Skip Navigation


AOBPreview originally published online on August 31, 2006
Annals of Botany 2006 98(5):1061-1072; doi:10.1093/aob/mcl190
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
98/5/1061    most recent
mcl190v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by LHUILLIER, E.
Right arrow Articles by BOUVET, J.-M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by LHUILLIER, E.
Right arrow Articles by BOUVET, J.-M.
Agricola
Right arrow Articles by LHUILLIER, E.
Right arrow Articles by BOUVET, J.-M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?


© The Author 2006. Published by Oxford University Press on behalf of the Annals of Botany Company. All rights reserved. For Permissions, please email: journals.permissions@oupjournals.org

Extensive Clonality and Strong Differentiation in the Insular Pacific Tree Santalum insulare: Implications for its Conservation

EMELINE LHUILLIER1, JEAN-FRANÇOIS BUTAUD2 and JEAN-MARC BOUVET1,*

1 CIRAD, Forestry Department, Research Unit 39 ‘Genetic Diversity and Breeding of Forest Tree Species’ Campus International de Baillarguet TA 10/C, 34398 Montpellier Cedex 5, France
2 Laboratory of Natural Products Chemistry, University of French Polynesia BP6570-98702 Faa'a, Tahiti, French Polynesia

* For correspondence. E-mail jean-marc.bouvet{at}cirad.fr

Received: 13 April 2006    Returned for revision: 8 June 2006    Accepted: 24 July 2006    Published electronically: 31 August 2006


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 

• Background and Aims The impact of evolutionary forces on insular systems is particularly exacerbated by the remoteness of islands, strong founder effects, small population size and the influence of biotic and abiotic factors. Patterns of molecular diversity were analysed in an island system with Santalum insulare, a sandalwood species endemic to eastern Polynesia. The aims were to evaluate clonality and to study the genetic diversity and structure of this species, in order to understand the evolutionary process and to define a conservation strategy.

• Methods Eight nuclear microsatellites were used to investigate clonality, genetic variation and structure of the French Polynesian sandalwood populations found on ten islands distributed over three archipelagos.

• Key Results It was found that 58 % of the 384 trees analysed were clones. The real size of the populations is thus dramatically reduced, with sometimes only one genet producing ramets by root suckering. The diversity parameters were low for islands (nA = 1·5–5·0; HE = 0·28–0·49). No departure from Hardy–Weinberg proportion was observed except within Tahiti island, where a significant excess of homozygotes was noted in the highland population. Genetic structure was characterized by high levels of differentiation between archipelagos (27 % of the total variation) and islands (FST = 0·50). The neighbour-joining tree did not discriminate the three archipelagos but separated the Society archipelago from the other two.

• Conclusions This study shows that clonality is a frequent phenomenon in S. insulare. The genetic diversity within populations is lower than the values assessed in species distributed on the mainland, as a consequence of insularity. But this can also be explained by the overexploitation of sandalwood. The differentiation between archipelagos and islands within archipelagos is very high because of the limited gene flow due to oceanic barriers. Delineation of evolutionary significant units and principles for population management are proposed based on this molecular analysis.

Key words: Clonality, conservation, endangered species, genetic diversity, genetic structure, insularity, nuclear microsatellites, Santalum insulare, French Polynesia


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 
Oceanic islands have long been considered as remarkable ecosystems, in particular since Darwin's theory of evolution led biologists to recognize them as ‘evolutionary laboratories’ (Darwin, 1859). Islands have several features that make them attractive for evolution studies: discrete entities; small areas but varied habitats; isolation from the mainland and other islands within archipelagos; and they are often geologically dynamic (Emerson, 2002). These factors characterize the insularity and are typically associated with a foundation bottleneck and small population sizes. The small population-size phenomenon particularly favours inbreeding depression and genetic drift which result in less genetic diversity and reproductive fitness, increasing the probability of extinction (Frankham, 1998; Frankham et al., 2002). The remoteness of insular systems and oceanic barriers also result in substantial differentiation between island populations potentially leading to speciation (Barton, 1989) and/or to endemism (Emerson, 2002). Often considered as ‘hot-spots of biodiversity’ (Myers, 2000), island ecosystems are vulnerable because of insularity combined with the negative impact of human activities, such as overexploitation, habitat fragmentation, pollution and induction of biological invasions. The latter could generate marked disequilibrium because of the lack of predators, parasites and diseases in island environments (Blondel, 1995). Invading species often have direct negative impacts on insular endemic species, but they can also create competition, alter the structure of trophic levels, modify disturbance regimes, and exacerbate the effects of fragmentation, finally disrupting whole ecosystem processes and structure (Vitousek et al., 1997; Mack and D'Antonio, 1998; Parker et al., 1999).

Although island ecosystems raise numerous scientific issues, there are a limited number of studies addressing the impact of evolutionary forces on within-species genetic diversity, especially for forest tree species which represent a significant component of tropical island ecosystems. Sandalwood species distributed on small islands of the Pacific constitute an excellent model for studying pattern of diversity in insular systems structured into archipelagos. This paper presents a genetic analysis of Santalum insulare, which is distributed on most of the archipelagos of Eastern Polynesia in the South Pacific and which is today considered as endangered (IUCN, 2004).

Santalum insulare is a root-hemiparasitic, medium-sized tree (15 m tall, 30 cm diameter at a height of 1·3 m) growing in a wide variety of habitats from 0 to 2200 m a.s.l., on very diverse host species characterizing each habitat (Butaud, 2004). Hermaphroditic flowers are pollinated by insects (wasps and bees), cross-pollination being the rule, but some self-pollination has been reported on isolated trees (Butaud et al., 2005), classifying S. insulare as an ‘often cross-pollinated species’ (Kulkarni and Muniyamma, 1998). Fruits are red drupes which are normally swallowed and then disseminated by birds.

Populations of S. insulare have been strongly affected by human colonization of the small islands and their future is now threatened. During the past two centuries, the overexploitation of Polynesian sandalwood by Europeans (initially for the Chinese incense market and later for the market of essential oil extracted from its fragrant heartwood; Butaud, 2004) dramatically reduced and fragmented the populations. This decline was also heightened by the invasion of alien species that arrived with humans: rats feeding on seeds; fruit and flower abortion due to insects; goats and other herbivora eating young shoots or seedlings; and plants modifying the understorey and choking mother trees (Butaud and Tetuanui, 2005). Moreover, because of human hunting activities and predation by rats, frugivorous bird populations have declined and seed dispersal is now very limited. Santalum insulare has persisted in Polynesian archipelagos thanks to its ability to resprout from root suckers or stumps of poached trees. Thus, it is expected to develop marked clonality, as observed in some threatened populations of S. lanceolatum (Warburton et al., 2000). As illustrated by the presence on the IUCN red list of six of the nine varieties of S. insulare (IUCN, 2004), there is now an urgent need to implement a strategy of conservation for this species.

