Content uploaded by Jean-François Butaud
Author content
All content in this area was uploaded by Jean-François Butaud on Oct 05, 2015
Content may be subject to copyright.
170
...
Sandalwood (Santalum insulare) formerly planted on Tubuai Island, May 2012 (Photo J.-F. Butaud).
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
...
171
Marotiri), despite several indications on Tubuai
(Aitken 1930; Sam Tahuhuterani pers. comm.
2002). Furthermore, surveys done under the pres-
ent scientific program did not identify any intro-
duced species of sandalwoods. However, it is likely
that the Indian sandalwood is present in gardens
on Tubuai and Rurutu, the most populated islands.
… History of the discovery
and exploitation of sandalwood
from the Austral Islands
In 1896, Jean Nadeaud was the first botanist to
obtain the information on the Austral Islands san-
dalwood, in this case that of Raivavae (Nadeaud
1896). However, these sandalwood populations
had been made known to the western world some
70 years earlier, at the time of the sandalwood trad-
ers and the triangular trade between the Anglo-
Saxon sailors, China, and the Pacific islanders. The
latter supplied wood to the sailors in exchange for
metal objects, fabric, whaleboats, weapons and
liquor. Sandalwood traders used the wood as an
exchange currency with China, mainly for tea
(Shineberg 1967; Dening 1980).
The stock of sandalwoods on the two (or three)
islands must have been low compared to Fiji, the
Marquesas, Hawaii, or New Caledonia as there are
very few traces left of this exploitation.
Raivavae
According to Edwards (2003, partially quoting
Maude 1968): "Sandalwood traders were coming to
Raivavae in the 19th century. Captain Fodger con-
vinced the locals to supply 1.5 tons of sandalwood in
1812 (Maude 1968: 196) and in 1819, Captain Lewis
agreed with the Tahitian King Pomare II to take him to
Tubuai and Raivavae in exchange of sandalwood from
Raivavae. There are very few records of visits of san-
dalwood traders on Raivavae but sandalwood was as
at the time the main commodity in eastern Polynesia.
It is therefore likely that other traders came. Oral
tradition confirms this allegation as the top of the
Taamaa peninsula (probably Taamora, the summit
of Vaianaua peninsula) was allegedly covered with
sandalwood forest. Today, it is mainly grasslands.
(…) In 1817, sandalwood was the main commodity of
the eastern Pacific but trade was declining due to the
depletion of sandalwood resources on many islands.
In addition, the Australian government imposed a tax
Sandalwood is a small, root hemi-
parasitic tree, prized worldwide for
the distinctive fragrance of its heart-
wood (its sapwood has no scent).
Sandalwood essential oil, extracted
by distillation of the wood, is used
in perfumery and aromatherapy. In
India and China, sandalwood has
been used in religious ceremonies
for thousands of years, particu-
lary as temple incense (Ehrhart &
Nasi1996).
There are 16 species of sandalwood,
exclusively distributed in the Indo-
Pacific region (Barrett & Fox 1995).
The most commonly known species,
the white sandalwood (due to the
pale color of its heartwood) or Indian
sandalwood (Santalum album), grows
naturally and is cultivated in India, Indonesia, and
Australia, but has also been introduced into many
countries including French Polynesia. Other known
species are the yellow sandalwoods (with a yellow-
ish heartwood) from New Caledonia (Santalum
austrocaledonicum) and Fiji (Santalum yasi).
…
The sandalwood in French Polynesia
There is only one native species of sandalwood
in French Polynesia, the eastern Polynesian san-
dalwood (Santalum insulare). This species is also
found on the Cook Islands (Mitiaro only) and on
the Pitcairn Islands (Henderson only), hence it is
endemic to eastern Polynesia. In contrast, the Indian,
New Caledonian, and Fijian sandalwoods were intro-
duced relatively recently to French Polynesia for
timber. Among these three species, the Indian san-
dalwood is now a naturalized species on the semi-
dry hills of the northwestern coast of Tahiti, on
Moorea and Hiva Oa (Butaud & Defranoux 2007).
Botanists have divided Santalum insulare into
nine varieties restricted to the different archipela-
gos, based on their anatomy (flowers, fruits, and
leaves) (Fosberg & Sachet 1985). Of these nine
varieties, two are found on the Austral Islands: the
first is endemic to Raivavae (var. raivavense) and
the second to Rapa (var. margaretae). Botanists
and foresters have never been able to demonstrate
the presence of sandalwood on the other Austral
Islands (Maria, Rimatara, Rurutu, Tubuai, and
Jean-François Butaud, Jean-Pierre Bianchini,
Jean-Marc Bouvet, Vincent Gaydou,
Émeline Lhuillier,
Phila Raharivelomanana & Fanny Rives
Current State of Knowledge
and Implications
for Conservation
and Enhancement
Sandalwood
172
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
schooner the year after”. However, she questions
the veracity of the statements made by sandalwood
traders who might have wanted to conceal a more
important source of sandalwood (New Hebrides).
Given that these two sources are concordant, it is
highly likely that part of the sandalwood exploitation
on Rapa (and, similarly, on Raivavae) occurred in the
mid-1820s. These dates and information are reported
in the Encyclopédie de la Polynésie (Toullelan 1986).
Based on an average weight of 50 kg of heartwood
per tree, the 70 tons of sandalwood harvested on
Rapa represented about 1 750 individual trees.
Assessement
Sandalwood was occasionally exploited on the Austral
Islands over two decades (from 1812 to 1827). This
consisted of uprooting the trees, stripping them
of their sapwood on the spot and transporting the
logs to the seashore. Tree populations were thus
overexploited. J. Nadeaud (1896) noted that sandal-
wood used to be abundant on Raivavae, but that the
populations were plundered for sale (or bartering).
However, Edwards (2003) indicated that tribal wars
interrupted the sandalwood trade on Raivavae and
that much wood remained during the first quarter of
the 19th century. Nevertheless, it is possible that the
remaining populations were exploited between 1830
and the end of the 19th century.
of 2.10 pounds per ton of sandalwood (and mother-
of-pearl) (Maude 1968: 200). An important reserve
of sandalwood remained on Raivavae and Tubuai, but
high demographic pressure on limited resources caused
frequent tribal wars which sidelined the trade and the
flow of highly desirable European goods".
Maude & Crocombe (1962) further mentioned
that Captain Michael Fodger of the Daphne
obtained 1.5 tons of sandalwood as ransom for the
release of a Raivavae chief he held hostage. This is
more proof of the brutality of sandalwood traders
throughout Polynesia.
Fodger was the first to mention the presence of
sandalwoods on Raivavae. Captain Lewis of The
Arab was persuaded by Pomare II to take him to
the island in exchange for sandalwood, following
the information provided by Fodger (Massal 1973).
Rapa
Hanson (1973) recounts: "In 1825, taking advantage
of a stopover on Rapa, the galley Snapper loaded sev-
eral tons of sandalwood. In 1826, an English trader
settled on the island for the exploitation of resources,
including sandalwood, a commodity highly-prized by
the Chinese at the time". Shineberg (1967) notes that
sandalwood had been found on the Austral Islands
in the 1820s: "Captain Ebrill brought back 40 tons
from his 1826 voyage, and 28 tons in his 50-ton
TRADITIONAL AND CONTEMPORARY USES OF SANDALWOOD
ON THE AUSTRAL ISLANDS
The Polynesian name for sandalwood on the
Austral Islands is the same on Raivavae and Rapa:
"ahi". Early botanists sometimes spelled it "eahi",
including the "e" meaning "the" or "this". In com-
parison, sandalwood is called "ahi" on the Society
Islands, "puahi" on the Marquesas Islands, and
"a'i" on the Cook Islands.
As elsewhere in the Pacific (Brown 1935; Shineberg
1967), the inhabitants of the Austral Islands, in
particular those of Raivavae (Nadeaud 1896), iden-
tified several varieties of sandalwood based on the
qualities of the heartwood (scent, color, hardness).
It appears that regardless of the species, there are
roughly three varieties:
• a high-quality dark sandalwood often called red,
• a relatively high-quality sandalwood called yellow,
• a low-quality sandalwood called white.
These names have led to some confusion between the
high-quality Indian white sandalwood and the low-
quality variety (white) of Polynesian red sandalwood.
As far as Raivavae is concerned, J. Nadeaud noted in
1896, from the account of one of his acquaintances,
a native of the island, the different varieties recog-
nized by the islanders:
• The "ahi mara aito" with hard and dark wood;
• The "ahi popouru" with soft and very fragrant
wood, the real "ahi";
• The "ahi tiare", called "ahi marea" by the
Tahitians, with a pale yellow wood.
While the inhabitants of Raivavae distinguished
the different varieties, we have no record of any
traditional use from the island even if it is likely
that many uses existed. An ethnobotanical study
of the island should help collect more information
even if some knowledge has been lost.
On Rapa, sandalwood was only used as fuel wood
due to its pleasant smell. It is likely that it was also
appreciated for its insect repellent properties.
