| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Population Biology |
University Herbarium, 1001 Valley Life Sciences Building, Department of Integrative Biology, University of California, Berkeley, Berkeley, California 94720 USA
Received for publication March 7, 2006. Accepted for publication March 28, 2007.
ABSTRACT
The Polynesian-introduced Cordyline fruticosa is used as a proxy for reconstructing human colonization patterns in Oceania. Because of its material, nutritional, medicinal, and religious importance, green-leaved C. fruticosa was transferred by Polynesian settlers to virtually every habitable Pacific island before European contact. Previous studies propose that green-leaved C. fruticosa is unable to reproduce sexually. To confirm sterility, crosses between fertile and putatively sterile forms were performed. To look for population structure in C. fruticosa that might confirm sterility as well as illustrate patterns of human migration, amplified fragment length polymorphism data were generated. Genotypic similarities were visualized using neighbor joining phenograms and analyses of molecular variance and principal components. The results from greenhouse crosses show that the Eastern Polynesian form is sterile; this finding is corroborated by a lack of genetic variability in Eastern Polynesian accessions. Sterile C. fruticosa appears to have been preferentially transferred throughout Eastern Polynesia; selection for the sterile form may be related to consumption of its rhizomes. Identification of a sterile form of C. fruticosa, possibly developed within Western Polynesia, may be significant to the systematics of Cordyline because it raises the possibility that the fertile form may actually be native to some Pacific islands.
Key Words: AFLP • Cordyline fruticosa • ethnobotany • Laxmanniaceae • Pacific • phylogeography • Polynesian colonization • selection
Organisms useful to humans have evolved in concert with human cultural development and movement throughout history. Recent studies have used phylogeographic information from such organisms to reconstruct human settlement patterns in Oceania (Matisoo-Smith et al., 1998
; Matisoo-Smith and Robins, 2004
; Zerega et al., 2004
). In natural systems, phylogeographic data are useful for understanding microevolutionary processes that have occurred relatively recently (Avise, 2000
). Genotypic patterns within species can reflect recent natural events, such as climate changes, addition or removal of barriers to gene flow, or selective pressures. Phylogeographic and population genetic models facilitate comparisons between different genotypic patterns and expectations under different population histories. Thus, it is possible to predict the specific genotypic patterns that might be expected under different demographic scenarios (Avise et al., 1984
; Brooks and McLennan, 1991
; Nielson and Wakely, 2001
). Co-distributed taxa and data sets can allow inferences to be made about shared community history using comparative phylogeographic principles (Platnick and Nelson, 1978
; Edwards and Beerli, 2000
).
Phylogeographic models can also be useful for understanding the genotypic structure of organisms that are dependent upon humans for their dispersal and persistence, which has implications for human movement in antiquity (Vanraamsdonk, 1993
). Studies of the population structure of cultivated plants also reflect characteristics of the plants themselves, providing insight into the nature of human selective forces that have influenced their genetic diversity, morphological diversity, and geographic distribution (Smith, 2001
). The islands of the Pacific are especially suited for phylogeographic studies of human-dispersed organisms. Islands, particularly islands that are geographically isolated by large expanses of water, naturally lend themselves to an endless number of interesting comparative problems (Darwin, 1839
; Wallace, 1880
). The recent human history in much of Oceania reduces the number of confounding historical factors, facilitating investigations of the selection and dispersal of introduced organisms and the historical movements of their vector of introduction: the Polynesians (Lebot and Lévesque, 1989
; Matisoo-Smith et al., 1998
; Zerega et al., 2001; Matisoo-Smith and Robins, 2004
).
The settlement of Polynesia took place after the settlement of much of the western and central Pacific region (Fig. 1) by a pre-Polynesian culture, termed Lapita (Kirch, 2000
). Lapita peoples, who are thought to have originated in Taiwan and the Philippines, began to move through Near Oceania (New Guinea, the Bismarck Archipelago, and the Solomon Islands) approximately 3500 yr ago (Bellwood, 1984
; Kirch, 1997
). These pre-Polynesian Lapita settlers then began to move rapidly into the previously uninhabited islands of Remote Oceania, through Fiji, and finally into Western Polynesia. Polynesian culture appears to have developed from Lapita roots in Western Polynesia from approximately 1000 B.C. to A.D. 1 in relative isolation from Fiji and Near Oceania (Green, 1967
). Following this so-called "long pause" in voyaging, remarkably rapid colonization of Eastern Polynesia was largely completed between A.D. 1 and A.D. 800 (Kirch, 2000
).
|
). C. fruticosa is associated with many cultural, material, and religious uses (Merlin, 1989
; Ehrlich, 1999
). As an ornamental, it has been planted around homes, on graves, and at religious sites. The leaves have been used for costumes, decorations, clothing, sandals, packaging, medicines, and cooking. The large rhizomes can be baked into a molasses-like food product and eaten, although this practice is largely restricted to Polynesia as the rhizomes are not generally eaten elsewhere in Oceania (Fornander, 1916–1920
; Ellis, 1963
; Handy and Handy, 1972
; Yen, 1973b
, 1974
; Abbott, 1992
; Ehrlich, 1999
).
