are greater now than in the past.

My commentary:

  • I infer that these deposits are about as old as those in heuweltjies in the Western Cape province of South Africa.
  • Peat may possibly act as a trap and bank for selenium, as well as iodine.

Salt tends to accumulate in the centres of the islands, to be eventually leached to the saline groundwater table, or redissolved, leading to the recovery of vegetation on the saline surfaces.

The result:
Although certain zones do become saline, there are many opportunities for geophagy of iodate and calcium, without excess salinity.

McCarthy does point out, in certain of his papers, that the calcium precipitated on islands is brought up (and concentrated?) by termites (mounds?).

Notes from McCarthy, McIver and Verhagen (1991):

I infer that the whole island fringe acts like a termite mound, in the sense that it is

  • elevated,
  • colonised by trees,
  • replete with calcium carbonate,
  • unleached,
  • perpetuated by the precipitation of solutes, by virtue of transpiration, and
  • ?fire-free.

The centres of the islands act like land, in which bare ground develops because of salinity.

The fringe of the island is raised by the precipitation of silica and calcium originating in the waters of the Okavango River. The centre of each island is at a lower altitude than not only the fringe of the island, but also the water level of the surrounding marsh/swamp. This is because, in contrast to the fringe lf the island, the centre is actually deflated. Obviously, the level of the groundwater beneath the centre of the island is likewise lower than the water level of the surrounding marsh/swamp. This is true for all of the islands, regardless if the stage in the flood-cycle. The saline water in the centre of the island may actually sink into the briney watertable there, because saline water is denser (heavier) than fresh water.

(writing in progress)

Posted on June 07, 2023 04:30 AM by milewski milewski


The word 'pan' has a long history in southern Africa, referring to a flat, bare landform on plains, which is base-rich/sodic but not necessarily saline. However, this word has not achieved enough international recognition to qualify for Wikipedia, partly because 'pan' has so many other meanings in the English language.

Southern African pans are not adequately described by the next-best terms in Wikipedia ( and

With this in mind:
My reference is Alison (1899,

There are many pans in the Highveld of Free State and North West provinces in South Africa, with diameters ranging from 100 m to 1.6 km.

Of these pans, "many have small springs near their edges" (Alison, 1899).

This author's explanation is based simply on the action of the hooves of wild ungulates which have long been naturally abundant in the regions concerned.

There are many natural springs in the Highveld. Each forms a small marsh, attractive to wild ungulates. The commuting results in mud being exported by means of adherence to the hooves. This trampling and depletion of material leads to the formation of pans, according to Alison (1899).

In other words, Alison's hypothesis is that pans are essentially a function of springs plus trampling.

Alison (1899) points out that the anomalous lack of any alluvial deposits in these pans suggests continual removal by large animals: "it is a well known fact that without exception these pans have no alluvial deposit, or so little, between two to three inches, just about the amount that would collect since the country became civilised and the game was destroyed and driven away."

Alison (1899) views pans and large wild ungulates as mutually dependent.

I note that there is a large pan on the summit of Platberg (, near Harrismith ( in Free State province, and several others in similar situations, e.g. 'Inquela Mountain', near Newcastle (,_KwaZulu-Natal and

My commentary:
The vital link, overlooked by Alison, is the role of nutrients, particularly megacatalysts.

Also see:

Posted by milewski about 1 year ago

Laloy (1905) and Passarge (1911) agreed with Alison (1899) about the role of hooves in creating pans in South Africa.

Other authors disagreed.

Verhagen B Th (1991) On the nature and genesis of pans - a review and an ecological model. Palaeoecology of Africa 21: 179-194.

This author returns to the hypothesis of Alison (1899). On page 184, he presents a nice map of the distribution of pans, which form a belt from the southwest to the central Kalahari, and reach a maximum density of one pan per 22 square kilometres.

On page 182, Verhagen (1991) points out that the floors of even saline pans can provide fresh water, thus attracting ungulates.

"Hand-dug wells in the pan floor and margins can encounter useful fresh water supplies".

The quality of the groundwater in the Karoo beds beneath the pans is consistently poorer than that under the surrounding sandsheet.

"The salinity of Kalahari pan floors is therefore being leached downwards through the permeability of the underlying material".

A key to the formation of pans is the prevention of vegetational succession, by trampling and export of topsoil.

Most pans in the Kalahari are many tens of metres above the bedrock. If one drills into the floor of the lab, one may encounter perched aquifers in lenses of sand in the sandy clay of the upper part (7-9 m deep) of the Kalahari beds, and a small yield of poor-quality water from the upper layer, 50-60 m deep, of the underlying Karoo rocks.

Surrounds of pans do not nave extensive calcrete layers. "Calcretes are far from omnipresent, even in the southern Kalahari...and often appear in the vicinity of pans...probably generated in association with pans, rather than predate them".

