Population structure and genetic diversity of wild Helianthus species from Mozambique

ISSN 1022�7954, Russian Journal of Genetics, 2010, Vol. 46, No. 8, pp. 967–975. © Pleiades Publishing, Inc., 2010.Original Russian Text © A. Ribeiro, M. Gouveia, A. Bessa, A. Ferreira, A.T. Magumisse, M. Manjate, T. Faria, 2010, published in Genetika, 2010, Vol. 46, No. 8, pp. 1086–1094.

967

1 Native to North America the genus Helianthus hasbeen spread throughout the world. The genus displayssubstantial phenotypic and genetic diversity and habi�tat variation and is composed of wild, weedy, anddomesticated species [1]. In Mozambique the specieswere presumably introduced accidentally during slavetrade, since 1645 and during at least two centuries, andhad become naturalized. Preliminary surveys [2],point for the presence of H. argophyllus Torrey andGray (silver�leaf sunflower) in the provinces ofMaputo and Gaza and of H. debilis Nutall (weak sun�flower) in the province of Maputo. Due to environ�mental disturbances caused by the civil war (1975–1992) and excessive urbanization (1995 – up to date),wild sunflower germplasm has been progressivelydecreased and scattered populations are found in someparts of the country, particularly in semiarid habitatsand sandy and salty soils reaching the beaches alongthe southeast coast of the country.

Aware of the current threatens and of the economicrelevance of sunflower, the Government of Mozam�bique in coordination with research institutions hasbeen concentrating efforts to re�launch its culture andthe oilseed industry since 1998. Amongst others, a sur�vey along the coastline was recommended, whichsought to establish a germplasm bank and to identify

1 The article is published in the original.

potential sources of resistance to biotic and abioticfactors for future breeding programs. Unfortunately,very little research has been done and only threepapers concerning fundamental aspects of Helianthusspecies in Mozambique have been published. Thesereported the introgression between Helianthus speciesin Inhambane Bay [3], the genetic diversity within andbetween populations of H. argophyllus in the Maputoarea [4], and the origin of H. argophyllus, H. debilisand their hybrids [5].

In this study, we assessed the genetic diversityamong 44 wild accessions of Helianthus species grow�ing along the Mozambique coastline (Fig. 1) throughRAPD markers, aiming at contributing for a betterunderstanding of the partitioning of genetic variationand population’s structure, which are essential stepsfor the design of a suitable conservation strategy.

MATERIALS AND METHODS

Plant material. In total 44 accessions of Helianthusspecies were collected in ten localities from five dis�tricts along the southeast coast of Mozambique, asillustrated in Fig. 1. The number of accessions, dateand topography were recorded for each population(Table 1). Accessions were named according to thenomenclature used at the germplasm bank from theFaculty of Agronomy and Forest Engineering/Edu�

Population Structure and Genetic Diversityof Wild Helianthus Species from Mozambique1

A. Ribeiroa, M. Gouveiab, A. Bessaa, A. Ferreiraa, A. T. Magumissec, M. Manjated, and T. Fariac

a Eco�Bio�Instituto de Investigação Científica Tropical, Quinta do Marquês (EAN), 2780�505 Oeiras, Portugalb Centra de Ciências da Vida, Universidade da Madeira Campus da Penteada, 9000�390 Funchal, Madeira, Portugalc Faculdade de Agronomia e Engenharia Florestal, Universidade Eduardo Mondlane, CP 257, Maputo, Mozambique

d Instituto National de Gestão de Calamidades, Rua da Resistência 1746–8°, Maputo, Mozambiquee�mail: [email protected] August 5, 2009

