**4. Discussion**

454 Ecosystems Biodiversity

Average # of bands per primer

Diversity Index (DI)

Effective Multiplex Ratio (E)

Correlation coefficient (r)

Marker Index (MI)

Method

RAPD 14.60 0.29 17.13 4.914 Within Female 0.492

**ISSR** 16.20 0.33 16.20 5.289 Within Female 0.606

**DAMD** 19.00 0.23 19.00 4.423 Within Female 0.245

**SLXY** 17.25 0.25 17.25 4.366 Within Female 0.489

Table 3. The comparison of different PCR methods for the assessment of genetic diversity in

Mantel Coefficient for 100 random iterations (Z)

Standard Normal

RAPD / ISSR 21.179\*\* 673.781 0.8432+++ RAPD / DAMD 9.968\*\* 970.869 0.5009+ RAPD / SLXY 18.021\*\* 828.749 0.7853+++ ISSR / DAMD 8.895\*\* 790.384 0.4619+ ISSR / SLXY 17.562\*\* 680.878 0.7854+++ DAMD / SLXY 6.399\*\* 978.962 0.3850+

Table 4. Correlations among the distance matrices for the RAPD, ISSR, DAMD and SLXY primer PCR band data in case of the gender based set of the betelvine landraces. A significant, (p = 0.005), standard normal variate **(g)** was obtained among the matrices

generated by all the four methods in all possible pairs of matrix comparison.

Variate (g)

**\*\*** Highly significant values (Critical value **p**0.005 = 2.575)

Average similarity index (SI)

Comparison

Between Male and

Between Male and

Between Male and

Between Male and

Methods compared Matrix 1 / Matrix 2

**+++**: High correlation

**<sup>+</sup>**: Moderate or low correlation

Within Male 0.544

Female 0.282 Within Male 0.671

Female 0.400 Within Male 0.389

Female 0.183 Within Male 0.545

Female 0.266

the gender set of betelvine landraces.

Dioecy is a widespread condition in flowering plants. Despite their recent evolutionary origin, 6% of the 240,000-angiosperm species are dioecious and 7% of 13,000 genera of angiosperms include dioecious species, suggesting that it has arisen many times during flowering plant evolution (Renner and Ricklefs, 1995). Dioecy is correlated with perennial climbing growth, wind or water pollination and has a preponderence in tropical flora. Natural selection, development of complex physiological and morphological traits, male fitness, ecological context, sex ratio, phylogenetic perspective and nuclear-cytoplasmic gynodioecy are some of the factors affecting distribution of dioecy. The plant taxa can offer insights into level of dimorphism that exists prior to the evolution of complete dioecy and the tradeoffs or constraints faced by hermaphrodites. The betelvine is one such interesting dioecious plant with an almost obligate vegetative propagation, lianaceous habit and perennial growth that provides a good system for studying molecular aspects of dioecy in general and functional dioecy in particular.

The study with PCR profiles was the first step towards resolving gender differences, if any, in betelvines. In an earlier study with 53 landraces that included only a few landraces with known genders, the RAPD method had provided a distinction between male and female betelvines (Verma *et al.*, 2004). In the present study with a larger sampling specifically amongst the betelvines with known genders, the bootstrap NJ tree for the RAPD data clearly differentiated the male and female landraces into two separate broad clusters, thereby supporting our original grouping of the landraces in terms of male and female vines based on known or actual flowering data. Banerjee *et al.* (1999) have reported a similar study of RAPD profile variation in another dioecious species, *Piper longum* L. and have further shown that at least two RAPD fragments were consistently associated with male plants. The clear separation of the betelvines on the basis of gender provides important leads for the identification and development of gender specific primers and probes. This work, however, has an important caveat that the economically most important product of the betelvines are the leaves and leaves of both male and female vines have a market value. Therefore the gender specific detection and or diagnosis will have only an academic value in case of betelvines unlike that in plants such as Papaya (Deputy *et al.*, 2002; Ma *et al.*, 2004). ISSR-PCR has been successfully utilized to distinguish gender or gender-specific markers in dioecious plants like hops and datepalm (Jakse et al. 2008; Younis et al. 2008). In the present

SPAR Profiles for the Assessment of Genetic Diversity Between

mechanisms underlying the variation measured.

now exploit the diversity judiciously.