The genus Cordyline has been grouped with a number of families; within the genus there are approximately 20 species (Beever, 1983
; Ehrlich, 1989
; Conran, 1998
). Species diversity is concentrated in Australia, New Zealand, and New Guinea, and also extends to South America and the Mascarene Islands. Cordyline fruticosa is generally considered to be a cultivated species with no "native" distribution of its own (Barrau, 1965
; D. Yen, Australian National University, retired, personal communication); others have suggested that the origin may be in New Guinea due to the tremendous diversity there in terms of foliage color, shape, and size (Ridley, 1924
). The morphological diversity of C. fruticosa attenuates toward the extremely isolated and remote islands of Eastern Polynesia (Ehrlich, 1999
). In the furthest reaches of the Polynesian Triangle, such as the Hawaiian Islands and New Zealand, only a single form is documented (Krauss, 1974
; Ehrlich, 1999
; Simpson, 2000
). Various Hawaiian floras describe this form as a sparingly branched woody plant reaching up to 3 m in height with a terminal cluster of large, green, spirally arranged leaves (Hillebrand, 1888
; Degener, 1930
; Wagner et al., 1990
). Species descriptions that match the Hawaiian form can be found in other regional Polynesian floras (Cheeseman, 1903
, 1925
; Setchell, 1926
; Brown, 1931
; Pétard, 1946
; Sykes, 1970
; Parham, 1972
; Smith, 1979
; Whistler, 1992
; Welsh, 1998
; Waldern et al., 1999
). Polynesian names for this form of C. fruticosa across Eastern Polynesia are generally cognates of the words "ti leaf"—"lau ti," "'auti," "la'i," "rau ti"—whereas in Western Polynesia, the qualifier describes its habitat: "ti vao" or "si vao," meaning "forest ti" (Cox, 1982
; Whistler, 1991
).
Despite the similarities in cultural uses and phenotypes of C. fruticosa across much of the Pacific, the mode of reproduction in green-leaved C. fruticosa is not completely understood. Results from studies of pollen fertility in Western and Eastern Polynesian C. fruticosa showed that Eastern Polynesian plants appear to be pollen-sterile; descriptions in regional floras suggest that they may be unable to set fruit, in contrast to fully fertile Western Polynesian and Fijian C. fruticosa (Yen, 1973b
; Hinkle, 2004
). The presence of fertile C. fruticosa at high elevations and in association with native vegetation has led some workers to consider it to be a separate, native species (Seemann, 1865
). Currently, the origin, distribution of various forms, and reproductive capacity of this ubiquitous and important Pacific species remain unclear.
In this study, the capacity for sexual reproduction in putatively sterile Eastern Polynesian C. fruticosa is tested by crossing field-collected plants under greenhouse conditions. Given that differences in reproductive capacity are likely to be reflected in population genetic structure, I investigate whether genotypes of the fertile form are distinct from the putatively sterile form. Further, I test whether sterility appears to have originated once or repeatedly, and if sterility in Eastern Polynesian plants is associated with reduced genetic diversity. The results are considered with respect to human settlement patterns in Polynesia. I discuss the significance of a sterile form in light of ethnobotanical uses of C. fruticosa and its possible geographic origin in the Pacific. Lastly, I consider the geographic range of the fertile form and the possibility that it is actually native to some archipelagoes in the central and western Pacific.
MATERIALS AND METHODS
Plant materials
Fresh leaf samples were collected from the Hawaiian, Society, Marquesas, Gambier, Cook, and Austral Islands in Eastern Polynesia, from Tonga and Samoa in Western Polynesia, and from Fiji and Vanuatu (Appendix). Due to logistical constraints, collection effort was concentrated in Polynesia, Fiji, and Vanuatu. Leaf material was collected from 2001 to 2005 and dried on silica gel. Stem cuttings from a subset of the leaf collections from Hawaii, the Society Islands, and Samoa were also collected and propagated under greenhouse conditions at the University of California, Berkeley, California, USA. Horticultural varieties with colored leaves were also collected for comparative purposes. Cordyline fruticosa with green leaves was collected in cultivated areas as well as in areas with native vegetation that were not clearly under cultivation.
Crosses
Crosses were performed to test the ability of C. fruticosa from Western and Eastern Polynesia to set fruit. Pollen stainability and germination were tested in previous studies (Hinkle, 2004
, 2005
). Western Polynesian and horticultural plants were in fruit at the time of field collection; Eastern Polynesian plants were collected in a sterile state. Six plants with fertile pollen (four from Samoa and two horticultural color varieties) and seven plants with sterile pollen (from the Hawaiian and Society Islands) were chosen based on simultaneous flowering times (Table 1, Appendix).
|
. Flowers were emasculated, bagged, and left to open overnight. The following 2 d, the stigmas of emasculated flowers were dusted with pollen from flowers that had opened the day that crosses were performed in the following experimental treatments: (1) pollen from Western and horticultural plants was introduced to the stigmas of fertile Western and horticultural plants (excluding self-pollinations) to establish normal fruit set proportions; (2) Western and horticultural plants were self-pollinated to test for self-compatibility; (3) pollen from Western Polynesian and horticultural plants was introduced to the stigmas of putatively sterile Eastern Polynesian plants; (4) pollen from Eastern Polynesian plants was introduced to the stigmas of Western Polynesian and horticultural plants to test for pollen viability; and (5) controls (emasculation but no pollination) were performed only in Western Polynesian and horticultural plants to test for incidental pollination or parthenocarpy. Treated branches were bagged; fruits that developed to maturity were counted 2 mo after treatment. In order to increase the chance of fertilization in treatment 3, the sample size of treated flowers was increased by treating and tagging, but not emasculating and bagging, flowers in Eastern Polynesian plants. A 
2 test was used to test for significance in the proportions of flowers that matured to fruits; multiple comparison tests were used to determine specifically which treatments differed from others (Zar, 1999
).