Abstract—The production of sunflower suffered a major decline in Mozambique after its independence in1975. Civil war, human activities and environmental damage subjected the species to an ecological stress con�tributing to reduce the number and size of wild populations. As this reduction is often related to a loss ofgenetic variation we estimated the genetic diversity within and among populations of wild Helianthus from fivedistricts of Mozambique using RAPD markers. The 44 accessions studied grouped into four major clustersexhibiting structured variability with regard to geographic origin. A high level of genetic diversity (He = 0.350and I = 0.527) was retained at the population level. The genetic variation among populations was high(59.7%), which is consistent with low gene flow (Nm = 0.338). The proportion of total genetic diversity resid�ing among these populations should be kept in mind to devise different conservation strategies in order to pre�serve these populations. Currently wild Helianthus genetic resources present in Maputo and Sofala are on theedge of extinction mainly due to excessive urbanization. Therefore, conservation of what remains of this plantgenetic diversity is essential for sustainable utilization and can be useful for breeding programs.

DOI: 10.1134/S1022795410080089

PLANT GENETICS

968

RUSSIAN JOURNAL OF GENETICS Vol. 46 No. 8 2010

RIBEIRO et al.

Niassa

CaboDelgado

TeteNampula

Zambezia

Sofala

Belra

Man

ica

Gaza Inhambane

Morrumbene

Inhambane

Xai�Xai

Map

uto

0 100 200 km

INDIAN

OCEAN

Maputo

Fig. 1. Map of Mozambique showing the provinces and districts where the wild accessions from Helianthus sp. were collected forthis study.

ardo Mondlane University (FAEF/UEM): G = Gaza,I = Inhambane, M = Maputo and S = Sofala, followedby a number ordered according to the available cata�logue, which included samples from other surveys.After harvesting, samples were either herborized orfrozen at –80°C for DNA extraction. Seeds were keptin the germplasm bank of FAEF/UEM.

DNA extraction and RAPD assay. Total genomicDNA was extracted from leaves with the DNeasy PlantMini kit (Qiagen, Germany) according to manufac�turer’s instructions. Average yield was calculated spec�trophotometrically (Lambda EZ201, Perkin Elmer,USA). RAPD assays [6] were performed as describedin [7], Forty�three decamer primers were initiallytested (Kits C, D and F, Operon Technologies Inc.,USA), from which 5 primers were used for further

analysis: OPC�04, OPD�05, OPD�15, OPD�19,OPF�09. For each primer a negative control, in whichDNA was replaced with water, was included. Dupli�cate amplifications of at least two different DNAextractions were performed for all samples to assessthe consistency of the fragments profile. PCR prod�ucts were resolved by electrophoresis in 1.2% agarosegels [7] and photographed under ultraviolet light usingthe BioDoc�It System (UVP, Inc. Upland, CA).

Data analysis. RAPD data was analyzed using thecomputer program NTSYSpc version 2.20e [8]. Anunweighted pair group arithmetic mean method(UPGMA) cluster analysis was performed based onthe DICE’s similarity coefficient. Phenograms weregenerated using the tree display option (TREE).Goodness of fit of the cluster analysis was determined


POPULATION STRUCTURE 969

by computing a cophenetic value matrix. A copheneticcorrelation of r > 0.9 was considered a very good fit.The genetic similarity data matrix between accessionswas utilized to perform principal coordinates (PCO)analysis via PCOORDA using DCENTER andEIGEN modules of NTSYSpc.

For statistical analysis the binary data matrix wastransformed to allele frequencies under the assump�tion that RAPD fragments behave as diploid, domi�nant markers with alleles being either present (ampli�fied) or absent (not amplified) and that each amplifiedfragment corresponds to a different RAPD locus.Assuming Hardy�Weinberg equilibrium POPGENEsoftware version 1.31 [9] was used to estimate geneticvariation within populations by calculating the follow�ing parameters: percentage of polymorphic loci (P),mean effective number of alleles per locus (Ne), Nei’sgene diversity (He) [10] based on allelic frequenciesand Shannon’s index of phenotypic diversity (I) [11]based on marker frequencies. The genetic structureand variability among the subpopulations was esti�mated by calculating Nei’s unbiased genetic distances[12], genetic differentiation (Gst) and gene flow (Nm).An UPGMA cluster analysis based on unbiased dis�tance between subpopulations [12] was constructedusing TFPGA (Tools. For Population Genetic Analy�ses) v. 1.3 [13], bootstrapping over loci with 1000 per�mutations. Correlation coefficients were determinedusing the STATISTICA program [14].