Aitken, an ethnologist who studied life on Tubuai
in the 1920s (Aitken 1930), reported several uses
of sandalwood on the island. Unfortunately, it can-
not be determined whether these applied to the
local sandalwood or to that from Raivavae. Nor is it
known whether the inhabitants practicing such uses
were those who originally came from other islands.
Indeed, Aitken’s inventory of the plants of the island
does not include the sandalwood or "ahi/eahi/iahi".
The reported uses included the following:
• Flowers and pieces of sandalwood or rosewood
are placed in the box used to store clean laundry.
• Sandalwood was used to perfume the tapa.
• "Raau vari": cure for painful periods or lack of
menstruation. Mix the water of four green coco-
nuts with a handful of dried sandalwood pow-
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
173
der. Dosage: half a glass of the warmed liquid
followed by the same quantity cold. Repeat
three times a day until healed.
• "Raau taria mai": cure for ear pains or ear abs-
cess. Crush six red leaves and six green leaves of
"tou" (Cordia subcordata); add a small quantity
of "monoi ahi" (coconut oil with sandalwood).
Pour a small quantity of juice directly in the ear,
several times a day. The mixture can be slightly
warmed before application.
It is interesting to note that both medicinal recipes
are very similar to Marquesan recipes.
There is clearly an imbalance of available information
between Tubuai, where there is little or no san-
dalwood left, and Raivavae and Rapa, where nat-
ural populations remain. It suggests that some
knowledge on the Raivavae and Rapa sandalwoods
must have been overlooked or lost. Ethnobotanical
studies of both islands are highly desirable in order
to preserve any remaining knowledge.
During the past ten years or so, a demand for san-
dalwood has arisen on Raivavae (no commercial
prospect for sandalwood exists on Rapa, where there
are only relict populations). Sandalwood is trans-
formed and sold as hairpins, powder to perfume the
monoi or even as raw wood on Raivavae or Tahiti.
BOTANICAL DESCRIPTION OF THE SANDALWOOD
FROM THE AUSTRAL ISLANDS
…
Botanical history
The first samples of sandalwood from the Austral
Islands were collected in 1921 by a couple of
American ethnologists, the Stokes, who spent sev-
eral months on both islands (mainly on Rapa). In
1935, the American botanist F.B.H. Brown used the
two samples to describe two new taxa:
• Santalum insulare var. raivavense for the
Raivavae sandalwood, considering the variety
to be very close to the Tahitian one (Santalum
insulare var. insulare),
• Santalum margaretae for the Rapa sanda-
lwood, considering that the anatomical diffe-
rences from the Tahitian variety were significant
enough to describe it as a different species.
In 1938, C. Skottsberg, using the samples collected
by H. St John and F.R. Fosberg during the 1934
Mangarevan Expedition, validated Santalum insulare
var. raivavense and reduced the Rapa sandalwood
to a simple variety of the Polynesian sandalwood,
Santalum insulare var. margaretae, considering that
the observed variation did not justify its naming as
a distinct species (Skottsberg 1938).
The lastest study, by Fosberg & Sachet in 1985,
maintained and repositioned both varieties within
the nine botanical varieties of Santalum insulare of
eastern Polynesia and considered the species as a
variable complex.
…
Recent morphological data
The botanical descriptions by Fosberg & Sachet
(1985) can be amended with our own observations
and collections of fruits (N = 36), seeds (N = 206),
and leaves (N = 520). While hundreds of measures
of seeds and leaves were taken during the different
missions, the former descriptions were based only
on three collections for the Rapa variety and eight
for the Raivavae variety. In addition, no descrip-
tion of the mature fruit was available for either
variety; the fruit described from Rapa was clearly
immature. The newly acquired data are presented
in Tables 41 and 42.
Fruits and Seeds
During the various expeditions, fruits, seeds (shell and
kernel), and shell pieces (endocarp) were collected
from the various populations. Sometimes, the only
parts available were leftover seeds gnawed by rats. The
figures in Table 41 relate to mature fruits and fully
developed seeds except the fruit collected on Rapa.
Only one green fruit, shaped as a double pear (nar-
rowed at the apex and base) was observed on Rapa
on a tree in the Karapoo Rahi population. Three old
nuts damaged by rats were found on the ground in
Anatakuri Nako and a single one in Karapoo Rahi. The
green fruit was close to maturity as its seed was well
developed and hard. However, it is likely that a mature
fruit would have been 1 or 2 mm bigger than the green
fruit, making it 28 mm long and 19 mm wide. The
dimensions and shapes of the fruits and seeds were
recorded. Average, maximum, and minimum values
for each population are noted in Table 41. Sandalwood
seeds have a more or less pointed apex and a more or
less developed excrescence at the base. These param-
eters were estimated, along with the roughness of the
seeds. Seeds have lines of weakness or ridges where
the shell cracks open upon germination; the number
of ridges, mainly between two and four, was recorded.
Raivavae fruits are round to oval, averaging about
25 mm in length and width while the only fruit
from Rapa is more oval and slender, reaching an
estimated length of 28 mm and an estimated width
of 19 mm at maturity. Based on these unbalanced
data, a net differentiating criterion would be the
length/ width ratio, which is higher than 1.4 on
Rapa and lower than 1.2 on Raivavae.
On Raivavae, seeds are oval, obtuse, and rough.
Most often, they have three dehiscence lines and
very little excrescence at the base. They have an
average length of 17.2 mm and an average width
of 15.7 mm with a length/width ratio lower than
1.2. The limited data available from Rapa indicate
a length of 19 mm and a width of 13.1 mm, with a
ratio higher than 1.3. The length/width ratio is also
a distinctive criterion between both varieties.
174
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
It should be noted that the dimensions of these
apparently viable fruits and grains can vary by a
factor of two within the same population. Seeds
from "motus" on Raivavae appear to be slightly
larger than those from the main island.
Leaves
The sample size was 10 adult leaves per tree.
Table 42 presents the average, maximum, and min-
imum values for each population and each island
of the length and width of the limb, the length of
the petiole, the nervation, and the angles of the
apex and limb base.
Measurements show that the leaves of Raivavae san-
dalwoods are elliptical to oval, with a petiole of (13)-8-
(4) mm long. The limb is acute to obtuse, 79 mm long
and 42 mm wide on average and has 6 to 14 pairs of
veins. Leaves of Rapa sandalwoods have similar char-
acteristics, but are more frequently oval, with an obtuse
apex on the limb. On average, the limb is 66 mm long
and 32 mm with 6 to 11 pairs of veins and the petiole
is (10)-7.4-(5) mm. The limb is smaller on average on
Rapa (which could be due to more restrictive weather
Island
Population
Nb.
Fruit
Shape Fruit
Length
Fruit
Width
Nb.
Seed
Nb.
Ridge
Seed
Length
Seed
Width Apex
Excrescence Roughness
Raivavae
Anatonu-Mouatapu area 2
Round
23
22-21.5-21
13 3.1 21-18.1-16 18-16.5-15 Fairly
pointed None to little Very rough
Mahanatoa-Pahonu area 0
Round
-
-
3 3 18-17.3-17 18-17-16 Not very
pointed None Fairly rough
Motu Araoo 4
Oval
27-22.8-15
25-20.5-10
4 3.8 19-15-9 17-13.8-7 Not very
pointed Fairly Not very
rough
Motu Haha 0
-
-
-
1 3 15 13 Not very
pointed None Fairly smooth
Motu Mano Iti 1
Round
27
27
1 3 20 19 Not very
pointed None Fairly rough
Motu Numiri Rahi 1
Oval
34
31
28 3.4 24-21.1-19 20-18.2-16 Pointed Little Fairly smooth
Motu Rani 24
Flattened
Oval
32-25.2-23
29-25.9-24
87 3.3 21-17.4-15 19-15.4-13 Not very
pointed Little Fairly rough
Motu Tahaupaetoru 1
Oval
30
26
2 3 19-18.5-18 17-16.5-16 Fairly
pointed Little Fairly rough
Motu Vaiamanu 1
Oval
31
27
2 3 22-20-18 20-19-18 Pointed None to little Fairly rough
Vaianaua peninsula 1
Round
22
22
48 3.1 20-15.8-13 18-14.6-12 Not very
pointed None to little Fairly to very
rough
Total 35 Oval
34-25.3-15
31-25.1-10
201 3.2
24-17.2-9
20-15.7-7
Not very
pointed
Little
Fairly to
very rough
Rapa
Karapoo Rahi
1
Double
Pear
26
18
2 2 19 14 Pointed A lot Rough
Anatakuri Nako
0
-
-
-
3 - - 12.5 - - -
Total 1
Double
Pear
26
18
5 2 19
14-13.1-12.5
Pointed
A lot Rough
Table 41: Morphological data of fruits and seeds of Santalum insulare from the Austral Islands (dimensions in mm, max-average-min).
Ripe sandalwood fruit on Motu Araoo, Raivavae,
January 2005
(Photo
J.-F. Butaud).