Male and Female Landraces of the Dioecious Betelvine Plant (*Piper betle* L.) 457

very interesting to determine the identity and sequence of genomic regions of betelvines that have resulted in the discrete multibanded profiles even under high PCR stringency. **Comparison of the different methods used to assess the genetic diversity in betelvines:** The gender based set of betelvine landraces was systematically analyzed with four different types of primers. For each set of primers used, the profile data were used for the calculation of the Diversity Index (DI), Effective Multiplex Ratio (E) and Marker Index (MI) and Mean probability (p). These calculations allowed a comparison of the four PCR-methods. The values calculated in each case, are given in Table 3. Parson *et al.* (1997) suggested that differences in the chromosomal location of the three types of markers could influence the diversity assessment. Kojima *et al.* (1998), indicated that in wheat RAPDs were more representative of chromosomal regions enriched in repeated sequences, while ISSRs were related, as RFLPs, to coding sequences. A similar case was seen in case of lentil (Sonnate and Pignone, 2001) where the authors could not find any congruence between the RAPD and ISSR method. Carvalho and Schaal (2001) also obtained different levels of polymorphism in cassava, where the SSR-primed markers showed less polymorphism than the RAPD markers. In their case also grouping of varieties from different geographical habitats varied between the RAPD and ISSR techniques. In absence of any pedigree information about the varieties, we could not address the issue of concordance between the molecular profiling based estimates of genetic similarity and pedigree but, we may expect greater genetic information about genetic similarity from the molecular profiling based estimates in accordance with Russell *et al.* (1997). From their result also it was apparent that the different techniques reveal information about distinct regions of the genome. Moreover, the rate of evolution of the primer target site sequence is most likely different for the two types of markers. So depending on this aspect, the divergence shown amongst the genotypes by the different techniques would also differ. Powell *et al.* (1996) suggested, that any estimation of genetic relationships between individual genotypes was affected by, the number of markers, the distribution of markers in the genome (genome coverage) and the nature of evolutionary

The four methods do not reveal polymorphism within gender set to the same extent. Under our experimental conditions ISSR method was found to have the higher Marker index as well as PIC. This is clearly reflected in the bootstrapped Neighbour Joining tree for the ISSR data where the entire female vines clustered together in one broad group while the male vines were separated in at least three distinct sub-groups. In general during the course of the present study we have observed that the male betelvines are invariably more heterogenous than the female vines. Such a result is of significance for the application of breeding methods for the improvement of betelvines. Unfortunately betelvine as a crop is cultivated by vegetative means ever since it was first domesticated. There have been only sporadic attempts at the improvement of betelvines through controlled process. The present study has resulted in the assessment of range of diversity in the betelvines for a breeder who can

In our study we found that ISSR method, which showed the highest Diversity and Marker Index, can be the method of choice for diversity analysis type of studies, in so far as polymorphism or Marker Index is the criterion. In the present study actually all four methods were almost equally useful for the analysis of the betelvine landraces (MI values in the range 4.366 to 5.289, Table 3). In fact the Mantel Test (distance matrix correlation)

