Ecological notes on Syzygium (Myrtaceae) in southern Africa, part 1

@fynbos @troos @botaneek @mark_smale @fynbosphil @vandalsen @aavankampen @andrew_hankey @graham_g @sedgesrock @geoffnichols @ricky_taylor @richardgill @magdastlucia @reubenheydenrych @adriaan_grobler @dustaway @craig-r @terra_australis @julian70

One of the obvious biotic differences between southern Africa and Australia is in the incidence of Myrtaceae (

Myrtaceous trees and shrubs cover much of Australia across a wide range of climates and soils. By contrast, they tend to be few and far-between - and therefore easily overlooked - in southern Africa.

A main ecological difference between these continents is the extreme nutrient-poverty of Australia, together with a limited incidence of herbivores and an extreme incidence of wildfires.

This raises the question:
How do the indigenous Myrtaceae of southern Africa relate to poor soils, herbivores, and wildfire?

The following notes may begin to shed some light on this topic, focussing on syzygiums (, a genus indigenous to both continents.

KWAZULU-NATAL (eastern South Africa, see

The relevant myrtaceous tree species here are Syzygium cordatum ( and Syzygium guineense (

These have succulent fruits and are dispersed and sown by birds and other frugivores, as well as (in the case of S. guineense) by flotation. In these ways they differ from wildfire-adapted Myrtaceae in Australia, as well as most woody plants adapted to poorly-drained substrates worldwide. The native people tend to spare S. cordatum owing to its edible fruits. The particular ecological relevance of S. cordatum is as one of the few indigenous members of its family to occur in savanna/grassland (as opposed to forest/woodland) in Africa.
The reference is: Tinley K L (1976) The ecology of Tongaland. Published by the Natal branch of the Wildlife Society, Durban. The fieldwork for this report was done in 1958.
In Kwazulu-Natal, my coverage starts near Kosi Bay ( in the far northeast of Kwazulu-Natal, moving southwards through the area of Lake Sibhayi to the Mkuze area.

Kosi Bay area:

Syzygium cordatum is a predominant tree in the tallest forest, found on permanently inundated substrates, in this area.

Swamp forest is about 10m high, reaching 18 m, occurring where the inundation consists of running rather than stagnant water. It reaches only 7.5 m high along minor streams.

Syzygium cordatum is the most frequent tree in swamp forest, where it is typically 15 m high. Associated trees include Ficus hippopotami, F. capensis, Ilex mitis, Schefflera umbellifera, Bridelia macrantha, Voacanga thouarsii, Morella serrata, Rapanea melanophloeos, Erythrina caffra, Sapium ellipticum, Barringtonia racemosa, Halleria lucida, Afrocarpus falcatus, and Rauwolfia caffra, and the palm Raphia australis is present. An abundant epiphyte is the fern Stenochlaena tenuifolia, which virtually clothes the tree trunks from ground level to the crowns. Lianes include Canthium gueinzii, Lygodium kerstenii, Smilax kraussiana, Ipomoea digitata, and Cissampelos torulosa.
Now let us account for the lowest vegetation in the Kosi Bay area, and work our way through the various intermediates between grassland and forest, including savannas and woodlands, and the seral stages of these.
Grassland in this area also features S. cordatum, this time as a stunted plant only knee-high. Other woody species occurring in grassland are Diospyros lycioides, Parinari mobola (< 0.75m), Dalbergia obovata (< 0.2m!), Ochna arborea (< 0.3m), Strychnos innocua (< 1.2m) and S. spinosa (< 1.2m).

Where the palms Phoenix and Hyphaene are conspicuous, leading the re-conversion of grassland to forest on what I take to be rather poorly-drained substrates, S. cordatum persists through the succession. It takes the form of e.g. a short tree, 2 m high, in patches of scrub in the grassland. Here it is associated with e.g. Vachellia kosiensis, Sclerocarya caffra, Canthium ventosum, Trichilia emetica, and Strychnos madagascariensis.

Syzygium cordatum remains common in woodlands 4-10 m high, containing also Terminalia sericea, Antidesma venosum, Peltophorum africanum, Strychnos spinosa, Vachellia kosiensis, Combretum molle, Alizia adianthifolia, Vachellia nilotica, Ficus stuhlmannii, S. caffra, T. emetica, S. madagascariensis, Albizia versicolor, Apodytes dimidiata, Manilkara discolor, and many others.

These woodlands have various heights and densities and are, at least partly, successional to dense forest other than swamp forest or dune forest (see

Syzygium cordatum is absent from these 'climax' forests inland of the littoral belt, whether short or tall, in which it is replaced by e.g. Scolopia, Mimusops, Sideroxylon, Ekebergia, Hymenocardia, Apodytes, Manilkara, Diospyros, and many others. One species of Myrtaceae is present: an uncommon tall shrub or low tree in the mid-stratum, namely Eugenia natalitia (

These dense, wildfire-free forests have many species of lianes of at least 20 genera, and correspond at least partly to what in Australia would be called ‘vine thicket’. Epiphytes including Usnea are present. Where the dense forest is tallest, the predominant tree is Manilkara discolor. An example is Mabibi Forest, in which the emergent trees reach 15 m high. I find this dominance interesting in view of the recent revelation that Manilkara may be the only ectomycorrhizal genus in the study area.

The large snail Metachatina kraussi ( is common in these forests.