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
175
conditions on Rapa). In addition, the angles of the
apex and limb base help distinguish the two varieties:
the leaves on Rapa are more triangular or slender at
the apex (65 grads on Rapa and 90 on Raivavae) and
slightly less rounded at the base (95 grads on Rapa and
110 on Raivavae).
There was no significant variation between the popu-
lations of the high island and those of the "motus"
on Raivavae. However, the leaves from the low-alti-
tude sheltered populations appear to be larger than
those from the exposed populations at higher altitude.
Thus populations at Mouatapu (high altitude cliff),
Mahanatoa Pahonu (exposed rocky area at medium
altitude), and Tapioi (boulder at mid-altitude) have
smaller limbs. These observations indicate the existence
of ecotypes by adaptation to site-specific conditions.
Assessement
More information is now available, allowing a better
characterization of these two varieties. Differentiating
criteria are clearer for seeds and leaves but there is an
important variability within each variety and popula-
tion. In addition, these variations could be correlated
with ecological conditions such as altitude and climate,
showing that the sandalwood adapts to its habitat.
Island Population
Nb.
trees
Shape
(E = Elliptical,
O = Oval)
Pétiole Lenth Width
Apex
(A = Acute,
O = Obtuse)
Nervation Apex
angle
Base
angle
Raivavae
Anatonu-Mouatapu area 3E > O 7-5.8-5 88-71-60 50-40-31 A11-9-7 130-95-70 150-120-100
Anatonu temple 3E > O 12-9.4-6 123-83-62 50-39-29 O > A 12-9-6 100-75-60 150-110-70
Hotuatua islet 3E = O 11-6.7-5 109-81-65 52-45-34 A12-9.5-8 120-85-60 150-120-90
Mahanatoa-Pahonu area 1E > O 8-7.1-6 58-52-43 32-29-26 O = A 9-7.5-6 130-90-70 130-100-90
Anatonu-Mahanatoa area
1 E 11-9.8-8 87-76-72 50-40-30 O8-7-7 120-85-70 100-90-80
Mouatapu 1E = O 8-6.4-6 58-53-47 31-27-20 A8-6.5-6 100-70-60 130-115-90
Tapioi 1 E 12-11.3-10 72-63-57 47-38-32 A10-8.5-7 120-95-80 120-100-80
Vaianaua peninsula 8E > O 12-8-5 100-70-50 61-39-22 A > O 12-9-6 160-105-60 160-105-70
Motu Araoo 3 E 13-9.6-6 131-98-79 74-51-43 O13-9.5-6 110-90-70 120-95-80
Motu Haha 2E = O 11-8.3-5 91-77-66 58-49-31 O > A 11-9-7 120-90-70 150-125-80
Motu Mano Iti 2 E 9-6.3-4 100-90-71 48-40-31 O13-12-10 90-80-60 120-100-80
Motu Mano Rahi 3 E 9-7.6-6 94-83-40 63-44-35 O > A 14-11-10 130-95-70 120-110-80
Motu Niupapa Rahi 1 E 9-7.3-6 94-84-72 45-38-31 A10-9-8 100-80-70 160-120-100
Motu Numiri Rahi 3 E 11-7.7-4 99-83-66 57-45-30 O > A 13-10-7 140-105-70 150-115-80
Motu Rani 3 E 11-8.3-5 113-87-66 54-47-30 O > A 12-9.5-8 120-90-70 120-100-80
Motu Vaiamanu 5E = O 12-8.5-4 107-85-64 58-43-31 O = A 14-10.5-8 130-85-50 140-110-80
Total 43 E > O 13-8.0-4 131-79-40 74-42-20 A = O 14-9.5-6 160-90-50
160-110-70
Rapa
Karapoo Rahi 7O > E 8-6.5-5 91-67-46 50-33-22 O11-8-6 100-65-40 130-95-70
Anatakuri Nako 2O > E 10-7.8-5 78-64-50 37-31-22 O12-8.5-6 80-60-50 120-90-60
Total 9O > E 10-7.4-5 91-66-46 50-32-22 O11-8-6
100-65-40
130-95-60
Table 42: Morphological data of leaves of Santalum insulare from the Austral Islands (dimensions in mm and angles in grads, max-
average-min).
Inflorescence of sandalwood on Motu Numiri Rahi,
Raivavae, November 2002
(Photo
J.-F. Butaud).
176
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
This assessment is based on information from
herbarium samples, inhabitants of the islands of
Tubuai, Rapa, and Raivavae, and recent surveys.
…
Tubuai
As mentioned above, no botanist has ever found
any sandalwood on Tubuai. The locals simply
recall that "previously", sandalwood was present
on the island (Butaud 2006). Recently, only one
person could attest the presence of sandalwood
ten years earlier on Tubuai (S. Tahuhuterani pers.
comm. 2002). It seems reasonable to assume that
sandalwood still exists on the island based on
this account and the configuration of the island.
The area mentioned includes Mt Tavaetu and Mt
Hanareho, their western slopes, and their east-
ern foothills where a tree supposedly used to be
exploited near the old "tou" (Cordia subcordata)
or "miro" (Thespesia populnea). In 2003, a one-day
survey focused on the area but there was no sign of
sandalwood. If there are any sandalwoods remain-
ing on Tubuai, it is likely that there are fewer than
ten relict individuals. However, it is likely that a
few trees were recently introduced from Raivavae
and grow on Tubuai near settlements.
…
Rapa
Before our surveys, sandalwood populations on
Rapa Island were known to botanists from two
sites: the historical site of Mt Tanga, where the first
two samples were collected in 1921 and 1934, and
Anatakuri Nako bay, near the pass leading to Akao
bay, discovered in 1984. Our surveys focused on
these sites, as well as on two other sites mentioned
by islanders, in particular by Benjamin Pukoki,
the local SDR manager. The latter sites are located
respectively on the left side of Anarua valley and
on Karapoo Rahi islet. The findings of the surveys
are presented in Figure 31 and in the following
paragraphs (Butaud 2006).
The summit of Mt Tepiahu to Mt Tanga east of
Haurei (Ahurei) was surveyed, as well as the slopes
northwest of the pass between these two mounts. No
sandalwood tree was found. The vegetation is very
degraded today, consisting only of a bush of Cattley
guavas (Psidium cattleianum), ferns (Dicranopteris
linearis), and "reeds" (Miscanthus floridulus). This
degradation is probably due to fires and overgraz-
ing by goats. In addition, the site is relatively visible
and accessible from the current quay from which
the inhabitants or possibly visitors could have easily
spotted any sandalwood. Another Mt Tanga exists
north of Area and was visited. The vegetation there is
very low and again no sandalwood tree was observed.
Anatakuri Nako bay
The islanders, in particular B. Pukoki, know this
sandalwood site. The botanists J. Florence and
N. Hallé also visited the site in 1984. According to
J. Florence, it contained about 30 trees in 1984; 10
years ago (around 1992), only about 15 trees were
known from the site, according to B. Pukoki. The
survey shows that this historical population now
includes only three live trees and 13 dead trees, still
standing or washed away by a landslide. The site is
exposed to the north and has a slope of 70 %.
This stand is located in Anatakuri Nako bay, under
the pass to Akao bay. The main tree was found at
an altitude of 205 m and had a diameter at breast
height of 13 cm (stump diameter of 24 cm) for a
height of 4.5 m. The two other trees had a stump
diameter of respectively 13 and 26 cm. On the
observation date, March 26, 2002, the leaves just
came out, with a pale to soft green color and were
visible from far away. Only a few flower buds
were observed. No fruit was found. Three old nuts
damaged by rats were found on the ground: the
width of the shells was measured and was close to
12.5 mm.
The surrounding vegetation was very poor as
Freycinetia and Macropiper were the only semi-
woody plants directly found near the sandalwoods.
What appeared to be a Sophora had been dead for
several years and a little further, a few Cyathea and
Metrosideros were observed.
Several assumptions can be made to explain the
extremely rapid decline of this population. First,
no regeneration is possible due to the presence of
STATE OF THE RESOURCE
Figure 31: Status of Santalum insulare populations on Rapa.
© Pacific-image - From Saquet ill.
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
177
also present as indicated by an old eaten nut on
the ground.
On March 26, 2002, the trees had recent leaves of
light to soft green color. Some trees seemed to have
been affected by the sea spray. The phenological
stage was more advanced than on Anatakuri Nako
as most trees were in full bloom. A single fruit was
observed; a large double-pear shaped green fruit of
a length of 26 mm and a width of 18 mm, with a
19 mm long and 14 mm wide seed. The nut found
on the ground had the same dimensions. The
largest tree was 4 m tall with a diameter at breast
height of 17 cm.
The surrounding vegetation was very degraded and
poor due to overgrazing by goats. It was primarily
herbaceous layer including Paspalum conjugatum,
Commelina diffusa, Bidens pilosa, Ageratum conyzoi-
des, Macropiper sp. and Gahnia sp. There were sev-
eral scattered shrubs and trees, including Cyathea
medullaris (many dead individuals), Pandanus sp.,
Xylosma sp., and Celtis pacifica.