RESULTS

Taxonomic Classification

A great variability of phenotypic characteristics wasobserved among the 44 accesses suggesting that theyrepresent hybrids of H. argophyllus × H. annus,H. annus × H. debilis, and H. argophyllus × H. debilis .

Phenetic Analysis

The phenogram generated by UPGMA analysis isshown in Fig. 2. The cophenetic correlation used as ameasure of goodness�of�fit for cluster analysis wasvery high (r = 0.903), indicating an excellent represen�tation of the original data. Four major clusters wereidentified from the phenogram. The first (M�I) andsecond (M�II) clusters comprised accessions collectedfrom Maputo. The third (In�I) was a tight cluster andcomprised 7 accessions collected from Morrumbene.The fourth major cluster was subdivided into threesubclusters at more than 70% level of similarity. Sub�cluster 1 (In�11) included accessions from Morrum�bene. Subcluster 2 (In�III) contained the single acces�sion collected from Beira (S367) and 8 accessionsfrom Inhambane. The third subcluster (G�I) groupedall accessions from Xai�Xai and one accession fromInhambane. Accessions from Cocane (I.385 andI.386) were dissimilar and clustered independent ofother accessions. According to Spearman’s rank cor�relation a highly significant negative correlation (r =⎯0.961, p < 0.001) was observed between clusters andgeographical origin of the accesses, whereas no corre�lation was detected between topography and geo�

Table 1. Provenience and number of accessions of wild Helianthus species studied

Population Subpopulation Date of collection Topography Accessions

Beira Vaz April, 2002 Flat S367

Morrumbene Morrumbene August, 2002 Plain I409, I410, I411, I412, I413, I414, I415, I416, I417, I420, I421, I423, I424, I425

Cocane April, 2002 Undulated I385, I386

Inhambane Inhambane April, 2002 Undulated I368, I369, I370, I370a

Est. IB April, 2002 Flat I377, I379, I382

Undulated I374, I384

Xai�Xai Maciene April, 2002 Hill G387, G390, G391, G394

Maputo B. Pescadores April, 2002 Hill M1, M2, M3

Flat M4, M5

Pnte C. Sol April, 2002 Flat M6, M7, M8, M9

FAEF April, 2002 Plain M10, M14, M15

Av. J. Nyerére April, 2002 Undulated M12, M13

970


RIBEIRO et al.

graphical origin of the accesses (r = 0.189, p = 0.219)or topography and clusters (r = 0.071, p = 0.646).

Principal Coordinates Analysis

Principal coordinates (PCO) analysis was per�formed to view the relationships between the acces�sions studied (Fig. 3). The first two principal compo�nents accounted for 42.4% of the total variation (26.0and 16.4%, respectively) in the genetic distancematrix. PCO analysis results indicated four distinctgroups. Accessions from Maputo were found to beindependent and included into two separated groups(M�I and M�II), which perfectly coincided with thecluster analysis (Fig. 2). The third group was formedexclusively by accessions from Morrumbene. Thisgroup could be divided into two subgroups, coincidingwith the clusters In�I and In�II, with the exception ofaccession I.412 that diverges from In�I to In�II. The

fourth group comprised a mixed distribution of theaccesses from all districts but Maputo.

Genetic Variation

Estimates of genetic variation based on RAPDmarkers were relatively high in wild accessions ofHelianthus species from Mozambique (Table 2). At thesubpopulation level, the percentage of polymorphicloci ranged from 11.6 in Maciene and Inhambane to69.8 in Morrumbene, although accessions fromMaputo were the most polymorphic (ranging from44.2 to 60.5).

