#05 2020

What is the potential of native vegetation for the development of agroforestry systems on the Maputaland Coastal Plain? *

By Allister Starke, PhD Candidate, SAEON Grasslands-Forests-Wetlands Node

The Indian Ocean Coastal Belt Biome of Maputaland is a subtropical, fire-regulated ecosystem characterised by a matrix of forest, woodland, grassland and wetland vegetation which intergrade across an undulating sandy coastal plain. 

The ecosystem is well-suited to research under the SAEON Grasslands-Forests-Wetlands Node because these vegetation types are largely contiguous and are some of the last significant examples of untransformed coastal vegetation in KwaZulu-Natal.

Furthermore, natural vegetation in Maputaland is faced by ongoing transformation pressure from multiple competing land uses and a changing climate. An understanding of relations among vegetation composition, environment and anthropogenic disturbance is therefore key for successful natural resource management in the region going forward.

Trade-off between land use and the environment

A major trade-off between land use and the environment for local communities in Maputaland relates to the tension between the economic and resource benefits gained by plantation forestry versus the loss of ecosystem services provided by natural vegetation. These choices are made more difficult owing to the complex relationship between the region’s vulnerable groundwater supply and forestry plantations, and the fact that many native plant species provide resources which contribute towards local livelihoods.

The Maputaland region also supplies major urban centres such as eThekwini with medicinal plants and other plant products such as fibre and oils, while its coastal rangelands provide year-round grazing for local livestock and serve as a forage reserve for inland areas during severe drought seasons.

Can agroforestry serve as a framework for integrating useful native species into contemporary land use?

The research conducted during my PhD formed part of a broader Water Research Commission study aimed at developing an approach to land use in the region that considered the tension among water resources, natural vegetation and plantation forestry.

Allister Starke (right), Bhekani Mdhluli (left) and Xolani Ngubani sampling grassland in Maputaland against the backdrop of a plantation. (Photo: A Starke)

Specifically, I looked at the composition of forest and grassland ecosystems, asked how the floristic composition of different vegetation types varied according to environmental gradients, how natural vegetation was affected by forestry plantation disturbance, especially abandoned plantations, and what products and services were offered by vegetation. I then considered using agroforestry as a framework for integrating useful native species into contemporary land use.

Forestry plantations tend to promote forest expansion processes due to their shaded structure and fire-suppressed environments, but I wanted to find out if the composition of woody species differed according to variation of these conditions, and whether this difference could be used as a screening tool to identify native species for use in agroforestry systems.

In conjunction, we expect for grasslands that floristic heterogeneity will be greatest in nutrient-poor environments which tend to occupy drier sandy ridges, than in fertile, mesic low-lying environments, and that the recovery of grassland composition following disturbance is usually extremely slow. I sought to find out whether these expectations held, and if they could assist with decision-making about where best to locate forestry plantations in the communal rangelands of Maputaland.


In summary, clear-felled and frequently burnt abandoned plantation compartments suited the regeneration of common woody savanna species, while forest pioneers preferred the understory of unfelled plantation compartments. However, and curiously, some generalist forest pioneers showed little preference to either open or closed canopy secondary environments.

Findings from the grassland component of the study confirmed that floristic richness was greatest in nutrient-poor grassland types, which consequently supported a greater diversity of plant resources than mesic sites, which were the least diverse but provided greater forage resources.

Coupled to the issue of diminishing water resources, grassland afforestation in Maputaland is likely to have unintended long-term consequences for the forage, medicinal and other plant resources available in the region.

Considerable resource diversity was provided by woody savanna species, including dual keystone ecological and biocultural species such as Sclerocarya birrea (marula) and Hyphaene coriacea (ilala). These and other multi-purpose woody species occurred in forest margins, woodlands, grasslands and in secondary environments. They are ideal candidates for agroforestry systems and should be on a preferred list for tree-planting projects that may occur in the region.

Silvopasture using combinations of native woody species, native forage grasses and/or forestry species planted in low-density stands, was considered an appropriate land use for hydrologically vulnerable areas, while the inclusion of agroforestry practices (such as windbreaks, shelterbelts and fodder-banks) into plantations using native species would also serve to increase the range of ecosystem services provided by forestry plantations.

Some potential agroforestry products that occur in Maputaland: (a) Hypoxis hemerocallidea (inkomfe) used medicinally as an emetic for dizziness and mental disorders. (b) Fire-wood stack of Syzygium cordatum (umdoni). (c) Hardwoods, possibly Brachylaena huillensis (iphahlaehlati) sold on the roadside as firewood. (d) Furniture clad with Phragmites australis (umHlanga). (e) Hymenocardia ulmoides (itsatsalatsane) coppice regrowth. (f) Sclerocarya birrea (marula) fruit sold at a roadside market. (g) Ephemerally flooded wetlands are characteristic in Maputaland and are a good source of forage. (h) Sap collection of Hyphaene coriaceae (ilala). (i) The edible fruit Salacia kraussii (ibhonsi). (j) Phragmites australis awaiting collection after being harvested from wetlands near the Pongola River. (k) The fruit of Vangueria infausta (umvelo). (l) Fruit of Syzygium cordatum and Phoenix reclinata (isundu). (m) Within the natural forests at Manzengwenya plantation, the bark of various trees species is harvested for the medicinal properties it contains. (n) Elevated grasslands provide seasonal grazing resources and are a source of medicinal plants. (Collage: A Starke)

There is a way to go before commercial silvopasture systems in Maputaland are practically realised. However, the fact that a diverse pool of culturally relevant and useful woody species was suitable for use in agroforestry systems points towards an interesting future of developing multi-purpose land uses in the region.

I thank the Water Research Commission for funding this study and SAEON for support received.

*  Supervised by Tim O’Connor, Colin Everson and Coert Geldenhuys