This population is likely the most important on
Rapa and apparently the only hope for safeguard-
ing this endemic variety if conservation measures
are taken. It is crucial to eradicate the goats from
this island of high value for flora (J.-F. Butaud
pers. obs.) and fauna. The second step would be to
eradicate the rats for the benefit of birds, insects,
and the regeneration of some plant species.
Assessement
There are very few data available on the original
range of the Rapa sandalwood. Therefore, it is
difficult to assess the extent of the populations of
sandalwood before the arrival of sandalwood trad-
ers or herbivores about 200 years ago. However,
several assumptions can be made:
rats and their predation on fruits, as well as the
presence of goats and their predation on any plant-
lets and suckers. Mature trees are constantly under
the attack of goats (and maybe cows) that eat the
leaves. This is obvious as no leaf was found below
one meter fifty from the ground. Furthermore,
goats are causing the disappearance of host plants,
leading to a nutritional stress for the sandalwoods.
And last, it can be reasonably assumed that soil
compaction due to the presence of herbivores has
a negative impact on sandalwoods (root asphyxia)
and that, associated with the loss of neighboring
trees, contributes to landslides on steep slopes as
observed for this population. The landslide has
washed at least three sandalwoods and four other
live or dead trees are located at its edge. It is
remarkable that anthropogenic exploitation of san-
dalwood has played no role in the decline of this
population. There was no sign of exploitation on
this site well known by the islanders.
A potential conservation effort for this very precari-
ous stand would be to build a fence impassable to
herbivores and to plant native trees near the san-
dalwoods and the landslide.
Anarua bay
Thirty years ago, B. Pukoki noticed at least three
sandalwoods on the southern side of Anarua bay,
above Akaeke point. The survey area is located
at an altitude of 150 m, with a northern expo-
sure and an average slope of 80 %. This site is
a relict semi-dry forest including the following
woody species: Metrosideros collina, Nesoluma poly-
nesicum, Streblus anthropophagorum, Psydrax odo-
rata, Cyclophyllum barbatum, Myrsine sp., Meryta
sp., Myoporum rapense, Boehmeria virgata, Celtis
pacifica, Maytenus pertinax, Allophyllus rapensis,
Scirpus sp., and Freycinetia sp. No sandalwood,
dead or live, was found. The forest was highly
degraded due to overgrazing by goats and cows.
Several trees were observed, particularly on the
lower part. The sandalwoods observed 30 years ago
might have been exploited and thrown to the ocean
from the neighboring cliff.
Karapoo Rahi islet
B. Pukoki and several other inhabitants knew this
population of sandalwoods. It is found on an islet
located south of the island, separated by a stretch
of shallow water of about thirty meters wide. The
side where the sandalwoods are found has a west-
ern and northwestern exposure, an altitude of 80 to
110 m and a slope of 80 %.
This population comprised 11 distinct adult trees
and four dead trees. There was no sign of regenera-
tion and no indication of exploitation by man. The
sapwood was still intact on the dry trees, which
means that they died not long ago. Dry leaves
were even present on one of the dead trees. Goats
(about thirty) have devastated this islet. Rats were
Dead sandalwood on Karapoo Rahi islet,
Rapa, due to the loss of shrubby host plants
and overgrazing by herbivores, April 2002
(Photo
J.-F. Butaud).
178
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
the local inhabitants, nor the Hawaiian botanists
who visited the area for a month during the first
half of 2002 were able to find other trees.
The 14 trees found in 2002 should be compared to
the estimated 1 750 trees exploited by sandalwood
traders in the beginning of the 19th century. This
clearly shows that exploitation by man was the
main factor leading to the depletion of sandalwood
on the island of Rapa.
…
Raivavae
Before our various surveys, seven populations of
sandalwood were known through herbarium sam-
ples. Three populations were from coral "motus"
(Vaiamanu, Tahaupaetoru, and Haamu), one from a
volcanic islet (Hotuatua), and the three others from
the high island itself (Vaianaua, northern slopes of
Mt Hiro and south side of Mt Hiro). Samples high-
lighted one of the characteristics of the Raivavae
sandalwood: it grows both on limestone on "motus"
and on the volcanic soils of the high island. Among
the seven populations, six were found during our
recent surveys. The population on the southern side
of Mt Hiro was not located but is not necessarily
lost as the survey in the area was limited. In addi-
tion, 13 new populations were found. All sites were
surveyed and studied (Fig. 32) with the exception
of Anatonu stadium — the population was observed
through binoculars —, the south side of Mt Hiro,
and Motu Haamu, which was not surveyed during
the expeditions but still has many sandalwoods (W.
Tetuanui, SDR pers. comm.). Table 43 presents the
altitude and population size for each site.
• The exposure is significantly similar for the
four sites visited: west, northwest to north. It
appears that the sandalwoods on Rapa prefer the
sites most exposed to the sun and heat, with the
driest conditions. This confirms the observations
made on the Society Islands (Butaud 2004).
• The altitude range is also relatively restricted
for the various sites (between 80 and 250 m).
The sandalwood seems to grow at a lower alti-
tude, with a lower level of precipitations.
• Sandalwood populations were observed on
steep slopes (70 to 80 %), which are common on
Rapa. This does not preclude the growth of san-
dalwoods on gentler slopes (< 50 %). However,
populations on gentle slopes are easier to access
and exploit, which might have led to their loss
on such sites and their restriction to steeper
slopes today.
• Finally, while highly degraded, the vegetation
on theses sites could be considered as a low
altitude mesophilous vegetation, but not supra-
littoral. Most tree species found on these sites
do not grow at higher altitudes (above 300 m)
where the vegetation varies from hygrophilous
to ombrophilous.
It is highly likely that the original range of the Rapa
sandalwood covered areas of mesophilous vegeta-
tion, at an altitude between 10 and 300 m, with a
slope higher than 50 %, and facing west, northwest
and north. Potential surveys should focus on this
type of habitat to search for new populations of
sandalwood. The chances are very small as neither
Figure 32: Status of Santalum insulare populations on Raivavae. © Pacific-image - From Saquet ill.
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
179
In addition to the "motus" in
the aforementioned table, the
Papararuu, Ruahune, Arae,
Ruahota Rahi, Ruahota Iti,
Opunui, Numiri Iti, Tou,
Otaha 4, and Otaha 3 "motus"
were visited but no sandalwood
was observed. It appears that
the size of the "motu" might
affect the presence or absence
of sandalwood. Below a certain
surface threshold, sandalwoods
are not present. Small "motus"
are often submerged by the sea,
which could impede the growth
or germination of sandalwoods.
Larger "motus" are higher and
often protected against sea rise.
The originality of Raivavae
compared to the other islands
where sandalwood can be
found in French Polynesia lies
in the diversity of its soils. The
Raivavae sandalwood can be
found both on the "motu" coral
soils and on the volcanic soils
of the high island. Sandalwoods
on "motus" are peculiar, as
are some other plant species
(Myoporum stokesii, Nesoluma
polynesicum) unusual for this
type of habitat. Precipitations
might be a factor as they are
well distributed throughout the
year and might wash out the salt from the sols
(150 mm per month on average with a minimum
of 108 mm in September).
Therefore, there are two main types of populations
(Butaud 2006).
Populations on "motus"
"Motus" are flat limestone areas, varying from
sand to gravel with more or less organic mat-
ter. The surrounding vegetation is rather sparse
with a grass cover. The main trees found near
sandalwoods include the "tafano" (Guettarda spe-
ciosa), the "fara" (Pandanus tectorius), the coco-
nut tree (Cocos nucifera), the "pu'atea" (Pisonia
grandis), and the "'aito" (Casuarina equisetifolia).
It should be noted that no other species, includ-
ing the sandalwood, is found on the pure stands
of "'aito", "pu'atea", and "fara". The shrub layer
can contain the "pohotura'aua" (Sophora tomen-
tosa), the "nanie" (Suriana maritima), the "ouru"
(Scaevola sericea), Hedyotis romanzoffiensis and the
"miki miki" (Pemphis acidula). The herbaceous
layer includes the "naveoveo" (Lepturus repens),
the "tonene" (Cassytha filiformis), Psilotum nudum,
Chamaesyce fosbergii, Boerhavia tetrandra, the
Location Population Altitude
(m)
Size
(nb. individuals
)
Volcanic high island
Vaianaua peninsula 40-80 20
Mahanatoa-Pahonu area 10 0 20
Border Mahanatoa-Anatonu 50-80 70
Anatonu temple 50-125 80
Anatonu stadium 70 > 10
Anatonu-Mouatapu area 30-90 90
Mouatapu 235 1
Mt Hiro 400 < 10 ?
Mt Tapioi 120 1
Hotuatua islet 3-45 15
Coral "motu"
Motu Vaiamanu 41 000
Motu Numiri Rahi 245
Motu Rani 160
Motu Tahaupaetoru 250
Motu Araoo 3250
Motu Niupapa Rahi 2300
Motu Haamu 4> 100
Motu Mano Rahi 2150
Motu Mano Iti 217
Motu Haha 250
Total 1-400 > 2 340
Table 43: Description of the populations of Santalum insulare on Raivavae.
