GENETIC RESEMBLANCE

A. V. VRÂNCEANU and F. M. STOENESCUResearch Institute for Cereals and IndustrialCrops, 8264 Fundulea, Romania

INTRODUCTION

Genetic diversity is considered one of themost important factor securing crop stabiiity.A large genetic diversity would reduce theyield fluctuation in a certain area because dif-ferent genotypes react differently to the envi-ronmental variation. Also, it is well known thatthe appearance and spreading of the new,more virulent races of the pests are positivelycorrelated with the area occupied by the cor-responding genes for resistance.

A greater genetic diversity at territorial Ie-vel may be achieved in sunflower by meansof a higher number of different hybrids butthe numbey per se does not constitute an ade-quate solution, taking into account that certainhybrids could be genetically related and thecultivation of more alike genotypes may beequivalent to the extension of a single one.

Presently sunflower breeders seeking yieldimprovement are generally faced with a rela-tive vast array of potential parent genotypesamong which only a limited number of crossescan be made in order to obtain new valuableinitial breeding material. The choice is usuallybased on a combination of yield data, visualassessment of agronomic characteristics andpest resistance, and very seldom the pedigreeknowledge. Genetic investigations based onlestcrosses are costly and so quite rare. It is gene-rally assumed that maximum improvement ofthe initial material can be obtained from cros-ses which release the maximum amount ofvariations between parents which are geneti-cally diverse.

From this point of view, our attempts toevaluate the genetic relationship among hy-brids, which constitute the aim of this paper,could be of a certain usefulness.

MATERIALS AND METHODS

Thc identification of the genotype structuresfitting properly the ecological conditions ofdifferent regions has been one of the mainpurposes of the F.A.O. Research Network onSunflower. All sunflower eultivars tested since

OF SUNT'I.,OWER CT]LTIYARS TESTEI)11\ INTBR]\ATTONAL TRIALS

the Network's establishment have representedthe latest achievements of the main sunflowerbreeding centres from all over the world(Annex 1). Trials have been conducted with101 Fl hybrids and 13 open pollinated varie-ties (OPV) in four biennial cycles, startingfrom 1976, under a wide range of environmentsfrom Europe, Near and Middle East, Africa,Latin America and U.S.A. and their resultswere published in the Information BulletinHELIA, numbers I, 2, 3, 5 and 7.

The estimation of the relationship degree ofsunflower cultivars on genealogy basis requiresthe knowledge of parental genetic origin, whichis not always accessible. Such an approach hasalso the inconvenience that divergent inbredscould be often selected from the same sourcematerial, while similar ones mav arise fromdistinct origins.

The genetic resemblance estimation on thebasis of common allele percentage is also diffi-cult and relative for this study, because thenumber of the identified genes in sunflower isstill small and the existence of their alleles inthe tested genotypes has been only approxi-mately investigated.

A more complete assessment of the geneticresemblance has been done for some groups ofcereal cultivars, on the basis of their vield res-ponse to a large range of environments.. {ohnson (1977), working with corn hy-brids, elaborated a model which providêsweighted estimates of hybrid means, coeffi-cients of the regression of each hybrid on anenvironmental index and coefficients of ther:egression of each hybrid on a set of ortho-gonal residu,al vectors. Genotypic similaritywas defined as the euclidian distance betweengenotypes in the space whose coordinate axeswere the locations. A claster analysis arrangedthe hybrids into groups that were differeniia-ble in terms of means and stabilitv.

Habgood (19?7) proposed a method ofestimating the genetic diversity of self-fertili-zing cereal cultivars based on genotype-envi-ronment interactions. The degree of similarityor diversity between each pair of genotypes inan array is estimated by the correlation over

Sunflower cultivars tested intrials in the period

Supplyingcountry

Argentina(Continental)

Bulgaria(GeneralToshevo)

France(INRA)

Hungary(Iregszemcse

andSzeged)

Italy

Poland(Poznan)

Romania(Fundulea)

Spain(INTA)

Biennialcycles

1980-1981

1976-1977

1978-19?9

1982-1983

1976-1977

1978-1979

1980-1981

1982-1983

1978-1979

1980-1981

1982-1983

1978-19?9

1980-1981

1982-1983

1978-1979

1980-1981

1976-19?7

1976-1977

1978-1979

1980-1981

1982-1983

1976-197?