Using all polymorphic loci obtained, the effectivenumber of alleles per locus across all populations esti�mated Ne = 1.581 ± 0.262 (ranging from 1.077 to1.422). Maputo population displayed the highest val�ues for all parameters (mean of 1.501).

The genetic diversity estimated by Nei’s genediversity (He) ranged from 0.087 to 0.241 in Morrum�

G.387G.394G.390G.391I.382I.368I.369I.370I.370aI.374I.384I.377I.379S.367I.385I.425I.409I.411I.410I.413I.417I.416I.412I.414I.415I.420I.421I.423I.424I.386M.1M.2M.5M.6M.7M.3M.12M.15M.14M.4M.10M.9M.13M.8

G�1

In�III

In�II

In�I

M�II

M�I

0.34 0.50 0.66 0.82 0.98Coefficient

Fig. 2. Phenogram showing relationships among 44 accessions of Helianthus sp. based on diversity of RAPD markers using theUPGMA algorithm and the Dice similarity coefficient.



bene subpopulations while similar gene diversity wasobserved within subpopulations from Maputo (0.183–0.236) or Inhambane (0.044–0.097). The overallmean gene diversity was 0.350 ± 0.113.

The Shannon’s diversity index as a measure of thedegree of variation within each population is given inTable 2. Shannon’s index of phenotypic diversity waslower in Xai�Xai (I = 0.068) population but revealedhigher variability in Maputo (I = 0.447) population.Total genetic diversity across populations was 0.527 ±0.134, suggesting high genetic differentiation betweenpopulations.

A positive Pearson correlation was detectedbetween percentage of polymorphism and Shannonindex (r = 0.956, p < 0.001) and between effectivenumber of alleles per locus and Nei’s gene diversity(r = 0.997, p = < 0.001), suggesting that RAPD mark�ers were unevenly distributed.

Due to the fact that only one accession was avail�able, Beira/Vaz was taken out of this, analysis.

Genetic Structure

The distribution of genetic diversity among andwithin populations was also assessed with genotypicdata obtained from RAPD loci (Table 3). Total genediversity (Ht) varied between 0.086 (Inhambane) and0.294 (Maputo) with a mean of 0.340 ± 0.017. Withinpopulations genetic diversity (Hs) ranged from 0.070(Inhambane) to 0.214 (Maputo) with a mean of0.137 ± 0.005. In the subdivided populations limitedgenotypic diversity and little evidence of subdivisionwere observed in accessions from Inhambane. In theMorrumbene population genetic diversity was higherin total than within populations suggesting a partition�ing between populations. The coefficient of gene dif�ferentiation (Gst) in all wild sunflower accessions cal�

G.387G.394

G.390G.391

I.382 I.368

I.369

I.370

I.370a

I.374I.384

I.377

I.379

S.367

I.385

I.425

I.409

I.411I.410

I.413

I.417

I.416

I.412

I.414

I.415

I.420

I.421I.423

I.424

I.386

M.1

M.2M.5 M.6

M.7

M.3

M.12M.15

M.14

M.4

M.10 M.9

M.13

M.8

M�II

M�I

0.4

0.6

–0.4–0.7

Dim

�3

Dim�2

Dim�1

0.4

Fig. 3. Principal coordinates (PCO) analysis using RAPD data of 44 accessions of wild Helianthus sp. populations from five dis�tricts of Mozambique: Beira (�), Morrumbene (�, �), Inhambane (�, �), Xai�Xai (×), Maputo (+, , �, �) (see Table 1 fordetails).

972


RIBEIRO et al.

culated by population genetic analysis was 0.597, indi�cating that the genetic variation among populationswas 59.7%. In Inhambane and Maputo variation wasmainly within population as nearly 18% and 27% ofthe total genetic variation can be explained by RAPDsdifference, respectively. The estimated Nm fromInhambane and Maputo populations points to a mod�erate gene flow (2.290 and 1.343, respectively).According to the estimated Nm Morrumbene’s acces�sions showed geographical differentiation.

Beira and Xai�Xai populations were excluded fromthis analysis because only one sub�population wasavailable.

