Content uploaded by Daniel Lowy

Author content

All content in this area was uploaded by Daniel Lowy on Dec 17, 2015

Content may be subject to copyright.

Age determination of large live trees with inner cavities:

radiocarbon dating of Platland tree, a giant African

baobab

Patrut, Karl Reden, Pelt, Diana Mayne, Daniel Lowy, Margineanu

To cite this version:

Patrut, Karl Reden, Pelt, Diana Mayne, Daniel Lowy, et al.. Age determination of large live

trees with inner cavities: radiocarbon dating of Platland tree, a giant African baobab. Annals

of Forest Science, Springer Verlag (Germany), 2011, 68 (5), pp.993-1003. <10.1007/s13595-

011-0107-x>.<hal-00930675>

HAL Id: hal-00930675

https://hal.archives-ouvertes.fr/hal-00930675

Submitted on 1 Jan 2011

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of sci-

entiﬁc research documents, whether they are pub-

lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destin´ee au d´epˆot et `a la diﬀusion de documents

scientiﬁques de niveau recherche, publi´es ou non,

´emanant des ´etablissements d’enseignement et de

recherche fran¸cais ou ´etrangers, des laboratoires

publics ou priv´es.

ORIGINAL PAPER

Age determination of large live trees with inner cavities:

radiocarbon dating of Platland tree, a giant African baobab

Adrian Patrut &Karl F. von Reden &Robert Van Pelt &

Diana H. Mayne &Daniel A. Lowy &Dragos Margineanu

Received: 9 October 2010 / Accepted: 23 January 2011 / Published online: 1 July 2011

#INRA and Springer Science+Business Media B.V. 2011

Abstract

&Introduction For large trees without a continuous

sequence of growth rings in their trunk, such as the African

baobab (Adansonia digitata L.), the only accurate method

for age determination is radiocarbon dating. As of today,

this method was limited to dating samples collected from

the remains of dead specimens.

&Methods Our research extends significantly the dating of

such trees to large live specimens with inner cavities. The

new approach is based on collecting samples from the

cavities and their subsequent radiocarbon dating.

&Results The giant two-stemmed Platland tree, also known

as Sunland baobab, was investigated by using this new

approach. AMS radiocarbon dates of the oldest sample

segments originating from the two inner cavities indicate

that the large stem I (364.5 m

3

) is 750±75 years old, while

the much smaller stem II (136.7 m

3

) has 1,060±75 years.

Results also show that stem I is still growing very fast,

while the older stem II slowed down consistently its growth

over the past 250 years. The complete mapping of Platland

tree determined an overall wood volume of 501.2 m

3

.

&Conclusions Dating results demonstrate that the size–age

relation cannot be used for estimating accurately the age of

African baobabs.

Keywords Adansonia digitata .Radiocarbon dating .Age

determination .Growth rate .Accelerator mass spectrometry

1 Introduction

For trees which exhibit annual or seasonal growth rings,

ring counting, possibly refined by cross-dating, represents

the most accurate and reproducible method for age and

growth rate determination. Ring counting is typically

performed on the remaining stumps of dead trees. Never-

theless, there are several cases in which this procedure

cannot be utilized. Relevant examples are trees with growth

rings that are not strictly annual or seasonal, with no well-

defined rings or without a continuous sequence of rings, as

a consequence of their hollow parts. In such cases, ring

counting is replaced by alternative direct (radiocarbon

dating) or indirect dating methods (relation size/diameter-

age, projections of short-term growth data, projections

based on mortality rates, etc.) (Worbes 2002; Patrut et al.

2010a). The most investigated species of this type is the

African baobab (Adansonia digitata L.), the angiosperm

with the stoutest trunk. For large and old baobabs, a

Handling Editor: Erwin Dreyer

A. Patrut (*):D. Margineanu

Department of Chemistry, Babes-Bolyai University,

Arany Janos 11,

400028 Cluj-Napoca, Romania

e-mail: apatrut@gmail.com

K. F. von Reden

NOSAMS Facility, Department of Geology & Geophysics,

Woods Hole Oceanographic Institution,

Woods Hole, MA 02543, USA

R. Van Pelt

College of Forest Resources, Box 352100,

University of Washington,

Seattle, WA 98195, USA

D. H. Mayne

Baobab Trust,

Parklands 2121,

Johannesburg, South Africa

D. A. Lowy

FlexEl, LLC,

College Park, MD 20742, USA

Annals of Forest Science (2011) 68:993–1003

DOI 10.1007/s13595-011-0107-x

hypothetically accurate ring counting is not possible, as

growth rings may no longer be observed in certain areas

ofthetrunkandtheyarealsomissingintheareaoflarge

cavities.

As indirect methods often provide questionable results,

radiocarbon investigation represents the only alternative

method for the accurate dating of such trees. Given its

higher costs, so far, radiocarbon dating was not used on the

large scale. Several noteworthy investigations were per-

formed, however, on different tropical species, for deter-

mining the age of trees and/or growth rates, for accurately

dating the observed growth rings, for checking and

correcting ring counting, or for providing climate informa-

tion (Martinez-Ramos and Alvarez-Buylla 1998; Worbes

and Junk 1999; Poussart et al. 2006). Traditionally, for large

trees, such as the African baobab, this research was

limited to dating wood samples collected from the

remains of dead specimens, which decay very fast (Swart

1963;Patrutetal.2007). Work reported here extends

considerably the possibility of aging African baobabs, by

introducing a methodology which allows, to date standing

and live specimens. This new approach is based on

radiocarbon dating of very small wood samples collected

from the inner cavities of live baobabs. The first live

specimen investigated by this methodology was the huge

two-stemmedPlatlandtree.