The first study of S. insulare by Butaud et al. (2005) pointed to a strong genetic structure of the species using plastid markers. The present study used nuclear microsatellites specifically developed for S. insulare in order to complement the previous ones. Using these new molecular markers the objectives of this study were (a) to assess the level of clonality within the populations, (b) to analyse the distribution of microsatellite diversity within the natural range, and (c) to define management units based on the molecular data according to the Moritz approach (Moritz, 1994). The ultimate aim was to propose adapted and feasible strategies for conservation of S. insulare, which is a significant natural resource on the small, isolated islands of French Polynesia.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 
Sampling
Santalum insulare Bertero ex A.DC. (Santalaceae) is endemic to Eastern Polynesia and is divided into nine botanical varieties found on the Cook Islands (Mitiaro Island), French Polynesia (ten islands) (Table 1) and the Pitcairn Islands (Henderson Island) (Fosberg and Sachet, 1985). These botanical varieties are based on a limited number of morphological traits and are contested today. Some of them are now considered as ‘ecotypes’ by Polynesian botanists (J. Florence, IRD, France, pers. comm.) because they are found in specific environments. In French Polynesia, S. insulare is present on the Marquesas, Society and Austral archipelagos, but is absent from the low-lying islands of the Tuamotu-Gambier archipelago (Fig. 1). The leaves were collected from 384 individuals on the ten islands of French Polynesia where S. insulare is found (Table 1 and Fig. 1). On each island, the number of sampling sites varied. Populations were generally small and scattered, except for Nuku Hiva, which has numerous individual trees in a large area. Sampling was very unbalanced, the number of individuals per island varying from three in Fatu Hiva to 77 in Tahiti (Table 1). This imbalance results from different factors, including the original size of the sandalwood population on each island, the impact of 19th century exploitation, the contemporary degradation by fire and overgrazing, and the difficulty of exploring these rugged and remote islands. Where the size of the population was sufficient, trees separated by >10 m were selected to avoid selection of the same clone (assuming that resprouting by root suckers can create patches of the same clone). However, to analyse the clonality in some populations, an exhaustive sampling was conducted. Three leaves were collected from each tree and immediately dried using silica gel.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Description of Santalum insulare populations and sampling

 

Figure 1
View larger version (10K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 1. Distribution of Santalum insulare in French Polynesia.

 
DNA extraction and microsatellite analysis
Total DNA was extracted from 100 mg of dried leaf material using a MATAB method derived from Bousquet et al. (1990), with one additional chloroform-isoamyl alcohol (24 : 1) extraction. The eight nuclear microsatellites used for genetic analyses were specifically designed for Santalum insulare and their characteristics are described in Lhuillier et al. (2006): mSiCIR33, mSiCIR39, mSiCIR42, mSiCIR44, mSiCIR139, mSiCIR148, mSiCIR153 and mSiCIR185.

Data analysis
Detection of clones and clonality evaluation
Asexual reproduction leads to a clonal structure, in which one clone (genet) may consist of several trees (ramets). Individuals with the same multilocus genotype can be members of the same clone or else arise by chance from independent instances of sexual reproduction. The approach developed by Parks and Werth (1993) and the program MLGsim by Stenberg et al. (2003) were used to calculate the probability of the latter. This program first calculates a Psex-value for genotypes that are found more than once, i.e. the probability that these genotypes occur the observed number of times in a sexual population with the observed allele frequencies, and assuming Hardy–Weinberg and linkage equilibria. In a second step, the program uses Monte Carlo simulation (here with at least 10 000 simulations) to obtain the critical values of Psex for the desired significance level (here 0·05), allowing the genotypes that are significantly over-represented and thus probably members of the same clone to be identified (Stenberg et al., 2003). If the Psex-value is smaller than the critical Psex-value, then the hypothesis that the multilocus genotype results from panmictic sexual reproduction is rejected at the 0·05 level. The genotype is then considered as a clone. Once all clones had been identified, parameters of clonal diversity were calculated per sampling site:

G/N = the proportion of distinguishable genotypes, where N is the number of sampled trees and G the number of distinct multilocus genotypes

D = the genotypic diversity (also called the complement of Simpson index, derived from Pielou, 1969), D = 1–{Sigma}ni(ni – 1)/N(N – 1), where ni is the number of samples with multilocus genotype i (Kennington and James, 1997).

Taking into account both the significance of Psex-values and the information supplied by the field observations (geographic position, observation of root suckering system between two trees), an estimate of the number of genets (NGe) was given. Finally, the resolving power of the set of nuclear microsatellites was assessed by examining the relationship between G/N and the number of combined loci according to the method developed by Arnaud-Haond et al. (2005).

Definition of populations within each island
To define populations within islands, the approach taken consisted of the detection and location of genetic discontinuities between populations. It was based on Bayesian theory, which tends to be much used in plant conservation biology (Marin et al., 2003). A Bayesian model implemented in a Markov Chain Monte Carlo scheme which infers the location of genetic discontinuities from individual geo-referenced multilocus genotypes, without a priori knowledge of population units and limits was used (Guillot et al., 2005a). The main assumptions of the method are (a) the number of populations is unknown and all values are considered a priori to be equal, (b) populations are spread over areas given by the union of some polygons of unknown location in the spatial domain, (c) Hardy–Weinberg equilibrium is assumed within each population, and (d) allele frequencies in each population are unknown and treated as random variables following the so-called Dirichlet model.

The computer package GENELAND was used to infer the number of populations and the spatial locations of genetic discontinuities between those populations (Guillot et al., 2005b). Using the Dirichlet model, the first run with 100 000 iterations was made to define the mode of posterior distribution of the number of populations. Using this mode as a fixed value the model was rerun for between 100 000 to 500 000 iterations (depending on the quality of the results) to assess the posterior probability of any pixel of the domain belonging to each population.

To measure the impact of small population sizes on the quality of the assignment, simulated populations comprising 10 or 30 individuals, and structured in three subpopulations characterized by a differentiation parameter FST = 0 and FST = 0·1, were created with GENELAND. This experimental design is representative of the populations in the present study. The correct assignment of the individuals in the three subpopulations was then tested with 100 replications, using the same Dirichlet model.

Estimation of genetic diversity and departure from random mating
One individual tree per clone was used for genetic analysis. Allele frequencies, mean number of alleles per locus (nA), observed heterozygosity (HO) and expected heterozygosity (HE) (Nei, 1978) per island and archipelago were computed with GENETIX 4.03 (Belkhir et al., 2001).

To check if the differences in sample sizes and the various spatial scales over which individuals were pooled into populations affected the diversity estimates, the allelic richness per population and island (containing at least eight genets) was calculated This index takes into account the dependence on sample size with an adaptation of the rarefaction index of Hurlbert (1971) (El Mousadik and Petit, 1996), named ‘R’, using FSTAT version 2.9.3.2 [EC] (Goudet, 1995). The principle is to estimate the expected number of alleles in a subsample of 2n genes, given that 2N genes have been sampled (N > n). In FSTAT, n is fixed as the smallest number of individuals typed for a locus in a sample. A test using the Pearson coefficient was conducted to find a correlation between the diversity parameters and the island size. An estimation by Weir and Cockerham (1984) of the fixation index FIS, assessing the departure from Hardy–Weinberg equilibrium, was calculated using GENETIX 4.03 and was tested using 5000 permutations.