Genetic Distance between Populations

Based on Nei’s unbiased measure of genetic iden�tity and genetic distance, the genetic distance amongthe 5 populations studied ranged from 0.0528 to0.5401 (Table 4). The largest genetic distance was

Table 2. Estimates of genetic variation in wild Helianthus sp. from the southeast coast of Mozambique using RAPD markers. Per�centage of polymorphic loci (% P), effective number of alleles (Ne), Nei’s gene diversity (He) and Shannon’s information index (I)

Population Subpopulation % P Ne He I

Morrumbene Cocane 20.9 1.148 ± 0.291 0.087 ± 0.171 0.127 ± 0.249

Morrumbene 69.8 1.405 ± 0.359 0.241 ± 0.194 0.361 ± 0.276

Mean 83.7 1.431 ± 0.332 0.263 ± 0.171 0.403 ± 0.236

Inhambane Inhambane 11.6 1.077 ± 0.240 0.044 ± 0.129 0.065 ± 0.186

Est IB 23.3 1.171 ± 0.330 0.097 ± 0.181 0.141 ± 0.261

Mean 30.2 1.133 ± 0.236 0.088 ± 0.146 0.140 ± 0.224

Xai�Xai Maciene 11.6 1.080 ± 0.235 0.046 ± 0.132 0.068 ± 0.191

Maputo B. Pescadores 58.1 1.422 ± 0.410 0.236 ± 0.217 0.344 ± 0.308

Pnte C. Sol 60.5 1.370 ± 0.360 0.220 ± 0.197 0.330 ± 0.284

FAEF 53.5 1.388 ± 0.414 0.217 ± 0.216 0.317 ± 0.308

Av. J. Nyerere 44.2 1.312 ± 0.355 0.183 ± 0.208 0.267 ± 0.304

Mean 88.4 1.501 ± 0.349 0.296 ± 0.170 0.447 ± 0.226

Overall 100 1.581 ± 0.262 0.350 ± 0.113 0.527 ± 0.134

Table 3. Estimates of genetic diversity in wild Helianthus sp. from Mozambique. Total gene diversity (Ht), gene diversitywithin population (Hs), genetic differentiation (Gst) and gene flow (Nm)

Population n Ht Hs Gst Nm

Morrumbene 16 0.259 ± 0.033 0.164 ± 0.180 0.369 0.857

Inhambane 9 0.086 ± 0.020 0.070 ± 0.014 0.179 2.290

Maputo 14 0.294 ± 0.028 0.214 ± 0.019 0.271 1.343

Overall 44 0.340 ± 0.017 0.137 ± 0.005 0.597 0.338

Table 4. Estimate values of Nei’s unbiased measures of genetic identity (above diagonal) and genetic distance (below diagonal)for wild Helianthus sp. accessions from five districts of Mozambique, based on [12]

Beira Morrumbene Inhambane Xai�Xai Maputo

Beira – 0.8127 0.9119 0.8280 0.5827

Morrumbene 0.2074 – 0.8616 0.8045 0.7885

Inhambane 0.0922 0.1490 – 0.9486 0.6814

Xai�Xai 0.1887 0.2176 0.0528 – 0.6617

Maputo 0.5401 0.2376 0.3837 0.4130 –



between the population of Beira and the population ofMaputo, while the lowest genetic distance wasobserved between the populations of Inhambane andXai�Xai. The genetic distances between these popula�tions are consistent with the results of genetic identityanalysis (Table 4).

A dendrogram was generated using UPGMAmethod based on the Nei’s unbiased genetic distancematrix (Fig. 4). The 10 subpopulations of Helianthusspecies were clearly divided into two major groups:group 1 comprises the four subpopulations fromMaputo, whereas group 2 includes the other 6 subpop�ulations. This result is in accordance with the clusteranalysis that was performed using RAPD markers(Fig. 2).

