L

7th International ~r

•in—.-

5 - 8 June

•y^flljjj

fcillESL u&lkl

•&IK : \\i

pniA prui^RroiH B ••fiL.'

v. ,4R

Organized by

Le Groupe de Physioiogie etPathologieVegetales

Faculte des S c i e n c e s

- Nant

AN ANALYSIS OF GENETIC VARIATION IN NATURAL POPULATIONS OF OROBANCHE FOETIDA FROM SPAIN AND TUNISIA B. Roman1, C. Alfaro1, A. Torres1, Z. Satovic2, M. Kharrat3 A. Pujadas4, and D. Rubiales5 1 CIFA. Dep. MejorayAgronomia.Apdo. 3092,14080 Cdrdoba, Spain, [email protected] 2 Faculty ofAgric. Dep. Seed Science and Technology. Svetosimunska 25,10000 Zagreb, Croatia. 3 WRAT, Food Legume Lab., BP10, 2049, Ariana, Tunisia 4 UCO-ETSIAM Dep. Ciencias y Recursos Agricolas y Forestales., Apdo 3048, 14080. Cdrdoba, Spain. 5 CSIC'Instituto de Agricultura Sostenible, Apdo. 4084,14080 Cdrdoba, Spain. Keywords: Orobanche foetida, RAPD, AMOVA, genetic diversity SUMMARY RAPD markers were used to study variation among Orobanche foetida and Orobanche foetida ssp. lusitanicus populations growing either on wild or cultivated hosts. O. foetida populations from agricultural environments were collected on chickpea (Cicer arietinum L.) and faba bean (Vicia faba L.) fields in Tunisia. O. foetida and O.foetida ssp. lusitanicus populations on wild hosts were sampled in Spain. Jaccard's coefficient of similarity was calculated considering 126 RAPD amplification products and populations were grouped by cluster analysis using the UPGMA method. Although the AMOVA analysis revealed a substantial internal variation among individuals within Tunisian O. foetida populations (75.4%), there was a significant divergence between the two hosts considered (24.6 %). The dendrogram also shows a clear;differentiation between O. foetida populations infecting different hosts (C. arietinum and V. faba) and among populations infecting crops and wild flora. INTRODUCTION The existence of O. foetida Poir. in legume fields in the central North of Tunisia is of relevant importance because this species has now attacked V. faba and C. arietinum while it was only parasitizing wild hosts until recently (Kharrat et al., 1992). The parasite is more aggressive on faba bean than on other food legume crops and only peas (Pisum sativum L.) escapes to O. foetida attack. Since the appearance of O. foetida in crop fields in Tunisia some studies have been developed in order to identify sources of resistance among an ICARDA germplasm collection (Kharrat and Halila, 1994), to test faba bean varieties against different inoculum levels (Kharrat et al., 1994) or to evaluate different chemical control strategies (Kharrat and Halila, 1996). However the pattern of molecular diversity among these populations have not been studied yet. In Andalusia (Southern Spain), where O. crenata causes large yield losses in faba bean fields, O. foetida only attacks wild legumes such as Astragalus lusitanicus, Scorpius muricatus, Trifolium angustifolium, Lotus spp. and Melilotus spp. (Pujadas, 1999). The allogamous behaviour of O. crenata could lead to interspecific crosses that could be monitored by molecular markers. The relatedness among the Orobanche species occurring on natural plant species and Orobanche attacking crop plants needs to be determined in order to have a better knowledge of the evolution of this species from wild parasitic plants into aggressive parasitic weeds. The aim of the present study is to determine the genetic relationships among O. foetida populations on V. faba and C. arietinum fields from Tunisia and O. foetida and O. foetida ssp. lusitanicus populations on wild genotypes from Southern Spain using RAPD markers.