Dating results for segments extracted from samples

originating from the two cavities of Platland tree, which

correspond to a depth of over 20 cm in the wood, are

presented and discussed. As these segments consist of

the original old wood, dating results made it possible to

determine the age of the two stems of Platland tree,

which was the main goal of our research. We also present

for the first time results of the complete mapping of a

large African baobab. The age difference of the two

stems is discussed in the framework of the questionable

size–age relation.

2 Materials and methods

2.1 Dating live African baobabs

The main characteristics of the new approach for age

determination of large live trees via dating samples

collected from their inner cavities, applied to the case of

the African baobab, can be summarized as follows:

1. Collection of samples from inner cavities through

cavity walls and subsequent radiocarbon dating of

these samples define the new methodology.

2. The extant wood collected with an increment borer

from cavities, at depth values exceeding 20 cm, is the

original old wood which corresponds to the respective

positions. Such a depth is needed for excluding

successive regrowth layers triggered by cavity fires

and also by a somewhat limited ability of the baobab of

repairing its wounds, i.e., for closing its cavity. A 60-

cm increment borer is more than sufficient for reaching

the original old wood and preventing collection of any

regrowth layer.

3. The age of an investigated baobab is calculated by

extrapolating age values of the oldest radiocarbon-dated

samples to the symmetry/geometric center of the trunk

at that height. Extrapolation is done by using growth

rate values of the trunk, calculated from different

sample ages, as well by considering growth rate values

reported in the literature.

4. The symmetry center at a certain height above ground

level can be accurately determined from a cross-section

of the trunk at this height.

5. In order to evaluate more accurately both the age and

the growth rate dynamics, one collects additional

(younger) samples from the exterior (through the bark);

these samples are also radiocarbon dated.

6. Usually, the pith of a tree, including the African

baobab, is not perfectly centered. However, with the

exception of certain individuals with a highly

irregular trunk, it is reasonable to assume that in

single-stemmed baobabs, the pith is relatively close

to the symmetry center of the trunk. If one assigns

the symmetry center to the pith and one considers the

fast growth rate of young baobabs, an additional

error of ±50 years for the final age value would

cover for all these assumptions.

7. In the case of baobabs with a multi-stemmed trunk,

the current symmetry center of a stem cannot be

assigned to its pith. One should consider that, after

fusion, the growth of stems continued only in

directions where there was room for accommodating

their growth. Thus, one must calculate the symmetry

center of a stem prior to fusion, which can be

reasonably approximated by the pith. The moment of

fusion must be estimated from the age of samples

collected from the fusion area.

2.2 The study area

The Platland tree (Fig. 1) is located on the Sunland Nursery

of the Platland farm, 10 km from Modjadjiskloof (formerly

Duiwelskloof) and 25 km from Tzaneen in the Limpopo

Province, South Africa. Its GPS coordinates are 23°37.255′

S, 030°11.884′E, and the altitude is 719 m. Mean annual

rainfall in the area is 492 mm (Mooketsi station; 1925–

2009).

994 A. Patrut et al.

2.3 Measurements

The common external measurements of the Platland tree and the

measurements inside the two inner cavities were performed by

using a Bosch DLE 70 Professional laser rangefinder (Robert

Bosch GmbH, Stuttgart, Germany) and graduated tapes.

Cross-sections of the Platland tree at ground level, 1, 2, 2.5,

and 5 m were mapped by setting up a frame around the tree

with a graduated tape. A compass and an Impulse 200 laser

rangefinder (Laser Technology, Inc., Centennial, CO, U.S.A.)

were used to map the cross-sections. Additional cross-sections

on the largest section were mapped at 6.5 and 8 m. All of the

Fig. 1 (top) The very impressive Platland tree, which is also called

Sunland baobab. General view taken from the east; (bottom) The

image taken from the north shows the two-stemmed trunk of Platland

tree, with the larger and taller stem I to the left, while stem II is on the

right. The fusion area of the stems can also be observed

Age determination of large live trees 995

mostly round branch and stem sections above or around this

had their basal diameters estimated by using a Criterion 400

survey laser (Laser Technology, Inc., Centennial, CO, U.S.A.).

System lengths were either measured directly or interpreted

from detailed photos of the tree structure without leaves.

Parabolic or conic equations were used for these smaller

systems based on how robust and foliated each system was.

2.4 Sample collection

Seven wood samples (numbered 1 to 7) were collected from

different positions of the walls of the two cavities of Platland

tree, according to the previously described methodology. Due

to the irregular profile of the cavities' walls, the inner samples

originate from heights of 0.40–1.70 m above cavity level. A

number of three additional samples (numbered 11 to 13)

were collected from the exterior of the two stems through the

bark. The sampling was performed at heights of 1.20–2.75 m

above ground level. All samples were obtained by using a

Haglöf CO600 increment borer (60 cm long, 0.43 cm inner

diameter) (Haglöf Sweden AB, Langsele, Sweden).

