Adansonia madagascariensis, a marine hydrochory hypothesis

Abstract

Adansonia madagascariensis fruits were found in May 2011 on the beach of Anjiabe in the north of Madagascar. Marks found on the fruits and the absence of this species on this coast indicated that the fruits had been in the sea for a long time. The viability of seeds contained in these fruits was assessed and compared to the viability of seeds collected from under trees. The results show that the time spent in the sea did not affect their germination potential and that germination is possible, and even improves, in tidal zones. These initial results confirm the hypothesis of marine hydrochory for this species. They show that baobab fruits can remain in the sea without affecting the viability of seeds to colonize new areas. This could explain the frequent occurrence of this species along the coast of Madagascar and, more rarely, on the Mayotte coast.

Full text

Photo 1.
Adansonia madagascariensis is often found in thalwegs, bordering temporary and permanent water courses and by the sea,
at the edge of tidal zones. In this photo, a mature fruiting baobab by the sea in the bay of Diego-Suarez.
Photo C. Cornu.
Cyrille Cornu1, 2
Wilfried Ramahafaly2
Pascal Danthu3, 2
1Cirad
Umr Tetis
Campus international de Baillarguet
34398 Montpellier Cedex 5
France
2Cirad
DP Forêts et Biodiversité
BP 853, Antananarivo
Madagascar
3Cirad
Ur Bsef
Campus international de Baillarguet
34398 Montpellier Cedex 5
France
Adansonia madagascariensis,
a marine hydrochory hypothesis
BOIS ET FORÊTS DES TROPIQUES, 2014, N° 320 (2) 7
AD ANS ON IA M ADAG AS CA RIE NS I S

RÉSUMÉ
ADANSONIA MADAGASCARIENSIS, UNE
HYPOTHÈSE D’HYDROCHORIE MARINE
Des fruits d’Adansonia madagascariensis
ont été trouvés en mai 2011 sur la plage
d’Anjiabe au Nord de Madagascar. Les
traces présentes sur les fruits et l’absence
de peuplements de l’espèce sur cette
côte indiquent qu’ils ont séjourné long-
temps en mer. La viabilité des graines
contenues a été évaluée et comparée à
cell e de grai nes collectées sous des
semenciers. Les résultats montrent que
leur potentiel germinatif n’est pas altéré
par un séjour en mer et que leur germina-
tion est possible, voire favorisée dans les
zones tidales. Ces premiers résultats
confirment l’hypothèse d’une hydrocho-
rie maritime pour cette espèce. Ils mon-
trent que les fruits de baobab peuvent
séjourner en mer et conserver des graines
viables pour colonis er de nouveaux
espaces. Ils pourraient ainsi expliquer la
fréquente présence de cette espèce le
long du littoral malgache et celle, bien
que plus rare, sur les côtes mahoraises.
Mots-clés: Adansonia madagascarien-
sis, baobabs, fruits, hydrochorie marine,
graines, germination, biogéographie,
Madagascar.
ABSTRACT
ADANSONIA MADAGASCARIENSIS,
A MARINE HYDROCHORY HYPOTHESIS
Adansonia madagascariensis fruits were
fo und in May 2011 on the bea ch of
Anjiab e in the n orth o f Ma dagasca r.
Ma rks fo und on the frui ts and th e
absence of this species on this coast
indicated that the fruits had been in the
sea for a long time. The viability of seeds
contained in these fruits was assessed
and compared to the viability of seeds
collected from under trees. The results
show that the time spent in the sea did
not affect their germination potential and
that germination is possible, and even
improves, in tidal zones. These initial
results confirm the hypothesis of marine
hydrochory for this species. They show
that baobab fruits can remain in the sea
without affecting the viability of seeds to
colonize new areas. This could explain
the frequent occurrence of this species
along the coast of Madagascar and, more
rarely, on the Mayotte coast.
Keywords: Adansonia madagascariensis,
baoba bs, fr uits, mar ine h ydro chor y,
se eds, ger mination, biogeog raphy,
Madagascar.