study one ISSR primer each with dinucleotide and trinucleotide motifs and three primers with tetranucleotide motifs were tried in PCR and revealed distinct profiles that were however broadly similar across all the landraces. This result indicated that the ISSR regions were apparently conserved at least in length if not in sequence. The polymorphic bands helped to resolve the NJ tree into two major clusters with two groups of male and female betelvine clearly separating out. DAMD method has not been specifically used for any gender distinction studies in plants. For betelvine landraces, the NJ tree separated the male and female genders of the landraces and at the same time the DAMD method also revealed a relatively greater diversity amongst the male betelvines. The primers based on sequences specific to known X and Y chromosome of a dioecious plant *Silene latifolia* were tested with the dioecious betelvines in the gender set of landraces. Here the expectation was that the X and Y specific primers would reveal clear differences in the PCR profiles of the female and male betelvines respectively. These primers when used singly as well as in combination, however, did not reveal any such discrete profile differences. Surprisingly, these primers actually resulted in RAPD like multibanded profile and hence the band data for this was also scored in the same way as for the other three methods. This kind of study has never been done for any other plant to our knowledge. The conclusion about the most similar or most dissimilar landraces was based on cumulative data for all the SLXY primers. The result is actually interesting. Primarily the results indicate that multiple primer binding sites were present in all the DNAs. If these primers were generating multiple products it would appear that they have no co-relation with gender determining sequences in betelvines. This however is not entirely true since the NJ tree from the cumulative data can be clearly resolved into two major clusters, for the male and female betelvine respectively. Thus the PCR products resulting from these primers seemingly differs between the two genders and since several products are formed. It would appear that these primers are amplifying sequences from more than one gender determinant. Further since primers from both X and Y specific sequences gave multiple amplification products from the both male and female betelvines, our results indicate that gender determination in betelvine may either not be dependent on specific sex chromosomes or if such chromosome do exist male and female betelvines have similar sets of sequences on these chromosomes. Considering that all dioecious plants do not always have heteromorphic and distinct sex chromosomes, on the basis of our results with SLXY primers we infer that betelvine is one of those dioecious plants that lack distinct sex chromosomes.

In a novel approach, the fourth group of primers used was actually a heterologous set of primers were derived from X and Y chromosome specific sequences of the dioecious plant *Siliene latifolia*. The use of such primers for the dioecious betelvines was expected to reveal specific information about chromosomal basis for dioecy, if any. Interestingly, the SLXY primers resulted in multibanded profiles of several distinct bands even under stringent PCR conditions clearly indicating that several dispersed sequences homologous to the primers used were present. Though the primers collectively generated data that segregated the male and female vines, no single primer gave a sharply defined dimorphic profile for the two genders. This observation leads us to the conclusion that dioecy in betelvine may not follow the same chromosomal basis as in the case of *Silene latifolia*. Of course it is possible that such a lack of dimorphic profile could also be attributed to lack of strong homology of the primers to the appropriate regions of the betelvine. In this situation however, it would be

study one ISSR primer each with dinucleotide and trinucleotide motifs and three primers with tetranucleotide motifs were tried in PCR and revealed distinct profiles that were however broadly similar across all the landraces. This result indicated that the ISSR regions were apparently conserved at least in length if not in sequence. The polymorphic bands helped to resolve the NJ tree into two major clusters with two groups of male and female betelvine clearly separating out. DAMD method has not been specifically used for any gender distinction studies in plants. For betelvine landraces, the NJ tree separated the male and female genders of the landraces and at the same time the DAMD method also revealed a relatively greater diversity amongst the male betelvines. The primers based on sequences specific to known X and Y chromosome of a dioecious plant *Silene latifolia* were tested with the dioecious betelvines in the gender set of landraces. Here the expectation was that the X and Y specific primers would reveal clear differences in the PCR profiles of the female and male betelvines respectively. These primers when used singly as well as in combination, however, did not reveal any such discrete profile differences. Surprisingly, these primers actually resulted in RAPD like multibanded profile and hence the band data for this was also scored in the same way as for the other three methods. This kind of study has never been done for any other plant to our knowledge. The conclusion about the most similar or most dissimilar landraces was based on cumulative data for all the SLXY primers. The result is actually interesting. Primarily the results indicate that multiple primer binding sites were present in all the DNAs. If these primers were generating multiple products it would appear that they have no co-relation with gender determining sequences in betelvines. This however is not entirely true since the NJ tree from the cumulative data can be clearly resolved into two major clusters, for the male and female betelvine respectively. Thus the PCR products resulting from these primers seemingly differs between the two genders and since several products are formed. It would appear that these primers are amplifying sequences from more than one gender determinant. Further since primers from both X and Y specific sequences gave multiple amplification products from the both male and female betelvines, our results indicate that gender determination in betelvine may either not be dependent on specific sex chromosomes or if such chromosome do exist male and female betelvines have similar sets of sequences on these chromosomes. Considering that all dioecious plants do not always have heteromorphic and distinct sex chromosomes, on the basis of our results with SLXY primers we infer that betelvine is one of those dioecious