My commentary:
I see a critical factor to be the salinity of the pans. Australia shows that salinas form, regardless of the absence of large animals. However, in South Africa, non-saline pans have formed. Hence, I have invoked the crucial role of megacataylsts.

Please see next comment for continuation...

Posted by milewski about 1 year ago

...continued from the last comment, w.r.t. Verhagen (1991)

This author portrayed pans in the Kalahari as generally 'rooted' in pre-Kalahari geological substrates. He found pan sediments under the pan, all the way down to the Karoo beds.

Pans are 'deep', not just superficial, landforms, but they are also ongoing, not just 'fossil' landforms.

"The sequence clearly had its origins in the pre-Kalahari depression", at least in the case of some pans. Salinity is maximal at the centre of the pan, a few metres deep. By contrast, carbonate is most concentrated at the edge of the floor of the pan.

Nitrate is concentrated in some pans.

Verhagen (1991) suggested a feedback loop in which ungulates mediate a cycle in which a pan becomes vegetated, then becomes bare again.

On page 190, he reiterates more articulately Alison's original argument, thus bolstering it:

Pan leads to salinity/alkalinity, which facilitates the deposition of calcium carbonate. When a certain threshold of salinity is reached, ungulates are disincentivised, and the pan tends to be filled in and vegetated. During unusually rainy periods, the saline water table rises, precluding the ponding of fresh water on the pan, disincentivising ungulates - as happened on farms in the Kalahari in Northern Cape and/or North West provinces of South Africa. Vegetation tends to decontaminate the pan w.r.t. salinity over time, leading to vegetational succession. In due course, fresh water resumes its attraction of ungulates, which tend to bare the floor of the pan, reincarnating the original bare pan.

My commentary:
Why are pans located where they are, in the first place? Why did they arise in these precise locations, millions of years ago? I hypothesise that the nutrient-richness of the spring water is crucial.

Stromatolites ( occur at one pan, namely Urwi Pan, which was probably lacustrine during last pluvial periods ( and

My commentary:
I invoke a link between calcium carbonate and iodate. Hypothetically, the formation of the pan leads to deposition of calcium carbonate, which leads to the accumulation of iodate, which leads to geophagy, which leads to the perpetuation of the pan.

Pans are associated with sources of fresh water.

Etosha Pan has small depressions, surrounding artesian springs, along its southern perimeter. Wild ungulates drink from these springs, many of which are situated on low mounds of tufa. This indicates that the calcrete at pans has a consequential, not causal, relationship to the pan.

My commentary:

I see a resemblance to the mound springs of the artesian basin in east-central Australia (

Posted by milewski about 1 year ago
Posted by milewski about 1 year ago

Jones (1990,

This is a good reference for a) the thorough decomposition of food by fungus-culturing termites, so that negligible carbon is deposited in the mounds, and b) the productivity of fungus-culturing termites relative to other termites..

My commentary:
This contrasts with heuweltjies inhabited by Microhodotermes viator, which are rich in carbon.

Macrotermes forages throughout the matrix among the mounds, and its tunnels greatly affect the whole area, not just the mounds.

My commentary:
One cannot understand miombo ecosystems without realising that its soil is tilled intensively by termites. Effectively, fungus-culturing termites concentrate litter from the miombo into the mounds, and decompose it rapidly and thoroughly, saving the nutrients from leaching.

page 296: The mounds of fungus-culturing termites are enriched in nitrogen, despite lacking organic matter.

My commentary:
Is this a consequence of the exclusion of fire?

page 298: Each worker/soldier of Macrotermes apparently consumes 0.5 g of carbon per gramme of biomass of termites per day. This is "an order of magnitude greater than previous carbon consumption estimates for termites".

My commentary:
Is the biomass of termites here expressed was fresh mass, or dry mass? How does this intake rate compare to ungulates?

Posted by milewski about 1 year ago

Watson J P (1974,

Calcium carbonate under the mounds of termites may extend as deep as 6 m, and can total up to 20 tons of calcium carbonate per mound.

Once alkalinity is established in the core of the mound, it tends to beget further precipitation of calcium carbonate, a quirk of chemistry. This applies only to calcium, and not to anions other than carbonate or bicarbonate.

It seems sure that groundwater is the ultimate source of calcium accumulated in mounds.

Posted by milewski about 1 year ago

Lee and Wood (1971), page 27: The only clade of termites using much saliva (as opposed to faeces = excrement) in construction is Macrotermitinae. This clade is equally unusual in not using faeces in construction, the faeces instead being recycled into the fungus cultures until completely oxidised.

Posted by milewski about 1 year ago

Add a Comment

Sign In or Sign Up to add comments