Analyses of population differentiation
To investigate the genetic structure, several analyses of molecular variance (AMOVA) were run (Excoffier et al., 1992) using ARLEQUIN (Schneider et al., 2000) with 1000 permutations. The structure was tested by archipelagos and islands within the species distribution area, the structure due to islands within each archipelago, and the structure by populations within islands when some have been detected by GENELAND. The extent of genetic differentiation was estimated from the theta estimator of FST (Weir and Cockerham, 1984). The pairwise FST were also calculated among the populations and the P-values were corrected using the sequential Bonferroni procedure (Rice, 1989).

To relate the dispersal ability of Santalum insulare to its geographical distribution, a Mantel test (Mantel, 1967) was conducted in GENETIX 4·03. The procedure assesses the significance of the correlation between pairwise FST/(1 – FST) estimates and the logarithm of the Euclidian distance (in kilometres) between pairs of islands (Rousset, 1997), with the Spearman rank coefficient as the statistical test, using 1000 random permutations of the matrix. Isolation by distance was also sought within the Marquesas and Society archipelagos.

Pairwise genetic distances between islands were computed using Cavalli-Sforza's chord measure (Cavalli-Sforza and Edwards, 1967), obtained from the GENDIST program (PHYLIP version 3·6; Felsenstein, 1993). This distance matrix was used to construct a neighbour-joining tree (Saitou and Nei, 1987) using the NEIGHBOR program of PHYLIP. The robustness of each node was evaluated by bootstrapping data over loci for 1000 replications using the SEQBOOT program of PHYLIP. The consensus tree obtained by the CONSENSE program of PHYLIP was displayed with TREEVIEW software (Page, 1996).

Detection of bottlenecks
To detect the occurrence of a recent bottleneck event, the heterozygosity excess method described by Cornuet and Luikart (1996) and implemented in the BOTTLENECK software was employed (Piry et al., 1999). The proposed tests are based upon the fact that populations that have experienced a recent reduction in effective size should exhibit a more rapid reduction of allelic diversity than heterozygosity at polymorphic loci. Hence, in a recently bottlenecked population, gene diversity is higher than the equilibrium heterozygosity estimated from the observed allele numbers assuming mutation drift equilibrium (Luikart and Cornuet, 1998). Three tests are available in BOTTLENECK: the standardized differences test requires at least 20 loci, so only the sign test and Wilcoxon's signed-rank test were run. As recommended by Piry et al. (1999), in the case of microsatellite markers, the model used was a two-phase mutation model with 95 % stepwise mutation model and 5 % multistep mutations.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 
The extent of clonality
The total sample of 384 trees exhibited only 156 distinct multilocus genotypes which are not shared by different islands. Among the 156 genotypes, 61 were in multiple copies. Among the 61 genotypes, 52 presented a Psex-value smaller than 0·05, suggesting that it is very unlikely that these genotypes were identical by chance. However, cross-checks with field observations indicated that some of these clones could be made up of several genets identical at the loci examined. This was the case when trees were found on motu (little coral islands of Raivavae) >2 km distant, or when trees were several hundred metres distant and separated by vegetation or geographic barriers. Even if the hypothesis of long distance (several kilometres) propagation by root suckering cannot be excluded, it seems very unlikely because several centuries are needed, plus a decrease in sea level in the case of motu populations. In addition, no biological mechanism, such as long-distance vegetative propagule dissemination, is known in S. insulare to explain how these physical barriers could have been crossed. Concerning the nine supposed clones which had a Psex-value higher than 0·05, they were made up of only two elements. Once they were compared with field observations, they appeared as part of a unique clump comprising dozens of ramets, whose connections to the stump are observable. It seems that the sampling of only two trees did not allow the detection of clones for particular multilocus genotypes. Thus, only one genet was counted for each of these genotypes. As a result, 162 genets were considered for the diversity analysis.

Eighty-one per cent of the populations tested were concerned by clonality and 42 % presented one unique multilocus genotype. The clonal diversity parameters are given in Table 2 for sites where sampling was appropriate for the evaluation of clonality (exhaustive or random sampling). It appeared that clonality did not affect equally the different populations of sandalwood. The mean proportion of distinct genotypes (G/N) ranged from 0·06 in Maauu (Nuku Hiva) to 1·00 in Taiohae Vaioa (Nuku Hiva). The mean genotype diversity (D) was nil in several populations on different islands and peaked in Taiohae Vaioa. Low G/N values are generally correlated with low D values, which tends to show that marked clonality reduces diversity within populations. However, some populations, as Rotui or Pic Vert, exhibited high D values and quite low G/N values. The clonality did not affect all the small stands in the sandalwood populations. Taiohae Vaioa, for example, did not seem to have clones. The mean values of G/N and D were calculated by averaging the values estimated in Table 2. Values of the estimated number of genets (NGe) used to define populations and to compute diversity parameters are given in Tables 3 and 4.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Measures of clonal diversity in some sampling sites of Santalum insulare

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Detection of populations within each island using the Bayesian model (Guillot et al., 2005a)

 

View this table:
[in this window]
[in a new window]

  		TABLE 4. Genetic diversity estimates obtained with eight nuclear microsatellites for Santalum insulare

 
Detection of populations within islands
The spatial statistical model used to define genetic discontinuities was applied to islands with more than five genets. The number of populations given by the model varied between 1 and 3 (Table 3).

Nuku Hiva and Tahiti exhibited a marked difference in the posterior probability for assigned and non-assigned individuals, showing that two populations could be detected in each island. In Nuku Hiva one population is present on the west coast at low altitude (lowland population) and the other is located on the mountain zone and on the high plateaus of the island (highland population). In Tahiti the lowland population is distributed on the periphery of the island at a medium altitude. The other one is situated on the highest summits and is identified as the highland population.

For the other islands the differences in posterior probabilities for assigned and non-assigned individuals were very low and suggested that each island was composed of a single population. The correct assignment of the individuals in the three subpopulations tested with 100 replications, using the same Dirichlet model, showed that, on average, 92 % (65 %) of the 30 individuals were correctly assigned, versus 77 % (63 %) of the ten individuals for FST = 0 and FST = 0·1, respectively. Thus, the assignments of individuals in small population such as Hiva Oa or in Tahuata should be considered with caution.

Genetic diversity analysis
The eight microsatellite loci were polymorphic and the number of alleles per locus ranged from three for mSiCIR139 to 15 for mSiCIR148 (Lhuillier et al., 2006). Considering each locus in each population separately, the distribution of allele frequencies was highly unbalanced (results not shown). Twenty-two alleles were private, the largest number being 12 in Tahiti. Only two alleles were shared by all the islands. In Table 4, it was decided to give also diversity parameters for small populations (but containing at least eight genets) despite the lack of accuracy of the estimates (see standard errors).