RESUMEN
ADANSONIA MADAGASCARIENSIS, UNA
HIPÓTESIS DE HIDROCORIA MARINA
En mayo de 2011 se encontraron frutos de
Adansonia madagascariensis en la playa
de Anjiabe en el norte de Madagascar. Las
marcas presentes en los frutos y la ausen-
cia de rodales de la especie en esta costa
indican que éstos permanecieron mucho
tiempo en el mar. Se evaluó la viabilidad
de las semillas de estos frutos y se com-
paró con la de las semillas recolectadas
en portagranos. Los resultados muestran
que su potencial germinativo no se ve
afectado por el tiempo pasado en el mar y
que su germinación no sólo es posible,
sino que incluso se ve mejorada en las
zonas inte rmareales. Estos primeros
resultados confirman la hipótesis de
hidrocoria marina de esta especie y mues-
tran que los frutos de baobab pueden per-
manecer en el mar manteniendo semillas
viables para colonizar nuevos espacios.
Esto podría explicar la frecuente presencia
de esta especie a lo largo del litoral de
Madagascar y, de forma más escasa, en
las costas de Mayotte.
Palabras clave: Adansonia madagasca-
ri e nsis, baobabs, fr utos, hi drocori a
marina, semillas, germinación, biogeo-
grafía, Madagascar.
C. Cornu, W. Ramahafaly, P. Danthu
8
BOIS ET FORÊTS DES TROPIQUES, 2014, N° 320 (2)
AD ANS ON IA M ADAG AS CA RIE NS I S

Introduction
Madagascar is a biodiversity hotspot (MYERS et al.,
2000). The fauna and flora present on the island is unique
throughout the world and around 83% of its plant and land
vertebrae are endemic (GOODMAN & BENSTEAD, 2005).
This uniqueness can be explained partially by the past
isolation of the island, as well as the ecological diversity of
its natural environments (WILMÉ et al., 2006). Yet this
biodiversity is threatened by human activity, both directly
(deforestation, slash and burn agriculture) and indirectly
(climate change) (GADE, 1996; TANDROSS et al., 2008).
Baobabs feature amongst the emblematic species of
Madagascar (WICKENS & LOWE, 2008; PETIGNAT & JASPER,
2012). Of the nine species currently known in the world
(BAUM, 1995; PETTIGREW et al., 2012), six are endemic to
Madagascar. Three (Adansonia suarezensis, A. perrieri,
A.grandidieri) are on the IUCN Red List of Endangered Species
(2012), whilst the other three, A. madagascariensis, A. za and
A. rubrostipa appear for the moment to be under less threat.
Preservation of this heritage relies on improving knowl-
edge of the biological mechanisms of the species concerned.
And yet questions remain regarding modes of dispersal and
germination. Previous studies have shown that baobab seeds
are adapted to arid climates and to dispersal by zoochory1, and
more specifically, that they have hard seed coats which require
extensive scarification to remove the obstacle of seed coat
inhibition for germination to occur (BAUM, 1995; ANDRI-
ANTSARALAZA et al., 2010; RAZANAMEHARIZAKA et al., 2006).
In historic times, seed dispersal was probably performed by
large vertebrates such as giant tortoises (Dipsochelys sp.)
(ANDRIANTSARALAZA et al., 2013), or even elephant birds
(Aepyornis sp.), extinct in Madagascar for many centuries
(GRANDIDIER, 1905; BAUM, 1996; ANDRIANTSARALAZA et al.,
2010; PEDRONO et al., 2013).
Whilst this syndrome of dispersal by zoochory applies to
A. madagascariensis, a number of observations relating to this
species indicate a possible dispersal by hydrochory, dispersal
by water. The species, present across a band covering the
north-west of the island (figure 1), is in fact often present in the
thalwegs and bordering temporary or permanent water
courses. It is equally often found close to the sea, at the edge
of tidal zones (BAUM, 1995) (photo 1). In the same way as
A. suarezensis and A. rubrostipa, A. madagascariensis can be
found along the coast where sea water flooding occasionally
occurs (BAUM, 1996). Hydrochorous dispersal along rivers or
during flooding in the rainy season is a likely occurrence in
species with a hard pericarp (A. digitata, A. madagascariensis,
A. za and A. perrieri) (BAUM, 1996). And finally, A. madagas-
cariensis is also not completely endemic to Madagascar as
some individual specimens, which appear not to have been
transported by man, can be found along the shores of Mayotte
(figure 1) (CHARPENTIER, 2006) at Dapani (photo 2a) and
Mliha (photo 2b).