DISCUSSION

Considering the economic importance of sun�flower as an oilseed crop, the Government of Mozam�bique has been concentrating efforts to re�launch itsculture and the oilseed industry. Integrated in a con�servation program to elaborate a systematic survey ofthe available genetic resources, ten wild sub�popula�tions of Helianthus species were studied in order toestimate the level and distribution of genetic variation.Most populations collected in five districts at thesoutheast coast of Mozambique were found to be smalland consisted of only a few individuals. Therefore, thelow number of populations, sub�populations and indi�viduals was reflected on the number of accessions usedin the present study. In Beira, for instance, a singleaccession was found in the wild.

In spite of the progressive lost of wild sunflowergermplasm, the overall genetic variation determined inthis study was remarkably high. Such variability mayresult from the fact that the 44 sunflower accesses mayrepresent hybrids of H. argophyllus × H. annus,H. annus × H. debilis, and H. argophyllus × H. debilis.The results are in agreement with the observations ofT. Faria (unpublished) and [15], based on the greatvariability of phenotypic characteristics. In fact, taxo�nomic units, with high frequency of genetic materialexchange, also defined as compilospecies, have beenfirst reported in Helianthus by [16–18].

The distribution of Helianthus accessions by clus�ters followed a pattern closely related to their geo�graphical origin (Fig. 2). Accessions from Maputo aregenetically more dissimilar than those collected fromother districts of Mozambique. This result is furthersupported by PCO analysis, which combined accessesfrom Maputo into two distinct groups. Furthermore,estimation of gene diversity (Table 2) showed that theavailable genetic variability in Maputo and Morrum�bene is higher than that of Xai�Xai and Inhambane,Low variation in the later population might be due toHelianthus introgression hybrids between H. argophyl�lus and H. debilis ssp. cucumerifolius, as previouslyreported by [3].

Total genetic diversity (Ht) was higher among thanwithin populations (Table 3). This could be due togenetic drift, small sample size or to sampling errorscaused by the fact that accessions were selected tomaximize diversity. Even so, the proportion of totalvariation associated with interpopulational differenti�ation was relatively high (Gst = 0.597), indicating that

5.000 4.000 3.000 2.000 1.000 0

Inhambane

Est. IB

Maciene

Vaz

Morrumbene

Cocane

Pnte C. Sol

FAEF

Av. J. Nyere�re

B. Pescadores

Fig. 4. Genetic relationships between the 10 subpopulations of wild Helianthus sp. from Mozambique based on Nei’s (1978) unbi�ased genetic distance.

974


RIBEIRO et al.

a substantial proportion of genetic diversity (59.7%)lies among individual accessions. These results con�trast with the high level of genetic variation (71%)observed within populations of wild and cultivatedsunflower from Maputo using AFLP markers [4] andmay reflect the loss of genetic resources over the years.However, because RAPD markers are dominant wecannot exclude that the amount of genetic variation isunderestimated.

The estimated gene flow for Inhambane andMaputo populations (Nm = 2.290 and 1.343, respec�tively) shows that a moderate gene flow might be suffi�cient to account for the similarity within and amongthese populations. Gene flow between wild Helianthusspecies has been widely reported and occurs regularlyat long distances [19, 20]. Contrarily to Inhambaneand Maputo, in Morrumbene a clear population par�titioning was observed (Table 3). Moreover, the overallgene flow (Nm) was 0.338, which indicates that thegene permutation and interaction among populationswas relatively low.

Additionally, the results showed that wild sunflowerexhibits geographically structured genetic variation(Table 4), as accessions from Maputo were geneticallydivergent from the others. In Maputo the sampledpopulations were distanced not more than 5 km. Thus,some degree of genetic flow is likely to happen. Wehypothesize that the disturbed environment that hasoccurred in Maputo due to human activities and urbanpressure has contributed to increase genetic differenti�ation.