DNA extraction and genotyping using amplified fragment length polymorphisms (AFLP)
Genomic DNA was extracted from 20 mg of dry leaf material or 100 mg of fresh leaf material using Qiagen DNeasy (Valencia, California, USA) or Invitrogen Easy DNA (Carlsbad, California, USA) extraction kits. Genomic extractions were quantified using a NanoDrop 1000 Spectrophotometer (NanoDrop Technologies, Wilmington, Delaware, USA). The AFLP method developed by Vos et al. (1995)
was used for genetic analyses, with modifications as follows. Restriction and ligation reactions were performed simultaneously in a 50 µL solution containing approximately 250 ng genomic DNA, 5 U EcoRI, 0.5 U MseI, 5 µL 10x EcoRI Buffer I, 1 U T4 DNA ligase, 10 nmol ATP, 2.5 µg BSA, 5 pmol EcoRI adapter, and 50 pmol MseI adapter. The restriction-ligation product was diluted with 125 µL water. Preamplification was performed in a 25 µL solution containing 3 µL of diluted restriction–ligation product, 2.5 µL 10x PCR Buffer, 50 nmol MgCl2, 4 µg of BSA, 5 nmol dNTPs, 7.5 pmol of each of the preamplification primers, and 0.5 U Taq polymerase. The EcoRI primary amplification primer was identical to the adapter sequence, whereas the MseI primer had an extra "C" as a selective nucleotide. The PCR reaction was performed on a MJ PTC-0200 thermal cycler (Bio-Rad Laboratories, Waltham, Massachusetts, USA) using the following parameters: 18 cycles of 30 s at 94°C, 1 min at 60°C, and 1 min at 72°C. The primary amplification product was then diluted with 200 µL water. Selective amplification was performed with a similar cocktail, except that 6.5 µL diluted preamplification product was added in place of the restriction-ligation product, 7.5 pmol of the MseI selective primer was added in place of the Mse + C primer, and 6 pmol EcoRI fluorescently labeled selective primer was added in place of the Eco + 0 primer. Thirty selective primer pairs were screened, and two pairs were chosen that had large numbers of peaks with appropriate levels of polymorphism (EcoRI-ACG/MseI-CTC and EcoRI-AGG/MseI-GTC). Selective amplification had two cycle steps: 13 cycles of 30 s at 94°C, 30 s at 65°C initially and then lowered –0.7°C at each cycle, then 1 min at 72°C, followed by 18 cycles of 30 s at 94°C, 30 s at 56°C, and 1 min at 72°C. The selective amplification product was dried down in an 80°C incubator to 5 µL, and 0.5 µL concentrated product was mixed with 1.2 µL loading buffer (formamide, EDTA buffer, and ROX-500 size standard) (Applied Biosystems, Foster City, California, USA). Samples were run on an ABI Prism 377 automated sequencer using Genescan 3.1 software (Applied Biosystems). Negative controls were run at each step from restriction-ligation to selective amplification to detect any systematic contamination. All samples were duplicated from the restriction-ligation step; 25% of the duplicates came from separate extractions of the same leaf sample, and 35% of samples were run three or more times from the restriction-ligation step.
Data analysis
Raw data were visualized using Genotyper 2.5 software (Applied Biosystems). Samples that did not consistently amplify well were removed from the analysis. In total, 101 samples were used. Categories of fragments from 100 to 500 base pairs were generated in Genotyper, and each category was then evaluated manually. Presence or absence of peaks was scored based on agreement of replicates at each locus; conflict or uncertainty at a given locus was scored as missing data.
Neighbor joining (NJ) phenograms of all individuals were constructed using PAUP* version 4.0b10 (Swofford, 1998
) based on the distance measure by Nei and Li (1979)
. The NJ tree was bootstrapped using 1000 replicates. Cordyline banksii, endemic to New Zealand, was treated as the outgroup taxon. In MacClade version 4.07 (Maddison and Maddison, 2000
), two binary characters were mapped onto the topology: "geography" and "fertility." For the character geography, samples from Fiji, Tonga, and Samoa were considered to be Western, and all Eastern Polynesian archipelagoes were considered to be Eastern. For the character fertility, plants that produced fruit were considered to be fertile, and those that produced sterile pollen were scored as sterile (Hinkle, 2005
). Flowering is uncommon in Eastern Polynesian plants, and Western Polynesian plants flower and fruit at various times throughout the year. Therefore, the fertility status was known for only 57 of the 101 accessions.
Analysis of molecular variance (AMOVA; Excoffier et al., 1992
) implemented in the program Arlequin (Schneider et al., 1997
) was used to partition the variance within and among levels of hierarchical structure when grouped by either geographical origin or fertility status. Taxa with a population size of one (C. fruticosa from Vanuatu and C. banksii) were excluded, as were taxa of uncertain geographic origin (i.e., horticultural color varieties). Taxa were grouped in two ways: as Eastern Polynesia and Western Polynesia and Fiji or fertile and sterile, which were the larger hierarchical units (called groups in Arlequin). Taxa were divided further into archipelagoes (e.g., Hawaii, Societies, etc.) and were considered the smallest hierarchical unit (called populations in Arlequin). Western Polynesian and Fijian samples that were reconstructed as sterile on the NJ phenogram were placed into a Western Polynesian/Fijian sterile population within the sterile group. Within each group, AMOVAs were run to see how much of the variance could be partitioned among populations; in other words, to see if samples collected from the same archipelago are more similar. With each AMOVA, the significance of the F statistics was determined from 1000 random permutations of the data. FCT was calculated by permuting populations among groups; FSC was calculated by permuting genotypes among populations within groups; and FST was calculated by permuting genotypes among populations among groups (Schneider et al., 1997
).
A principal component analysis (PCA) was performed to visualize the distribution of the data in three dimensions. AFLP data were analyzed using an uncentered variance-covariance matrix in PC-ORD version 3.01 (McCune and Mefford, 1997
). Finally, a mean character difference distance matrix of Fijian, Western Polynesian, horticultural, and Eastern Polynesian plants was calculated in PAUP* and expressed as a histogram.
RESULTS
Crosses
Results of all crosses are summarized in Table 2. When Western Polynesian and horticultural plants with viable pollen were crossed (Treatment 1), 58% of the flowers set fruit. The results from Treatment 1 differed significantly from all other treatments (P < 0.0001). Fruiting was very low or entirely absent in self-crosses and the control, consistent with the low proportion of selfing and parthenocarpy seen in other species of Cordyline (Beever, 1983
; Beever and Parkes, 1996
). No fruits developed in any of the Eastern Polynesian plants despite treatment with pollen that had resulted in a high incidence of fruit set in Treatment 1. Eastern Polynesian pollen appeared to be ineffective in producing fruit in fertile Western Polynesian and horticultural plants that set fruit in Treatment 1. This is consistent with the low proportion of stainable pollen and pollen tube formation in Eastern Polynesian pollen (Table 1).