The aim of this article is to provide some elements
towards testing the hypothesis of a marine hydrochory for
A. madagascariensis, attempting to distinguish factors
associated with fruit dispersal, seed physiology and the
environmental context of seeds washed up on the shore.
■
●
●
Anjiamangirana
Ambatonjanahary
Anjiabe
Comoro
Islands Mayotte
0 100 20050 Km
●
Collection site of ground fruits (GF)
■
Collection site of floating fruits (FF)
Adansonia madagascariensis
Main sea currents
Mozambique
Channel
Madagascar
BOIS ET FORÊTS DES TROPIQUES, 2014, N° 320 (2) 9
AD ANS ON IA M ADAG AS CA RIE NS I S
Figure 1.
Area of presence of A. madagascariensis, collection site
of fruits and surface circulation scheme of water in the
Mozambique Channel from DONGUY and PITON (1991).
Photo 2.
Adansonia madagascariensis on the beaches at Mliha (a)
and Dapani (b) in Mayotte.
Photos J.-P. Lumaret & P. Danthu.
1The dispersal of seeds, spores, or fruit by animals.

Materials and methods
Our study initially assessed the germination capacity of
seeds from fruits that had spent a long time at sea. These
fruits are referred to hereafter as “floating fruits” or FF. The
results were then compared with the germination capacity of
seeds obtained from fruits collected under trees in natural
populations (“ground fruits” or GF). The study subsequently
assesses the influence of the environmental context and, in
particular, of the salt content of substrate on the germination
of seeds obtained from fruit washed-up on the shore.
Twenty-nine fruits were picked off the shore in June
2011 from the beach at Anjiabe to the northeast of the town
of Diego-Suarez (figure 1). The fruits were found on the
fo reshore am ong st a lar ge amoun t o f other detr itus
(driftwood, shells, fruit, plastic waste), indicating that
they’d been washed- up during a particularly low-tide.
Biometric analysis of the fruits and seeds was carried out to
identify the species to which the fruits belong.
Similarly, thirty fruits were collected under A. mada-
gascariensis in Anjiamangirana and Ambatonjanahary, two
geographically distant populations with contrasting ecologi-
cal contexts (figure 1).
The fruits were measured (weight, volume, length).
Their external appearance (wear and tear, cracks) and their
internal condition (state of seeds, fibres and pulp) (photo 4)
has been described. The seeds were then extracted and
counted. They were separated from the surrounding pulp,
cleaned, measured (photo 5) and weighed. Their viability
was visually assessed and mechanically tested using finger
pressure. Damaged seeds were discarded. The intact ones
were stored in a dry place at an ambient temperature
between 15-25°C until they were used in November 2012.
Germination tests were carried out according to the
protocol described by RAZANAMEHARIZAKA et al. (2006), in
plastic germination boxes (17cm x 11cm x 5cm), in the dark,
at a temperature of 30 ± 2°C, in a closed atmosphere, on a
bed of sand, sterilised and moistened with distilled water.
Aseed was considered to have germinated when the radicle
had emerged from the seed coat. On completion of the
experiment, th e viability of ungerminated seeds was
assessed by dissection.
Two distinct tests were performed. The aim of the first
was to evaluate the germination capacity, as defined by CÔME
(1968), of seeds that had spent time in the sea compared with
seeds from the fruit collected under the trees. Germination
capacity is the percentage of seeds able to germinate over a
given period under defined conditions. This involved scarify-
ing all the seeds with a treatment of concentrated sulphuric
acid (95%) for six hours according to RAZANAMEHARIZAKA et
al. (2006) to remove their seed coat inhibition. Four repeti-
tions were carried out using ten seeds extracted from six fruits
(three each from the sea and the ground).