1978-19?9

1980-1981

10

Annex I

the F.A.O. internationalof 19?6-1983

Cultivars (SH:single hybrids,'IH:three-way hybrids,

)PV:open pollinated varieties)

Contiflor (SH)

Helios 322 (SH)

Hemus (OPV), HB-451(SH), Peredovik (OPV)

HB-?83 (SH)

Airelle (SH), Relax (SH),Remil (SH) Issanka (OPV)

Remil (SH), Luciole (SH),rNRA-7702 (SH)

Primasol (SH)

H9P1(SH),H9P2(TH),H9P4(1H)Sorex (SH), Olga II (SH)

H-27-77 (SH), Marika (SH),Vera r(SH)

G 9-76 (SH), G 19-?7G 24-77 (SH)

IH-10 (OPV), Iregi(OPV), Iregi Csikôs

Gahib 6 (SH)

(SH),

816 B(oPv)

HNK-81 (SH), HNK-84(SH), Koflor 1 (SH), IH-56(SH), IH-155 (SH), Gahib ?(SH), Citosol 2 (SH), Cito-sol 3 i(SH)

Argentario (OPV)

Ala (OPV)lffielkopolski (OPV)

Romsun 52 ,(SH), Romsun53 (SH), Romsun 59 (SH),Sorem B0 (SH), Sorem Bz(SH), Sorem HT-64 (TH)RO-18 (SH), RO-20 (SH),Romsun 301 (SH).

Record (OPV), Romsun 59(SH), Sorem HT-111 (TH),Sorem HT-116 (TH), So-rem HT-117 (TH), Romsun90 (sH).

RO-18 (SH), RO-19 (SH),RO-33 (SH), RO-26 (SH),RO-27 (TH), RO-34 (SH),RO-29 (TH), RO-40 (SH),RO-45 (SH), RO-46 (SH),Ro-loo (SH), RO-130 (SH)

RO-25 (SH), RO-36 (SH),RO-70 (SH), RO-44 (SH),RO-131 (SH), RO-134 (SH),RO-141 (TH), RO-150 (SH)

H-23 (SH)

Sepasol (OPV), HS-1161(sH), HS-72 M (SH)

Halcon (SH), Pinzon (SH),sH-s-690 (sH), sH-3000x2(HT), SH-72 MX1161(HT)

U.S.A.

Yugoslavia(Novi Sad)

P.O.I. 301 A(sH), H-2413BO A (SH)

H-894Sungro

Sunbred 265 (SH), Interst.77?5 (SH), DO-704 (SH)

Sunbred 254 (SH), Cargill205 (SH), Seedtec S-315(SH), Stauffer 3101 (SH)

YU-NS-65 (SH), YU-NS-I(SH)

Novi Sad 20 (OPV), NoviSad 61 (OPV), VNIIMK8931 (OPV), NS-H-27 (SH),NS-H-34 (SH), NS-H-63RM (SH)

NS-H-10 (SH), NS-H-l1(sH), NS-H-13 (SH), NS-H-17 (SH), NS-H-33 (SH),NS-H-36 (SH),

NS-H.3 (SH), NS-H-4 (SH),NS-H-5 (SH), NS-H-40(sH), NS-H-42 (SH), NS-H-43 (SH).

a range of environments of their respectivedeviations from the mean yield of aII genoty-pes in each environment.

Sàulescu et 41. (1981) used the coeffi-cient of correlation (r) and the coefficient ofdetermination (r2) between the yields of eachcultivar pair in two large sets of winter wheatyield tests. The coefficient of correlation hasthe advantage of indicating directly the pairsof genotypes with the highest yield stability.It is known that the mean variance of twovariables depends not only on the variance ofeach of them but also on the magnitude of thecoéfficient of correlation between them :

' s,f"+I]:ls1*ls?.1-l'.S* Sv. so, ast2l4."^4*r2

"rrr.U", than 1 is the correlation coefficient (r)

between the yields of two cultivars, the betteris compensated their yield variation in diffe-rent environments and the variance of the ave-rage yields is smaller.

In the present paper, the genetic resem-blance of sunflower cultivars was estimatedin accordance \Mith their seed yields in diffe-rent years and locations, using the square-ofthe càrlelation coefficients between seed yieldsof each pair of cultivars, i.e. the coefficient ofdetermination (12'100), which can be conside-red a weighted estimation for the percentageof identical alleles for aII genes controlling theyield in the given set of environments." Three replications of each trial were usedin calculatiôn and the assessment of the sig-nificance of differences between correlationcoefficients and further between the coeffi-cients of determination lvvas performed by

computing the value, Z: : 1 n (lfr) - t n

(1-r), where 1 n is the na-tural logarithm andr the correlation coefficient.

