Trees for health – forever: Implementing sustainable medicinal bark use in Southern Africa. Regional Workshop, Johannesburg, 1 – 3 November 2005.
Impact of uncontrolled bark harvesting on the resource base
1 Department of Forest Science, University of Stellenbosch, c/o Forestwood cc, P O Box 228, La Montagne 0184, Pretoria, South Africa
2 School of Animal, Plant & Environmental Sciences, University of the Witwatersrand, Private Bag 3, Wits, 2050, South Africa
Commercial harvesting of medicinal plants is a significant cause of resource depletion. Harvesting impacts depend on the plant part removed and the level of harvesting effort and efficiency by the gatherers. In South Africa, bark is the most popular medicinal product harvested from trees and comprises at least 31% of the plants harvested and traded annually in KwaZulu-Natal. In the Gauteng markets for traditional medicine, centred in Johannesburg, approximately 33% of the species estimated to be sold are trees, 51% of which are exclusively harvested for bark and not other products such as roots, leaves or fruits. Plant populations can be driven to extinction if increases in harvesting effort chase an ever diminishing yield. Damage assessments for trees can be based on the presence of debarked tree trunks. We report here on studies to assess the impact of uncontrolled bark use on the resource base in South Africa.
Part 1. Using a resource inventory to assess the impacts on a forest: Umzimkulu forests, Eastern Cape Thirteen state forests in the Umzimkulu District were sampled, using 388 circular plots (0.04 ha) at 40 m intervals on 37 randomly located transects running across the altitudinal gradient through each forest. Within each plot all trees of all species from which bark had been stripped were measured, and the number of stems of all other species was recorded. For bark-stripped trees data were recorded to determine what proportion of the bark on the stem up to the main branches was removed, what proportion of the bark around the stem was removed, and its crown condition (health status). Results from this inventory showed that
Bark harvesting occurred along the entire length of all transects, i.e. was not limited by steep slopes, but the steep slopes may constrain environmentally friendly timber extraction practices.
Bark harvesting focussed on specific species. Bark was harvested from 36 of the 95 species recorded, and 6.2% of the 7280 stems recorded. Although this impact was small, some species was particularly severely stripped: Curtisia dentata (50 stems recorded, 60.0% harvested); Ocotea bullata (359 stems, 57.4%); Prunus africana (10 stems, 70.0%); Pterocelastrus rostratus (29 stems, 86.2%); Rapanea melanophloeos (124 stems, 38.7%).
On average, they removed 20-40% of the bark on the main stem of harvested trees of the most intensively harvested species. Bark stripping of O. bullata was particularly severe with some trees stripped up to 12 m, with 20% of the trees with more than 50% of the bark on the main stem removed, and on average 31.4% of the total bark on the main stem of the stripped trees removed. Many large O. bullata trees were cut to harvest the bark higher up the stem, but the cut trees were not used for their valuable timber.
The population structure of the targeted species was severely affected. Stem diameters of O. bullata, P. africana and R. melanophloeos showed strong bell-shaped distributions, but with most of the trees above 30 cm DBH harvested with few to no seedlings, saplings or plants smaller than 15 cm stem diameter within the forest to replace the dying trees.
The extent to which a tree was ring-barked (percent of bark removed from the circumference of the stem) had a greater effect on the decline in crown condition than the percent of total bark removed from the main stem.
Tree species responded differently to bark harvesting. O. bullata and C. dentata easily developed coppice shoots at the base of the stem or around the debarked wound. Developing shoots of O. bullata trees were particularly heavily browsed. If the parent tree was severely debarked, it and its coppice shoots died. However, if it was cut then strong coppice shoots developed and the stump would survive. Most stripped standing trees of O. bullata were dead or dying. Debarked P. africana trees rapidly regrew bark through cambium development on the wound, and also developed coppice shoots. R. melanophloeos and Cryptocarya myrtifolia did not easily develop basal or stem sprouts, nor callus tissue around the debarked wound.