The aim of the second experiment was to assess the
influence of environmental factors, in particular, the salinity
of the germination substrate, on the germination capacity of
seeds. Three germination substrates were collected in situ:
(i) from the beach at Anjiabe; (ii) bordering the tidal zone
(upper tidal limit); (iii) twenty-five metres beyond that limit.
Photo 3.
Fruit washed up on the beach at Anjiabe.
Photo C. Cornu.
Photo 4.
Fruits and seeds from Adansonia madagascariensis:
(a) outer aspect of ground fruit; (b) outer aspect of floating
fruit; (c) inner aspect of ground fruit; (d) inner aspect of
floating fruit.
Photos C. Cornu.
10
BOIS ET FORÊTS DES TROPIQUES, 2014, N° 320 (2)
AD ANS ON IA M ADAG AS CA RIE NS I S

The control substrate consisted of river sand (collected from
the banks of the Ikopa at Antananarivo). The sodium chlo-
ride content (expressed in equivalent g-NaCl per kg soil) of
the four germination substrates was measured by electrical
conductivity. It was 6.40g/kg for the beach sand, 3.43 for
the sample collected on the upper limit of the tidal zone,
0.63 for the sample collected 25 metres beyond and 0.13
for the control sand sample. Six fruits were used for this
experiment, three FF and three GF. The germination capacity,
as defined by CÔME (1968), of forty seeds from each fruit
was tested on each type of substrate. Germination rates
were checked daily.
The two experiments were considered completed
when no new germinations were recorded for three consec-
utive days. The experiments were completed in twelve days.
The confidence intervals for each mean were calculated
at a threshold of 5%. Student’s t-testswere performed to com-
pare paired means (at a threshold of 5%). A series of ␹² tests
at a threshold of 5% were carried out on the proportions to
test the independence of the different classification criteria.
Results
Species identification by biometric analysis of fruits
In the absence of formal taxonomic identification data
for the floating fruits, a comparative biometric analysis of
the fruits and seeds from the FF was undertaken. Table I
shows that the biometric values, length and width of fruits,
size and weight of seeds collected from Anjiabe were signif-
icantly different to those known for the two other baobab
species in the north of Madagascar (A. suarezensis and
A. perrieri). However, they are clearly roughly the same size
as those reported in the litera ture ( BAUM, 1 995;
RAZANAMEHARIZAKA et al., 2006), as well as those meas-
ured from the ground fruits of A. madagascariensis collected
at Ambatonjanahary and Anjiamangirana (CT). Therefore, it
can be concluded that the fruits collected on the beach at
Anjiabe (FF) belong to the A. madagascariensis species.
Effects on the fruits of time spent in the sea..
The outer part of the epicarp of the baobab fruit has a
velvety layer (photo 4a). Whilst this layer was absent from
27% of the FF, in the others it was still scantily present and
had deteriorated severely (photo 4b). This damage to the
epicarp of the fruits is one of the directly noticeable conse-
quences of their prolonged period at sea.
Another noticeable effect was the presence of cracks in
80% of the fruits with only 20% remaining intact. However,
on opening, all the fruits were devoid of pulp, with a very
small amount of fibre and the frequent presence of sand,
indicating that sea water had entered the fruits (photo 4d).
The average volume of fruits ranged from 0.42 ± 0.03g/cm3
at Anjiamangirana and only 0.27 ± 0.02 at Anjiabe, confirm-
ing a loss of volume of FF in relation to GF. The sea water
which had entered the FF had most likely dissolved all or
part of the pulp and destroyed a large majority of the fibres.
Photo 5.
Reniform seed from Adansonia madagascariensis.
Photo C. Cornu.
BOIS ET FORÊTS DES TROPIQUES, 2014, N° 320 (2) 11
AD ANS ON IA M ADAG AS CA RIE NS I S
Table I.