57

Plant Material • 9'«t# K*,** * ^ O. foetida populations from agricultural environments were^coil^^^cnickpea (C. arietinum L.) and faba bean (V. faba L.) fields in Beja (Tunisia). Spanish (Xijoetida and O. foetida ssp. lusitanicus populations on wild hosts were sampled in the province of Cordoba and Granada respectively. The number of Orobanche plants per populations analysed varied from 10 to 21 depending on the availability of specimens in each case. RAPD analysis Floral buds were used for DNA extraction using the method proposed.by Lassner et al. (1989), modified by Torres et al. (1993). For RAPD analysis, approximately 20 ng of genomic DNA was used as a template in a 25 /d volume per PCR reaction. Mixture composition and reaction conditions were as described by Williams et al. (1990) with slight modifications (Torres et al., 1993). A total of 5 RAPD primers (OPB-03, OPE-17, OPJ-01, OPJ-13 and OPV-09), purchased in commercially available kits from OPERON Technologies (Alameda, USA), were analysed. Statistical analysis The analysis of molecular variance (AMOVA) was used to partition the total phenotypic variance into within populations and among populations. The AMOVA was performed using WinAMOVA ver. 1.55 software (Excoffier, 1992). Distance among individuals was measured as an Euclidean metric distance that was calculated between all possible pairwise combinations of molecular genetic markers (RAPD bands) for individual plants. The variance components were tested statistically by non-parametric randomisation tests using 1000 permutations. A non-parametric test for the homogeneity of molecular variance (HOMOVA) based on Bartlett's statistics was performed to test variance homogeneity among populations (Stewart and Excoffier, 1996). Bartlett's statistics null distributions were obtained after 1000 permutations. Jaccard's similarity coefficient was computed using SYSTAT® 7.0 software package. A cluster analysis based on the similarity matrix was performed using UPGMA method and the dendrogram was obtained in order to visualise the relationships among single individuals. RESULTS AND DISCUSSION The 5 RAPD primers analysed generated 126 polymorphic, clear and reproducible bands that were used in the population genetic analyses. The number of bands per primer varied from 14 to 39 with an average of 25.2 bands/primer. The proportion of polymorphic loci varied among populations with the highest value in O. foetida population infecting faba bean (0.921) and chickpea (0.571) and the lowest in O. foetida (0.198) and O. foetida spp. lusitanicus (0.114) on wild hosts. None of the populations displayed unique bands. Hierarchical analysis of phenotypic diversity using AMOVA was performed to analyse the partitioning of the variation between the two regional groups (Tunisia and Spain) as well as among populations within the groups and among individuals (Table 1). In spite of a substantial internal variation among individuals within Tunisian O. foetida populations (75.40%), there was significant divergence among them (24.60 %;<)> = 0.246; p < 0.001). This fact shows a clear differentiation between O. foetida populations infecting different hosts (C. arietinum and V. faba). The corresponding HOMOVA analysis reveals tat the molecular variances were also significantly heterogeneous among populations (Bp = 1.438, p < 0.001). In the case of the Spanish populations the phenotypic divergence between them was higher (67.90%) since they belonged to different | = 0.679; p < 0.001) as well as the significant Bartlett's test subspecies. Highly significant § value (<> (B = 0.432, p < 0.001) were a clear evidence for the existence of presumed levels of genetic differentiation between populations. Jaccard's similarity coefficient varied from 0.027 to 0.851 between different pairs of individuals. The dendrogram obtained by the UPGMA method revealed similar relationships among samples showing a clear separation between different cultivated hosts parasitized of Tunisian O. foetida populations and among Tunisian and Spanish populations. 58

In this study we determined whether individual broomrape plants collected from one host were more closely related to one another than to broomrape plants of the same species collected from a different hosts in the same region. We have found a clear and significant host-differentiation between O. foetida populations infesting faba bean and chickpea fields in Tunisia. These results differ from those obtained by Paran et al. (1997) where all the diversity found within O. aegyptiaca and O. crenata populations was among individuals and not among hosts. Verkleij and Pieterse, (1994) also suggested that geographic separation was more important than host specialisation with respect to allozyme variation in O. crenata populations from Syria and Spain, and by Bharathalaksmi et al (1990) in Striga hermonthica from sorghum and maize. Similar results were obtained with ISSR markers by Roman et al (2001) with O. crenata populations. However, Musselman (1986) suggested that the host of a parasitic plant such as Orobanche may play a role in the variability of the parasite. This fact was evidenced by Musselman and Parker (1982) where plants collected from one host and grown on a second one differed from each other in their morphology. Table 1. AMOVA and HOMOVA analysis for the partitioning of RAPD variation among regions (Tunis vs. Spain), among populations (within regions) and within populations as well as among and within each of two regions Source of variation Df Variance % Total $P-value Bartlett' P-value componen varianc statistics s index ts e (B) Among regions (Tunisia vs. 1 3.38 < 0.001 < 0.001 6.590 0.180 18.02 Spain) 5.01 Among populations / Within 2 0.325 26.68 < 0.001 15.257 < 0.001 regions 55.30 0.447 53 10.37 Within populations < 0.001 Among populations (Tunisia) Within populations

1 34

4.79 14.67

24.60 75.40

0.246

< 0.001

1.438

< 0.001

Among populations (Spain) Within populations

1 19

5.70 2.69

67.90 32.10

0.679

< 0.001 0.432

< 0.001

The within population variation was clearly higher in the O. foetida populations attacking crops (75,4%) than on O. foetida populations on wild flora (32.1%). The higher phenotypic diversity within Tunisian O. foetida populations than between them is consistent with the distribution of variation in allogamous species (Shoen and Brown, 1991) whereas the lower intrapopulation divergence in O. foetida samples on wild hosts is consistent with an autogamous behaviour. These results are expected since an autogamous species has a higher chance to survive and maintain in isolation than an allogamous one. Although the distance between the wild populations and the agricultural ones may hint to possible source of the-weedy populations, the geographical distance (Tunisia and Spain) could be masking the genetic distances among them. O. foetida populations parasitizing various wild legumes in the Tunis and Cap-Bon areas have been reported by Pottier-Alapetite (1981). In future studies it will be of great interest to consider populations on these hosts in order to study the evolution from wild flora to crops parasites in this species.