The projections of sampling points on a cross-section of

the trunk with the cavities included are shown in Fig. 2.

One should mention that all seven samples collected from

the two large cavities, as well the three samples collected

from the exterior, were shorter than 60 cm, having lengths

between 22 and 51 cm. These values indicate the existence

of additional hollow parts between the open cavities and the

exterior of the stems. Segments with a length of 0.5 cm

were extracted from determined positions of the original ten

samples. The segments were processed and investigated by

accelerator mass spectrometry (AMS) radiocarbon dating.

2.5 Sample/segment preparation

The standard acid–base–acid pretreatment method (Olsson

1986) was used to remove soluble and mobile organic

components. The resulting cellulose samples were combusted

to CO

2

by using the closed tube combustion method (Sofer

1980). Then, CO

2

was reduced to graphite on iron catalyst,

under hydrogen atmosphere (Vogel et al. 1984). Eventually,

the resulting graphite samples were analyzed by AMS.

2.6 AMS measurements

Radiocarbon measurements were carried out at the National

Ocean Sciences AMS Facility of the Woods Hole Ocean-

ographic Institution with the Pelletron ® Tandem 500 kV

AMS system (Roberts et al. 2010) and the Tandetron 3MV

AMS system (von Reden et al. 1994). The surface of the

graphite samples was sputtered with cesium ions, and the

secondary negative ions were extracted and accelerated in

the AMS system.

12

C and

13

C ions were measured in

Faraday cups, where a ratio of their currents was recorded.

Simultaneously,

14

C ions were recorded in a particle

detector, so that instantaneous ratios of

14

Cto

12

C were

also recorded. These raw signals were compared to ratios

obtained with a known standard material (Oxalic Acid I,

NIST-SRM-4990) and converted to a fraction modern

value, which was corrected for isotopic fractionation with

the normalized δ

13

C value of −25

0

/

00

. Fraction modern

values were ultimately converted to a radiocarbon date.

2.7 Calibration

Fraction modern values were calibrated and converted into

calendar ages with the OxCal v4.1 for Windows (Bronk

Ramsey 2009), by using the IntCal09 atmospheric data set

(Reimer et al. 2009).

3 Results

3.1 Mapping results

The trunk of Platland tree, which consists of two stems (I

and II), has a total circumference at breast height (cbh;

1.30 m above ground level) of 34.11 m and a footprint of

67.9 m

2

, which corresponds to a functional diameter of

9.30 m at ground level. It is to mention that the huge trunk

is covered by a very thin silver colored bark, which is less

than 1 mm thick. The canopy dimensions are of 37.7 m (in

direction NS) and 32.4 m (in direction WE).

The larger stem I has a height of 18.9 m, a cbh value of

24.20 m, a footprint of 43.9 m

2

, and a cross-sectional area

Fig. 2 Cross-section of the two-stemmed Platland tree (at 1 m above

ground), showing the two connected cavities (displayed at cavity

level), the positions of the ten sampling points and the sampling

directions. Because the samples were collected at various heights,

several sampling points are not located right on the contour of the

cavity or of stems, which are figured at a well-defined height

996 A. Patrut et al.

at bh of 42.7 m

2

. The formal diameter at bh calculated from

the cbh value in circular approximation is 7.70 m, while the

much more accurate functional dbh derived from the cross-

sectional area at bh is of 7.37 m. The smaller stem II has the

height of 15.8 m, a cbh value of 18.34 m, a footprint of

24.0 m

2

, and the cross-sectional area at bh of 23.3 m

2

.In

this case, the formal dbh is 5.84 m, while the functional dbh

becomes 5.45 m.

The two stems are connected by a fused section, which

covers a shared cbh of 4.10 m and has a maximum height of

2.20 m. The trunk comprises two interleading cavities on either

side of the fused section, connected by a small opening/

doorway. The very large cavity inside stem I has a maximum

length of 4.60 m, a width of 4.81 m, and a height of 4.88 m. The

irregular cavity inside stem II has a maximum length of 1.67 m,

amaximumwidthof2.50m,andaheightof2.47m;italsohas

an elevated extension towards the north, with a length of

2.24 m. The basal surface of cavity inside stem I is 15.9 m

2

,

while the cavity inside stem II has the surface of 2.8 m

2

.

There is also an elevated cavity located inside stem I, at the

height of 5.80 m. This cavity with two entrances, which is

visited and used by birds as a night shelter, has the maximum

dimensions of ca. 2×2 ×1.5 m.

For hollow baobabs, the large cavities are located at ground

level or, more seldom, at a certain height above ground.

However, for the Platland tree, the level of cavities is at 0.78 m

below the current ground level, suggesting that the inside level

corresponds to the original ground level. The increase of the

ground level over the past centuries is very probably the result

of mud and sediment left behind by several flood episodes of a

creek located in the close proximity of the tree. One can state

that today, the trunk/base of the Platland tree is buried in the

soil up to the height of 0.78 m. After 1990, the new owners of

the Platland farm opened a pub in the very large cavity inside

stem I. Consequently, the tree has become heavily promoted

as the Sunland “pub”baobab.