Comparative biometric values of floating fruits and seeds collected from the beach at Anjiabe with the biometric reference
data for Adansonia suarezensis, A. perrieri and A. madagascariensis from BAUM (1995) for fruits (1in the table) and from
RAZANAMEHARIZAKA et al. (2006) for seeds (2 in the table), and the measurements collated from the two control lots
sampled at Ambatonjanahary and Anjiamangirana.
Fruits Seeds
Length (cm) Width (cm) Length (mm) Volume (mg)
Floating fruits Anjiabe 6.7 ± 0.7 7.5 ± 0.6 11.4 ± 0.3 410 ± 30
Adansonia suarezensis 20 - 4018 - 14116.5 ± 0.521717 ± 482
Adansonia perrieri 15 - 2518 – 13112.9 ± 0.22710 ± 62
Adansonia madagascariensis ≤ 101-110.9 ± 0.42366 ± 92
Adansonia madagascariensis Ambatonjanahary 7.9 ± 0.9 6.8 ± 0.4 11.5 ± 0.2 350 ± 30
Adansonia madagascariensis Anjiamangirana 6.2 ± 0.4 7.9 ± 0.4 9.5 ± 0.2 210 ± 20

The average number of apparently intact seeds in the
FF was far less than in the GF: 32 ± 13 at Anjiabe, compared
to 102 ± 17 at Anjiamangirana. More than 40% of the seeds
from the GF collected at Ambatonjanahary were infected
with weevils, whereas none of the apparently intact seeds
from FF showed any signs of infection by predators.
Germination capacity of seeds
Germination capacity of the FF seeds was not signifi-
cantly different from that of the GF seeds, 72% to 62%
(t=1,424 < t0.975=2,776) (figure 2a) (photo 6). Figure 2b
shows that the germination rate of seeds from FF was greater
than that of the GF seeds: on the fourth day the percentage
of germinated seeds (relative to the total number of seeds
germinated at the end of the experiment) was 67% for the FF
and only 39% for the GF (significant difference: t=3,180 >
t0.975=2,776). This would indicate that contact with sea
water enables more rapid germination of seeds of A. mada-
gascariensis. This observation reflects the results of various
authors who have shown that subjecting seeds to osmotic
treatment, known as priming, can encourage and ultimately
accelerate their germination (HEYDECKER et al., 1973; HEY-
DECKER & GIBBINS, 1978). In experimental conditions, prim-
ing can be obtained by soaking seeds in solutions of poly-
ethylene glycol or saline solutions (AFZALL et al., 2008), in
limited concentrations, for example, 100mM for Prosopis
juliflora (NASR et al., 2012). In the floating fruits, this
increase in osmotic pressure could have two origins: solubil-
isation of the sugar rich pulp (OSMAN, 2004) or penetration
of sea water into the fruits. However, it is not possible from
the results of this study to make such hypotheses.
Effect on germination of the site where fruits wash up
The germination substrate has a strong influence on the
germination capacity of seeds. Figure 3 shows that very salty
substrates (beach sand and tidal zone limit) are unfavourable
to the survival of seeds of A. madagascariensis, whether FF or
GF: 12 days after sowing no germination had been recorded
and a large proportion of seeds were necrosed, significantly
more for seeds of FF than GF (␹²obs = 9.77 ≥ ␹²0.95 = 3.84).
The only germinations recorded were from sowings
made in soil collected 25m beyond the tidal zone limit and
the control substrate, with two observations:
 seeds from FF germinated significantly better than those
from GF seeds (␹²obs = 18.24 ≥ ␹²0.95 = 3.84).
 seeds from FF germinated significantly better in the soil
collected 25m beyond the tidal zone limit with a light salt
content (0,63g/kg equivalent NaCl) than in that of the
control substrate (␹²obs = 11.32 ≥ ␹²0.95 = 3.84).
It would appear that seeds from FF germinate better than
seeds of GF. These observations support the hypothesis that a
priming effect which encourages germination of seeds devel-
ops during the period of time the pods spend drifting at sea. It
would also seem that a slightly saline substrate is favourable
to the germination of A. madagascariensis. These conditions
seem to come together at the strandline, just above the high
tide zone, yet are subject to marine influences.