The potential value of the forest capital was calculated but only for the bark and timber in Malowe and Nzimankulu forests (total area of 1114 ha), for selected bark-harvested tree species (C. myrtifolia used as substitute for O. bullata, O. bullata, P. africana and R. melanophloeos). The estimated timber volume for trees of these species ≥30 cm DBH totaled 57 957 m³ (76% O. bullata) with a potential timber value of US$15.17 million (88% O. bullata). Estimated bark volume for trees ≥10 cm totaled 254 604 bags with 35 kg/bag (79% O. bullata) with a potential bark value of US$30.53 million (80% O. bullata). However, only an estimated 63 519 bags were harvested worth US$7.61 million (85% O. bullata). Not all harvested bark reached the market due to the illegal activities and bags left in the forest and not found (25 bags with O. bullata bark in one spot, covered with fern leaves). Due to the destructive bark harvesting, much of the capital was lost.
In a resurvey of two of the forests in January 2004, using 60 plots in each forest on similar transects, showed that one of the 9 trees of P. africanaharvested, died with no coppice regrowth. Of the 126 trees of O. bullata recorded, 80 were harvested standing (63.5%) and a few were cut, 66 of the standing harvested trees (82.5%) were dead (82.5%) of which only 34 (51%) showed some sign of sprouting. This meant that all the potential additional bark was lost, the timber was lost and that about 50% of the trees of this protected tree species were lost from the system.
Part 2. Using market information to assess the condition of the resource: Faraday market, Johannesburg Data on the estimated mass of bark sold through the medicinal plant markets informs the magnitude of resource depletion. However, it is difficult to translate these figures into actual impacts on plant population numbers, partly due to the paucity of data on population sizes and structures (i.e. age/size distributions). If, however, the number of trees debarked annually could be estimated, then the degree of damage to wild populations and the scale of cultivation required to replace harvested stocks could be gauged. Similarly, if the size of the trees targeted by harvesters could be determined from the thicknesses of bark sold in the muti (traditional medicine) markets, then one could assess the stem size-classes of trees preferentially targeted and/or available to the gatherers directly from quantitative surveys in the markets. The question that therefore arises is: how does one translate records of bark thicknesses sold in the muti markets to stem size-class availability, potential population structures, temporal changes in resource availability and the sustainability of harvesting practices?
To answer these questions, research was started in 1998 to examine the relationship between bark thickness and stem diameter for six commonly used tree species, namely Albizia adianthifolia, Acacia xanthophloea, Balanites maughamii, Elaeodendron transvaalense, Rhus chirindensis and Warburgia salutaris. An aim was to be able to use the prevalence of bark thickness size-classes found in the muti markets and shops as a surrogate for knowing what stem size-classes were exploited by commercial bark harvesters. In addition, the study undertook to estimate the mean harvestable mass of bark per stem size-class so that the number of trees debarked could be approximated from the quantity of bark estimated to be traded annually in the muti markets. The prevalence of bark of varying thicknesses sold in the Gauteng’s traditional medicine markets, and the quantity traded annually, was obtained from three studies conducted in Johannesburg between 1995 and 2001.
Harvesters tend to select plants in the larger size-classes to maximise their returns, and a decrease in the availability of thick bark is indicative of the decline in availability of larger trees. Results showed a decreased availability of bark estimated to be harvested from trees from the larger size-classes between 1995 and 2001. For example, in 1995, most E. transvaalense bark in the markets was 6–14 mm thick and estimated to come from trees 25–60 cm diameter at breast height (but mostly 40–49 cm). By 2001, most of the bark was 3–8 mm thick and estimated to come from trees <25 cm in diameter. Depending on the extent of bark removal (e.g. 10 or 100%) from stems from a range of size-classes, then between 409–16,167 E. transvaalense trees were estimated to be debarked to obtain the 11.2 tonnes of bark sold in Johannesburg’s Faraday street market in 2001.
Besides the harvesting intensity and the number of trees debarked, what further impacts a species’ risk to bark harvesting is the ability of the wound created by bark removal to recover. E. transvaalense is highly utilised but slow growing with little wound recovery. W. salutaris has few mature trees existing outside of nature reserves and harvesting has shifted to Mozambique. The long-term consequences of high volume harvesting of such species are serious. Bark that continues to be harvested from the remaining smaller individuals will be thinner, and the market will reflect this prevalence of smaller trees. Harvesting may then shift to a different location, and the medicinal markets will reflect these newly discovered localities by the occasional presence of thick bark.25>