Formally, one can consider the Platland tree composed of

two sections: Platland I (stem I and the corresponding

branches) and Platland II (stem II and the corresponding

branches). The total wood volume of the Platland tree

above the present ground level was found to be 448.2 m

3

,

i.e., 273.1 m

3

below 5 m and 175.1 m

3

over 5 m. This total

volume of 448.2 m

3

above ground level consists of the

volume of Platland I, which is 330.2 m

3

, and the volume of

Platland II, which measures 118.0 m

3

. Nevertheless, the

overall wood volume of stems and branches should be

calculated above the cavities level, which is the initial

ground level. Thus, the overall wood volume of the

Platland tree becomes 501.2 m

3

, out of which 364.5 m

3

for stem I and 136.7 m

3

for stem II.

Cross-sections of the Platland tree at different heights are

displayed in Fig. 3. The measured and calculated charac-

teristic parameters of the Platland tree are summarized in

Table 1.

3.2 AMS results and calibrated ages

Fraction modern values and radiocarbon dates of the oldest

segment of each of the 10 samples collected from the

cavities and from the exterior, which were labeled by the

sample code/number followed by the additional index x, are

listed in Tables 2and 3. Radiocarbon dates and errors were

rounded to the nearest year.

For calibration, we used the general IntCal09 data set

(Reimer et al. 2009), rather than the SHCal04 data set for

the Southern Hemisphere (McCormack et al. 2004). Our

choice is justified by the fact that the SHCal04 curve does

not yet contain information for lower southern latitudes and

does not include results from Africa for the time frame

corresponding to our sample ages. The offset between

calibrated ages obtained when using the two calibration

data sets is of several decades.

Calibrated (cal) ages are also displayed in Tables 2and 3.

The 1-σprobability distribution was chosen to derive

calibrated age ranges. For three segments (1x,3x,7x), the

1-σdistribution is consistent with only one range of

calendar years (marked in italics), while for the other seven

segments (2x,4x,5x,6x,11x,12x,13x), the 1-σdistribution

corresponds to several ranges of calendar years. For these

segments, the confidence interval of one range (marked in

italics) is much greater than of the others; therefore, it was

selected as the cal AD range of the segment for the purpose

of this discussion.

Fig. 3 Cross-sectional areas of the trunk/stems of Platland tree at

different heights (ground level, 1 m, 2 m, 2.5 m, and 5 m)

Age determination of large live trees 997

For obtaining single calendar age values of segments, we

derived a mean calendar age of each segment from the 1-σ

range with the highest probability. Calendar ages of segments

represent the difference between AD 2010 and the mean value

of the selected 1-σrange, with the corresponding error.

Calendar ages and errors were rounded to the nearest 5 years.

Table 2 AMS radiocarbon dating results and calibrated calendar ages of the oldest segments of samples collected from the cavities

Segment code

[stem]

Depth

a

[height

b

]

(10

−2

m)

Fraction modern

[error]

Radiocarbon date

[error]

(

14

C years BP)

Cal AD range(s)

1-σ

[confidence interval]

Segment age

[error]

(calendar years)

1x[I] 23 [148] 0.9401 [±0.0034] 496 [±27] 1415–1438 [68.2 %] 585 [±10]

2x[I] 21 [145] 0.9316 [±0.0036] 569 [±29] 1320–1350 [39.8 %] 675 [±15]

1391–1412 [28.4%]

3x[I] 25 [90] 0.9501 [±0.0034] 411 [±27] 1441–1485 [68.2%] 545 [±20]

4x[I] 48 [50] 0.9601 [±0.0022] 327 [±18] 1515–1529 [10.4%] 440 [±20]

1542–1599 [44.8%]

1618–1634 [13.0%]

5x[II] 45 [102] 0.9597 [±0.0033] 330 [±27] 1495–1529 [18.2%] 440 [±30]

1541–1602 [38.2%]

1616–1634 [11.8%]

6x[II] 21 [170] 0.8930 [±0.0029] 909 [±23] 1045–1095 [39.6%] 940 [±25]

1120–1142 [16.0%]

1147–1164 [12.6%]

7x[II] 23 [40] 0.9003 [±0.0030] 845 [±24] 1168–1221 [68.2%] 815 [±25]

a

Depth in the wood of the oldest dated segment (from the cavity wall)

b

Height above cavity level (which is 0.78 m below ground level)

Parameter (unit) Platland tree Platland I Platland II

Height (m) 18.9 18.9 15.8

Circumference at bh (m) 34.11 24.20 18.34

Formal diameter at bh (m) 10.86 7.70 5.84

Footprint (m

2

) 67.9 43.9 24.0

Basal functional diameter (m) 9.30 7.48 5.43

Cross-sectional area at 1 m (m

2

) 66.8 43.2 23.6

Cross-sectional area at bh (m) 66.0 42.7 23.3

Functional diameter at bh (m) 9.17 7.37 5.45

Cross-sectional area at 2 m (m

2

) 63.2 40.4 22.9

Cross-sectional area at 2.5 m (m

2

) 57.3 36.2 21.1

Cross-sectional area at 5 m (m

2

) 39.7 29.7 10.0

Wood volume below 5 m (m

3

) 273.1 186.8 86.3

Wood volume over 5 m (m

3

) 175.1 143.4 31.7

Total wood volume above ground level (m

3

) 448.2 330.2 118.0

Overall wood volume above cavities level (m

3

) 501.2 364.5 136.7

Cavity/cavities area (m

2

) 18.7 15.9 2.8

Table 1 Measured and calculated

main parameters of the

Platland tree

998 A. Patrut et al.

3.3 Dating results of samples collected from the cavities

The segments of samples collected from the cavities of Platland

tree, which correspond to a depth in the wood up to 15–20 cm,

consisted exclusively of new growth layers triggered by

successive fires. The dating results revealed six major fire

events that affected the cavities of the Platland tree, which were

dated around AD 1550, 1650, 1780, 1900, 1955, and 1990 (as

established by using the general IntCal04 or IntCal09

calibration); however, this investigation did not enable to

determine the age of the two stems (Patrut et al. 2010a).