|
|
|
Results of the AMOVA (Table 4) showed that differences between Western and Eastern Polynesian samples accounted for 60.7% of the variance (P < 0.005, df = 1) when grouped by geography and 73.5% when grouped by fertility (P < 0.009, df = 1). The increase in among-group variance had little effect among populations; within populations, however, the variance in Fijian and Western Polynesian/fertile populations was reduced from 28.5% to 17.1% (P < 0.00001, df = 89, 83), reflecting the better "fit" of the five sterile Western Polynesian and Fijian samples within the sterile Eastern Polynesian group. Although samples are highly differentiated by both fertility status and geographic origin, fertility status has slightly more power to differentiate samples.
|
In the PCA, the first three principal components accounted for 89.8% of the cumulative variance of the samples (Fig. 3). The first principal component (76.3%) separates C. fruticosa from C. banksii and somewhat distinguishes Eastern Polynesian from Western Polynesian and Fijian samples. The second principal component (10.0%) isolates the tightly clumped Eastern Polynesian samples while the third component (3.6%) separates the Fijian samples from the Polynesian samples.
|
|
Sterility in Eastern Polynesian C. fruticosa
Green-leaved C. fruticosa in Eastern Polynesia appears to be incapable of both successful fruit set and ovule fertilization, which may largely explain the low genetic diversity among Eastern Polynesian samples. If Eastern Polynesian plants are incapable of sexual reproduction, it begs the question of whether or not they represent a widespread clonal population. In order to reliably distinguish clones from nonclones using AFLP data, some studies have used similarity thresholds of genetic distance. Using Nei–Li distances, the threshold was 2% in Populus (Arens et al., 1998
) and 0.985 (1.5%) in Salix (Douhovnikoff and Dodd, 2003
), slightly lower than C. fruticosa. These thresholds may be raised given the intrinsic error rates of AFLP and the presence of uninformative loci that may contribute to increased genetic distances of 2–5% (Huys et al., 1996
; Tohme et al., 1996
). Genetic distances between members of group 1 of the Eastern Polynesian samples are lower than any pairwise comparison within other groups (Fig. 4); genetic distances that also included group 2 were only slightly overlapped with other groups. There appears to be a similarity threshold that divides sterile and fertile plants, which lends some support to the idea that sterile plants are likely to be clones.
Sterility in C. fruticosa may also be significant in the context of the numerous asexually propagated root, tuber, and tree crops that characterize Polynesian agricultural systems (Barrau, 1965
; Yen, 1973b
). In the case of breadfruit (Artocarpus altilis), a Polynesian-introduced staple starch crop, virtually all cultivars in Eastern Polynesia are seedless despite the presence of seeded varieties in Western Polynesia and all other areas of Near Oceania (Ragone, 2001
; Zerega et al., 2004
). The concordant pattern seen in C. fruticosa at the divide between Western and Eastern Polynesia may signify an apparent preference for sterile C. fruticosa plants due to improved portability, rhizome flavor or texture, increased ecological tolerance or size, or other characteristics that sterility could potentially confer.
Bottleneck at Western–Eastern Polynesian split
Numerous studies have shown reduced diversity in terms of human genotypes, genotypes of introduced organisms, languages, cultural practices, and artifacts among Eastern Polynesian archipelagoes when compared with other archipelagoes to the west (Burrows, 1939
; Bellwood, 1989
; Murray-McIntosh et al., 1998
; Austin, 1999
; Gray and Jordan, 2000
; Kirch, 2000
; Ragone, 2001
; Matisoo-Smith and Robins, 2004
). This pattern is again confirmed by the clear geographic partitioning of genetic variance in C. fruticosa found at the boundary between Western and Eastern Polynesia (Table 4, Figs. 2, 3). The striking similarity among C. fruticosa genotypes across the vast Eastern Polynesian region suggests a genetic bottleneck and provides yet another example supporting a scenario of rapid but limited eastbound dispersal out of Western Polynesia. However, the lack of population structure within the Eastern Polynesian group does not provide sufficient resolution for the interpretation of fine-scale patterns relevant to colonization events within Eastern Polynesia.
Models for the introduction of C. fruticosa into Polynesia
The presence of both fertile and sterile genotypes in the west but only sterile genotypes in the east is consistent with the concept that newer populations often have only a subset of the allelic diversity found in older source populations (Hewitt, 1996
). The nearly identical genotypes in the Eastern Polynesian group are also characteristic of recent expansion over a short time period (Pannell, 2003
). Two competing models of population divergence and/or colonization may provide a useful framework for understanding how this pattern may have developed. Assuming these individuals have not been exchanged during historic times, either (1) there was west to east colonization with no back migration or (2) there was west to east colonization with subsequent back migration.
If the first hypothesis is correct, it would be expected that both the sterile and fertile genotypes existed in a homeland area in the west (i.e., Fiji and Western Polynesia), but the sterile genotypes were the only ones brought to Eastern Polynesia. The absence of fertile plants in Eastern Polynesia could be explained in two ways: one possibility is that sterile and fertile genotypes were introduced to the east but fertile genotypes did not persist there. For example, when transporting cuttings by mail for the present greenhouse study, the success rate for propagating plants from the Hawaiian and Society Islands was over 90%, but less than 20% for fertile Samoan plants. Despite this isolated observation, a second and more likely possibility is that sterile plants were preferentially moved and fertile plants were intentionally left behind.
The second hypothesis of a divergence/colonization model includes the added possibility of back-migration of sterile genotypes to archipelagoes west of Polynesia (i.e., Fiji) subsequent to the colonization of Eastern Polynesia. There is ample evidence of extensive borrowing of words, technologies, plants, and other materials in the central Pacific region (Hocart, 1929
; Parham, 1943
; Geraghty, 2004
). For example, Polynesian outlier islands to the west of Fiji are reported to have forms of C. fruticosa with distinctly Eastern Polynesian names and uses, further supporting a back-migration scenario (Burrows, 1936
; Firth, 1961
; Yen, 1973a
). Although the reproductive and genetic data for C. fruticosa presented here may be equivocal in favoring either model, evidence from the anthropological, archaeological, and linguistic literature is consistent with a back-migration scenario.