1978-1979

1980-1981

1982-1983

1976-197?

1978-1979

1980-1981

1982-1983

RESULTS AND DISCUSSION

Data from Tables 1 to B give the coefficientsof determination (12.100) calculated for eachof the eight biennial trials conducted co-ope-ratively in the seasons 1976_1977, lgTB-1g?g,1980-1981 and 1982-1983. Taking into ac-count the similarities in seed yields of eachpair of cultivars under various environmentalconditions (16-41 locations), one could observea large amplitude of the estimations of geneticresemblance, going from 8-84% in the firstbiennial cycle to 52-gï0lo in the fourth bien-nial cycle.

Both similar responses and therefore a grea-ter genetic resemblance and uncorrelated res-ponses and therefore great genetic differencesamong the pairs of cultivars can be establis-hed. An important number of cultivars fallsinto an intermediate group of a less strikingresemblance.

It should be noted that almost all values ofthe coefficients of determination were signi-ficantly different from 100, meaning that anypair of cultivars will give more stable averageyields than each cultivar taken individually.

Diagram representation of the coefficients ofdetermination, in which their magnitude isconventionally marked by the thickness of thelines, allows to obtain a better image of thegenetic relationships among the cultivars tested(Figures 1 to B). Values of the coefficients ofdetermination less than 7b were considered asindicating uncorrelated responses and there-fore more or less distinct genotypes.

Helros 3 2 2Sorpm HT- 64

Sorem 82 o' r Airetle

R0-59 o

R0-s3R0-52

lssonkoa

Romsun 301 ?-_ Wietkopolski

i

R0-?0i/

YU- NS -1

oR0-18

Geretic simitority:

tovt ('/5 <r.2, 100<gO)

Fig. 2 - Diagram representation of genetic relation_ship of sunflorver cultivars from Trial No. 2, 19?6-19??

NS-61

Argenlorio

Gerntic simitority:

low (75<rZ . 100<80)

-

rædrum ^(01 a 72 . rcg < g5J, high (r2 . loo > 86t --r

F-iq. 3 -- Diagram representation of genetic relation-ship of sunflower open pollinated varieties from Trial

No. 1, 1978-19?9

Surgro380 A

H-2t a

H-094

Fig. 4 - Diagramship of sunflower

HS-1161CrerPiic sinrilority,

..-.--.-,-- low (75 < f2 . ilo< 80]

-,

medium {81< f 2 100 < 85}

-

high ( 12. 100>86)

representation of genetic relation-hybrids from Trial No. 1. 19?8-19?9

Genefic simitorify,

___-_..--._ low (75<r2 100<g0l._--.-- medrum (81 <12.100<B5l

Fig. 1 - Diagram representation of genetic relation-ship of sunflower hybrids from Trial No. 1, 19?6-19??

HB -1s1

R0- 22:"'-------:i;*o./-\" NS-H-13

R0-26 1t:::-'!""--'--'

R0-34

NS-H-10 iNS - r1-17

oerntic similoriiy:

--..------. low (?S < f2, 100 <90 )

-

medium (81 <r'. 100<85)

-

hrgh (1. 100>86)

îig, 5 - Diagram representation of genetic relation-ship of sunflower cultivars from Trial No. 1' 1980-1981

d 265

NS-H-36

lnierst 77?5

R0-46m'701+

-sH-3000 x2

'1161 6enetic simitority:_ f;ilïî,,i;d3;'l*."'îig, 6 * Diagram representation of genetfc relation-ship of sunflower hybrids from Trial No. 2, 1980-1981

A similar response and 'therefore a greatergenetic resemblance showed the following pairsand groups of cultivars :

Trial No. 1 (19?6-19?7) :

YU-NS-65/Relax, Romsun 52lRomsun 53, Rom-sun 53/Sorem HT-64 (Table 1).

Trial No. 2 (1976-7977) :

YU-NS-1/H-23, Wielkopolski/RO-30 (labie 2).

Trial No. 1 (i978-1979) :

Argentario/Sepasol/Record, NS-61/Peredovik/Argentario, NS-2O/Argentario/Sepasol (Table 3).

Trial No. 2 (1978-1979) :

H-894/Romsun 90/NS-H-63/HS-1161/P.O.I. 301

A, Sorex/Olga II, Sorem HT-l1?/Sorem HT-l16/HB-451, NS-H-34/Sungro-80 A (table 4).