Here, we present dating results of the oldest segment of

each of the seven cavity samples, labeled by an additional

index x. For a given sample, the oldest segment corresponds

to depths in the wood from 21 to 48 cm. The AMS results

and calibrated ages are listed in Table 2. The results are in

good agreement with the original positions of the respective

segments/samples in the two stems, showing that these

segments consist of the original old wood. The sequence of

segment ages demonstrates that the trunk of Platland tree

comprises two stems which fused partially some time ago.

The radiocarbon date of the oldest segment 2xoriginat-

ing from the cavity inside stem I was of 569± 29 BP (before

present, i.e., before AD 1950), which corresponds to a

calibrated age of 675±15 years. However, the oldest dated

segment 6xoriginates from the cavity inside stem II; its

radiocarbon date was of 909±23 BP, corresponding to a

calibrated age of 940±25 years.

The radiocarbon dates of the deepest segments collected

from the two cavities toward the fusion/common area, i.e.,

4xand 5x, were 327±18 and 330±27 BP, respectively. These

values correspond to calibrated ages of 440 ±20 and 440±

30 years. The practically identical values indicate that the

two stems of Platland tree fused partially at least 440 years

ago. Consequently, one can consider that the stems of

Platland tree fused ca. 450 years ago, around AD 1560.

3.4 Dating results of samples collected from the exterior

The dating results and calibrated ages of the deepest and

oldest segments of the three samples collected from the

exterior of stems are shown in Table 3. The two deepest

segments (45 and 51 cm) originating from stem I, i.e., 11x

and 12x, at different heights above ground (1.34 and

2.75 m), were radiocarbon dated to 49±27 and 122±25 BP.

The corresponding calibrated ages are 120± 25 and 150±

25 years.

The deepest segment 13x(20 cm) originating from stem

II, at a height of 1.20 m above ground, had a radiocarbon

date of 170±22 BP, with a corresponding calibrated age of

250±25 years.

3.5 Growth rates of the stems

Calibrated ages and positions of segments 11xand 12x,

which originate from samples collected from the exterior of

stem I, were used to derive the mean radial increase. Thus,

stem I had grown by 3.75×10

−3

m year

−1

over the past ca.

120 years to the height of 1.34 m, and by 3.40×

10

−3

m year

−1

over the past ca. 150 years to the height of

Table 3 AMS radiocarbon dating results and calibrated calendar ages of the oldest segments of samples collected from the exterior

Segment code

[stem]

Depth

a

[height

b

]

(10

−2

m)

Fraction modern

[error]

Radiocarbon date

[error]

(

14

C years BP)

Cal AD range(s)

1-σ

[confidence interval]

Segment age

[error]

(calendar years)

11x[I] 45 [134] 0.9939 [±0.0034] 49 [±27] 1706–1720 [12.1 %] 120 [±25]

1819–1833 [10.4%]

1862–1915 [45.7%]

12x[I] 51 [275] 0.9849 [±0.0031] 122 [±25] 1685–1707 [12.1%] 150 [±25]

1719–1732 [7.1%]

1808–1826 [9.2%]

1833–1885 [31.5%]

1913–1928 [8.2%]

13x[II] 20 [120] 0.9791 [±0.0028] 170 [±22] 1669–1682 [11.6%] 250 [±25]

1736–1781 [40.1%]

1799–1805 [5.5%]

1931–1945 [11.0%]

a

Depth in the wood of the oldest dated segment (from the bark)

b

Height above ground level

Age determination of large live trees 999

2.75 m. Such high values indicate that the very large stem I

is still growing very fast.

However, as reported elsewhere (Patrut et al. 2010b), the

mean radial increase expresses only an apparent growth rate

of trunk/stems. Trees grow in three dimensions and their size

is accurately expressed by the wood volume of the trunk/

stems, which is in proportion to the cross-sectional areas at

different heights. Consequently, the area increase and volume

increase are much better estimates of the growth rate than the

radial increase. If one considers in this respect the dating

results and positions of all six samples/segments collected

from the cavity and the exterior of stem I, one can state that

stem I accelerated its growth over the past ca. 150 years.

This result corresponds to our previous research on African

baobabs (Patrut et al. 2010b), as well to other research done

on giant tree species (Sillett et al.2010).

The calibrated age and position of segment 13x, which

originates from the sample collected from the exterior of

stem II, shows that this stem had grown by only 0.80 ×

10

−3

m year

−1

over the past ca. 250 years to the height of

1.20 m. This low value demonstrates that the smaller and

older stem II slowed down its growth considerably and that

it could be close to the final stage of the African baobab's

life cycle, when trees almost stop growing.