If the sterile form originated in Western Polynesia, intermediate genotypes may point to its possible geographic origin. Reconstruction of fertility status in Fig. 2 suggests that sterility arose only once. The two basal genotypes within the Eastern Polynesian group are from Samoa and the westernmost Eastern Polynesian archipelago, the Cook Islands. This corroborates findings by Matisoo-Smith et al. (1998)
using mitochondrial DNA data from the Polynesian-introduced Rattus exulans: their basal haplotype group includes samples from Samoa and the Cooks. For C. fruticosa, intermediate genotypes from these archipelagoes support the idea of west-to-east migration and could indicate that the sterile form originated in Samoa and was subsequently introduced to the west.
Origins of C. fruticosa in the Pacific
AFLP data are commonly used to confirm taxonomic classifications (Martínez-Ortega et al., 2004
; Whittall et al., 2004
; Gottlieb et al., 2005
). Criteria for recognizing lineages at particular taxonomic ranks include genetic cohesion of taxa within clades, restricted gene flow, long branches leading to clusters of taxa, and geographic isolation (De Queiroz, 1998
). In Fig. 2, the branch length of the Eastern Polynesian group translates to approximately 0.05 changes per character. In other studies (using Nei–Li distance), between-species distances were 0.06–0.08 (Veronica), 0.1–0.25 (Ilex), 0.07–0.16 (Lycopersicon), and 0.1 (Brachyelytrum); in Potamogeton, distances between cryptic species were >0.25 changes per character (Saarela et al., 2003
; Martínez-Ortega et al., 2004
; Nuez et al., 2004
; Whittall et al., 2004
; Gottlieb et al., 2005
). Distances between subspecies were 0.02–0.06 (Veronica), and approximately 0.04 between varieties (Lycopersicon) (Martínez-Ortega et al., 2004
; Nuez et al., 2004
). The Eastern Polynesian group in Fig. 2 meets the criteria of De Queiroz (1998)
, suggesting that it represents a diverging lineage. Potentially recognizable taxonomic structure within C. fruticosa has been suggested before in Seemann's (1865) Flora Vitiensis, where he described the sterile form as "C. terminalis (= C. fruticosa)" and treated the fertile form as a new species native to Fiji, C. sepiaria (Seem.).
The question then becomes whether or not this lineage could have originated in Western Polynesia (i.e., a divergence/colonization model with back migration). Considering that humans arrived in Fiji/Western Polynesia less than 3500 yr ago, it may be difficult to reconcile the amount of genetic change with such a short time span unless there was strong selection. Although it is not known exactly what Polynesians may have been selecting for, the strongest evidence may come from the use of C. fruticosa rhizomes as a food source.
The consumption of Cordyline rhizomes is largely restricted to Polynesia; consumption of rhizomes is uncommon in most areas of the Western Pacific (Yen, 1974
). The importance of Cordyline as a food source may relate to the theory that increased geographic isolation and small island size result in an increased risk of local extirpation of resources (MacArthur and Wilson, 1967
). The presence of a hardy and readily available food source such as Cordyline would have been of increased significance to Polynesian settlers and may have led to increased selection for certain features related to rhizome consumption. Preference for features associated with sterility in the genus Cordyline by Polynesians is apparently not unique to C. fruticosa, which did not grow well in temperate New Zealand. There, Polynesian settlers used native Cordyline species as C. fruticosa is used elsewhere in Polynesia, most notably as a source of food.
In New Zealand, the development of a different sterile Cordyline cultivar of the endemic C. australis (T. Armstrong, Landcare Research, personal communication) is relevant to the question of a possible Western Polynesian origin of sterile C. fruticosa. This C. australis cultivar, known as ti para or ti tawhiti, is a dwarfed, weak-stemmed plant whose soft rhizomes were reportedly preferred by New Zealand Maori for cooking and eating (Harris and Heenan, 1991
; Simpson, 2000
). This form is not known to flower and was propagated asexually, evidence that, by the time of first contact with Europeans, it had become a Polynesian cultigen. In other words, within the time span of roughly 500–700 yr, it appears that Polynesians were capable of transforming a native, outcrossing, long-lived, perennial tree in the genus Cordyline into a vegetatively propagated food plant whose rhizomes were easier to cultivate, process, and consume. This example provides compelling evidence that a sterile form of C. fruticosa could have been developed in a short period of time, concomitant with the emergence of other characteristic Polynesian innovations in Western Polynesia between 1000 B.C. and A.D. 1.
Considering the origin of the fertile form, could the genetic distances, genetic cohesion within clusters, apparent lack of gene flow, and geographic isolation between the Fijian and Western Polynesian groups suggest that they are separate lineages? If so, this is evidence that the fertile form of C. fruticosa may not be a human introduction at all, but actually a native species to the Pacific whose range ends at Western Polynesia. The assumption has always been that the fertile form has escaped cultivation; however, it certainly could not have escaped from cultivation of the sterile form. Furthermore, the fertile horticultural varieties with colored leaves do not appear to have escaped cultivation because they are not found in association with native vegetation.
The recognition of a sterile cultivated form releases the fertile form from the restriction of being a human introduction. Certainly many botanists have remarked that C. fruticosa grows as if it were native in undisturbed and high-elevation areas that typically have few introduced species (Seemann, 1865
; Christophersen, 1935
; Whistler, 1978
; Smith, 1979
). Proper treatment of the idea that fertile C. fruticosa is potentially indigenous to the Western Pacific would require more sampling in other archipelagoes such as Vanuatu, New Caledonia, the Solomon Islands, and New Guinea, with the expectation that each archipelago would continue to fit the criteria of being genetically and geographically sorted. In addition, phylogenetic information about the genus, currently unavailable, would be highly relevant.