The Austral archipelago showed the lowest diversity (R = 3·23), whereas the Marquesas and Society archipelagos exhibited more diversity and were quite equivalent (R = 5·72 and 5·69, respectively). The other diversity parameters, more affected by sample size, followed the same pattern.

The rarefaction index ranged from 2·04 in Raivavae to 3·35 in Nuku Hiva. Observed and expected (in parentheses) heterozygosity values ranged from 0·29 (0·28) in Raivavae to 0·48 (0·44) in Hiva Oa, and the differences between islands were lower than the standard error. The correlations between diversity parameters and island areas were not significant at the 5 % level, except for nA with a Pearson coefficient of correlation r = 0·838 (P < 0·05).

In Nuku Hiva, the R diversity index tended to be slightly greater in highlands than in lowlands, whereas in Tahiti the R-values were close. The other diversity parameters did not follow the same trend which can be explained by the sample size effect.

Concerning Hardy–Weinberg equilibrium, no significant heterozygote deficit was detected when considering the elementary populations, except for the highland population of Tahiti (FIS = 0·11). The FIS value was also significantly positive in Tahiti island.

The detection of a recent bottleneck was applied to the population of the islands of Nuku Hiva, Moorea, Tahiti and Raivavae because their population size was higher than 15 which is suitable for running the test. For none of the populations did the Wilcoxon's signed-rank test indicate departure from mutation drift-equilibrium, and recent bottlenecks were not detected. The probability associated with heterozygosity excess was markedly higher than 0·05.

Analysis of population differentiation
All FST values estimated by AMOVA were significant. Twenty-seven per cent of the variation in the species distribution area was explained by the variation among archipelagos and 50 % by the variation among islands (Table 5). Global FST among islands within archipelagos was 0·24 in the Marquesas, 0·34 in Society and 0·66 between Raivavae and Rapa in the Austral archipelago. The AMOVA for Nuku Hiva and Tahiti revealed a significant differentiation between lowlands and highlands. FST was higher in Nuku Hiva (0·22) than in Tahiti (0·02).


View this table:
[in this window]
[in a new window]

  		TABLE 5. Results of AMOVA testing the genetic structure among archipelagos and among islands for Santalum insulare

 
Among archipelagos, pairwise FST values were highly significant (P < 0·001) and were 0·30 between Marquesas and Society, 0·41 between Marquesas and Austral, and 0·49 between Society and Austral. Among all the populations, all but two of the pairwise FST values ranged from 0·19 (between Fatu Hiva and Raiatea) to 0·66 (between Raivavae and Rapa). However, not all of them were significant (after Bonferroni correction), particularly when they were calculated for islands with population sizes that were small (Fatu Hiva, Raiatea, Rapa). Only two pairs of islands exhibited low and non-significant FST values: Tahuata/Hiva Oa (0·07) and Moorea/Raiatea (0·07).

The neighbour-joining tree (Fig. 2) did not clearly separate the three archipelagos. The tree topology is quite consistent with the geographic position of the samples but the bootstrap values were lower than 50 %, except for Society versus all others (bootstrap value of 80 %). The respective highland and lowland populations of Tahiti and Nuku Hiva were found on the same branch with bootstrap values higher than 50 %. Tahuata and Hiva Oa, as well as Moorea and Raiatea, were gathered, as expected, by the low and non-significant FST.


Figure 2
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 2. Neighbour-joining tree based on Cavalli-Sforza chord measures for Santalum insulare. Nuku Hiva and Tahiti are divided into lowlands and highlands. Bootstrap values in percentages are shown at the base of each branch.

 
The Mantel test revealed a significant correlation between geographical and genetic distances considering all the islands (r = 0·58; P = 0·01), but the relationship was not explained by a simple linear model (Fig. 3A). Below 500 km, which is the maximum distance between two islands within an archipelago, except in the Austral, the FST values tended to be lower. Above 500 km, the highest values were reached, although low values were found as well. Within the Society archipelago, no isolation by distance was detected (r = 0·18; P > 0·05), whereas the Mantel test revealed a significant correlation (r = 0·48; P < 0·01) in the Marquesas archipelago. This trend is illustrated in Fig. 3B.


Figure 3
View larger version (13K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIG. 3. Relationship between genetic differences (FST) and geographical distances (A) between all island pairs; (B) between pairs of populations of the Marquesas archipelago (open circles, broken line) and pairs of populations of the Society archipelago (closed circles, continuous line).

 


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 
Clonality, anthropogenic pressures and adaptation strategies
The suspected clonality from root suckering observed during the sampling was confirmed by the results and was more marked than expected. Fifty-eight per cent of sampled trees proved to be ramets with multilocus genotypes identical to those of other trees. Without DNA techniques, the simple observation of root systems gives a very inaccurate estimate of the relationships between individual trees, as some underground links between trees disappear with time. This highlights the effectiveness of this molecular technique in assessing the genetic constitution and spatial distribution of potentially clonal populations. In addition, the combination procedure (Arnaud-Haond et al., 2005) showed that the eight loci studied have sufficient resolving power in the populations (results not shown). Nevertheless, a larger set of microsatellites would be helpful to validate the distinction of genets with the same multilocus genotype but separated by a very long distance.

In terms of diversity parameters, the inclusion of the ramets in preliminary analyses had biased the estimates: particularly strong negative values of FIS resulting from an excess of heterozygotes due to the clonal effect were found (results not shown).

The mean estimates of the proportion of distinguishable genotypes (G/N = 0·35) and genotypic diversity (D = 0·43) for the 20 assessed populations of Santalum insulare revealed a high degree of clonality, as indicated by the comparison with former studies. For example, Smith et al. (2003) found higher values (G/N = 0·53; D = 0·72) for threatened populations of Eucalyptus curtisii using microsatellites. Kennington and James (1997) found lower values for populations of the endangered species E. argutifolia (G/N = 0·20; D = 0·41), but they used allozymes which have lower resolution than microsatellites in detecting distinct genotypes. Nevertheless, comparisons between populations should be considered with caution because D and G/N parameters are correlated with the sampled site area S. In the present study, a significant correlation was found between D and S (r = 0·48; P < 0·05), and between G/N and S (r = 0·49; P < 0·05).

Santalum insulare clonality peaked in some populations: Maauu (Nuku Hiva), Tiapa (Tahiti) and some populations in Hiva Oa, Tahuata, Raiatea and Rapa. They are composed of a single genet (D = 0) and spread only by clonal growth. Warburton et al. (2000) found the same extreme situation in threatened populations of another species of the Santalum genus, S. lanceolatum.