Photo 6.
Seedlings of A. madagascariensis from seeds
of washed up fruits.
Photo C. Cornu.
Figure 2.
Germination power of A. madagascariensis seeds from
floating fruits (FF) and ground fruits (GF): a) cumulative
germination rate relative to number of seeds sown;
(b) germination dynamic: cumulative germination rate
relative to number of seeds germinated at end of
experiment.
0
20
40
60
80
1 2 3 4 5 6 7 8 9 10 11 12
Germination (%)
Number of days
FF
GF
0
25
50
75
100
1 2 3 4 5 6 7 8 9 10 11 12
Germination (%)
Number of days
FF
GF
12
BOIS ET FORÊTS DES TROPIQUES, 2014, N° 320 (2)
AD ANS ON IA M ADAG AS CA RIE NS I S
a
b

Discussion
Hydrochory has been studied in tropical forest species but
in general only taking into consideration the floatability of seeds
and their subsequent germination (LOPEZ, 2001), whereas the
present analysis concerning Adansonia madagascariensis
relates to the dispersal of the fruits. The present observations in
situ can nonetheless not exclude that the seeds, whose floata-
bility has been verified, might be diaspores.
The results of the present experiments show that
marine dispersal is possible for A. madagascariensis. The
geographic coastal distribution of the species (photo 7, fig-
ure 1), its presence along water courses indicates, at least
partially, that it is dependant on water. The floatability
potential of its fruits associated with a low volume, its
shape adapted for barochory2and hydrochory, the salt-
resistance of its seed coat associated with the aptitude of
its seeds to germinate in a slightly saline soil are all strong
indicators that the fruits and seeds of A. madagascariensis
are adapted to marine hydrochory. The results obtained in
this study limit this hypothesis. They show that the fruits of
A. madagascariensis are apt at floating and therefore capa-
ble of ensuring the transport of seeds over long distances.
During this drifting phase, the environment of the seeds
within the fruits evolves: entry of sea water, solubilisation of
pulp. These changes can have positive effects on the viabil-
ity of the seeds: elimination of parasites and predators like
weevils, elimination of parasite-infected seeds, improve-
ment of germination by priming effect due to the increase in
osmotic pressure within the pod. However, and as pointed
out by SCARANO et al. (2003) with Carapa guianensis, it is
very likely that an extended period of time immersed in sea
water might prove harmful to seeds, as prolonged contact
with salt could become toxic.
Nonetheless, the presence of a few species A. mada-
gascariensis on the beaches of Mayotte at Dapani and
Mliha (CHARPENTIER, 2006) would indicate that the seeds
may be able to survive long enough for the dominant cur-
rents in the north part of the Mozambique Channel (figure 1)
(DONGUY & PITON, 1991; SÆTRE & DA SILVA, 1984) to
ensure the transport of the fruits and viable seeds to the
shores of the Comoro Islands. This observation supports the
conclusions of PASCAL et al. (2001) who showed that the
flora of Mayotte most likely resulted from several successive
waves of migration either from Africa, or from Madagascar,
and that various introductions are recent, particularly those
coming from Africa. From the results of the present study, it
is possible to supplement the list of recently introduced
species to Mayotte established by PASCAL et al. (2001),
with the addition of A. madagascariensis.
Whilst current estimates of the phylogeny of Adansonia
by BAUM (2003), based on DNA sequence data and morphol-
ogy, indicate that African, Australian and Malagasy baobab
groups have derived from a common ancestor, the dates are
nonetheless approximative. This evolution occurred between
9.4 and 10.5 million years ago, thus eliminating the hypothe-
sis of a Gondwanian origin of baobabs, whose recent emer-
gence occurred 58 million years ago (BAUM, 2003). This
recent evolution militates for a marine hydrochory.
BOIS ET FORÊTS DES TROPIQUES, 2014, N° 320 (2) 13
AD ANS ON IA M ADAG AS CA RIE NS I S
Photo 7
Adansonia madagascariensis on the shoreline at Ankify.
