Appendices Appendix I: Age of Baobabs Adanson (1771) suggested the larger specimens of baobab seen by him in West Africa in 1749 and 1750 with diameters greater than 9 m might be more than 5,000 years old, stating a tree of one year of age had a diameter of 0.41 m and height 1.624 m. Escayrac (1853) questioned his extrapolation roughly having measured a specimen in Kordofan with a trunk diameter of 8.446 m taller than Adanson’s, which was 23.714 m for a diameter of 9.745 m. Walter (1971) pointed out age of the largest individuals can be easily overestimated, quoting a tree planted in Khartoum (annual rainfall less than 200 mm) which reached a diameter of almost 1 m in 30 years (3.14 m circumference or 10.47 cm/year), although Wickens (1982) considers an age of 2,000 years for the largest trees plausible. A baobab transplanted to a mission station at Bagamoyo in 1869 (mean annual rainfall at Dar es Salaam 1,049 mm/annum) estimated at 2–4 years of age, increased in circumference at an average of 2.973 cm/year over 87 years to 1956, although the average annual increase steadily declined indicating trees may be much older than their circumfer- ence suggests (Guy 1970). A baobab at Saadani on the Tanzania coast (Mbuyu kinyonga ¼ tree where people were hanged) was measured by Watt in 1886 with a circumference of 2,290 cm. Measured at breast height in 1975 the circumference was 1,500 cm (Kwagilwa 1975). Extrapolating from the Bagamoyo tree it would have been 505 years old in 1975, or 770 years old when Watt measured it. The difference may be due to Watt having measured it at ground level although Guy demonstrated the tree can shrink, and may shrink in dry seasons and expand in wet. “Livingstone’s baobab” in Botswana (rainfall 450 mm) measured by Chapman (1868) in 1852 as 2652 cm in circumference, and by Livingstone in 1853 as 2591 cm, was still a vigorous tree in 1988, but as it has six stems it is difficult to know how Chapman or Livingstone arrived at their measurements (Plate Appx. 1). Guy (1970) measured it in 1966 as 2,038 cm and including the convolutions and buttresses as 2,786 cm. This could suggest an apparent 6% shrinkage from 2,591 cm in 1,853 to 2,446 in 1966. However if Livingstone and Chapman used linen measuring tapes these may well have shrunk if they had been wetted and may not have been giving true readings. Chapman (1852) had recorded several baobabs in the vicinity but only the one remained. The “Big Tree” recorded by Baines and C.A. Spinage, African Ecology - Benchmarks and Historical Perspectives, 1405 Springer Geography, DOI 10.1007/978-3-642-22872-8, # Springer-Verlag Berlin Heidelberg 2012 1406 Appendix I: Age of Baobabs Chapman between Ghanzi and Lake Ngami in Botswana, and measured as 1,524 cm in circumference in 1861, Guy identified with one in 1966 measuring 1,708 cm at breast height, and 1,798 cm at 4.5 m from the ground. A tree recorded by Holub (1881) near the Makgadikgadi Pans as “almost” 1,524 cm Guy recorded as 1,585 or 1,666 cm. A tree at Gutsaa Pan has dates of 1771 to 1859 carved into it, but there is no record of early measurement although it must have been a large tree already in the eighteenth century (Plate Appx. 2, Appx. 3). There is a number of other large baobabs in Botswana with nineteenth century dates carved into them, and possibly some eighteenth century ones, while “Baines’s baobabs” in 1988 looked almost exactly as Baines painted them in May 1862 (Plate Appx. 4, Appx. 5). Another tree measured by Livingstone, at Chiramba in Mozambique mean rainfall 800 mm, increased at an average of 1.68 cm/year between 1858 and 1965 (Guy 1970). Although this is apparently a 43% slower growth per annum than the Bagamoyo tree, rainfall is 24% lower and the area is subject to a cold dry winter regime also, factors which could slow growth. During bush clearing for Lake Kariba, at the confluence of the Sengwe and Zambesi Rivers, annual rainfall about 500 mm, a sample was taken from the heart of a 4.5 m diameter baobab tree in 1960 which according to the formula would have been 300 years old. Samples were taken also mid-way to the centre and just under the bark. C14 dating gave a date of 1010 Æ 100 years for the core and the mid-way sample 740 Æ 100 years. The indications were that the tree grew more slowly over the outer 2.25 m of diameter, “there appears to be no reason why some of the