In a novel approach, the fourth group of primers used was actually a heterologous set of primers were derived from X and Y chromosome specific sequences of the dioecious plant *Siliene latifolia*. The use of such primers for the dioecious betelvines was expected to reveal specific information about chromosomal basis for dioecy, if any. Interestingly, the SLXY primers resulted in multibanded profiles of several distinct bands even under stringent PCR conditions clearly indicating that several dispersed sequences homologous to the primers used were present. Though the primers collectively generated data that segregated the male and female vines, no single primer gave a sharply defined dimorphic profile for the two genders. This observation leads us to the conclusion that dioecy in betelvine may not follow the same chromosomal basis as in the case of *Silene latifolia*. Of course it is possible that such a lack of dimorphic profile could also be attributed to lack of strong homology of the primers to the appropriate regions of the betelvine. In this situation however, it would be

plants that lack distinct sex chromosomes.

very interesting to determine the identity and sequence of genomic regions of betelvines that have resulted in the discrete multibanded profiles even under high PCR stringency.

**Comparison of the different methods used to assess the genetic diversity in betelvines:** The gender based set of betelvine landraces was systematically analyzed with four different types of primers. For each set of primers used, the profile data were used for the calculation of the Diversity Index (DI), Effective Multiplex Ratio (E) and Marker Index (MI) and Mean probability (p). These calculations allowed a comparison of the four PCR-methods. The values calculated in each case, are given in Table 3. Parson *et al.* (1997) suggested that differences in the chromosomal location of the three types of markers could influence the diversity assessment. Kojima *et al.* (1998), indicated that in wheat RAPDs were more representative of chromosomal regions enriched in repeated sequences, while ISSRs were related, as RFLPs, to coding sequences. A similar case was seen in case of lentil (Sonnate and Pignone, 2001) where the authors could not find any congruence between the RAPD and ISSR method. Carvalho and Schaal (2001) also obtained different levels of polymorphism in cassava, where the SSR-primed markers showed less polymorphism than the RAPD markers. In their case also grouping of varieties from different geographical habitats varied between the RAPD and ISSR techniques. In absence of any pedigree information about the varieties, we could not address the issue of concordance between the molecular profiling based estimates of genetic similarity and pedigree but, we may expect greater genetic information about genetic similarity from the molecular profiling based estimates in accordance with Russell *et al.* (1997). From their result also it was apparent that the different techniques reveal information about distinct regions of the genome. Moreover, the rate of evolution of the primer target site sequence is most likely different for the two types of markers. So depending on this aspect, the divergence shown amongst the genotypes by the different techniques would also differ. Powell *et al.* (1996) suggested, that any estimation of genetic relationships between individual genotypes was affected by, the number of markers, the distribution of markers in the genome (genome coverage) and the nature of evolutionary mechanisms underlying the variation measured.

The four methods do not reveal polymorphism within gender set to the same extent. Under our experimental conditions ISSR method was found to have the higher Marker index as well as PIC. This is clearly reflected in the bootstrapped Neighbour Joining tree for the ISSR data where the entire female vines clustered together in one broad group while the male vines were separated in at least three distinct sub-groups. In general during the course of the present study we have observed that the male betelvines are invariably more heterogenous than the female vines. Such a result is of significance for the application of breeding methods for the improvement of betelvines. Unfortunately betelvine as a crop is cultivated by vegetative means ever since it was first domesticated. There have been only sporadic attempts at the improvement of betelvines through controlled process. The present study has resulted in the assessment of range of diversity in the betelvines for a breeder who can now exploit the diversity judiciously.