3.6 Ages of the stems

The oldest segments 2xand 6xfrom the cavities inside stem

I and stem II consist of the original old wood, which

corresponds to the respective positions in the stems. The

extrapolation of their calibrated ages to the presumptive

center/pith of each stem enabled us for determining the ages

of the two stems of Platland tree.

The segments 2xand 6xoriginate from heights of 1.45

and 1.70 m above cavity level, corresponding to values of

0.67 and 0.92 m above ground. For extrapolation, we used

field measurements, as well a cross-section of the two

stems at 1 m, which is close to the segment heights

(see Fig. 4).

The cross-sections at ground level and at 1 m show that

both stems of the Platland tree have relatively symmetrical

transversal sections, with quasi-ellipsoidal shapes. In this

case, for determining the presumptive center/pith of each

stem, i.e., the symmetry center, it is sufficient using the axes

which touch the center of the fusion section.

Age of stem I In the cross-section of stem I, we displayed

the minor axis of the ellipse, from an external point Y to the

fusion point O, which is the conjugate diameter from SE to

NW. Its length is YO=7.40 m. The present center of stem I

can be approximated with the midpoint of the axis YO,

denoted as C

I

. In this case, the two minor semi-axes/minor

radii are YC

I

=C

I

O=3.70 m.

It is mandatory, however, to determine the presumptive

position of the original center/pith of stem I, labeled C

I(o)

,

which can be approximated by the midpoint of the conjugate

diameter at the fusion moment, that occurred ca. 450 years

ago. After fusion, stem I had no more room for growing in

direction NW. Therefore, the original center C

I(o)

is shifted

toward NW relative to the present center C

I

. Therefore, we

calculated the linear increase in radius of stem I over the past

450 years, in direction SE. The age of segment 11xindicates

that stem I had grown toward SE by 0.45 m over the past ca.

120 years. In a very conservative estimate, if we consider

that this relatively large mean radial increase was constant

over the past 450 years, the radial increase toward SE over

this period was of ca. 1.70 m. In this case, the corresponding

axis/diameterofstemIwasof7.40−1.70= 5.70 m at the

fusion time, with the two semi-axis/radii, including C

I(o)

−O,

of 2.85 m. Thus, the distance between the present center C

I

and the original center C

I(o)

becomes C

I

−C

I(o)

=0.85 m.

The point 2x, which represents the intercept with the YO

axis of the virtual annual ring corresponding to the

position of segment 2x, is located at 3.75 m from the

external point Y and at 0.80 m from the original center

C

I(o)

,i.e.,Y−2x=3.75 m and 2x−C

I(o)

=0.80 m.

In order to determine the age of stem I, one should

add to the age of the oldest segment 2x(675± 25 years)

the time needed by stem I for growing from 0 to a radius

of 0.80 m. The growth rate of African baobabs, especially

during their early life stage, is very different. However, if

we consider that the stem I of Platland tree had grown fast

and it is still fast growing, then, based on its current

dimensions, sample ages and also our previous research

(Patrutetal.2010b), one can estimate that the time needed

tostemIforreachingaradiusof0.80mwasofca.

75 years. All these approximations sum to a maximum

Fig. 4 Cross-section of the two stems (at 1 m above ground) showing

the accurate positions of the oldest dated segments from each cavity

(2x,6x), along with the present center (C

I

,C

II

) and the original center

(C

I(o)

,C

II(o)

) of each stem

1000 A. Patrut et al.

error of ±75 years for the final age of stem I, including the

age error of segment 2x. Hence, the calculated final age of

stem I is 750±75 years. One can state that it started

growing around AD 1260.

Age of stem II In this case, we showed in the cross-section

of stem II the major axis of the ellipse, from an external

point Z to the fusion point O. This is also the transverse

diameter from NW to SE, with a length ZO =6.44 m. The

present center of stem II can be approximated by the

midpoint of the axis ZO, marked as C

II

. The two major

semi-axes/major radii are ZC

II

=C

II

O=3.22 m.

The original center/pith of stem II, labeled C

II(o)

, can be

approximated by the midpoint of the transverse diameter at

the fusion moment. After fusion, stem II had no more room

for growing in direction SE. Consequently, the original

center C

II(o)

is shifted toward SE relative to the present

center C

II

. For calculating the increase in radius of stem II

over the past 450 years toward NW, we used the age of

sample 13x, which indicates a growth of 0.20 m over the

past 250 years. As stated, this low value indicates that the

growth of stem II slowed down considerably over this time

frame. If one considers that the mean radial increase was

around twice faster from the fusion moment up to 250 years

ago, the linear increase toward NW over the past 450 years

was of ca. 0.52 m. Thus, the corresponding axis/diameter of

stem II was of 6.44−0.52=5.92 m at the fusion time, with the

two semi-axis/radii, including C

II(o)

−O of 2.96 m. The

distance between the present center C

II

and the original

center C

II(o)

becomes C

II

−C

II(o)

=0.26 m. The segment 6x

was practically positioned on the axis ZO, at 2.02 m from

the fusion point O and at 0.94 m from the original center

C

II(o)

,i.e.,O−6x=2.02 m and 6x−C

II(o)

=0.94 m.

For determining the age of the stem II, one should add to

the age of the oldest segment 6x(940±25 years) the time

needed to stem II for growing from 0 to a radius of 0.94 m.