Photo C. Cornu.
Figure 3.
Influence of substrate salinity on the germination capacity of
seeds from A. madagascariensis. Condition of seeds 12 days
after sowing on different substrates of varying salt contents,
sampled from river (control), 25 metres from the tidal zone,
at the tidal zone limit and on the beach: (a) ground fruits;
(b) floating fruits.
0%
25%
50%
75%
100%
0,13 0,63 3,43 6,4 Salt content
Necrosed
Ungerminated, viable
Germinated
0%
25%
50%
75%
100%
0,13 0,63 3,43 6,4 Salt content
Necrosed
Ungerminated, viable
Germinated
a
b
2The dispersal of seeds, spores, or fruits by gravity.

Conclusion
This study shows that time spent at sea by fruits of
A.madagascariensis does not affe ct the germination
potential of their seeds and that germination is possible,
and even enhanced, in tidal zones. These results confirm
the hypothesis of a marine hydrochory for this species. They
show tha t baobab fruits can remain in th e sea whilst
maintaining viable seeds capable of colonizing new areas.
Thus explaining the frequent occurrence of this species
along the coast of Madagascar and, less frequently, on the
Mayotte coast. It would be interesting to study the marine
hydrochory potential of other species of genus Adansonia,
attempting to provide some understanding of the very
specific current geography of this genus across the world.
The presence of a native species in Australia, A. gregorii, far
from the eight other species may also be explained by an
ancient marine hydrochory.
Acknowledgements
This study was carried out as part of the ECOBAO project
financed by FSP PARRUR. The authors wish to thank Lucien
Rasonaivoson, Roméo Randriamalala, Daniel Verhaegen
and Voninavoko Rahajanirina for their participation in this
study. Soil analyses (conductometry) were carried out by LRI
(Laboratoire des Radio-Isotopes) at Antananarivo.
Bibliographical references
AFZALL I., RAUF S., BASRAL S. M. A., MURTAZA G., 2008.
Halopriming improves vigor, metabolism of reserves and
ionic contents in wheat seedlings under salt stress. Plant
Soil and Environment, 54: 382-388.
ANDRIANTSARALAZA S., PEDRONO M., TASSIN J., EDMOND R.,
DANTHU P., 2010. Baobabs de Madagascar : un anachronisme
de la dispersion ? Bois et Forêts des Tropiques, 306: 7-15.
ANDRIANTSARALAZA S., PEDRONO M., TASSIN J., ROGER E.,
RAKOUTH B., DANTHU P., 2013. The role of extinct giant tor-
tois e in the ge rmination of extant baobab Adansonia
rubrostipa seeds in Madagascar. African Journal of Ecology,
DOI:10.1111/aje.12101.
BAUM D. A., 1995. A systematic revision of Adansonia
(Bombacaceae). Annals of the Missouri Botanical Garden,
82: 440-470.
BAUM D. A., 1996. Ecology and Conservation of the Baob-
abs. Primate Report, 46-1: 311-327.
BAUM, D. A., 2003. Bombacaceae, Adansonia, Baobab,
Bozy, Fony, Renala, Ringy, Za. In The Natural History of
Madagascar. Eds S.M. Goodman & J.P. Benstead. The Uni-
versity of Chicago Press: 339-342.
CHARPENTIER M., 2006. Le baobab, cet arbre étrange et
monstrueux. Univers maoré, 5: 28-31.
CÔME D., 1968. Problèmes de terminologie posés par la
germination et ses obstacles. Bulletin de la Société fran-
çaise de Physiologie végétale, 14: 3-9.
DONGUY J., PITON B., 1991. The Mozambique Channel revis-
ited. Oceanologica Acta, 14: 549-558.
GADE D. W., 1996. Deforestation and its effects in highland
Madagascar. Mountain Research and Development, 16:
101-116.
GOODMAN S. M., BENSTEAD J. P., 2005. Updated estimates of
biotic diversity and endemism for Madagascar. Oryx, 39: 73-77.