APPENDIX. Voucher information and accession numbers for taxa used in this study. AH numbers are specimens collected by the author. AG numbers refer to genomic extractions of specimens for vouchers in collection series of other collectors, or when vouchers were not collected by the collector. Collection numbers preceded by other initials are vouchers from collection series of other collectors. Vouchers are housed at the University Herbarium at UC Berkeley (UC), Bishop Museum (BISH), or Musée de Tahiti et Ses Isles (PAP). PAP specimens have not yet been assigned accession numbers (NYA); PH indicates photo voucher only, dashes indicate that vouchers are missing
FOOTNOTES
1 The author thanks J.-Y. Meyer, M. Prebble, E. Claridge, A. Larsen, V. Garcia, W. A. Whistler, P. Kirch, and G. McCormack for their help with collections; the University of the South Pacific, American Samoa Community College Department of Agriculture, National Tropical Botanical Garden, Lyon Arboretum, Waimea Arboretum, Richard B. Gump South Pacific Research Station, and the Bishop Museum for assistance in the field; and A. Nettel-Hernandez, R. Dodd, and S. Lynch for assistance with AFLP. Special thanks go to E. Lasso, C. Epps, M. Hickerson, and the late C. Ehrlich for their help in improving this paper. This research is part of a doctoral dissertation supervised by B. Mishler, T. Carlson, B. Baldwin, and P. Kirch and was funded by Botany in Action of the Phipps Conservatory and Botanical Gardens, the Department of Integrative Biology at the University of California, Berkeley, the University of California Pacific Rim Research Program, the Richard B. Gump South Pacific Research Station, and the National Science Foundation (Doctoral Dissertation Improvement Grant DEB-0407975). 
2 Current address: Highlands Biological Station, P.O. Box 580, Highlands, NC 28741 USA (ahinkle{at}email.wcu.edu
)
LITERATURE CITED
Abbott I. A.. 1992. La'au Hawai'i: traditional Hawaiian uses of plants Bishop Museum Press, Honolulu, Hawaii, USA.
Arens P. Coops H. Jansen J. Bosman B.. 1998. Molecular genetic analysis of black poplar (Populus nigra L.) along Dutch rivers. Molecular Ecology 7: 11-18.[Medline]
Austin C. C.. 1999. Lizards took express train to Polynesia. Nature 347: 113-114.
Avise J. C.. 2000. Phylogeography: the history and formation of species Harvard University Press, Cambridge, Massachusetts, USA.
Avise J. C. Neigel J. E. Arnold J.. 1984. Demographic influences on mitochondrial DNA lineage survivorship in animal populations. Journal of Molecular Evolution 20: 99-105.[CrossRef][ISI][Medline]
Barrau J.. 1965. Witness of the past: notes on some food plants of Oceania. Ethnology 4: 282-294.[CrossRef][ISI]
Beever R. E.. 1983. Self-incompatibility in Cordyline pumilio (Agavaceae). New Zealand Journal of Botany 21: 93-95.[ISI]
Beever R. E. Parkes S. L.. 1996. Self-incompatibility in Cordyline australis (Asteliaceae). New Zealand Journal of Botany 34: 135-137.[ISI]
Bellwood P. S.. 1984. Hypothesis for Austronesian origins. Asian Perspectives 26: 107-117.
Bellwood P. S.. 1989. Colonization of the Pacific: some current hypotheses. In A. Hill and S. Serjeantson [eds.], Colonization of the Pacific: a genetic trail 1-59 Clarendon Press, Oxford, UK.
Brooks D. R. McLennan D. A.. 1991. Phylogeny, ecology, and behavior: a research program in comparative biology University of Chicago Press, Chicago, Illinois, USA.
Brown F. B. H.. 1931. Flora of Southeastern Polynesia, vol. 1, monocotyledons. Bernice P. Bishop Museum Bulletin 84 Honolulu, Hawaii, USA.
Burrows E. G.. 1936. Ethnology of Futuna. Bishop Museum Bulletin 138 Bishop Museum Press, Honolulu, Hawaii, USA.
Burrows E. G.. 1939. Western Polynesia: a study in cultural differentiation. Etnologiska Studier 7 Göteborg, Sweden.
Cheeseman T. F.. 1903. The flora of Rarotonga, the chief island of the Cook group. Transactions of the Linnaean Society, Series 2, Botany 6: 261-313.
Cheeseman T. F.. 1925. Manual of the New Zealand flora Government Printer, Wellington, New Zealand.
Christophersen E.. 1935. Flowering plants of Samoa. P. Bernice Bishop Museum Bulletin 128 Bishop Museum Press, Honolulu, Hawaii, USA.
Conran J. G.. 1998. Lomandraceae. In K. Kubitzki, H. Huber, P. Rudall, P. Stevens, and T. Stützel [eds.], The families and genera of vascular plants, vol. III: monocotyledons 354-365 Springer-Verlag, New York, New York, USA.
Cox P. A.. 1982. Cordyline ovens (umu ti) in Samoa. Economic Botany 36: 389-396.[ISI]
Darwin C.. 1839. Journal of the researches into the geology and natural history of various countries visited by H.M.S. Beagle, under the command of Captain Fitzroy, R.N. from 1832 to 1836 Henry Colburn, London, UK.
De Queiroz K.. 1998. The general lineage concept of species, species criteria, and the process of speciation. In D. Howare and S. Berlocher [eds.], Endless forms: species and speciation 19-31 Oxford University Press, Oxford, UK.
Degener O.. 1930. Ferns and flowering plants of Hawai'i National Park Honolulu Star Bulletin, Ltd., Honolulu, Hawaii, USA.
Douhovnikoff V. Dodd R. S.. 2003. Intra-clonal variation and a similarity threshold for identification of clones: application to Salix exigua using AFLP markers. Theoretical and Applied Genetics 106: 1307-1315.[ISI][Medline]
Edwards S. V. Beerli P.. 2000. Gene divergence, population divergence, and the variance in coalescence time in phylogeographic studies. Evolution 54: 1839-1854.[CrossRef][ISI][Medline]
Ehrlich C.. 1989. Special problems in an ethnobotanical literature search: Cordyline terminalis (L.) Kunth, the "Hawaiian ti plant.". Journal of Ethnobiology 9: 51-63.