In conclusion, the genetic variation in wild sun�flower species was relatively high. The proportion oftotal genetic diversity residing among these popula�tions should be kept in mind to devise different con�servation strategies in order to preserve these popula�tions. Human activities and subsequent environmen�tal disturbance can potentially alter the spatialdistribution of genetic diversity and accelerate its loss,which should be prevented. Presently wild Helianthusgenetic resources present in Maputo and Sofala are onthe edge of extinction mainly due to excessive urban�ization. Therefore, conservation of what remains ofthis plant genetic diversity is essential for sustainableutilization.

RAPD markers are likely to provide adequate levelsof resolution for comparison among closely relatedaccessions or populations [19, 21–23] and to charac�terize plant genomes [24–26]. Our results were clearand reproducible, showing that this simple and quickmolecular technique is reliable for purposes like ours.To our knowledge this is the first work that responds tocurrent and practical concerns about sunflower inMozambique, integrating the establishment and eval�uation of a small germplasm bank and recovering thescarce wild Helianthus resources. These can be furtherevaluated for their ability to cope with environmentalconstraints and used in future breeding programs.

ACKNOWLEDGMENTS

The authors thank, Dr. Luís Neves, Director ofBiotechnology Center/Universidade Eduardo Mond�lane for providing the lab facilities, Paula Alves andPatrícia Santos, from ECO�BIO/Tropical ResearchInstitute for technical help and Miguel Sequeira, Uni�versidade da Madeira, for helping with the PCO. Thiswork was financed by Gabinete de Relacões Interna�cionais da Ciência e do Ensino Superior/UniversidadeEduardo Mondlane (Proc) no. 4.1.3/UEM) and Insti�tuto Português de Apoio ao Desenvolvimento (Proj.no. 0136/MA/01).

REFERENCES

1. Seiler, G.J. and Rieseberg, L.H., Systematics, Origin,and Germplasm Resources of the Wild and Domesti�cated Sunflower, Sunflower Technology and Production,Schneiter, A.A., Ed., Madison: ASA/CSSA/SSSA,1997.

2. Faria, M.T. and Manjate, M., Evaluation of Wild Sun�flower Populations: Preliminary Study, Proc. Interna�tional Symposium on Sunflower and Other Oil Seed Cropsin Developing Countries, Maputo, 1999, pp. 174–177.

3. Vischi, M., Nonino, F., and Olivieri, A.M., AFLPMarkers for the Study of Introgression between WildSunflower Species (H. argophyllus and H. debilis ssp.cucumerifolius) and in Relation with H. annuus, Helia,2002, vol. 25, pp. 47–58.

4. Quagliaro, G., Vischi, M., Tyrka, M., and Olivieri, A.M.,Identification of Wild and Cultivated Sunflower forBreeding Purposes by AFLP Markers, J. Hered., 2001,vol. 92, pp. 83–42.

5. Vischi, M., Di Bernardo, N., Scotti, I., et al., Compar�ison of Populations of Helianthus argophyllus andH. debilis subsp. cucumerifolius and Their Hybrids fromthe African Coast of the Indian Ocean and the USAUsing Molecular Markers, Helia, 2004, vol. 27,pp. 123–132.

6. Williams, J.G.K., Kubelik, A.R., Livak, K.J., et al.,DNA Polymorphisms Amplified by Arbitrary Primersare Useful as Genetic Markers, Nucleic Acids Res.,1990, vol. 18, pp. 6531–6535.

7. Gouveia, M.M.C., Ribeiro, A., Várzea, V.M.P., andRodrigues C.J., Jr., Genetic Diversity in Hemileia vas�tatrix Based on RAPD Markers, Mycologia, 2005,vol. 97, pp. 393–404.

8. Rohlf, F.J., NTSYSpc, Numerical Taxonomy and Multi�variate Analysis System: Version 2.2, New York: ExeterSoftware, 2005.

9. Yeh, F.C., Yang, R.C., and Boyle, T., POPGENE. TheUser Friendly Software for Population Genetic Analysis,Molecular Biology and Biotechnology Center, Univ.,Allerta, Canada,, 1997.