In our study we found that ISSR method, which showed the highest Diversity and Marker Index, can be the method of choice for diversity analysis type of studies, in so far as polymorphism or Marker Index is the criterion. In the present study actually all four methods were almost equally useful for the analysis of the betelvine landraces (MI values in the range 4.366 to 5.289, Table 3). In fact the Mantel Test (distance matrix correlation)

SPAR Profiles for the Assessment of Genetic Diversity Between

Male and Female Landraces of the Dioecious Betelvine Plant (*Piper betle* L.) 459

Fig. 4. The NJ tree for the cumulative band data for all the four PCR methods after 1000 replicate bootstrap analysis clearly depicts the separation of the out group taxon, *Piper hamiltonii* (501) from the betelvines which in turn are clearly distributed into separate clusters of male and female landraces. Interestingly each cluster of male or female landraces is further resolved into two sub clusters each, demarked in the figure with smaller labeled parenthesis. The scale at the bottom of the figure is for the Jaccard coefficients. Each landrace is identified by its number and color code as in Table 1. The dashed line through

the figure separates the NJ tree into the respective gender halves.

analysis (Table 4) also resulted in highly significant normal standard variate g > p0.005 = 2.575 in all cases. However, here RAPD and ISSR have resolved the landraces almost equally similarly since the correlation (r = 0.8432, Table 4) was strong between the distance matrices of the two methods. This is an interesting observation and it is suggestive of the possibility that the landraces have more differences amongst themselves in their minisatellite and related tandem repeat sequences. This is actually supported by the observation that all DAMD bands in betelvine were polymorphic.

Sex detemination systems based on heteromorphic X and Y sex chromosomes are particularly interesting to study from both a developmental and evolutionary perspective. There are many paralleles between the sex chromosomes, in different species even between animals and plants. The evolution of heteromorphic sex chromosome systems in widely differing species suggests that similar forces may have been involved in each case (Charlesworh, 1992; Ellis, 1998; Charlesworth and Guttman, 1999). The *Silene* genus is an example of how the evolution of an XY system contributes to morphological change and speciation. The Y chromosome differs from all other chromosomes not only in that it is the only chromosome that does not recombine along majority of its length, but also is being present only in the male sex in a permanent haploid condition (Y genetic isolation), in having a common ancestry and persistent meiotic relationship with the X, and the tendency of its genes to degenerate during evolution (Y genetic erosion). The Y becomes a specialized male chromosome, which essentially behaves like a single recombination unit. The functional coherence of the Y can be achieved relatively early during Y evolution, which might be an essential condition for the maintenance of an XY system. Filatov *et al.* (2001) found several differences in polymorphism of the regions of X and Y chromosomes. In another study done by Lebel-Hardenack *et al.* (2002) for genetically mapping of the sexdetermination loci on the male-specific Y chromosome, it was found that *S.latifolia* has three dispersed male-determining loci on the Y chrosome. The sorrel *Rumex acetosa* (Polygonaceae) is a perennial dioecious weed, which possesses sex chromosomes (XX in females, XY1 Y2 in males). These studies indicated that it is only in the recent years that some details are emerging about molecular mechanisms and profiles vis-à-vis gender discrimination in plants.

The four methods tested with the gender set of betelvines have resulted in separation of the male and female betelvines. Though the individual sub-clusters in the four methods are not congruent however the primary separation of the male and female betelvines is clear and unambiguous. Thus the gender distinction of the betelvine landraces in terms of flowering is strongly supported by molecular profiling with four different PCR methods. This is an important and an interesting result. The four methods result in discrete profiles that reflect different genomic regions and in spite of that, the methods allow the landraces to be segregated on the basis of gender. Thus, it appears that there are several levels of genomic differences between the male and female betelvines. In other words, it appears that gender distinction in betelvine may not be confirmed to a few chromosomes or chromosomal regions. This inference is actually supported by lack of male specific and female specific PCR profiles when Y and X chromosome specific primers were respectively used. This further strengthens our inference that dioecy in betelvines is not apparently determined by distinct heteromorphic sex chromosomes. In order to have a comprehensive distribution of landraces, the data generated by all four methods combined and considered cumulatively.

analysis (Table 4) also resulted in highly significant normal standard variate g > p0.005 = 2.575 in all cases. However, here RAPD and ISSR have resolved the landraces almost equally similarly since the correlation (r = 0.8432, Table 4) was strong between the distance matrices of the two methods. This is an interesting observation and it is suggestive of the possibility that the landraces have more differences amongst themselves in their minisatellite and related tandem repeat sequences. This is actually supported by the observation that all