If we consider that the stem II of Platland tree has grown

much slower than stem I, its smaller dimensions, the older

sample ages and also our previous research, we can

estimate that the time needed to stem II for reaching a

radius of 0.94 m was of ca. 120 years. All approximations

are within the final error of ± 75 years. The calculated final

age of stem II becomes 1,060 ±75 years, indicating that it

started growing around AD 950.

4 Discussion

4.1 The size of Platland tree

The Platland tree is listed in the South African National

Register of Big Trees and is included in the South African

List of Champion Trees (Esterhuyse et al. 2001; Depart-

ment of Water Affairs and Forestry 2008). The only

accurate parameter of the true size of a tree is the wood

volume. The overall wood volume of Platland tree

(501.2 m

3

) is considerably larger than that of the Sagole

tree (414.1 m

3

), which is generally considered to be the

biggest baobab (Van Pelt, unpublished results). Therefore,

according to the volume values reported here, the Platland

tree becomes the largest known African baobab.

4.2 The relation size–age in the case of Platland tree's stems

Commonly, one considers that the life span/longevity of a

tree specimen, in our case of an African baobab, depends

on several factors: (1) its genetic potential; (2) the

environment (including mean annual rainfall, altitude, mean

annual temperature, number of rainy seasons per year,

number of frosty days per year, nature of the soil, slope of

the land, availability to underground water sources, etc.);

(3) the number and severity of dangerous events and

disease episodes during its life cycle (fire episodes, fungi

attack, insects, baobab disease, elephant damage, human

damage, drought periods, flood, frost episodes, etc.).

Theoretically, individuals having all these factors iden-

tical should grow over time at the same rate. Consequently,

under identical conditions, the biggest/largest trees may

also be the oldest. Such considerations justify the common

fallacy about trees that size is in direct proportion to their

age. Certain baobab researchers noted, however, consider-

able size differences between individuals of identical age,

which were attributed to site differences. Moreover,

important size differences between specimens growing on

the same site were reported; these dissimilarities were

considered to be mainly of genetic origin (Breitenbach

1985; Wickens and Lowe 2008).

Our study on the Platland tree demonstrates, once again,

how questionable the size–age relation is and the large

errors it can generate. According to complete measurements

and dating results, the very large stem I is considerably

younger than the smaller stem II (overall wood volume

364.5 vs. 136.7 m

3

; age 750 vs. 1,060 years). A preliminary

genetic research (Catana and Patrut, unpublished results)

suggested that the two stems of Platland tree possess

identical DNA and belong to a single individual. The two

stems have also grown on the same site. In this case, we

need an explanation for the unexpected finding that a stem

which is almost three times bigger is, however, over

300 years younger than the smaller stem of the same tree.

Researchers of tall and large tree species around the

world learned from their field experience that the largest

specimens are usually not also the oldest. Typically, the

largest specimens are those which had grown very fast

when they were young and, eventually, continued their

Age determination of large live trees 1001

rapid growth (Hartesveldt et al. 1975). The research

reported here demonstrates that such statements might be

also valid for the African baobab. Our research on

radiocarbon dating of baobabs (Patrut et al. 2010c), as well

dating results presented by other researchers (Swart 1963;

Woodborne et al. 2010) show that very large specimens are

not necessarily among the oldest trees and that medium-

sized individuals can also be very old. If one considers that

the largest baobabs are those which grew very fast over

their first ca. 100 years of life and maintained their rapid

growth, we may have an explanation for our results on the

two stems of Platland tree. In this case, stem I had probably

grown much faster and is also considerably bigger, as the

climate conditions in the area over its first century of life

(ca. AD 1260–1360) were probably much more favorable

for baobabs than over the first century of stem II (ca. AD

950–1050).

4.3 The Platland tree system

According to dating results, the present trunk of the

Platland tree comprises two stems of different ages.

Therefore, morphologically, it can be considered a double

tree. On the other hand, according to the preliminary

genetic analysis, the Platland tree is a single individual.

Both stems sprouted very probably from the same

rootstock at different times and, therefore, they have

identical DNA.

However, in the case of multi-stemmed and multi-

generation baobabs, such as the Platland tree, it is a difficult

to state whether they are single or multiple trees. They

should be rather considered “tree systems”, with a complex

development over a long life cycle. Over time, stems might

die and collapse, while new stems can sprout at any

moment from the roots or from the other broken or

unbroken stems.

Based on the dating results, stem II of Platland tree had

grown over 300 years without being disturbed by the

presence of another stem. Nevertheless, it has an important

lean toward NW (up to 40°) and the cross-sections show an

obvious miss in its ellipsoidal shape in the present fusion

area. These aspects suggest that stem II could have been

pushed constantly toward NW after it started growing,

possibly by a larger and older stem 0, located approximate-

ly on the site of stem I. In this hypothetical scenario, stem

0 died and collapsed at a certain moment and was

“replaced”by the present stem I, which might have

emerged from its remains. However, it is improbable that

stem I would still comprise somewhere old wood traces

originating from stem 0. If this scenario had happened,

then the multi-stemmed and multi-generation Platland tree

system should be much older than the age we derived from

radiocarbon dating.

Acknowledgements This work was fully funded by the Romanian

Authority CNCSIS-UEFISCDI under grant PN II-IDEI 2354, Nr.