Dune forest, which occurs in places in the littoral belt, also lacks S. cordatum. Its main trees are: Ptaeroxylon obliquum, Ziziphus mucronata, Celtis africana, Inhambanella henriquesii, Ficus quibeba, Mimusops caffra, Canthium obovatum, Ficus natalensis, Euclea sp., Warburgia salutaris, Croton zambesicus, Vachellia kosiensis, Diospyros sp., Sideroxylon inerme, Cassipourea gerrardii, Teclea gerrardii, and Cassine aethiopica.
To summarise so far:
Syzygium cordatum is common in the inland part of the Kosi Bay area, being generally associated with a combination of poor drainage and anthropogenic disturbance. It probably tolerates some wildfire as well, although Tinley (1976) does not mention this aspect. This syzygium occurs as plants ranging in heights from grass-high to > 15 m, depending partly on the successional stage, and seems capable of persisting in low, recovering vegetation provided that the subsoil is (seasonally?) waterlogged. Tinley (1976) describes this low vegetation as grassland, but it contains woody plants and is possibly more precisely described as open low savanna of anthropogenic origin.

In the narrow swamps along watercourses, S. cordatum not only grows into a medium-size tree (about 15 m high), but becomes a major component of the forest, just short of dominant.

Other, better-drained forests in the Kosi Bay area, whether inland or on the coastal dunes, lack this syzygium.
Lake Sibhayi area:

Syzygium cordatum occurs in stunted form in the area of Lake Sibhayi ( This a nutrient-poor part of Amatongaland ( and
According to Tinley (1976), a main habitat of S. cordatum in this area is the level just above that of the lake edges, where the upper soil is not waterlogged although the subsoil probably is. In such situations S. cordatum is the only tree, growing with the grasses and sedges Urelytrum squarrosum, Bulbostylis contexta, Monocymbium ceresiiforme, Eragrostis sp., Imperata cylindrica, Mariscus sp., Andropogon sp., Ischaemum arcuatum, Eragrostis capensis and E. chapelieri, Perotis patens, and Dactyloctenium geminatum.
Another main habitat of S. cordatum is woodland/savanna. One variant is ‘umdoni veld’, in which S. cordatum shares the tree stratum with many other taxa, e.g. Strychnos spinosa and S. madagascariensis, Dichrostachys cinerea, Albizia adianthifolia, Mundulea sericea, Hyphaene natalensis, Brachylaena discolor, Sclerocarya caffra, Ziziphus mucronata, Trichilia emetica, Apodytes dimidiata, Ficus stuhlmannii, and Vachellia nilotica ssp. kraussiana. “In the eastern section [of the Lake Sibhayi area] Syzygium cordatum is the dominant woodland species changing to Terminalia sericea in the west, where the latter is dominant. The central part of Sibayi has an overlap of both types of vegetation.”

Tinley (1976) notes that “This country is commonly used by hippos for grazing purposes”. Although nobody expects S. cordatum to be eaten by the hippopotamus (, the habitat is surely affected by this megaherbivore in Amatongaland. This defies resemblance to Australia, where megaherbivores are absent.
Swamp forest occurs in the Sibhayi area as in other parts of Amatongaland. The tallest trees are Ficus hippopotami, Macaranga capensis, Voacanga thouarsii, Morella serrata, and Halleria lucida, and the main plant near ground level is the scandent fern Stenochlaena tenuifolia. Syzygium cordatum occurs only at the edges, in this area.
Syzygium cordatum also occurs at the edges of two other types of forest, namely coastal forest and dune forest, in the Sibhayi area (see; bushpig, bushbuck, cane rat, red duiker, and suni occur in both types, but the nyala is absent from dune forest.)

Syzygium cordatum in the Sibhayi area is thus absent from vegetation that is free of wildfires (coastal and dune forest as well as swamp forest), or subjected to permanent inundation at the surface (swamp forest). On the other hand, it also seems to be excluded from considerable areas of grassland owing to wildfire and/or good drainage. Where the intermediate conditions suit it, it can be the most prominent species of tree in savanna.

to be continued in ...

Publicado el mayo 26, 2022 09:23 TARDE por milewski milewski


Very interesting observations and analysis!

Publicado por jeremygilmore hace alrededor de 2 años

@jeremygilmore Many thanks to you, Jeremy.

Publicado por milewski hace alrededor de 2 años

Very interesting! I took a walk at the estuary in St Lucia yesterday and noticed the white mangroves died during the water logged period, before the estuary opened to the sea. The Brazilian Pepper is thriving. Vachellia kosiensis (you refer to Acacia karroo) died back, but most of the trees seem to be sprouting new leaves now. I also noticed a black mangrove sprouting new leaves, after dying back. I will have to go back and check what the Syzygium cordatum trees did.

Publicado por magdastlucia hace alrededor de 2 años

@magdastlucia Many thanks for your comment, and I will correct the name accordingly...

Publicado por milewski hace alrededor de 2 años

Very thought-provoking thank you, as a taxonomist I always have many questions about the precise habitat preferences of particular taxa.
Incidentally the sedge genus Mariscus is now generally included in Cyperus (now a very large genus!).

Publicado por sedgesrock hace alrededor de 2 años

A very interesting and informative read. Thanks @milewski
There are quite a few Myrtaceae in Australia that are not fire adapted as well (they grow in rainforest type environments). Many are being pushed towards extinction because of myrtle rust

Publicado por craig-r hace alrededor de 2 años

Agregar un comentario

Acceder o Crear una cuenta para agregar comentarios.