GRANDIDIER G., 1905. Les animaux disparus de Madagascar
– gisements, époques et causes de leur disparition. Revue
de Madagascar, 8: 112-128.
HEYDECKER W., HIGGINS J., GULLIVER R. L., 1973. Accelerated
germination by osmotic seed treatment. Nature, 246: 42-44.
HEYDECKER W., GIBBINS M., 1978. The priming of seeds.
Acta Horticulturae, 83: 213-224.
IUCN, 2012. IUCN Red List of Threatened Species. Version
2012.2. [On line] www.iucnredlist.org Consulted 2013 March
the 30th.
LOPEZ O. R., 2001. Seed flotation and postflooding germina-
tion in tropical terra firme and seasonally flooded forest
species. Functional Ecology, 15: 763-771.
MYERS N., MITTERMELER R. A., MITTERMELER C. G., DA FON-
SECA G. A. B., KENT J., 2000. Biodiversity hotspots for con-
servation priorities. Nature, 403: 853-858.
NASR S. M. H., PARSAKHOO A., NAGHAVI H., KOOHI S. K. S.,
2012. Effect of salt stress on germination and seedling
growth of Prosopis juliflora (Sw.). New Forests, 43: 45-55.
OSMAN M. A., 2004. Chemical and Nutrient Analysis of
Baobab (Adansonia digitata) Fruit and Seed Protein Solubil-
ity. Plant Foods for Human Nutrition, 59: 29-33.
PASCAL O., LABAT J. N., PIGNAL M., SOUMILLE O., 2001.
Diversité, affinités phytogéographiques et origines présu-
mées de la flore de Mayotte (Archipel des Comores). Sys-
tematics and Geography of Plants, 71: 1101-1123.
PEDRONO M., GRIFFITHS O. L., CLAUSEN A., SMITH L. L.,
WILME L., BURNEY D. A., 2013. Using a surviving lineage of
Madagascar’s vanished megafauna for ecological restora-
tion. Biological Conservation, 159: 501-506.
PETIGNAT A. U., JASPER L. D., 2012. Baobabs de Madagas-
car : Les arbres à l’envers. Madagascar, Arboretum d’Antso-
kay, Carambole, 95 p.
PETTIGREW J. D., BELL K. L., BHAGWANDIN A., GRINAN E., JIL-
LANI N., MEYER J., WABUYELE E., VICKERS C. E., 2012. Mor-
phology, ploidy and molecular phylogenetics reveal a new
diploid species from Africa in the baobab genus Adansonia
(Malvaceae: Bombacoideae). Taxon, 61 (6): 1240-1250.
RAZANAMEHARIZAKA J., GROUZIS M., RAVELOMANANA D.,
DANTHU P., 2006. Seed storage, behavior and seed germi-
nation in A frican an d Malagasy baoba bs ( Adansonia
species). Seed Science Research, 16: 83-88.
SÆTRE R., DA SILVA A. J., 1984. The circulation of the
Mozambique Channel. Deep-Sea Research, 34 (5): 485-508.
SCARANO F. R., PEREIRA T. S., RÔÇAS G., 2003. Seed germi-
nation during floatation and seedling growth of Carapa
guianensis, a tree from flood-prone forests of the Amazon.
Plant Ecology, 168: 291-296.
TANDROSS M., RANDRIAMAROLAZA L., RABEFITIA Z., KI YIP
Z., 2008. Climate change in Madagascar; recent past and
future. World Bank, 18 p.
WICKENS G., LOWE P., 2008. The baobabs: pachycauls of
Africa, Madagascar and Australia. Kew, United Kingdom,
Royal Botanic Gardens, 498 p.
WILMÉ L., GOODMAN S. M., GANZHORN J. U., 2006. Biogeo-
graphic Evolution of Madagascar’s Microendemic Biota. Sci-
ence, 312: 1063-1065.
14
BOIS ET FORÊTS DES TROPIQUES, 2014, N° 320 (2)
AD ANS ON IA M ADAG AS CA RIE NS I S