The only natural plantlet of sandalwood
observed in all of Polynesia on Motu Rani,
Raivavae, November 2002
(Photo
J.-F. Butaud).
180
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
"metuapua'a" (Microsorum grossum), Achyranthes
aspera var. velutina, the "pia" (Tacca leontopet-
aloides), Davallia solida, and Nephrolepis sp.
Regenerated sandalwoods were only observed on
the Motu Rani. Most trees are from root or stump
suckers, following harvest but mainly due to acci-
dental or deliberate (clearing) fires that regularly
devastate some islets such as Motu Mano Rahi. The
biggest sandalwoods are 6 m tall with a diameter of
15 cm at the base. On average, they are about 3 m
tall with a diameter of 8 cm. The heartwood is only
formed in the roots when they are not unearthed
for sale or for monoi.
Populations on the high island
They are found mainly on dry crests with a north-
ern exposure (populations from Anatonu to
Mahanatoa) as well as near rocky areas (Tapioi,
Mouatapu, Vaianaua Peninsula, and Hotuatua
islet). These topographical conditions match those
of the sandalwoods found on the other islands
of French Polynesia. The slope is not necessarily
steep and some trees can even be found on a flat
terrain (Mt Taamora on the Vaianaua peninsula).
The surrounding vegetation is usually low (less
than 5 m), open, and degraded by fires or feral
goats. The tree and shrub layers include "purau"
(Hibiscus tiliaceus), "fara" (Pandanus tectorius),
Xylosma suaveolens, Celtis pacifica, Premna obtusi-
folia, Glochidion raivavense, Psidium guajava, and
even Serianthes rurutensis. The herbaceous layer
consists of Microsorum grossum, Ocimum gratis-
simum, Miscanthus floridulus, Elephantopus mollis,
Dianella intermedia, Nephrolepis sp., Dioscorea sp.,
Davallia solida, and Dicranopteris linearis. The lower
Sandalwood from a root sucker on Motu Mano Iti,
Raivavae, November 2002
(Photo
J.-F. Butaud).
Shrubby sandalwood on low coastal vegetation on
Motu Rani, Raivavae, January 2005
(Photo
J.-F. Butaud).
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
181
The sandalwood from the Austral Islands is faced
with a contrasting situation on the two islands
where it is still found. On Rapa, the remaining
populations are relict, quite old, in a moderately
anthropized environment. On Raivavae, the popu-
lations consist essentially of numerous young trees
surviving in a highly anthropized and modified
habitat. Regeneration, both seed-propagated and
vegetative, has not happened for dozens of years on
Rapa. On Raivavae, vegetative propagation through
root and stump suckers is intense and seed-propa-
gated reproduction, although marginal, still exists.
These dynamics are related to relatively similar
but slightly different ecological and anthropogenic
conditions among the islands. These conditions
are presented below and are not fundamentally
different from the other archipelagos (Butaud &
Tetuanui 2005).
…
Poaching
While direct human impact is not the main cause
for the diminishing numbers of sandalwoods in
the Austral Islands today, it was in the case in
the past and some populations did not recover
from the abusive harvests of the past centuries.
Poaching is still effective today and continues to
decrease the populations and to limit the growth
of large fruit-bearing trees. Focusing on the best
trees, the harvest contributes to a negative selec-
tion among varieties and limits the enhancement
opportunities under a production plantation pro-
gram. Human exploitation is particularly visible
on Raivavae due to a recent demand for sandal-
wood, arising partly from the regular meetings
in Tahiti, during craft shows, of craftsmen from
all over French Polynesia. On these occasions,
Marquesans, who are large users of sandalwoods,
were able to engage with the Raivavae and buy
sandalwood at a lower price than in the Marquesas
Islands. The Raivavae realized the importance of
the available resource and the potential revenues.
In "handicraft fare" from Rairua to Raivavae, sev-
eral sandalwood objects can be found including
hairpins (between 1 000 XPF and 1 300 XPF),
pieces of sandalwood more or less stripped of sap-
wood (2 000 XPF for a 60 cm long by 5 cm wide
branch or 3 000 XPF for a hollow branch of 80 cm
in length and 10 in width), and sandalwood pow-
der (2 000 XPF for a 150 ml jar), which is often of
low quality as it includes the sapwood or is mixed
with "ngaio" (Myoporum stokesii), another fragrant
wood. In Tahiti, the "craftsmen’ Heiva" offered in
2001 some sandalwood powder at 1 100 XPF per
bag (20 cm long, 5 cm wide) and in 2002 a non-
carved piece of sandalwood from Raivavae (30 cm
by 12 cm) at 50 000 XPF. At the 2005 agricultural
fair in Tahiti, a 35 cm piece with a 10 cm diam-
eter was offered at 3 000 XPF and a 35 cm piece
with an 8 cm diameter was sold at 2 500 XPF. Due
to the growing demand, we have observed many
trees stripped of their branches (particularly on
"motus") and others that are totally uprooted and
exploited (Anatonu, Vaianaua). While the san-
dalwood variety from Raivavae is not protected
through the regulations on nature protection, har-
vesting is regulated under the Polynesian Forest
Code, which requires felling permits.
…
Fire-induced degradation
Fire represents the second major threat on Raivavae.
Burning is deliberate on "motus" for clearing and
accidental on the heights of Anatonu. Land clear-
ing by fire is forbidden by the Polynesian forest
regulations but is still practiced. It leads to the loss
of populations of fruit-bearing sandalwoods, creat-
ing, in the best cases, numerous suckers from the
scars on the roots.
…
Predation of seeds by rats
The main limiting factor of the seed-propagated
reproduction on both islands is the predation of most
seeds by rats feasting on their kernel. Caused by
the Polynesian rat (
Rattus exulans
) on both islands
and by the Black rat (
Rattus rattus
) on Rapa, many
part of the "Anatonu temple" stand is located on an
unusual raised limestone area, where some more
calcicolous species such as the "tafano" (Guettarda
speciosa) are found. The sandalwoods on the high
island are generally shorter than 5 m, including the
numerous suckers from the Anatonu populations.
Assessement
The original range of the Raivavae sandalwood was
very broad. Given its good dynamics (suckers and
natural regenerations) on coral "motus", it is cer-
tain that its presence is natural and not due to man.
The original range was likely restricted to the edge
of the "motu", on the seaside, on rocky sols among
Scaevola or on the lagoon side on sandy soils among
Sophora, as well as in areas cleared by storms. The
extension of the range was due to the presence of
men and the anthropogenic clearing on the "motus"
since the Polynesian period. Today, recurrent fires
contribute to its vegetative propagation through
root and stump suckers. On the northern coast of
the high island, sandalwoods probably grew on
most medium-to-steep slopes of ridges and sides, on
rocky soils or scree, with the driest exposure, from
sea level to the summit. Its presence on the more
humid southern coast must have been limited to the
steepest slopes. It was likely absent from the coastal
plains and the large valleys with gentler slopes.
DYNAMICS AND THREATS
182
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
…
Background, sampling and methods
The chemical study was based on a 37 samples of
wood (Table 44). Dead trees were sampled on both
islands. Therefore, the total number of samples can
be higher than the number of known individuals
(in particular for Anatakuri Nako on Rapa). The
data presented below were produced by Gaydou
(2004) in addition to new samples in 2005.
The chemical studies aim at characterizing the
essential oil of the Austral Islands sandalwood
and estimating its variability based on geographic
CHEMICAL CHARACTERISTICS OF THE SANDALWOOD
FROM THE AUSTRAL ISLANDS
gnawed and open shells were found. Unlike in the
Society Islands and the Marquesas Islands (except
in Tahuata), predation is not absolute on Raivavae
where several dozens of mature seeds were found.
This could be explained by the absence or scar-
city of Black rats on Raivavae (J.-C. Thibault pers.
comm. 2002). The Polynesian rat is less arboreal and
only eats a portion of the seeds, probably the most
accessible ones. The near lack of seed reproduc-
tion is mainly due to the successive introductions
(Polynesian and modern) of rats on the islands.
…
Extinction of disseminators
In the same vein, the likely extinction of several
avian disseminators such as the Imperial pigeons
and Fruit doves (except the "koko" Ptilinopus hut-
toni on Rapa) on the Austral Islands (Steadman
1997) would have also limited the chances of ger-
mination of sandalwood seeds. The uneaten ripe
fruits fall and rot quickly on the ground, resulting
in the rotting of the kernel (J.-F. Butaud pers. obs.).
…
Overgrazing by herbivores
The current main threat on the Rapa sandalwood is
overgrazing by introduced herbivores, principally
goats but also cows. The feral goat populations on
Rapa are only restricted by the available food. Their
size is increasing every year, as they are not hunted
enough, leading to the loss of natural forest. Goats
are highly present on both populations of sandal-
woods on Rapa. Leaves are only found at the top of
trees and no sucker bigger than 20 cm was found.
