Stem diameter and bark surface area of the fluted trunk of Balanites maughamii ( Balanitaceae )

Balanites maughamii Sprague (Balanitaceae) is a woodland tree used and harvested for bark products in the traditional medicine trade o f South Africa. The tree has a distinctively fluted and buttressed stem, especially in mature individuals. This short communication quantifies the relationship between two diameter measurements D1 and D2 that respectively exclude and include the bark surface contained in the convolutions o f the flutes at five height intervals up the stem to 2 m. Regressions show D1 to be an accurate predictor o f D2 (r̂ = 0 .97-0.99) over a range o f tree sizes, hence obviating the necessity to measure both D1 and D2. The circumference and bark surface area on the stem was determined to estimate the quantity o f bark that can potentially be harvested. At least 69% o f the stem circumference and bark surface area was estimated to be contained within the convolutions o f the flutes. Bothalia 37,2: 211-214 (2007)


INTRODUCTION
Balanites maughamii Sprague (Balanitaceae) is a medium to large, slow-growing deciduous tree ranging from 8-20 m tall (Pooley 1993).The stem is straight and the trunks of older trees are distinctively fluted and buttressed (Pooley 1993;Van Wyk & Van Wyk 1997).
The grey bark has medicinal value and is harvested and sold to consumers in traditional medicine markets in KwaZulu-Natal (KZN), Gauteng and Mpumalanga (Botha et a l 2001;Grace 2002;Williams 2003) (Figure 1).Based on the total amount of bark harvested (m"), B. maughamii was ranked third out of 36 tree spe cies harvested for bark in the woodlands of southern Maputaland, KZN (Twine 2004).A detailed population study there revealed that 55% of all individuals [diam eter at breast height (dbh) > 10 cm] had harvest wounds, and the mean amount of bark harvested per individual was 1.09 m^ (Twine 2004).
In KZN, the species is classed as declining and consid ered to be heavily exploited for bark products (Cunning ham 1988;Netshiluvhi 1999;Grace 2002).Its legal sta tus in KZN is described as 'controlled' by Von Ahlefeldt et a l (2003), i.e. written permission is required from the land owner/holder for this species to be harvested or collected from the wild.The turnover from 23 trad ers in the Isipingo and Victoria Street informal herbal medicine markets in Durban was estimated to be 187 fifty kg bags per annum (± 1995) (Netshiluvhi 1999).On the Witwatersrand, 56% of the muti shops sold the bark (Williams et al. 2001), and a volume equivalent to ~ seven 50 kg bags were present between 17 of the 100 traders surveyed in the Faraday Street traditional medi cine market in Johannesburg in January 2001 (Williams 2003).On the western boundary of the Kruger National Park, 29% of the vendors sold Balanites maughamii bark and considered it a readily available resource (Botha et a l 2001).The mean price per 50 kg bag of B. maughamii bark bought by muti shops in Johannesburg in 1995 was R66.70 ± R33.50 [± standard deviation (SD)] (n = 15), and in 2001 a bag cost ~ RIOO.
As part of an extensive investigation into the rela tionship between tree size and bark thickness of six tree species, including Balanites maughamii'. 1, to determine the size of trees targeted by commercial bark harvest-ers from the thickness of the bark sold in the muti mar kets; 2 , the mean wet and oven-dry bark thickness per tree size class; and 3, the mean harvestable bark mass per stem (Williams et al. 2005, in press a, b), various aspects o f the tree stem profile were measured.These aspects included: 1, approximate height of the tree and branch-free bole length; and 2 , diameter of the stem at five height intervals.Bark thickness was also measured.Data collected for B. maughamii are a subset of the origi nal study.This short communication describes specific aspects of the B. maughamii tree stem profile related to the fluted trunk, including: 1, the relationship between two measurements around the stem that respectively include and exclude the bark surface area contained in the convolutions o f the flutes; 2 , the number of flutes observed at 1.3 m above ground (where dbh is normally measured); and 3, the percentage of the stem enclosed within the flutes.