Ehrlich C.. 1999. The ethnobotany of Cordyline fruticosa (L.) A. Chev.: the "Hawaiian ti plant" Ph.D. dissertation, State University at Buffalo, Buffalo, New York, USA.
Ellis W.. 1963. Journal of William Ellis: narrative of a tour of Hawaii or Owhyhee; with remarks on the history, traditions, manners, customs and language of the inhabitants of the Sandwich Islands Advertiser Publishing, Honolulu, Hawaii, USA.
Excoffier L. Smouse P. E. Quattro J. M.. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131: 479-491.[Abstract]
Firth R.. 1961. History and traditions of Tikopia The Polynesian Society, Wellington, New Zealand.
Fornander A.. 1916–1920. Fornander collection of Hawaiian antiquities and folk-lore. In T. Thrum [ed.], Bernice P. Bishop Museum Memoir 5. Bishop Museum Press, Honolulu, Hawaii, USA.
Geraghty P.. 2004. Borrowed plants in Fiji and Polynesia—some linguistic evidence. In J. Tent and P. Geraghty [eds.], Borrowing: a Pacific perspective 65-98 Pacific Linguistics, Canberra, Australia.
Gottlieb A. M. Gilbert G. C. Poggio L.. 2005. Molecular analysis of the genus Ilex (Aquifoliaceae) in southern South America, evidence from AFLP and ITS sequence data. American Journal of Botany 92: 352-369.
Gray R. D. Jordan F. M.. 2000. Language trees support the express-train sequence of Austronesian expansion. Nature 405: 1052-1055.[CrossRef][Medline]
Green R. C.. 1967. The immediate origins of the Polynesians. In G. Highland, R. Force, A. Howard, M. Kelly, and Y. Sinoto [eds.], Polynesian culture history 215-240 Bishop Museum Press, Honolulu, Hawaii, USA.
Handy E. S. C. Handy E. G.. 1972. Native planters in old Hawaii: their life, lore, and environment. P. Bernice Bishop Museum Bulletin 233.: Bishop Museum Press, Honolulu, Hawaii, USA.
Harris W. Heenan P. B.. 1991. Cordyline ‘Ti Tawhiti' and its relationship to Cordyline ‘Thomas Kirk'. Horticulture in New Zealand 2: 2-5.
Hewitt G. M.. 1996. Some genetic consequences of ice ages, and their role in divergence and speciation. Biological Journal of the Linnean Society 58: 247-276.[CrossRef][ISI]
Hillebrand W.. 1888. Flora of the Hawaiian Islands B. Westermann, New York, New York, USA.
Hinkle A. E.. 2004. The distribution of a male sterile form of ti (Cordyline fruticosa) in Polynesia: a case for human selection?. Journal of the Polynesian Society 113: 263-290.[ISI]
Hinkle A. E.. 2005. Population structure and reproductive biology of the Ti plant (Cordyline fruticosa) with implications for Polynesian prehistory Ph.D. dissertation, University of California, Berkeley, Berkeley, California, USA.
Hocart A. M.. 1929. Lau Islands, Fiji. Bishop Museum Bulletin 62 Bishop Museum Press, Honolulu, Hawaii, USA.
Huys G. Coopman R. Janssen P. Kersters K.. 1996. High-resolution genotypic analysis of the genus Aeromonas by AFLP fingerprinting. International Journal of Systematic Bacteriology 46: 572-580.
Kirch P. V.. 1997. The Lapita peoples: ancestors of the oceanic world Blackwell, Cambridge, Massachusetts, USA.
Kirch P. V.. 2000. On the road of the winds: an archaeological history of the Pacific Islands before European contact University of California Press, Berkeley, California, USA.
Krauss B. H.. 1974. Ethnobotany of Hawai'i Department of Botany, University of Hawai'i, Honolulu, Hawaii, USA.
Lebot V. Lévesque J.. 1989. The origin and distribution of kava (Piper methysticum Forts. F., Piperaceae): a phytochemical approach. Allertonia 5: 223-281.
MacArthur R. H. Wilson E. O.. 1967. The theory of island biogeography Princeton University Press, Princeton, New Jersey, USA.
Maddison D. R. Maddison W. P.. 2000. MacClade: analysis of phylogeny and character evolution, version 4.07 Sinauer, Sunderland, Massachusetts, USA.
Martínez-Ortega M. M. Delgado L. Alback D. C. Elena-Rosselló J. A. Rico E.. 2004. Species boundaries and phylogeographic patterns in cryptic taxa inferred from AFLP markers: Veronica subgen. Pentasepalae (Scrophulariaceae) in the western Mediterranean. Systematic Botany 29: 965-986.[CrossRef][ISI]
Matisoo-Smith E. Roberts R. M. Irwin G. J. Allen J. S. Penny D. Lambert D. M.. 1998. Patterns of prehistoric human mobility in Polynesia indicated by mtDNA from the Pacific rat. Proceedings of the National Academy of Sciences 95: 15145-15150.
Matisoo-Smith E. Robins J. H.. 2004. Origins and dispersals of Pacific peoples: evidence from mtDNA phylogenies of the Pacific rat. Proceedings of the National Academy of Sciences 101: 9167-9172.
McCune B. Mefford M. J.. 1997. Multivariate analysis of ecological data, version 3.01 MjM Software, Gleneden Beach, Oregon, USA.
Merlin M.. 1989. The traditional geographical range and ethnobotanical diversity of Cordyline fruticosa (L.) Chevalier. Ethnobotany 1: 25-39.
Murray-McIntosh R. P. Scrimshaw B. J. Hatfield P. J. Penny D.. 1998. Testing migration patterns and estimating founding population size in Polynesia by using human mtDNA sequences. Proceedings of the National Academy of Sciences 95: 9047-9052.
Nei M. Li W.-H.. 1979. Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences 186: 5269-5273.
Nielson R. Wakeley J.. 2001. Distinguishing migration from isolation: a Markov Chain Monte Carlo approach. Genetics 158: 885-896.