The surrounding vegetation has been grazed and
replaced by unpalatable weeds. Therefore, very
few host plants remain, soils are compacted due to
grazing, and sandalwoods decline at an accelerated
rate. Both seed-propagated and vegetative regen-
erations are impossible due to the pressure from
herbivores. While there are other reasons for the
scarcity of the Rapa sandalwood, goats will hasten
its loss. Goats are less abundant on Raivavae but
still present on the altitude stands and on Vaianaua
Peninsula where the absence of suckers is the first
impact of their presence.
…
Growth of invasive plants
Natural habitat changes due to more or less inva-
sive naturalized plants are also a concern on both
islands. Currently, there is no plant considered as
invasive in the sandalwood populations on both
islands but it is likely that in a near future, Psidium
cattleianum (Myrtaceae) on Rapa and Leucaena
leucocephala (Mimosaceae) on Raivavae will grow
and suffocate the mature trees, the suckers, and the
potential plantlets. In addition, many naturalized
grasses form the understory of sandalwoods and
certainly impair the germination of falling seeds.
The dynamics of sandalwood on the Austral Islands
has largely diverged from its natural path. While on
Raivavae the sandalwood seems to adapt to changing
conditions through a significant vegetative propaga-
tion, it might disappear soon from Rapa without any
voluntary effort. Threats are varied and accumula-
tive. Sandalwood is a good example of the challenges
faced by most natural habitats in French Polynesia.
Sandalwood branch debarked by rats on Mahanatoa,
Raivavae, December 2002
(Photo
J.-F. Butaud).
(islands, populations) and site-specific (type of
soil) criteria (Butaud et al. 2004).
Wood samples were collected mainly on dead wood
to prevent any injury to the rare trees remain-
ing in some populations. As most individuals, in
particular on Raivavae, are small, the high qual-
ity heartwood could not be collected. Often, only
false heartwood (duraminization due to injury) or
wood with many impurities were collected. While
experience has shown that in most cases the com-
position of false heartwood is very similar to actual
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
183
duramen, the nature and the
state of the wood were taken
into account in the result.
Wood extracts were ana-
lyzed by gas chromatogra-
phy. Concrete was obtained
through solvent extraction
from a small quantity of
wood and the essential oil
or essence by hydrodistilla-
tion of a larger quantity. In
practice, solvent extracts less
volatile (heavier) molecules
compared to hydrodistilla-
tion: the heavy fraction of the
concrete will not be found
(and no traces either) in the
essence. The proportion of
key molecules in the concrete
will be lower than in the
essence. Therefore, it is not
possible to accurately calcu-
late the yield of essential oil
as well as the proportion of
the most important essence
molecules on an olfactory
point of view.
The identification of mol-
ecules represented by chro-
matogram peaks was based
on the calculation of Kovats
indices (using retention
times) and on the coupling
with the mass spectrom-
etry and comparison with
the existing spectra data-
base. Twenty-six peaks or
groups of peaks could be
individualized. Some peaks
were grouped, as they could
not be separated in a suffi-
cient number of individuals.
They generally included one
majority compound and sev-
eral minority compounds.
…
Chemical composition of the concrete
Table 45 presents the results of the chemical analy-
ses of the Austral Islands sandalwood. For each
island, data are presented for all samples and for
samples of heartwood with no or few impuri-
ties (false heartwoods and impure duramens were
removed).
The molecules producing the characteristic smell
of sandalwood are mainly the (+)-(Z)-α-santalol
(α-santalol) and the (-)-(Z)-β-santalol (β-santalol),
on the dark green of the table. They form the main
compounds of the Austral Islands sandalwood con-
crete, ranging from over 35 % on average for all sam-
ples from Raivavae to about 60 % for quality samples
Table 44: Sampling of Santalum insulare on the Austral Islands for chemical and genetic
studies.
Island Location Known
numbers
Leaf
samples
Wood
samples
Raivavae
Vaianaua peninsula 20 8 3
Mahanatoa-Pahonu area 20 4 1
Border Mahanatoa-Anatonu
70 5
Anatonu temple 80 3 3
Anatonu stadium > 10
Anatonu Mouatapu 90 4 1
Mouatapu 1 1
Mt Hiro < 10?
Mt Tapioi 1 1
Hotuatua islet 15 3 2
Motu Vaiamanu 1 000 8 7
Motu Numiri Rahi45 3 1
Motu Rani 60 5 4
Motu Tahaupaetoru 50 3
Motu Araoo 250 5 3
Motu Niupapa Rahi 300 3 1
Motu Haamu > 100
Motu Mano Rahi 15 0 3 1
Motu Mano Iti 17 2 1
Motu Haha 50 3
Total
Raivavae
At least
2 340 64 (< 3 %) 80
Rapa Anatakuri Nako 33 (100 %) 4
Karapoo Rahi 11 11 (100 %) 5
Total
Rapa 14 14 (100 %) 9
Total
Australs
Over
2 350 78 (3 %) 37
from Rapa. Important variations were observed
among the various samples but most variations
seemed to be explained by the poor quality of the
heartwood collected. When only the purest samples
are considered (duramens with no or few impuri-
ties), the proportion of both molecules is about 49 %
on Raivavae and 59 % on Rapa.
The Rapa sandalwood appears to be of better qual-
ity than on Raivavae with a higher content in
α- and β-santalol. This needs to be put into per-
spective as the number of samples is relatively
low, in particular from Rapa, and the structure
of sandalwoods on Rapa is significantly different
than on Raivavae. Sampled trees on Rapa had a
184
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
Island* Raivavae (%) Rapa (%)
Samples All (28)
Pure heartwood (10)
All (9) Pure heartwood (5)
Indices Avg. Stand. dev. Avg. Stand. dev. Avg. Stand. dev. Avg. Stand. dev.
Concrete yield 7. 7 4 5.06 10.62 3.86 7.95 3.03 8.76 2.92
α-santalene 1.10 0.60 1.0 8 0.42 0.96 0.33 0.93 0.31
(Z)-α-trans-bergamotene 1.52 0.94 1.42 0.49 0.96 0.49 0.92 0.29
(+)-(Z)-α-santalol 26.58 11.6 9 35.18 7.33 33.19 14.96 40.60 12.52
(-)-(Z)-α-trans-bergamotol 4.69 1.87 5.80 1.46 4.64 0.81 4.92 0.46
(+)-(Z)-epi-β-santalol 2.04 0.82 2.55 0.57 2.86 1.19 3.53 1.06
(-)-(Z)-β-santalol 9.90 5.01 13.77 3.47 14.48 7. 51 18.33 6.65
(Z)-nuciferol 8.07 8.49 5.37 3.32 4.26 5.41 4.02 6.74
(-)-(E)-β-santalol 2.82 1.33 2.97 1.5 4 2.53 1.34 2.22 0.82
(E)-nuciferol 2.53 1.4 6 2.47 1.9 0 2.68 2.34 2.17 3.12
(Z)-α-santalyl acetate 4.09 4.21 3.86 4.90 1.63 2.24 1.6 8 2.89
Dihydroxy-bisabolene 4.22 3.67 3.90 3.05 3.34 2.75 2.83 2.52
α-santaldiol 0.94 0.70 0.87 0.32 1.24 0.72 1.14 0.33
Table 45: Composition of the Concrete of Santalum insulare on Rapa and Raivavae.
well-developed trunk, with an average diameter of
10 to 20 cm, and formed heartwood in branches,
trunks, and roots. On Raivavae, most individuals
were stump and root suckers, with a shrub or bush
structure, reaching less than 10 cm in diameter and
with heartwood formation only near the stump or
the root. The difference of quality should be related
to the age as well as to the location of the sampled
trees. Also, the difference of climate between both
islands (chillier and more humid on Rapa) has cer-
tainly an impact.
Compared to the Indian sandalwood (Santalum
album), the world reference in terms of heartwood
quality, the Austral Islands sandalwood is of good
quality. Standards for the essential oil of Indian san-
dalwood indicated α- and β-santalol contents respec-
tively between 41 and 55 % and between 16 and 24 %
(ISO 2002). As the data available are only for the
concrete in the case of the Austral Islands sandal-
wood (between 35 and 41 % for the α-santalol and
between 14 and 18 % for the β-santalol), the propor-
tion in the essential oil can only be roughly estimated
at 38-43 % for the α-santalol and 16-19 % for the
β-santalol. At first sight, the quality of the essential
oil from the Austral Islands sandalwood seems to be
slightly lower than the Indian sandalwood. For the
latter, extraction is done from trunks, branches, and
roots of large trees. On the contrary, for the Austral
Islands sandalwood, only small pieces of wood from
branches or trunks could be used. There is presum-
ably a high potential for the local varieties.