METHODS
Between March and May 1998, 39 Balanites maugh amii stems were measured at six sites in three South African provinces (Table 1).At each sample site, a population o f trees was located and individuals were selected from five stem diameter classes based on diam eter at breast height (dbh) ranging between 10 cm and 60 cm.A minimum of five and a maximum of ten trees were measured per diameter class (not per site).None of the individuals had suffered any prior harvesting dam age, and the bark on the bole was intact.Balanites indi viduals larger than 60 cm dbh were found in communal lands; however, these trees were not sampled as bark harvesters had previously removed whole sections of the bark, fluted stems and buttresses.The method used for assessing vertical height was a direct estimate using a 2 m height pole, with 0.5 m intervals.The number of pole lengths was counted by eye to estimate tree height and branch-free bole length.After the 22 Balanites stems were measured, the number of flutes at dbh were counted and their depth was categorized (subjectively, shallow or deep).
It is standard practise in forestry to measure tree stem girth with a forestry diameter tape.The tape is calibrated in K centimetres so that a circumference measurement is converted directly to a diameter measurement (Philip 1983), and the measurement is thus recorded as a diam eter dimension rather than a circumference.Two diam eter readings were taken at five height intervals (0.5 m, 1.0 m, 1.3 m, 1.5 m and 2.0 m, abbreviated as D,^, D ,j, D,5 and respectively) from the Balanites maughamii stem: 1, a circumference measurement around the stem that excludes the area inside the flutes (diameter 1, D l); and 2, a circumference measurement into the convolutions of the flutes, measuring along the entire bark surface (diameter 2, D2) (Figure 2).Hence Dl is the typical stem diameter measurement taken by foresters, usually at breast height (1.3 m, dbh), whereas D2 is a hypothetical diameter, where the flutes are pushed out to form a circle.Initially, only Dl was mea sured, but after six samples, D2 was also measured.

RESULTS A ND DISCUSSION
The Balanites maughamii individuals measured, ranged in height from 4 to 12 m, with a mean of 8 ± 2 m (SD).Branch-free bole length was 2.9 ± 1.4 m (SD).The dbh of the largest tree sampled was D l, 3 = 59.2 cm and D2j j = 260.0cm (circumference equals 186 cm and 817 cm respectively), from a site in a private protected area in KZN.
There was a very strong positive relationship between Dl and D2 at all height intervals up the stem (Figure 3A-E), especially at D^^ (r^ = 0.988, p < 0.0001, Figure 3A).No branching occurred on the stem below 0.9 m, hence results for D^^ were not affected by the response o f the tree to branching.The quadratic regressions were only slightly better fits than the linear regressions (results not shown).For example, r^ = 0.988 for the qua dratic equation at D^,j, whereas r = 0.979 for the linear equation at the same ifieight.
These results show that by measuring Dl at a par ticular stem height, D2 can be accurately estimated, hence obviating the necessity to measure both Dl and D2.When compared with the observed D2, the D2 pre dicted by the quadratic regression equations was slightly By converting the observed D2 measurements back to circumferences, the area of bark (m^) on the stem could be estimated.The mean amount of bark up to 2 m on the stem ranged from 3.3 ± 0.6 m^ (SD) (n = 10) on trees in the 10-19 cm diameter class (D l), to 16.1 ± 0.7 m" (SD) (n = 4) on trees in the 50-59 cm diameter class (Table 2).
As the dbh of Balanites maughamii individuals in creased, the number and depth of flutes at Dj ^ was observed to increase (Table 3), thus increasing the pro portion of the bark surface area within the convolu tions o f the flutes.Trees in the 10-20 cm and 20-30 cm stem diameter classes (D l) generally had two shallow flutes and one deep one.As tree size increased, the shal low flutes became deeper until there were 2 or 3 and 4 or 5 deep flutes in the 30-40 cm and 40-50 cm classes respectively.Trees larger than 50 cm had more than six deep flutes with sometimes as many as 10 per stem as the trees approached 60 cm dbh.Most o f the trunk circumference is contained within the concave sections of the flutes (Figure 4).At 0.5 m above ground, 73.0 ± 4.0% (SD) of the stem was within the flutes.The percentage decreased gradually with increasing height up the stem until it was 70.3 ± 4.3% (SD) at 2 m (Figure 4).Furthermore, in trees with larger dbh, a greater percentage of stem was enclosed within the flutes.Similarly, more than two thirds of the bark surface area is within the flutes [mean = 72 ± 3% (SD), n = 31].The proportion of the bark inside the flutes varied according to tree size, with up to 79% of the bark area found in the flutes of trees in the 50-59 cm stem diam eter class (D l), and decreasing to 69% in flutes of trees in the 10-19 cm stem diameter class.

CONCLUSION
Despite the buttresses in the Balanites maiighamii stems, it appears that the diameter measurement Dl is an acceptable predictor of D2.Hence, the bark surface area can be estimated as well as the amount o f bark that can potentially be removed from the stems.Because most of the bark area is contained within the convolutions of the flutes, the tree trunks are difficult to ring-bark.Even when harvesters remove whole sections of the flutes/but tresses, including the timber, they usually leave behind some of the bark at the base of the flute.This may poten tially enable wound recovery following harvesting and probably makes the species more resilient to harvesting.

TABLE 1 .-Sample sites and no. individuals sampled per site Province Area in the province Ownership and man agement regime n
2KwaZulu-