The extensive clonality in this species may be explained by two phenomena. The first is the overexploitation which induced root suckering by cutting the original tree. This hypothesis is supported by the correlation between the marked clonality and the overexploitation history of Polynesian islands. Marquesas sandalwood is known to have been particularly exploited, and populations of Maauu, Vaiteheii (Nuku Hiva) and Hiva Oa and Tahuata islands now appear as a single clump or a few clumps of trees. Populations in Raiatea, Rapa and Tiapa exhibited the same pattern of clonality (low G/N value and nil D value). All these populations share the fact that they were easily reached by humans because of their coastal proximity, whereas ‘highland’ populations like Tops of Terre Déserte or Crests of Toovii (Nuku Hiva) were located on remote and often abrupt ridges and slopes. A higher propensity to clone in populations threatened by human activities has also been reported in other potentially clonal tree species (Kennington and James, 1997; Warburton et al., 2000; Smith et al., 2003).

The second explanation for marked clonality concerns the strategy of the species in responding to harsh environmental conditions and stresses. In remote highland populations or on dry rocky inaccessible slopes, clonality was also present despite a low or nonexistent human impact (see, for example, populations of Rotui in Moorea or Pic Vert and Summit of Pic Vert in Tahiti). At these sites, the relative importance of clonal versus sexual recruitment may be strongly influenced by environmental factors: on steep and windswept ridges with skeletal soils, where rock falls and landslides are frequent, seed germination is unlikely, but wounded sandalwood roots can generate ramets by shoot production. In terms of clonal diversity, these populations are characterized by higher D values (>0·7) with G/N values that can sometimes be low (<0·4).

Whatever the reason, root suckering was likely to be a good means of survival for remnant populations in the absence of regeneration by seeds, most of them being eaten by rats in French Polynesia. As a result of the rat predation, genotypic diversity was not recovered by sexual reproduction, and populations were mainly composed of genets which survived the overexploitation.

Genetic diversity
The global genetic diversity of Santalum insulare given by the expected heterozygosity calculated over the entire distribution area, HE = 0·69, is quite high. But, as a result of the insular distribution of the species, a regional structure of the genetic diversity was found and the level of variability was not equivalent from one archipelago or island to another. For islands, the expected heterozygosity ranged from 0·28 to 0·49 (mean = 0·41) and the mean number of alleles per locus (nA) ranged from 1·50 to 5.

The previous study on Santalum insulare using plastid microsatellites (Butaud et al., 2005) concluded that the genetic diversity of this species was quite high and equivalent to that of other tree species, even though unidentified clones were kept in the analysis. However, the present study tends to moderate these results and shows very low diversity parameters compared with other equivalent studies on tropical tree species that conform to the predictions of Hamrick et al. (1992). For example, although fewer microsatellite loci were used in these studies, the present values are lower than those of Melaleuca alternifolia (nA = 9–19·4; HE = 0·50–0·76) (Rosetto et al., 1999) or Symphonia globulifera (nA = 3·7–16; HE = 0·67–0·85) (Aldrich et al., 1998), two mainland species. They also tended to be lower than those of a threatened mainland species Grevillea macleayana (nA = 3·2–4·2; HE = 0·42–0·53) (England et al., 2002), evaluated with six microsatellites. Finally, Santalum insulare showed lower diversity than Santalum austrocaledonicum (nA = 2–16; HE = 0·14–0·79, mean = 0·49), an insular species endemic to New Caledonia which has also been threatened by human activities (Bottin et al., 2005).

Three factors may explain the low level of diversity in Santalum insulare. First, the considerable remoteness of the Polynesian islands could result in a very marked founder effect. The probability of colonization of a species decreases with increasing distance between the source population and the new area (Blondel, 1995). The resulting small populations led to an increase in the rate of allele loss through genetic drift, resulting in a lower diversity than in the mainland populations. Second, it has to be considered that the more recent overexploitation by Europeans in the 19th century has virtually led to the total depletion of sandalwood resources (Cherrier, 1993), resulting in a second bottleneck. However, the present analyses did not detect recent bottlenecks. This result can be explained in two ways: (1) only eight loci were used, whereas Piry et al. (1999) reported that ten polymorphic loci with a sampling of 30 individuals are recommended to achieve a reasonably high statistical power (>0·80) to detect bottlenecks; (2) the populations are too recently bottlenecked and the quasi absence of generations after overexploitation could prevent the model from giving a significant test (Cornuet and Luikart, 1996). Low population turn-over can be explained by the third factor which is the predation of sandalwood seeds by the ship rat (Rattus rattus), introduced by Europeans. The impact of rats prevents natural regeneration by sexual reproduction and restoration of genotypic diversity.

Within French Polynesia, diversity parameters indicated that Society and Marquesas exhibit the most variability, in agreement with the results obtained using plastid microsatellites (Butaud et al., 2005). The more limited diversity in Austral may be explained by the small area of the islands, their remoteness, which prevents gene exchange, and the presence of only two genets in Rapa. The maximum diversity was present in Tahiti and Nuku Hiva, the two larger islands of Eastern Polynesia. Nevertheless, no significant correlation was detected with the rarefaction index R, which is not affected by sample size or island area.

Hardy–Weinberg equilibrium
Generally, a localized clonal spread may interfere with sexual reproduction by reducing pollen and mating between genets (Handel, 1985, in Eckert, 1999) and favour high levels of between-ramet self-pollination (geitonogamy) which may have a major impact on the evolution of the breeding system (Harder and Barrett, 1996, in Eckert, 1999). Consequently, significant heterozygote deficits, symbolized by positive FIS, were expected in populations of Santalum insulare, particularly in those severely affected by clonality. The lack of observed homozygote excess in most of the populations could be due to the absence of regeneration by sexual reproduction since the introduction 200 years ago of the ship rat, which eats such a high proportion of fruits that no seedlings are seen in the sandalwood stands (Butaud and Tetuanui, 2005).

Only two significant FIS values were found in the case of Tahiti island and its highland population. For the whole island, one explanation for the departure from Hardy–Weinberg proportions may be the Wahlund effect, which occurs when a population is divided into several subpopulations with different allelic frequencies. However, when FIS is due to the Wahlund effect only, it is expected to be more or less equivalent to the FST among pooled subpopulations, depending on the variance due to estimates. Since the FST observed in Tahiti is 0·02 (Table 5), and FIS is 0·12 (Table 4), it does not support this single hypothesis and involves others, such as inbreeding, assortative mating and/or selection (Nei, 1987).

For the highland population, the significant FIS value can be explained by the remoteness of the stand on the tops of high mountains of Tahiti and its very small number of individuals which favours inbreeding (Frankham et al., 2002). Because overexploitation by Europeans was undoubtedly absent in these remote populations, the past inbreeding, which occurred before the introduction of the ship rat, can now be observed via the present analyses.

Strong genetic structure
Within the entire distribution area, the FST value between islands (0·50) appears to be particularly high compared with values of other tropical tree species distributed in mainlands: 0·047 in Vitellaria paradoxa (Sanou et al., 2005), 0·08 in Vouacapoua americana (Dutech et al., 2004), 0·22 in Grevillea macleayana (England et al., 2002), but closer to species present on islands, such as Santalum austrocaledonicum (FST = 0·35) (Bottin et al., 2005).