The extraction yield by hydrodistillation for
the essence of Indian sandalwood reaches 3 to
6 % (Alpha 1997) and about 2.5 % for the New
Caledonian sandalwood (Butaud 2006). The labo-
ratory extraction protocol using solvent has a yield
of 8 to 10 % of concrete for the Austral Islands
sandalwood. The solvent extraction technique pro-
duces a higher yield in α- and β-santalol compared
to hydrodistillation. Furthermore, several mole-
cules (heavy fraction) not extracted by hydrodistil-
lation are only present in the concrete. The yields
obtained are not directly comparable but suggest a
good yield (likely to be higher than 4 %) of essential
oil under a long-term enhancement program (from
plantations) of the Austral Islands sandalwood
…
Variability of the concrete
To assess the chemical composition variability
of the concrete of Austral Islands sandalwood,
a Discriminant Factor Analysis (DFA) was per-
formed on all samples grouped in four populations
(the two from "Rapa" and a group each for the
"high island" and the "motu" on Raivavae), based
on the 12 molecules presented in Table 45. The
analysis maximizes the differences among groups.
Figure 33 illustrates the DFA.
The DFA highlights group segregation for α-, β- and
epi-β- santalol (positive F1 axis) and bergamotene
(negative F2 axis) and α-santaldiol (positive F2 axis).
The Rapa populations are relatively distinct with
higher concentrations of α- and β-santalol. There
is an apparent distinction between the two groups
from Raivavae, mainly based on the type of soil.
Individuals growing on the coral "motu" have lower
concentrations of α- and β-santalol compared to trees
on the volcanic high island. The tree marked with an
arrow on the graph also illustrates this statement: the
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
185
-5
-4
-3
-2
-1
0
1
2
3
4
5
-5 -4 -3 -2 -1 0 1 2 3 4 5
-- F2 axis (16.15 %) -->
-- F1 axis (76.33 %) -->
Individuals (F1and F2 axes: 92.48 %)
Anatakuri Nako
High island
Karapoo Rahi
Motu
On coral
Figure 33: Graphical representation of the Discriminant Factor Analysis of the chemical composition of the concrete of Santalum
insulare from the Austral Islands, by population group.
tree is from the high island but its profile corresponds
more to the "motu" as it grows on raised coral on the
main island (Anatonu temple population).
The various surveyed populations present variable
chemical characteristics, some considered to be an
adaptation to the habitat, as demonstrated by the
slight differences in composition according to the
type of soil for the same botanical variety.
GENETIC CHARACTERISTICS OF THE AUSTRAL ISLANDS SANDALWOOD
Two genetic studies were successively carried out
based on leaf samples collected on islands where
sandalwoods grow, in French Polynesia and on
the Cook Islands (Mitiaro). The studies used chlo-
roplast microsatellite (Butaud et al. 2005) and
nuclear (Lhuillier et al. 2006) markers and were
performed in the laboratory of genetics of CIRAD-
forêt in Montpellier. The results from these studies
for the Austral Islands are summarized here.
Table 44 describes the sampling for both studies.
In addition to the samples collected in the Austral
Archipelago, two samples from Mitiaro, Cook Islands
(collected by Lex Thomson, CSIRO) were added to
the analysis using chloroplast microsatellites.
The evaluation of genetic diversity was only done
at the level of the archipelago and the islands, but
note at the population level to assess population
dynamics. Exhaustive sampling of a population
(except for the two relict populations on Rapa)
was not possible and priority was given to cover a
maximum number of populations.
…
Chloroplast microsatellites
Throughout French Polynesia, 17 different chloro-
types were found on about 500 collected samples
from 10 islands where sandalwoods are found, in
addition to Mitiaro, Cook Islands. No chlorotype
was shared among the archipelagos. In the Cook-
Austral Archipelago, only three chlorotypes were
identified: one specific to Rapa, one specific to
Raivavae, and one shared between Raivavae and
Mitiaro. There is no diversity on Rapa and Mitiaro
and an average diversity on Raivavae (Table
46),
compared to the existing data from the other
Polynesian islands. Values can be explained by the
small size of the islands and the associated small
number of habitats but also by the relict nature of
sandalwood populations due to exploitation, fires,
and introduced animals. In addition, the archipel-
ago has a strong genetic structure (69 % of the total
molecular variation observed among islands and
31 % within the islands) due to the high differentia-
tion between Raivavae and Rapa from the absence
of a shared chlorotype. Differentiation is likely due
to the distance (over 500
km) between the two
islands, which restricts the exchange of seeds.
The main finding is the notable differentiation
among the sandalwoods of each archipelago, with
distinctive chlorotypes. Within the Cook-Austral
archipelago, Rapa is well isolated by its specific
186
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
chlorotype, corroborating its
botanical (several endemic
genera) and geological (very
old age of the islets neighbor-
ing Marotiri) characteristics
(Florence 1997; Clouard &
Bonneville 2001). The fact
that the Mitiaro chlorotype is
shared with Raivavae might
suggest a same origin as well
as preferential movements
within a same geological archipelago rather than
between two islands that are close but belong to two
different archipelagos. In order to refine the recently
developed conservation strategy for the Mitiaro
sandalwood (Tangianau
et al
. 2004), a last resort
hybridization, with the introduction of seeds from
the genetically closer Raivavae sandalwood, should
be considered rather than from the geographically
closer Raiatea (as initially planned).
However, as shown by the minimum dissimilarity
network among the chlorotypes (Fig. 34), one of
the two populations studied at Raiatea (black solid
circle under Society) has a chlorotype that is rela-
tively close to the one from Mitiaro and Raivavae,
which suggests a past gene flow in the archipelago
that does not seem to exist anymore.
The network reveals three series of chlorotypes,
the ones from the Cook-Austral archipelago falling
between the Marquesas and the Society chloro-
types. This is surprising from a geographic point
of view (as the Marquesas are closer to the Society
than the Austral Islands) but is a reflection of the
important influence of geology and the distinct
ages and histories of the various archipelagos on
sandalwood colonization. The likely routes of
colonization are established but neither the direc-
tion (Marquesas-Austral-Society or the opposite,
or Austral-Marquesas and Austral-Society) nor the
origin (even if the Hawaiian Archipelago is the most
likely from a taxonomic point of view (Fosberg &
Sachet 1985; Harbaugh & Baldwin 2005)).
At the population level, the distribution of chloro-
types is not very structured on Raivavae. The two
chlorotypes are rather mixed on the "motu" and
only one seems to be present on the high island.
The lack of structure might be due to the small
size of the island and to the good dynamics of san-
dalwood, probably due to the scarcity or absence
of Black rats and the high number of individuals.
…
Nuclear microsatellites
Nuclear microsatellites are more polymorphic than
chloroplast microsatellites as they are subject to
random recombination of the genome during sexual
reproduction. Nuclear microsatellites help detect
diversity at a finer scale, down to the subpopulation
level, if used in sufficient numbers. In the present
case, eight markers were applied and were able to
define multilocus genotypes (allele combination of
Islands Nb. samples Nb. chlorotypes Effective nb.
chlorotypes
Diversity
(He)
Raivavae 48 21.65 0.40
Rapa 14 1 1 0
Mitiaro 2 1 1 0
Archipelago 64 32.42 0.59
Table 46: Parameters of genetic diversity for the Austral archipelago. He: heterozygosis.
8 locus or 16 alleles). Considering the significant
multiplication in situ of sandalwoods through suck-
ers, trees sampled as distinct individuals ("ramets"),
but with the same multilocus genotype, are pre-
sumed to be from the same genetic individual
("ortet") and form a clone. Diversity is calculated
based on allele frequencies as well as genetic struc-
ture indices. Genetic distances between genotypes
are calculated based on allele composition.
It appears from the study that a little over 50 %
of sampled sandalwood trees are clones, with the
same genetic heritage as all tested microsatellites
(Table 47). This is particularly surprising as care
was taken during sampling at Raivavae to focus on
scattered trees in order to limit the selection of root
suckers from the same parent plant. This means
that the actual number of individuals (genetic indi-
viduals) is at maximum half of the estimated trees
or a little less than 1 200 on Raivavae.
On Rapa, the same genetic individual ("genet")
constitutes both known populations in which all
trees were sampled. The situation appears to be
more critical than what was assumed by conser-
vationists. The situation on Raivavae is relatively
Society
Society
Marquesas
Rapa
Raivavae
Raivavae
Mitiaro
Figure 34: Minimum dissimilarity network among the chlorotypes
of Santalum insulare in Eastern Polynesia (Circle sizes are
proportional to the number of individuals - Branch lengths are
proportional to the genetic distance).
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
187
different as there are not many clones (less than
40 %) on the big "motus" and more frequent ones
(over 50 %) in the more concentrated populations of
the high island or the small "motus". The dynamics
of the sandalwood is therefore likely to be important
on the "motu" than on the high island.
This clonality is due to the capacity of the sandal-
wood to grow some suckers from stumps (after
exploitation) or roots (after a fire), all the more so
because of the slow or inexistent seed reproduction
caused by predation by rats and the extinction of
avian disseminators, thwarting all chances of seed
germination and dissemination.