really large baobabs should not be several 1,000 years old” (Swart 1963). The average annual increase in radius over the last 1.14 m of the total radius was calculated as 1.5 mm, while the average ring width over the last 20 cm of growth was 1.1 mm, thus not incompatible with the fact that rings may be laid down annually. But Hall (1974) advised caution in the interpretation of C14 dating which is prone to error, and not all are agreed on its time relations. Guy (1982) considered circumference was not a reliable indicator of age as trees measured in South Africa showed both increases and decreases in circumference. Caughley (1973), assuming the concentric rings of the pith were annual, based upon observations in Luangwa Valley under an annual rainfall of 813 mm used the formula: age in years ¼ 0:213 Â girth in cm; giving a mean annual increase of 0.75 cm. Whereas Barnes (1980) in Ruaha NP under an annual rainfall of 580 mm, obtained an annual increase in girth of 2.7 cm. The tree at Khartoum would have been 70 years by Caughley’s method compared to its actual 30. But this tree, and the Bagamoyo tree which was measured in 1913, 1927, and 1956, suggest there is a rapid initial growth rate. Hence in its first 30 years the Khartoum tree girth increased at a rate of 10.47 cm/year compared with the Bagamoyo tree in its first 44 years of 12.27 cm/year, despite the vastly different rainfall. The latter then declined to 11.24 in the next 14 years and 3.41 over the last 29 years, indicating rapid initial growth for the first 40–50 years followed by steady decline in annual increments. Appendix I: Age of Baobabs 1407 Plate Appx. 1. Livingstone’s baobab Botswana, September 1988. C. A. Spinage Plate Appx. 2. Baobab at Gutsaa Pan Botswana, September 1988. C. A. Spinage 1408 Appendix I: Age of Baobabs Plate Appx. 3. Eighteenth and nineteenth century signatures carved into a baobab, September 1988. Gutsaa Pan, Botswana. C. A. Spinage Plate Appx. 4. Baines’s baobabs as painted by Thomas Baines May 1852, Kudiakam Pan, Botswana. # Royal Geographical Society Appendix I: Age of Baobabs 1409 Plate Appx. 5. Baines’s baobabs, September 1988. C. A. Spinage In March 1975 the author measured girths at breast height of baobabs growing on the ruins of the abandoned Arab city of Gedi on the Kenya coast, annual rainfall 1,040 mm at Malindi. Gedi, founded in the late thirteenth or early fourteenth century, was abandoned in the early seventeenth (Kirkman 1970). In 1560 it was reported of Kilwa south of Gedi that its buildings showed that it used to be larger and more populous. Thus the trees found growing on and into the walls at Gedi had some 350 years in which to do so (Plate Appx. 6). Extrapolating from the Bagamoyo tree, I found that all of the largest trees had ample time in which to grow there, apparently having taken root about 1777–1794 (Table 1 Appx. I). Extrapolating from the Khartoum sample their ages would have ranged from 51–56 years only. Using Barnes’s (1980) annual increase of 2.7 cm/year the ages would have ranged from 199–218 years, but rainfall is much higher at Gedi than in Ruaha, nevertheless correspondence is reasonably close. Table 1 Girths and estimated ages of baobab trees in the Gedi ruins 1975 Locality Circumference at Estimated age (years) breast height (cm) Tomb of the Fluted Pillar 590 198 House on the Wall 564 190 House of the Venetian Bead 552 186 House of the Chinese Cash 540 182 The Palace 538 181 1410 Appendix I: Age of Baobabs Plate Appx. 6. Baobab tree growing on wall of the ancient city of Gedi, Kenya coast, 1977. C. A. Spinage Although Adansonia apparently exhibits distinct incremental growth rings, studies in South Africa rated it relatively low for dendrochronological analysis because of false and also coalescing rings, with a factor of 4 on a scale of 0–13 (Lilly 1977). Lees (1860) argued that in the tropics there was a double season, thus: “Every computation of the age of trees in the tropics must be made upon the basis of two concentric rings of wood being formed in the year; since there can be no reasonable doubt that every fresh set of leaves forms a new concentric ring. Otherwise, if made upon that of only one being formed, it will represent the trees to be twice the age they really are. But the computations that have hitherto been made ...have omitted to take into account this double formation.