These results were expected in an island system, where oceanic barriers limit gene flow between islands (and thus archipelagos) (MacArthur and Wilson, 1967), and confirm the results obtained in Santalum insulare using plastid microsatellites (Butaud et al., 2005) and in other species like Santalum austrocaledonicum (Bottin et al., 2005). The large distances between archipelagos do not allow gene flow through pollen which was expected to be limited even within islands, as S. insulare is insect-pollinated. Seed dispersal through bird ingestion can occur over longer distances and may theoretically allow gene flow between islands. However, some Columbidae species that were able to eat and disperse fruits of Santalum insulare in Eastern Polynesia (Ptilinopus and Ducula genera) are now extinct (Steadman, 1997; McConkey and Drake, 2002) and the remnant species are critically endangered (IUCN, 2004). Moreover, given that these species are not migratory birds and that migratory birds themselves can exceptionally disperse seeds (Carlquist, 1980), the probability of seed dispersal between remote islands should be very low.

The extent of differentiation between islands varies according to the archipelago. Although the FST should be considered with caution, as the low population size may have biased the estimations of allelic frequencies (especially for Rapa, Raiatea and Fatu Hiva), it is particularly high between the distant islands of Raivavae and Rapa in Austral (FST = 0·66), whereas it was lower between the closer islands of Marquesas (FST = 0·23) and Society (FST = 0·34). The differentiation between two distant islands, such as Rapa and Raivavae, can be explained by a geological phenomenon characterized by the disappearance of intermediate islands, while they evolve from high volcanic island to coral atoll before being recovered by the ocean. The emergence of this oceanic barrier in the middle of an archipelago results in fragmentation of this area which can be associated with a mechanism of pseudo-vicariance present in some parts of the Pacific (van Welzen et al., 2003). Indeed, the initial area was not continuous but the islands were closer and were sequentially colonized by sandalwood, necessarily implicating a mechanism of dispersion.

For the larger islands (Table 3), a Bayesian approach was used to define populations. Two separate populations were detected in Nuku Hiva and Tahiti. For the former they correspond to an a priori separation based on a geographical location. For the latter the Bayesian distinction of populations was different from the present a priori separation based on the altitude. Both larger islands, Tahiti and Nuku Hiva, displayed a significant differentiation between lowland and highland populations, which was lower in Tahiti (FST = 0·02) than in Nuku Hiva (FST = 0·22). Despite the relatively small area of these islands, the differentiation stresses the limited gene flow in this species due to topographic barriers which may be accentuated by the progressive extinction of frugivorous birds, limited pollen dispersion by insects, and loss of forest areas due to fire and overgrazing.

The Mantel test showed an isolation-by-distance pattern at the global population level. This phenomenon was also observed within the Marquesas archipelago but not within the Society islands. For the latter a more complex pattern prevails due to the genetic proximity of Raiatea and Moorea, but further research is needed to explain this mechanism.

The neighbour-joining tree summarizes this differentiation pattern (Fig. 2). The low bootstrap values do not clearly confirm the intermediate genetic position of the Austral archipelago between the two others, which was suggested by Butaud et al. (2005). These authors, using plastid microsatellites and the ages of these archipelagos (Table 1), proposed that the most likely hypothesis is that the Austral islands were first colonized by sandalwood and the source population of the founders that independently colonized the Marquesas and Society archipelagos. This assumption, based on the fact that dispersal is the preponderant mechanism of species distribution in Pacific islands (Cowie and Holland, 2006), remains to be confirmed by further phylogenetic studies with more suitable markers and a wider sampling.

Implications for conservation and management
As indicated in the Introduction, Santalum insulare is endangered throughout its natural range as a result of various human activities: overexploitation, fire and introduction of alien species (Butaud and Tetuanui, 2005). In this study, the aim was to provide broad guidelines for suitable and efficient conservation strategies using the genetic information obtained by microsatellite markers. For this species, the present approach to conservation is based on the definition of management units (MUs) using nuclear microsatellites as suggested by Moritz (1994). The best strategy is likely to lie in the combination of neutral and adaptive information (Moritz, 2002), but identification of molecular MUs seems to provide a first valuable framework (Cavers et al., 2003).

Four MUs were defined in a first study on Santalum insulare using plastid microsatellites (Butaud et al., 2005): two corresponding to the Marquesas archipelago and Society archipelago, and two corresponding to each of the Austral islands (Rapa and Raivavae). The present study tends to delineate more precisely the previously defined MUs. The high differentiation parameters obtained (0·19 < FST < 0·66), even though some of them should be considered with caution, encourage the association of an MU with each elementary population. However, a single MU for each pair Moorea/Raiatea and Hiva Oa/Tahuata can be considered.

In addition, the present analysis suggests some recommendations related to genetic diversity and gene flow within MUs. The small populations affected by clonality should be conserved in seed orchards. Their diversity should be restored by the reintroduction of protected seedlings (Moritz, 1999) and by using the most appropriate provenances, i.e. those that belong to the same MU in order to maintain good fitness (Wilkinson, 2001).

Other measures should accompany these first conservation activities. The eradication of rats should be a priority to improve the production of fruits and to restore regeneration by sexual reproduction. In addition, the reintroduction and protection of Columbidae should be implemented to increase gene flow and seed germination which could be facilitated by ingestion by birds.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
ACKNOWLEDGEMENTS
 REFERENCES
 
The present results with Santalum insulare confirm some of the general characteristics related to the pattern of diversity in insular systems. For example, a strong differentiation, higher than for mainland tree species, was found originating from oceanic barriers. Also a low genetic diversity was found but its origin is difficult to explain because of the confounded effects of insularity and past overexploitation. However, recent bottlenecks were not detected by the statistical approach used and inbreeding was not highlighted in the populations, as was to be expected in small islands (Frankham, 1998), especially with overexploitation which accentuated the decrease in population size. This suggests that sandalwood in Eastern Polynesia in the past formed abundant populations characterized by random mating. The present study has also revealed a strong clonality in this species. Although other studies do not demonstrate that clonality is an adaptation to insularity, it seems that this characteristic allows S. insulare to survive in such an environment. Finally this study, emphasizing the lack of population regeneration due to ship rats, stresses the negative impact of invasive species in insular systems.


   ACKNOWLEDGEMENTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
REFERENCES
 
We thank the programme ‘Expéditions scientifiques aux Iles Australes’, led by the Research Delegation in collaboration with the Malardé Institute, for the research trip on Raivavae, and the Ministère de l'Ecologie et du Développement Durable of France for funding part of the study through the research programme ‘Ecosystèmes Tropicaux’. We are grateful for the technical assistance of the staff of the genotyping platform of the Montpellier Languedoc-Roussillon Genopole located in the UMR PIA of CIRAD. We are grateful to Frédéric Mortier for his assistance in the implementation of the Geneland Software. We are also grateful to Alexandre Vaillant for the implementation of nuclear microsatellites in the Forest Department Molecular Laboratory of CIRAD Research Unit ‘Forest genetics’. Lastly, we acknowledge the staff of the Rural Development Service of French Polynesia for their help in gathering leaf samples in remote sandalwood populations and for their current management of sandalwood diversity in all the archipelagos.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 ACKNOWLEDGEMENTS
 REFERENCES
 

    Aldrich PR, Hamrick JL, Chavarriaga P, Kochert G. (1998) Microsatellite analysis of demographic genetic structure in fragmented populations of the tropical tree Symphonia globulifera. Molecular Ecology 7:933–944.[CrossRef][Medline]

    Arnaud-Haond S, Alberto F, Teixeira S, Procaccini G, Serrao EA, Duarte CM. (2005) Assessing genetic diversity in clonal organisms: low diversity or low resolution? Combining power and cost efficiency in selecting markers. Journal of Heredity 96:434–440.[Abstract/Free Full Text]

    Barton NH. (1989) Founder effect speciation. In Otte D and Endler JA (Eds.). Speciation and its consequences(Sinauer Press, Sunderland, MA) pp. 181–208.

    Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F. (2001) GENETIX 4·03, logiciel sous WindowsTM pour la génétique des populations.(Laboratoire Génome, Populations, Interactions, CNRS UMR 5000, Université de Montpellier II, Montpellier, France).

    Blondel J. (1995) Biogéographie Approche écologique et évolutive.(Masson, Paris).

    Bottin L, Verhaegen D, Tassin J, Olivieri I, Vaillant A, Bouvet JM. (2005) Genetic diversity and population structure of an insular tree, Santalum austrocaledonicum in New Caledonian archipelago. Molecular Ecology 14:1979–1989.[CrossRef][Medline]

    Bousquet J, Simon L, Lalonde M. (1990) DNA amplification from vegetative and sexual tissue of trees using polymerase chain reaction. Canadian Journal of Forestry Research 20:254–257.

    Butaud JF. (2004) Santalum insulare (Bertero ex A. DC.): distribution and ecology. Sandalwood Research Newsletter 19:1–4.

    Butaud JF and Tetuanui W. (2005) Sandalwood in French Polynesia. Proceedings of the Regional Workshop on Sandalwood Research, Development and Extension in the Pacific Islands and Asia7–11 October 2002In Thomson L, Bulai S, Sovea L (Eds.). , Noumea, New Caledonia. SPC Forest and Trees Programme Field document No 1. February 2005.

    Butaud JF, Rives F, Verhaegen D, Bouvet JM. (2005) Phylogeography of Eastern Polynesian sandalwood (Santalum insulare), an endangered tree species from the Pacific: a study based on plastid microsatellites. Journal of Biogeography 32:1763–1774.[CrossRef][Web of Science]

    Carlquist S. (1980) Hawaii: a natural history. geology, climate, native flora and fauna above the shoreline. 2nd edn Lawa'i Hawai'i: Pacific Tropical Botanical Garden.

    Cavalli-Sforza LL and Edwards AWF. (1967) Phylogenetic analysis: models and estimation procedures. Evolution 32:550–570.[CrossRef]

    Cavers S, Navarro C, Lowe AJ. (2003) A combination of molecular markers identifies evolutionarily significant units in Cedrela odorata L. (Meliaceae) in Costa Rica. Conservation Genetics 4:571–580.

    Cherrier JF. (1993) In McKinnel FH (Ed.). Sandalwood in French Polynesia. Proceedings of a symposium held on 2 June 1991 at the XVII Pacific Science CongressHonolulu, HawaiiSandalwood in the Pacific region pp. 24–25 ACIAR Proceedings No. 49.

    Clouard V and Bonneville A. (2001) Ages of seamounts, islands and plateaus of the Pacifique Plate version 2.0. Jeune équipe Terre-Océan, Université de Polynésie française.

    Cornuet JM and Luikart G. (1996) Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144:2001–2014.[Abstract]

    Cowie RH and Holland BS. (2006) Dispersal is fundamental to biogeography and the evolution of biodiversity on oceanic islands. Journal of Biogeography 33:193–198.[CrossRef][Web of Science]

    Darwin C. (1859) On the origin of species by means of natural selection(Watts, London) (reprint of 1st edn 1950).

    Dutech C, Joly HI, Jarne P. (2004) Gene flow, historical population dynamics and genetic diversity within French Guiana populations of a rainforest tree species, Vouacapoua americana. Heredity 92:69–77.[CrossRef][Web of Science][Medline]

    Eckert CG. (1999) Clonal plant research: proliferation, integration, but not much evolution. American Journal of Botany 86:1649–1654.[Free Full Text]

    El Mousadik A and Petit RJ. (1996) High level of genetic differentiation for allelic richness among populations of the argan tree (Argania spinosa (L.) Skeels) endemic to Morocco. Theoretical and Applied Genetics 92:832–839.[CrossRef][Web of Science]

    Emerson BC. (2002) Evolution on oceanic islands: molecular phylogenetic approaches to understanding pattern and processes. Molecular Ecology 11:951–966.[CrossRef][Medline]

    England PR, Usher AV, Whelan RJ, Ayre DJ. (2002) Microsatellite diversity and genetic structure of fragmented populations of the rare, fire-dependent shrub Grevillea macleayana. Molecular Ecology 11:967–977.[CrossRef][Medline]

    Excoffier L, Smouse PE, Quattro JM. (1992) Analyse of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131:479–491.[Abstract]

    Felsenstein J. (1993) PHYLIP: Phylogenetic inference package, Version 3.573c. (University of Washington, Seattle, WA).

    Fosberg FR and Sachet MH. (1985) Santalum in Eastern Polynesia. Candollea 40:459–470.

    Frankham R. (1998) Inbreeding and extinction: island populations. Conservation Biology 12:665–675.[CrossRef][Web of Science]

    Frankham R, Ballou JD, Briscoe DA. (2002) Introduction to conservation genetics(Cambridge University Press, Cambridge).

    Goudet J. (1995) FSTAT (vers. 1·2): a computer program to calculate F-statistics. Journal of Heredity 86:485–486.[Free Full Text]

    Guillot G, Estoup A, Mortier F, Cosson JF. (2005a) A spatial statistical model for landscape genetics. Genetics 170:1261–1280.[Abstract/Free Full Text]

    Guillot G, Mortier F, Estoup A. (2005b) GENELAND: a computer package for landscape genetics. Molecular Ecology Notes 5:712–715.

    Hamrick JL, Godt MJW, Sherman-Broyles SL. (1992) Factors influencing levels of genetic diversity in woody plant species. New Forests 6:95–124.[CrossRef]

    Handel SN. (1985) The intrusion of clonal growth patterns on plant breeding systems. American Naturalist 125:367–384.[CrossRef][Web of Science]

    Harder LD and Barrett SCH. Lloyd DG and Barrett SCH. (1996) Pollen dispersal and mating patterns in animal pollinated plants. Floral biology: studies on floral evolution in animal-pollinated plants.(Chapman & Hall, New York, NY) pp. 140–190.