Genetic structure is very important
among the various archipelagos (Fst
of 0.4 between the Marquesas and the
Austral Islands and 0.5 between the
Society and the Austral Islands), which
corresponds to the results of the prec-
edent study. Within the archipelagos, the
most important differentiation among
islands is between Raivavae and Rapa
(Fst = 0.648), confirming the recogni-
tion of the two botanical varieties even
if the markers used only measure neu-
tral diversity (Fig. 35). The Rapa gen-
otypes are very clearly distinguished
from those of Raivavae (100 % boot-
strap). Furthermore, the populations of
Raivavae present a significant differen-
tiation (Fst = 0.162; P < 0.001). This cor-
roborates the grouping of the population
genotypes from Vaianaua and the South
West Motu observed in figure 35.
Figure 36 indicates that the Austral Archipelago
has an intermediate position compared to the other
archipelagos in terms of variability of nuclear mic-
rosatellites, confirming the findings of chloroplast
microsatellites. Bootstrap values (the probability of
observing a branch at a given spot on a tree) are
low due to the small sampling on Rapa and other
islands outside of the archipelago but the structure
of the tree is well established.
…
Assessement
The situation is contrasted between both islands
with a significant differentiation between the two
varieties. Diversity is very limited on Rapa with
Figure 35: Neighbor-Joining tree of genotypes in the Austral archipelago.
Table 47: Genotypes and clones of Santalum insulare from the Austral Islands.
Populations Samples Nb. génotypes % of clones
Anatonu East 7 3 57
Anatonu West 13 562
Motu East 8 7 12
Motu South-West 8 5 38
Motu Vaiamanu 8 5 37
Petits Motu 10 550
Vaianaua 8 4 50
Total Raivavae 62 34 42
Anatakuri Nako 3 1 67
Karapoo Rahi 11 191
Total Rapa 14 286
Total Australs 76 36 53
188
...
.......
Sandalwood, Current State of Knowledge and
Implications for Conservation and Enhancement
observed in private homes or in the SDR nurseries on
Tahiti and Raiatea presumably came from Raivavae. A
pastor on the island was presumed to have developed
his own germination technique, applied today by the
SDR agents in Moorea (non scarified seeds are put to
germinate in moss collected from trees and the kernel
is manually released from the shell after germina-
tion). While no private plantations of sandalwoods
were found on Raivavae, this information shows that
the population is willing to cultivate and plant san-
dalwoods and that the seeds can be easily accessed.
The lack of a real nursery on the island also hinders
the efforts to develop this species.
…
Conservation and enhancement
strategies
Rapa
The situation of the Rapa sandalwood is very sim-
ple. There are only two populations left, in visible
decline, composed of a unique clone each and
presenting therefore major risks of inbreeding and
genetic depression. Trees bear very few fruits and
are directly threatened by grazing by feral goats.
Wood from this variety seems to of very high qual-
ity. The Rapa sandalwood variety is very differenti-
ated from the varieties from Raivavae and other
Polynesian islands. Conservation of Santalum insu-
lare var. margaretae is therefore a priority as it has
an original part of the genetic diversity of the spe-
cies, is at the verge of extinction, and constitutes a
high quality natural resource for the future.
…
Reproduction efforts
Rapa
On Rapa, the only known experiment on the germi-
nation of the Rapa sandalwood by local SDR agents
used very few seeds and was a relative failure: two
seeds sprouted, then quickly rotted. However, this
demonstrates the possibility of collecting viable
seeds and getting to the germination stage.
Raivavae
In May 1985, seeds of Raivavae sandalwoods were
sent to Tahiti by the Rural Economy Services (SER)
but we do not have any subsequent information.
In March and April 2002, an SDR agent sent about
200 sandalwood fruits from Raivavae, from Motu
Rani to the 3rd agricultural sector on Tubuai. The
fruits were pulped upon arrival on Tubuai. They
were put to germinate at the Mataura nursery after
scarification (complete removal of the shell). No
preliminary treatment using gibberellic acid was
performed as advised by the management tech-
nique for the production of sandalwood plants in
nurseries (Butaud 2001). The results of this germi-
nation were deceiving: some seeds sprouted but no
seedling was produced due to a lack of monitoring
at the nursery (rotten kernels and damping-off).
The local staff performed a similar germination, which
produced several seedlings, scattered here and there
on Tubuai. None of these seedlings was recently found.
It seems that some inhabitants on Raivavae succeeded
in reproducing the local sandalwood as several trees
REPRODUCTION EFFORTS AND CONSERVATION
AND ENHANCEMENT STRATEGY
only two "genets" and richer on Raivavae where
many populations do not have a defined struc-
ture. Diversity indices for the Austral archipelago
are significantly lower than on other archipelagos
(Expected heterozygosis (He) of 0.33 compared
to 0.51 on the Society and 0.55 on the Marquesas,
average number of alleles per locus (nA) of 3.25
compared to 6.50 on the Society and 6.13 on the
Marquesas). This means that diversity is lower
due to the degradation of the natural environ-
ment and the populations as well as the small
size of the islands compared to the other archi-
pelagos. Therefore, the sandalwood populations
on the Austral Islands are more sensitive to new
disturbances and require preferred and early con-
servation efforts.
On the biogeographical level, the Austral Islands
sandalwood could constitute a link between the
Marquesas and the Society sandalwoods.
Figure 36: Neighbor-Joining tree of the islands of French
Polynesia.
Terrestrial Biodiversity of the Austral Islands,
French Polynesia
.......
...
189
The proposed strategy would be to start with the
mandatory preservation of both populations (at
least one tree for each population) and the search
of new populations on favorable sites.
The Anatakuri Nako population should be fenced
(wire fence) in order to mitigate the impacts of graz-
ing by goats and to restore the surrounding vegetation.
The minimal area to be fenced should be 2 500 m2 (or
a 50 m by 50 m square and 200 m of wire fence),
centered on the three remaining trees and includ-
ing the higher and lower parts of the population.
This technique could help restore the adult trees and
obtain root suckers. The genetic diversity would not
be enhanced but the clone will have better chances of
preservation. In a second phase, local shrubs or trees
could be planted in the enclosure to restore the stock
of host plants. The trunk could be wrapped (without
any nails or perforating devices) with metal bands (as
used on coconuts) to prevent the destruction of seeds
by rats. This would allow the harvest of potential
fruits and germination experiments ex situ.
Goats should be eradicated (culling or capture) on
the islet of Karapoo Rahi. This eradication would
benefit both the sandalwoods and the overgrazed
vegetation of interest. The sandalwood population
could be restored in a similar way than in Anatakuri
Nako. Host plants can also be replanted. Regarding
rats (probably Black rats), metal bands could be used
but it seems more interesting to perform a complete
eradication of rats on the islet to restore the entire
ecosystem (seabirds, endemic insects, vegetation).
In both populations, the seed harvest should be
done on a regular basis (every two weeks) so that
the mature fruits do not fall on the ground (and
risk rat predation or rot). The maturation duration
is between five and six months (J.-F. Butaud pers.
obs.). The maturity stage of the fruit should be esti-
mated to determine when to visit the populations
that are far from the main village.
If the production and germination of seeds prove
successful, the seedlings can be used first to
enhance the existing populations and mainly to
diversify the gene pool. One tree from Anatakuri
Nako could be planted in Karapoo Rahi and vice
versa. In this way, tree fertility (reduction of con-
sanguinity) could be enhanced in case of a self-
incompatibility system.
Raivavae
The Raivavae sandalwood is not threatened by
extinction in the short run. The number of genetic
or apparent individuals is still important and genetic
diversity is relatively high. The priority is to estab-
lish a conservation plantation (seed orchard) and
to conserve in situ isolated or original populations.
There are two of the latter: the population of the
Vaianaua peninsula, which only has old trees and
is grazed by goats, and the altitude population(s)
near Mt Mouatapu and Mt Hiro, comprising very
isolated trees in areas inaccessible to goats. The
strategy is the same for these populations and
involves eradication or control of the goat popula-
tions. On the Vaianaua peninsula, sandalwoods are
easy to access and subject to poaching. Awareness
should be raised among the population or at least
the owner(s) of the area.
A conservation plantation requires the establishment
of an equipped nursery (fridge, automatic watering
system, protected germination area) and the training of
future nursery workers. This project is being developed
by SDR for fruit-bearing species. Sandalwood could be
included. The second step is to collect seeds from the
maximum number of different populations, as genetic
studies did not highlight any significant differentiation
among the various areas. Collections could be done
without using rat poison, in particular on "motus"
where rats are not abundant. Otherwise, rat eradica-
tion could be performed on some small "motus", well-
developed trees could be wrapped with metal bands, or
rats could be eliminated in some populations (use of
rat poison every two weeks at least). Harvested fruits
can be pulped, separated by origin, and stored in a cold
place. Germination could follow the protocol devel-
oped for the Marquesas (Butaud 2001). The conserva-
tion plantation will include at least 200 individuals
from a dozen of origins, or twenty some trees for each
population. In situ enrichment planting can be derived
eventually from the ex situ conservation plantation.
Arboreal sandalwood in coastal vegetation on Motu Vaiamanu,
Raivavae, November 2002 (Photo J.-F. Butaud).