REFERENCES BHARATHALAKSMI, WERTH CR and MUSSELMAN U (1990). A study of genetic diversity among host-specific populations of the withweed Striga hermonthica (Del.) Benth. (Scrophulariaceae) in Africa. Plant System and Evol 17: 1-12 EXCOFFIER L (1992). WinAMOVA ver. 1.55 http://anthropologie.unige.ch/LGB/software/win/amova/ 59

bean in Tunisia. FABIS Newsletter 30: 46-47 . , a 48&M&0i11l&§0#£& in Tunisia: problems and KHARRAT M and HALILA MH (1994). Orobanche species on feba bcanlWicit International Workshop on management. Biology and management of Orobanche. In: Prbceedin^^t^tne' Orobanche and related Striga research. Amsterdam, The Netherlah^,pp:!639-643l KHARRAT and HALILA (1996). Control of Orobanche foetida oiiVicidfdbaiC^^^isovL between different control measures. In: Proceedings of the 6th International Symposium on Parasitic Weeds, G6rdoba, Spain, pp: 734-737 LASSNER MW, PETERSON P and YODER JI (1989). Simultaneous amplificationof multiple D I ^ fragments by polymerase chain reaction in the analysis of transgenic plants and their progeny. Plant Mol Biol Rep 7:116-128 MUSSELMAN LJ ana* PARKER C (1982). Preliminary host ranges of some strains of economically important brpornrap^ V ^ MUSSELMAN LJ (198^). Taxonomy of Orobanche. In: Proceedings of the^Workshop on Biology and control of OrpibflTic/ie, Wageningen, The Netherlands, pp: 2-10 t «* PA^AN I,^GIDONI D and JACOBSOHN R (1997). Variation between and within broomrape (Orobanche) species revealecf by RAPD markers, Heredity 78:68-74 ^^ , || POTnmkAp^PETm (1981). Eloredela Tunisie. ^giospermes, D i c p t y l ^ Piiblications Scieritifiques Tunisiennes, Nfinistere de KAgjiculture. vk , PUJADAS A (1999):Resistancei to Orobanche: The state of the art, Junta Andalucia, Spain, pp: 187rl93 ROMAN B, SATOVIC Z, RUBLALES D, TORRES AM, CUBERO JI, KATZIR N and JOELfDM (2001). Genetic dicersityjn Orobanche :crenaia*$opu\ations Spain and Israel. In: 4th European Conference on Grain legumes, Cracow. SCHOEN DJ and BROWN HD (1991). Intraspecific variation in population gene diversity and effective population size correlates with the matting system in plants. Proc Natl Acad Sci USA <88:4494-4497 STEWART CN and EXCOFHER L (1996).^Assessing^ population stmcture and variability with RAPD data: Application to Vaccmiiun macrpcaropoh (American cranberry). J. Evol. Biol. 9:153-171 ";_ TORRES AM, WEEDEN .NF aria" MARTfN A(1993). linkage among isozyme, RFLP and RAPD markers in Vicia faba. TheorAppl Genet 85:937-945 > , VERKLEIJ JAC and PIETERSE AH (1994). Genetic variability of Orobanche (broomrape)and Striga (witchweed) in ^ ,;ro Workshop on Orobanche and related Striga research.-Arnkerdain^ TieiNemerlands, pp. 67-79 i WILLMMS K 5 ! | ^ a J B E I J K | ^ UYAK KJ, R A F A L ^ polimorphisms amplified by arbitrary primers arejuseful as genetic markers: Nucleic Acids Res 18:6531-6535 J :§

Figure 1: UPG^^SendrogTam using Jaccard-s genetic distances among O, foetida mdividuals from Tunisia (T-O^foetidd on Vicia faba; T-O. foetida on Cicer arietinum) and Spain ( S-O. foetida ssp. lusitanicus and S-O. foetida). T-O. foetida foetida foetida T-O. foetida T-Q. foetida 1-0.-foetida

on Vf on Vf on Vf on Vf on Vf on Vf

Vf

T-O. foetida on Vf T-O. foetida on Vf T-O. foetida on Vf T-O. foetida on Vf T-O. foetida on Vf T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-Q. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca T-O. foetida on Ca S-O. foetida ssp. lus. S-O. foetida ssp. lus. S-Q. foetida ssp. lus. S-O. foetida ssp. lus. S-O. foetida ssp. lus. S-O. foetida ssp. lus. S-O. foetida ssp. lus. S-O. foetida ssp. lus. S-O. foetida ssp. lus. S-O. foetida ssp. lus. S-O. foetida ssp. lus. S-O.foetida S-O. foetida -O.foetida -O.foetida S-O.foetida S-O.foetida S-O.foetida S-O.foetida S-O.foetida S-O.foetida

=9 ?3l

f

T 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

-•

60

1 0.9