Sex detemination systems based on heteromorphic X and Y sex chromosomes are particularly interesting to study from both a developmental and evolutionary perspective. There are many paralleles between the sex chromosomes, in different species even between animals and plants. The evolution of heteromorphic sex chromosome systems in widely differing species suggests that similar forces may have been involved in each case (Charlesworh, 1992; Ellis, 1998; Charlesworth and Guttman, 1999). The *Silene* genus is an example of how the evolution of an XY system contributes to morphological change and speciation. The Y chromosome differs from all other chromosomes not only in that it is the only chromosome that does not recombine along majority of its length, but also is being present only in the male sex in a permanent haploid condition (Y genetic isolation), in having a common ancestry and persistent meiotic relationship with the X, and the tendency of its genes to degenerate during evolution (Y genetic erosion). The Y becomes a specialized male chromosome, which essentially behaves like a single recombination unit. The functional coherence of the Y can be achieved relatively early during Y evolution, which might be an essential condition for the maintenance of an XY system. Filatov *et al.* (2001) found several differences in polymorphism of the regions of X and Y chromosomes. In another study done by Lebel-Hardenack *et al.* (2002) for genetically mapping of the sexdetermination loci on the male-specific Y chromosome, it was found that *S.latifolia* has three dispersed male-determining loci on the Y chrosome. The sorrel *Rumex acetosa* (Polygonaceae) is a perennial dioecious weed, which possesses sex chromosomes (XX in females, XY1 Y2 in males). These studies indicated that it is only in the recent years that some details are emerging about molecular mechanisms and profiles vis-à-vis gender

The four methods tested with the gender set of betelvines have resulted in separation of the male and female betelvines. Though the individual sub-clusters in the four methods are not congruent however the primary separation of the male and female betelvines is clear and unambiguous. Thus the gender distinction of the betelvine landraces in terms of flowering is strongly supported by molecular profiling with four different PCR methods. This is an important and an interesting result. The four methods result in discrete profiles that reflect different genomic regions and in spite of that, the methods allow the landraces to be segregated on the basis of gender. Thus, it appears that there are several levels of genomic differences between the male and female betelvines. In other words, it appears that gender distinction in betelvine may not be confirmed to a few chromosomes or chromosomal regions. This inference is actually supported by lack of male specific and female specific PCR profiles when Y and X chromosome specific primers were respectively used. This further strengthens our inference that dioecy in betelvines is not apparently determined by distinct heteromorphic sex chromosomes. In order to have a comprehensive distribution of landraces, the data generated by all four methods

DAMD bands in betelvine were polymorphic.

discrimination in plants.

combined and considered cumulatively.

Fig. 4. The NJ tree for the cumulative band data for all the four PCR methods after 1000 replicate bootstrap analysis clearly depicts the separation of the out group taxon, *Piper hamiltonii* (501) from the betelvines which in turn are clearly distributed into separate clusters of male and female landraces. Interestingly each cluster of male or female landraces is further resolved into two sub clusters each, demarked in the figure with smaller labeled parenthesis. The scale at the bottom of the figure is for the Jaccard coefficients. Each landrace is identified by its number and color code as in Table 1. The dashed line through the figure separates the NJ tree into the respective gender halves.

SPAR Profiles for the Assessment of Genetic Diversity Between

economically important plant part is the leaf.

*longum* L. *Current Science* 77(5): 693-695.

markers. *Euphytica* 120: 133-142.

*Genet.* 106: 107-11.

388-390.

Scientific Publishers, Oxford pp. 25-50.