Trial No. 1 (1980-1981) :

Citosot

Genetk simltority:

- rnedium^(80<r2' t00<85)

- high (rz 100>861

Fig,7 - Diagram representation of genetic relation-ship of sunflower hybrids from Trial No. 1, 1982-1983

NS-H-Il

NS.H.3

Stouffer 3101

Seedtec S-31

Corgill

HNK- 81

citosot 2

6enetic srmitoritY'

- ffiiïpl,'&ïl;i*'u''

Fig. I - Diagram representation of genetic relation-ship of sunflower hybrids from Trial No. 2' 1982-1983

Marika/Ver a/RO-27 / Ala/}J-27 -7 7, NS-H-I 0/NS-H-1?, NS-H-11/RO-34, RO-22/ALa/RO-19, RO-26lPrimasol, RO-18/RO-26 (Table 5).Trial No. 2 (1980-1981) :

IH-10/Iregi 816 B/Halcon/Gahib 6, HaIcon/SH-3000 X 2/Pinzon/SH-?2 M X 1161/SH-S-690/RO-l 30/Do-704/Interst. 7775, Sunbred 265/Pin'zon/DO-704/Interst. 7775, RO-46/SH-3 000 X 2'RO-29/SH-S-690 (Table 6).Trial No. 1 (1982-1983) :

H9-P2/KofIor 1/NS-H-4/NS-H-5/HNK-81, G9-76lGr\-77 /HNK-81/HNK-84, RO-36/G19-77 /G24-77, NS-H-3/CitosoI 3/Koflor l/NS-H-4/NS-H-5, RO-25/HNK-84 (Tab1e ?).Triai No. 2 (1982-1988) :

H I P 4/NS-H-42/NS-H-40/Cargill 205/Sunbred254lCitosol 2, IH-56/CargiII 205/Gahib 7, HB-783/Seedtec S-315/Gahib 7, Seedtec S-315/IH-155 (Table B).

R0-29 . R0-40

R0-150 R0- 141

Toble 1

Éstimation of genetic resemblance (r2.100) among 11 sunflower medium-late hybrids (F.A.O. Trial

Cultivars(rt

1. Helios 322

2. YU-NS-653. Airelle4. Relax5. Remil6. Romsun 52

7. Romsun 53

B. Romsun 59

9. Sorem B0

1.0. Sorem 82

11. Sorem HT-64

47

100

63

16

100

^a

B4

36

100

15

42

15

)l

60

100

62

49

49

57

44

4B

57

100

33

78

11

79

7B

59

76

53

100

37

50

15

39

55

56

67

47

100

33

I69

61

54

83

76

60

100

35

66

15

6B

64

82

100

Table 2

Estimation of genetic resemblance (12'100) among7 sunflower early cultivars (F.A.O. Trial No. 2,

L.S.D. 0.05:9

19?6-19??, 23 locations)

cultivârs(oPv+Fd 1 , 4 7

1. Issanka

2. Wielkopolski

3. YU-NS-1

4. H-23

5. RO-18

6. RO-20

7. Romsun 301

100 53

100

36

61

100

28

69

B6

100

52

63

66

66

100

5B

B4

72

65

62

100

68

79

70

73

66

80

100

A general look over the diagrams of geneticrelationships permits to appreciate that the ge-netic diversity of the present sunflower hy-brids is quite limited and has even decreasedin the last trials, indicating thai many hybridsmay have one inbred line in common or theirparents could originate from the same or simi-lar sources of germplasm such as high oil vari-eties or closely related pollen fertility restorers.

Making a connection between the correla-tions presented in this study and {ata on mor-pho-physiological traits of the respective sun-flower cultivars published in HELIA, numbers\,2,3,5 and 7, one can observe that, contraryto expectation the similitude of reaction toenvironment was not always gre4ter betweengenotypes with similar maturity and plantheight, suggesting that other iharacterlsticscould play an equal important role in the res-ponse of sunflower cultivars to the variationsof soil and climate factors.

Table S

Estimation of genetic resemblance(r2.100)among 11 sunflower open potlinated yarieties (F.A.O.