1092. AMS radiocarbon dating at the NOSAMS Facility is supported

by the U.S. National Science Foundation under Cooperative Agree-

ment OCE-0753487. We would like to thank Heather and Doug van

Heerden, the owners of Sunland Nursery, for granting permission for

on-site investigation and also for sampling the Platland tree.

References

Bronk Ramsey C (2009) Bayesian analysis of radiocarbon dates.

Radiocarbon 51:337–360

von Breitenbach F (1985) Aantekeninge oor die groeitempo van

aangeplante kremeteartbome (Adansonia digitata) en opmerkinge

ten opsigte van lewenstyd, groeifases en genetiese variasie van

die spesie. J Dendrology 5:1–21

Department of Water Affairs and Forestry (South Africa) (2008) List

of champion trees. Available at http://www2.dwaf.gov.za

Esterhuyse N, von Breitenbach J, Sőhnge H (2001) Remarkable trees

of South Africa. Briza, Pretoria, pp 155–161

Hartesveldt RJ, Harvey HT, Shellhammer HS (1975) The giant

sequoias of the Sierra Nevada. National Park Service,

Washington, pp 57–59

Martínez-Ramos M, Alvarez-Buylla ER (1998) How old are tropical

rain forest trees? Trends in Plant Science 3:400–405

McCormack FG, Hogg AG, Blackwell PG, Buck CE, Higham TFG,

Reimer PJ (2004) SHCal04 Southern Hemisphere calibration, 0–

11.0 cal kyr BP. Radiocarbon 46:1087–1092

Olsson IU (1986) Radiometric Methods. In: Berglund B (ed)

Handbook of Holocene palaeoecolgy and palaeohydrology.

Wiley, Chichester, pp 273–312

Patrut A, von Reden K, Lowy DA, Alberts AH, Pohlman JW, Wittmann

R, Gerlach D, Xu L, Mitchell CS (2007) Radiocarbon dating of a

very large African baobab. Tree Physiol 27:1569–1574

Patrut A, Mayne DH, von Reden KF, Lowy DA, Van Pelt R,

McNichol AP, Roberts ML, Margineanu D (2010a) Fire history

of a giant African baobab evidenced by radiocarbon dating.

Radiocarbon 52:717–726

Patrut A, Mayne DH, von Reden KF, Lowy DA, Venter S, McNichol AP,

Roberts ML, Margineanu D (2010b) Age and growth rate dynamics

of an old African baobab determined by radiocarbon dating.

Radiocarbon 52:727–734

Patrut A, von Reden KF, Lowy DA, Mayne DH, Elder KE, Roberts

ML, McNichol AP (2010c) Comparative AMS radiocarbon

dating of pretreated versus non-pretreated tropical wood samples.

Nucl Instr Methods B 268:910–913

Poussart PM, Mynemi SCB, Lanzirotti A (2006) Tropical dendro-

chemistry: a novel approach to estimate age and growth from

ringless trees. Geophys Res Lett 33:L17711

Reimer PJ, Baillie MGL, Bard E, Bayliss A, Beck JW, Blackwell PG,

Bronk RC, Buck CE, Burr GS, Edwards RL, Friedrich M, Grootes

PM, Guilderson TP, Hajdas I, Heaton TJ, Hogg AG, Hughen KA,

Kaiser KF, Kromer B, McCormac FG, Manning SW, Reimer RW,

Richards DA, Southon JR, Talamo S, Turney CSM, van der Plicht J,

Weyhenmeyer CE (2009) IntCal09 and Marine09 radiocarbon age

calibration curves, 0–50,000 years cal BP. Radiocarbon 51:1111–1150

Roberts ML, Burton JR, Elder KL, Longworth BE, McIntyre CP, von

Reden KF, Han BX, Rosenheim BE, Jenkins WJ, Galutschek E,

McNichol AP (2010) A high-performance

14

C accelerator mass

spectrometry system. Radiocarbon 52:228–235

Sillett SC, Van Pelt R, Koch GW, Ambrose AR, Carroll AL, Antoine

ME, Mifsud BF (2010) Increasing wood production through old

age in tall trees. For Ecol Manag 259:976–994

1002 A. Patrut et al.

Sofer Z (1980) Preparation of carbon dioxide for stable carbon isotope

analysis of petroleum fractions. Anal Chem 52:1389–1391

Swart ER (1963) Age of the baobab tree. Nature 198:708–709

Vogel JS, Southon JR, Nelson DE, Brown TA (1984) Performance of

catalytically condensed carbon for use in accelerator mass

spectrometry. Nucl Instr Methods B 5:289–293

von Reden KF, Schneider RJ, Cohen GJ, Jones GA (1994) Performance

characteristics of the 3MV Tandetron AMS system at the National

Ocean Sciences AMS Facility. Nucl Instr Methods B 92:7–11

Wickens GE, Lowe P (2008) The baobabs: pachycauls of Africa,

Madagascar and Australia. Springer, Dordrecht, pp 141–142

Woodborne S, Hall G, Basson S, Zambatis G, Zambatis N (2010) The death

of a giant: on the age of baobabs. Savanna Network Meeting, Skukuza

Worbes M, Junk WJ (1999) How old are tropical trees? The

persistence of a myth. IAWA J 20:255–260

Worbes M (2002) One hundred years of tree-ring research in the

tropics: a brief history and an outlook to future challenges.

Dendrochronologia 20:217–231

Age determination of large live trees 1003