**5. References** 

Male and Female Landraces of the Dioecious Betelvine Plant (*Piper betle* L.) 461

The NJ tree from such a combined data for all methods is given in Figure 4. This tree also clearly separates the male and female betelvines. Interestingly the tree can be resolved into two subclusters within both male and female clusters as in the figure. MA, MB, FA and FB are the subclusters for males and females betelvines respectively (illustrated best by a radial tree form of the NJ tree as shown in Figure 5). We infer possibility of having at least two distinct ancient lineages for the male and female betelvines. In the absence of historical data and chronology of cultivation of these landraces it is not possible to confirm that there were actually a few discrete lineages of the cultivated betelvines and that the present day betelvine landraces are descendants of these lineages. The primary interest in all these studies is fuelled by the fact that dioecy as a proportion of accounts for only a small fraction of the numbers of flowering plants, yet in distribution across the plant families it is wide. Further, at the applied level, for plants of economic importance, detection and diagnosis of plant sex as an early event is desirable when the economically viable and important plant part is gender associated such as for example the fruits. In this context, betelvine is an exception in that for both male and female plants, the

Abe J., Xu D. H., Suzuki Y., Kanazawa A. and Shimamoto Y. (2003). Soybean germplasm pools in Asia revealed by nuclear SSRs. *Theor. Appl. Genet.* 106: 445-453. Ajibade S. R., Weeden N. F. and Chite S. M. (2000). Inter simple sequence repeat analysis of

Banerjee N. S., Manoj P. and Das M. R. (1999) Male sex-associated RAPD markers in *Piper* 

Bhattacharya E. and Ranade S. A. (2001). Molecular distinction amongst varieties of

Carvalho L. J. C. B. and Schaal B. A. (2001). Assessing genetic diversity in the cassava

Charlesworth D. and Guttman D. S. (1999). The evolution of dioecy and plant sex

deKochko A. and Hamon S. (1990). A rapid and efficient method for the isolation of restrictable total DNA from plants of genus *Abelmoschus*. *Plant Mol. Bio. Rep.* 8: 3-7. Deputy J. C., Ming R., Ma H., Liu Z., Fitch M. M., Wang M., Manshardt R., Stiles J.I. (2002)

Filatov D. A., Moneger F., Negrutiu I. and Charlesworth D. (2000). Low variability in a Y-

(*Manihot esculenta* Crantz) germplasm collection in Brazil using PCR-based

chromosome systems in *Sex determination in plants* edited by C. C. Ainsworth. BIOS

Molecular markers for sex determination in papaya (*Carica papaya* L.). *Theor Appl* 

linked plant gene and its implications for Y-chromosome evolution. *Nature* 404:

mulberry using RAPD and DAMD profiles. *BMC Plant Biology* 1: 3.

Charlesworth D. (1992). The evolution of sex chromosomes. *Curr. Biol.* 2: 515-516.

Ellis N. A. (1998). The war of the sex chromosomes. *Nature Genet.* 20: 9-10.

genetic relationships in the genus *Vigna*. *Euphytica* 111: 47-55.

Fig. 5. The radial NJ tree for the cumulative band data for all the four PCR methods after 1000 replicate bootstrap analysis clearly depicts the separation of the out group taxon, *Piper hamiltonii* (501) from the betelvines which in turn are clearly distributed into separate clusters of male and female landraces. Each cluster of male or female landraces is very distinctly resolved into two sub clusters each. The scale at the bottom of the figure is for the Jaccard coefficients. Each landrace is identified by its number and color code as in Table 1. The dashed line through the figure separates the NJ tree into the respective gender halves.

The NJ tree from such a combined data for all methods is given in Figure 4. This tree also clearly separates the male and female betelvines. Interestingly the tree can be resolved into two subclusters within both male and female clusters as in the figure. MA, MB, FA and FB are the subclusters for males and females betelvines respectively (illustrated best by a radial tree form of the NJ tree as shown in Figure 5). We infer possibility of having at least two distinct ancient lineages for the male and female betelvines. In the absence of historical data and chronology of cultivation of these landraces it is not possible to confirm that there were actually a few discrete lineages of the cultivated betelvines and that the present day betelvine landraces are descendants of these lineages. The primary interest in all these studies is fuelled by the fact that dioecy as a proportion of accounts for only a small fraction of the numbers of flowering plants, yet in distribution across the plant families it is wide. Further, at the applied level, for plants of economic importance, detection and diagnosis of plant sex as an early event is desirable when the economically viable and important plant part is gender associated such as for example the fruits. In this context, betelvine is an exception in that for both male and female plants, the economically important plant part is the leaf.