Nuez F. Prohens J. Blanca J. M.. 2004. Relationships, origin, and diversity of Galápagos tomatoes: implications for the conservation of natural populations. American Journal of Botany 91: 86-89.
Pannell J. R.. 2003. Coalescence in a metapopulation with recurrent local extinction and recolonization. Evolution 57: 949-961.[CrossRef][ISI][Medline]
Parham B. E. V.. 1972. Plants of Samoa Whitcombe and Tombs Ltd. under authority of A. R. Shearer, Government Printer, Wellington, New Zealand.
Parham H. B. R.. 1943. Fiji native plants with their medicinal and other uses. Polynesian Society Memoir No. 16 The Polynesian Society, Wellington, New Zealand.
Pétard P.. 1946. Cordyline terminalis. Ethno-botanique et médecine Polynésienne. Journal de la Société des Océanistes, 2: 194-208.
Platnick N. I. Nelson G.. 1978. A model of analysis for historical biogeography. Systematic Zoology 27: 1-16.
Ragone D.. 2001. Chromosome numbers and pollen stainability of three species of Pacific Island breadfruit (Artocarpus, Moraceae). American Journal of Botany 88: 693-696.
Ridley H. N.. 1922–1925. Flora of the Malay Peninsula, vol. 4. L Reeve & Co., London, UK.
Saarela J. M. Peterson P. M. Soreng R. J. Chapman R. E.. 2003. A taxonomic revision of the eastern North American and eastern Asian disjunct genus Brachyelytrum (Poaceae): evidence from morphology, phytogeography and AFLPs. Systematic Botany 28: 674-692.[ISI]
Schneider S. Roessli D. Excoffier L.. 1997. ARLEQUIN. A software for population genetics data analysis, version 2.1.4 Genetics and Biometry Laboratory, University of Geneva, Geneva, Switzerland. Website http://anthro.unige.ch/arlequin.
Seemann B.. 1865. Flora vitiensis: a description of the plants of the Viti or Fiji Islands with an account of their history, uses and properties L. Reeve and Co., London, UK.
Setchell W. A.. 1926. Tahitian spermatophytes collected by W. A. Setchell, C. B. Setchell, and H. E. Parks. University of California Publications in Botany 12: 143-240.
Simpson P.. 2000. Dancing leaves: the story of New Zealand's cabbage tree, ti kouka Canterbury University Press, Christchurch, New Zealand.
Smith A. C.. 1979. Flora vitiensis nova: a new flora of Fiji, vol. 5 Pacific Tropical Botanical Garden, Lawaii, Hawaii, USA.
Smith B. D.. 2001. Documenting plant domestication: the consilience of biological and archaeological approaches. Proceedings of the National Academy of Sciences 98: 1324-1326.
Swofford D. L.. 1998. PAUP*: phylogenetic analysis using parsimony (*and other methods), version 4 Sinauer, Sunderland, Massachusetts, USA.
Sykes W. R.. 1970. Contributions to the flora of Niue. Department of Scientific and Industrial Research (DSIR) Bulletin 200 Government Printer, Wellington, New Zealand.
Tohme J. Orlando-Gonzalez D. Beebe S. Duque M. C.. 1996. AFLP analysis of gene pools of a wild bean core collection. Crop Science 36: 1375-1384.
Vanraamsdonk L.. 1993. Wild and cultivated plants—the parallelism between evolution and domestication. Evolutionary Trends in Plants 7: 73-84.[ISI]
Vos P. Hogers R. Bleeker M. Reijans M. ven de Lee T. Hornes M. Frijters A. Pot J. Peleman J. Kuiper M. Zabeau M.. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Research 23: 4407-4414.
Wagner W. L. Herbst D. R. Sohmer S. H.. 1990. Manual of the flowering plants of Hawai'i University of Hawaii and Bishop Museum Press, Honolulu, Hawaii, USA.
Waldern S. Weisler M. I. Hather J. G. Morrow D.. 1999. The non-native vascular plants of Henderson Island, south central Pacific Ocean. Atoll Research Bulletin no. 463 National Museum of Natural History, Smithsonian Institution, C.
Wallace A. R.. 1880. Island life, or the phenomena and causes of insular faunas and floras Macmillan, London, UK.
Welsh S. L.. 1998. Flora societensis E.P.S. Inc., Orem, Utah, USA.
Whistler W. A.. 1978. Vegetation of the montane region of Savai'i. Pacific Science 32: 79-94.[ISI]
Whistler W. A.. 1991. Polynesian plant introductions. In P. Cox and S. Banack [eds.], Islands, plants, and Polynesians: an introduction to Polynesian ethnobotany 41-66 Dioscorides Press, Portland, Oregon, USA.
Whistler W. A.. 1992. Polynesian herbal medicine National Tropical Botanical Garden, Lawa'i, Hawaii, USA.
Whittall J. B. Hellquist C. B. Schneider E. L. Hodges S. A.. 2004. Cryptic species in an endangered pondweed community (Potamogeton, Potamogetonaceae) revealed by AFLP markers. American Journal of Botany 91: 2022-2029.
Yen D. E.. 1973a. Agriculture in Anutan subsistence. In D. Yen and J. Gordon [eds.], Anuta: a Polynesian outlier in the Solomon Islands, 116. Pacific Anthropological Records No. 21. B. P. Bishop Museum, Honolulu, Hawaii, USA.
Yen D. E.. 1973b. The origins of Oceanic agriculture. Archaeology and Physical Anthropology in Oceania 8: 68-85.
Yen D. E.. 1974. Arborculture and subsistence of Santa Cruz, Solomon Islands. Economic Botany 28: 247-284.[ISI]
Zar J. H.. 1999. Biostatistical analysis, fourth edition Prentice-Hall, Upper Saddle River, New Jersey, USA.
Zerega N. J. C. Ragone D. Motley T. J.. 2004. Complex origins of breadfruit (Artocarpus altilis, Moraceae): implications for human migrations in Oceania. American Journal of Botany 91: 760-76.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