10. Nei, M., Analysis of Gene Diversity in Subdivided Pop�ulations, Proc. Natl. Acad. Sci. U.S.A., 1973, vol. 70,pp. 3321–3323.

11. Lewontin, R.C., The Apportionment of Human Diver�sity, Evol. Biol., 1972, vol. 6, pp. 381–398.



12. Nei, M., Estimation of Average Heterozygosity andGenetic Distance from a Small Number of Individuals,Genetics, 1978, vol. 89, pp. 583–590.

13. Miller, M.P., Tools for Population Genetic Analysis(TFPGA) 1.3: A Windows Program for the Analysis ofAllozyme and Molecular Population Genetic Data, 1997,Computer Software Distributed by Author.

14. STATISTICA for Windows (Data Analysis Software Sys�tem), Version 7.1, Tulsa: StatSoft, 1998.

15. Olivieri, A.M., Quagliaro, G., Vischi, M., and Amici, S.,Selection of Segregants of Sunflower from Varieties andHybrids Freely Crossed during Few Generation, Proc.Internat.Symposium on Sunflower and Other Oil SeedCrops in Developing Countries, Maputo, 1999, pp. 221–228.

16. Heiser, C.B., Hybridization in the Annual Sunflowers:Helianthus annuus × H. argophyllus, Am. Midl. Nat.,1951, vol. 85, pp. 64–72.

17. Heiser, C.B., Hybridization in the Annual Sunflowers:Helianthus annuus × H. debilis var. cucumerifolius, Evo�lution, 1951, vol. 5, pp. 42–51.

18. Heiser, C.B., Variation and Subspeciation in the Com�mon Sunflower, Helianthus annuus, Am. Midl. Nat.,1954, vol. 51, pp. 287–305.

19. Arias, D.M. and Rieseberg, L.H., Gene Flow betweenCultivated and Wild Sunflowers, Theor. Appl. Genet.,1994, vol. 89, pp. 660–665.

20. Faure, N., Serieys, H., and Bervillé, A., Potential GeneFlow from Cultivated Sunflower to Volunteer, Wild

Helianthus Species in Europe, Agric. Ecosyst. Environ.,2002, vol. 89, pp. 183–190.

21. Lawson, W.R., Henry, R.J., Kochman, J.K., andKong, G.A., Genetic Diversity in Sunflower (Helian�thus annuus L.) as Revealed by Random AmplifiedPolymorphic DNA Analysis, Aust. J. Agric. Res., 1994,vol. 45, pp. 1319–1327.

22. Popov, V.N., Urbanovich, O.Y., and Kirichenko, V.V.,Studying Genetic Diversity in Inbred Sunflower Linesby RAPD and Isozyme Analyses, Russ. J. Genet., 2002,vol. 38, no. 7, pp. 785–790.

23. Fedorenko, O.M. and Gritskikh, M.V., Diversity ofArabidobsis thaliana (L.) Heynh. Natural Populationsat the Northern Limits of the Species Range: RAPDAnalysis, Russ. J. Genet, 2008, vol. 44, no. 4, pp. 425–428.

24. Sossey�Alaoui, K., Serieys, H., Tersac, M., et al., Evi�dence for Several Genomes in Helianthus, Theor. Appl.Genet., 1998, vol. 97, pp. 422–430.

25. Levina, E.A., Adrianova, I.Yu., Reunova, G.D., andZhuravlev, Yu.N., Genetic Variability and Differentia�tion in the Larch Populations within the Range of Larixolgensis A. Henry in Primorye, Russ. J. Genet., 2008,vol. 44, no. 3, pp. 320–325.

26. Martirosyan, E.V., Ryzhova, N.N., Skryabin, K.G.,and Kochieva, E.Z., RAPD Analysis of Genome Poly�morphism in the Family Lemnaceae, Russ. J. Genet.,2008, vol. 44, no. 3, pp. 417–422.

Documents

Population structure and genetic diversity of wild Helianthus species from Mozambique