Modes of dispersal and sowing in woody vegetation in Tongaland, southern Africa

(writing in progress)

On 10 March 1993, I discussed the vegetation in Tongaland with Ken Tinley, who had extensive experience in the field in that region.

The following is the summary of my thoughts, that I noted at the time.

In general, sandynsoils tend to be more suitable for plants with fleshy fruits than are dense clays.

Sodicity and waterlogging are inimical to potassium nutrition and the bearing of succulent fruits.

Mangroves, the souks of which are acidic here, are devoid of succulent fruits.

Swamp forest on nutrient-poor soils (not on floodplains) has relatively few plants with succulent fruits.

Well-drained termite mounds, in which sodicity is restricted to the apron (basal ring), bear stands of plants thoroughly dominated by species with succulent fruits, except for trees, which enter late in succession, when potassium-demanding, smaller plants are already established. These soils are base-rich and presumably rich in potassium, and only mildly sodic overall. They are well-drained despite often having moist subsoils.

In thickets in Tongaland, endozoochorous plants contribute >90% to the vegetation, whether by species or by individuals. However, some of these are euphorbs, which lack fleshy fruits but seem to be devioisnin getting doves to eat them with no obvious benefit to the birds (possibly grind-stones?).

Where the matrix consists of dark cracking clay, which is sodic, even thickets on the mounds tend to lack species with succulent fruits.

Stratification is important, with trees tending not to have succulent fruits. In most vegetation types, the understorey has more plants with succulent fruits than does the canopy. It may be no coincidence that vegetation with an open canopy and a well-developed understorey, such as that dominated by caesalpinioid legumes, have few plants with succulent fruits, because the two strata compete for potassium.

(However, this is somewhat contradicted by the fact that, in savanna, plants withnsucculent fruits tend tomoccur around the bases of big trees.)

This leads to the prediction that big trees should generally only bear succulent fruits where the canopy is dense and shade-casting. This seems to fit the pattern in Sapotaceae, which are prominent in the study area.

The roles of nitrogen-fixing and parasitic plants seem typical. The former tend to have ballistic and/or anemochorous fruits. However, one species of Morella may have succulent fruits.

This study area sheds little light on the role of fire. I expect succulent fruits to be commoner in fire-free than in fire-prone vegetation. However, dry forest in Tongaland has few plants with fleshy fruits, possibly because the soil is leached and trees dominate.

By numbers of individuals, not numbers of spp.:

Tongaland dry forest

forms patches, surrounded by savanna (perhaps can be thought of as 'fire-free miombo')

rainfall 300-1000 mm

duplex sands or ultrafine sands


Flora rich and specialised

The incidence of succulent fruits decreases with height above ground.

A few spp. of ballistically dispersed trees dominate the canopy (60% of individual trees). Species with succulent fruits contribute little (4%) to the tree individuals here. However, more than half of the plants in the understory are zoochorous, despite the prominence of Acanthaceae and Euphorbiaceae, and the inclusion of juveniles of mid-stratum leguminous trees.

By numbers of individual plants

  • canopy 4% zoochorous (but many spp., e.g. of Mimusops and Ficus), 60% ballistic, 20% anemochorous
  • mid-stratum (dominated by Millettia and Craibia) 20% zoochorous (very many spp.), 80% ballistic
  • understorey (dominated by Acanthaceae and Euphorbiaceae, bit including juveniles of legumes) 57% zoochorous, 43% ballistic

By numbers of spp.

  • canopy 40% zoochorous, 20% ballistic, 40% anemochorous
  • mid-stratum 90% zoochorous, 10% ballistic
  • understorey 20% zoochorous, 78% ballistic

Fire-free thickets on termite mounds in a matrix of Senegalia nigrescens, adjacent to dry forest. These mounds have gravelly, sandy, non-sodic clays, and bear stands in which almost 90% of the plants have succulent fruits.

Small patches of lawn of Cynodon (or even Sporobolus, where sodic) occur around the semi-sodic perimeter. The more sodic mounds tend to have sparse thickets. Sodic rings at the bases of mounds tend to form on duplex montmorillonitic soils with a darkmappearance.

Zoochorous plants contribute >90% of spp. and >90% of individuals.

Typical: Mimusops, Cleistochlamys (, Trichilia, Diospyros, Cordia, Strychnos, Berchemia.

Anemochorous in the stands on termite mounds: Philenoptera, Combretum, Dalbergia, Markhamia

On most of the termite mounds there occur also various genera of Euphorbiaceae


In Tongaland, forest on termite mounds has a greater incidence of plants with succulent fruits than does the adjacent dry forest. Dispersal and sowing by birds can be achieved without succulent fruits (e.g apparent deception in euphorbias). Thus, the incidence of succulent fruits is not simply a matter of the mechanics of succession and colonisation. Within stands (e.g. dry forest), plants withnsucculent fruits prefer the understorey over the canopy. Even in termite miunds, I suspect that the largest plants tend not to bear succulent fruits.

I infer that termite mounds are near-ideal for zoochorous plants, because they are

  • base-rich (including potassium) and well-drained, but generally not sodic,
  • often protected from fire, but
  • unfavourable for nitrogen-fixing plants, partly because they tend to be relatively rich in nitrogen, and not particularly rich in phosphorus.

Dry forest seems to be too poor in potassium for plants with succulent fruits. In this type, the mkd-stratum is richer than the canopy, in plants with succulent fruits.

Posted on July 10, 2023 08:11 AM by milewski milewski


Croton: some spp. have elaiosomes, and are dispersed and sown by ants ( and

Otherwise, Euphorbiaceae, occurring on termite mounds in Maputaland, have ballistic dispersal of seeds.




Ken Tinley was of the opinion that the seeds, lying on the ground, are then eaten and further dispersed by columbids that forage on the ground.

This may also apply to some Acanthaceae.

Ken had observed that even the relatively soft seeds of Ficus may emerge intact in the faeces of columbids, despite the fact that these columbids depend on the grinding of seeds in their gizzards, for their sustenance.

Turtur chalcospilos ( eats the seeds of Euphorbiaceae after ballistic (explosive) dispersal.

Posted by milewski about 1 year ago

Add a Comment

Sign In or Sign Up to add comments