    Hurlbert SH. (1971) The non-concept of species diversity: a critique and alternative parameters. Ecology 52:577–586.[CrossRef][Web of Science]

    IUCN. (2004) 2004 IUCN red list of threatened species. www.iucnredlist.org. Downloaded on 10 April 2006.

    Kennington WJ and James SH. (1997) Contrasting patterns of clonality in two closely related mallee species from western Australia, Eucalyptus argutifolia and E. obtusiflora (Myrtaceae). Australian Journal of Botany 45:679–689.[CrossRef]

    Kulkarni HD and Muniyamma M. Radomiljac AM, Ananthapadmanabho HS, Welbourn RM, Satyanarayana Rao. (1998) Floral biology and breeding systems in sandal, Santalum album L. Sandal and its products: Proceedings of an International Seminar18–19 December 1997Bangalore, India 84: pp. 135–146 ACIAR Proceedings.

    Lhuillier E, Vaillant A, Butaud JF, Bouvet JM. (2006) Isolation and characterization of microsatellite loci in Santalum insulare, Santalaceae. Molecular Ecology Notes 6: pp. 653–655.[CrossRef]

    Luikart G and Cornuet JM. (1998) Empirical evaluation of a test for identifying recent bottleneck populations from allele frequency data. Conservation Biology 12:228–237.[CrossRef][Web of Science]

    MacArthur RH and Wilson EO. (1967) The theory of island biogeography.(Princeton University Press, Princeton, NJ).

    Mack MC and D'Antonio CM. (1998) Impacts of biological invasions on disturbance regimes. Trends in Ecology and Evolution 13:195–198.[CrossRef]

    McConkey KR and Drake DR. (2002) Extinct pigeons and declining bat populations: are large seeds still being dispersed in the tropical Pacific? In Levey DJ, Silva WR, Galetti M (Eds.). Seed dispersal and frugivory: ecology, evolution and conservation.(CABI Publishing, Wallingford) pp. 381–395.

    Mantel N. (1967) The detection of disease clustering and a generalized regression approach. Cancer Research 27:209–220.[Abstract/Free Full Text]

    Marin JM, Diez RM, Insua DR. (2003) Bayesian methods in plant conservation biology. Biological Conservation 113:379–387.[CrossRef]

    Moritz C. (1994) Defining ‘Evolutionarily Significant Units’ for conservation. Trends in Ecology and Evolution 9:373–375.[CrossRef]

    Moritz C. (1999) Conservation units and translocations: strategies for conserving evolutionary processes. Hereditas 130:217–228.[CrossRef][Web of Science]

    Moritz C. (2002) Strategies to protect biological diversity and the evolutionary processes that sustain it. Systematic Biology 51:238–254.[CrossRef][Web of Science][Medline]

    Myers N, Mittermeler RA, Mittermeler CG, da Fonceca GAB, Kent G. (2000) Biodiversity hotspots for conservation priorities. Nature 403:853–858.[CrossRef][Medline]

    Nei M. (1978) Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583–590.[Abstract/Free Full Text]

    Nei M. (1987) Molecular evolutinary genetics(Columbia University Press, New York, NY).

    Page RDM. (1996) TREEVIEW: an application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences 12:357–358.[Free Full Text]

    Parker IM, Simberloff D, Lonsdale WM, Goodell K, Wonham M, Kareiva PM, et al. (1999) Impact: toward a framework for understanding the ecological effects of invaders. Biological Invasions 1:3–19.

    Parks JC and Werth CR. (1993) A study of spatial features of clones in a population of bracken fern, Pteridium aquilinum (Dennstaedtiaceae). American Journal of Botany 80:537–544.[CrossRef][Web of Science]

    Pielou EC. (1969) An introduction to mathematical ecology.(Wiley-Interscience, New York, NY).

    Piry S, Luikart G, Cornuet J-M. (1999) BOTTLENECK: a computer program for detecting recent reductions in effective population size from allele frequency data. Journal of Heredity 90:502–503.[Free Full Text]

    Rice WR. (1989) Analyzing tables of statistical tests. Evolution 43:223–225.[CrossRef][Web of Science]

    Rosetto M, Slade RW, Baverstock PR, Henry RJ, Lee LS. (1999) Microsatellite variation and assessment of genetic structure in tea tree (Melaleuca alternifolia—Myrtaceae). Molecular Ecology 8:633–643.[CrossRef][Medline]

    Rousset F. (1997) Genetic differentiation and estimation of gene flow from F-statistics under isolation by distance. Genetics 145:1219–1228.[Abstract]

    Saitou N and Nei M. (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4:406–425.[Abstract]

    Sanou H, Lovett PN, Bouvet JM. (2005) Comparison of quantitative and molecular variation in agroforestry populations of the shea tree (Vitellaria paradoxa C.F. Gaertn) in Mali. Molecular Ecology 14:2601–2610.[CrossRef][Medline]

    Schneider S, Roessli D, Excoffier L. (2000) ARLEQUIN: a software for population genetics data analysis. User manual version 2·0. Genetics and Biometry Laboratory, Department of Anthropology, University of Geneva, Switzerland. Free program distributed by the authors over internet from lfb.unige.ch/arlequin/.

    Smith S, Hughes J, Wardell-Johnson G. (2003) High population differentiation and extensive clonality in a rare mallee eucalypt: Eucalyptus curtisii. Conservation Genetics 4:289–300.[CrossRef]

    Steadman DW. (1997) The historic biogeography and community ecology of Polynesian pigeons and doves. Journal of Biogeography 24:737–753.[CrossRef][Web of Science]

    Stenberg P, Lundmark M, Saura A. (2003) MLGsim: a program for detecting clones using a simulation approach. Molecular Ecology Notes 3:329–331.[CrossRef]

    Vitousek PM, D'Antonio CM, Loop LD, Rejmánek M, Westbrooks R. (1997) Introduced species: a significant component of human-caused global change. New Zealand Journal of Ecology 21:1–16.

    Warburton CL, James EA, Fripp YJ, Trueman SJ, Wallace HM. (2000) Clonality and sexual reproductive failure in remnant populations of Santalum lanceolatum (Santalaceae). Biological Conservation 96:45–54.

    Weir BS and Cockerham CC. (1984) Estimating F-statistics for the analysis of population structure. Evolution 38:1358–1370.[CrossRef][Web of Science]

    van Welzen PC, Turner U, Hovenkamp P. (2003) Historical biogeography of Southeast Asia and the West Pacific, or the generality of unrooted area networks as historical biogeographic hypotheses. Journal of Biogeography 30:181–192.[CrossRef]

    Wilkinson DM. (2001) Is local provenance important in habitat creation? Journal of Applied Ecology 38:1371–1373.[CrossRef][Web of Science]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
98/5/1061    most recent
mcl190v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by LHUILLIER, E.
Right arrow Articles by BOUVET, J.-M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by LHUILLIER, E.
Right arrow Articles by BOUVET, J.-M.
Agricola
Right arrow Articles by LHUILLIER, E.
Right arrow Articles by BOUVET, J.-M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?