No. I, 19?6-19??, l9 locations)

Trial No I, 19?8-19?9, 38 locations)

Cultivars(oPv) 2 t I o 10 11

1. Peredovik2. Hemus

3. IH-104. Iregi 816 B

5. Iregi Csik6s6. Argentario7. RecordB. Sepasol

9. Novi Sad 20

10. Novi Sad 61

11. VNTIMK 8931

100 lt

100

72

69

100

76

67

7l100

63

60

6B

61

100

B4

di

'/D

66

67

100

B3

66

tl

63

66

B9

100

B5

60

7l65

60

BB

91

100

B4

63

65

64

63

B9

82

8B

100

B9

tt

53

4B

40

B9

BO

82

82

100

74

82

5B

47

66

74

76

76

'lD

77

100

I,.S.D. 0.05:9

T3

19?8-1979, 41 locations)

Cultivars(l'1) I 4 5 ,l

B o 10 11 1' 15 16 L7 18 19

1. HB-4512. Remil3. Luciole4. INRA-7?025. Sorex6. Olga II7. Sorem HT-1l1B. Sorem HT-1169. Sorem HT-117

10. Romsun 9011. HS-116112. HS-?2 M13. P.O.I. 301 A14. H-89415. H-24116. Sirngro 380 A1?. NS-H-2718. NS.H.3419. NS-H-63 RM

100 66100

.IÔ

7B100

827069

100

657B7674

100

'/D

lalô?386

100

786665776767

100

807l63657T7072

100

B3706B76656963B4

100

7463647l4751157074

100

727l736854It47646678

100

74747l7356725970597781

100

73787866

75:t+7376667L72

100

7572727554525B74.IU

86817983

100

a^

5865

52644966736673616B66

100

?11 661 ool zr641 661 70l 63631 4el 631 66?Bl 6sl ztl zo37i| 47 | 50l 55551 531 431 514el 4Bl 4el 66641 60l 611 6?6el 70l 5el 6B

641 711751e2511 67174178b8l 68l ozl ze73l| 7r | 65 | 637rl 72 | ?6 | 8266 | T2 | 63.i 75

lool 711 831 68

1100 | 65 | 63| 11001 64

| | lloo

fable 4

Estlmation ol genetic resemblance 11b.fOÔ) âmong 1Ô sunflower meilium-iate hybrids (F.A.O. Trial No. 2,

L.S.D. 0.05:8

Table 5

Estimation of genetic resemblance (12.100) among 15 sunflower medium early cultivars (F.A.O. Trial No 1,

Cultlvars(F1+OPV)

1. RO-222. H-27-773. NS-H-134. Marika5. Vera6. NS-H-l?7. NS-H-1l8. PrimasolL RO-19

10. RO-2?11. Ala12. NS-H-1013. RO-3414. RO-2615. RO-18

4869

100

7g100

6179617966

76100

778755

100

B45B28tltti)t344L

100

4l7682505364BB7924614566

100

7L76777767B51269617L74

100

BBB4499088615B6490B6

100

79B45386B667626476

100

J'49594845484650456646645886

100

6277797l667672B5537B677Btt

100

Table 6

Estimation of genetic resemblance among l8 sunflower medium-late hybrids (F.A.O. Trial No. 2, 1980-1981'

1980-1981, 16 locations)

L.S.D. 0.05:?

741 49?61 77?61 B37Ll 64691 61

100 I 6e| 100

22 locatlons)

cultivars(F1) 1 3 4 o 6 B

q l0 ll 1' 13 14 IO 16117118

1. RO-292. RO-403. RO-1304. Sunbred 2655. NS-H-366. Interst. 7?757. DO-?048. NS-H-33L RO-100

10. sH-s-69011. SH-72 MX 116112. Pinzon13. sH-3000x214. RO-4615. RO-4516. Contiflor17. Gahib 618. Halcon

100 64100

7969

100

736279

100

7667867600

77698386B3

100

7966B5908692

100

5B66628261797S

100

67B4797678lt7674

100

8761B469828582627L

100

745882BO

8188B87l7L90

100

1976s2B6B5B89219BO

B1B4

100

7l7287BO

8B85BB778085B692

100

567l76597266675972626476B8

100

3B

5045564944567634355966

100

48l66l6B76170174?01541835Bl?Bl7e7218418568l72l8B68168186661741747715216870l78l84?21?61s87217618770l7elel63163180B4l4Bl51

1ool63l?1lloo I B6| 1100

L.S.D, 0.05:8

L4

îable Ihstimation of geneilc resemblance (r'.roo) among 15 meitlum-early hybrirls (F.a.ô. Trial No. t, toSz-t988,

L.S.D. 0.05:6

Estimation of genetic resemblance 6r.100) among 18 medium-late hybritts (F.A.O. TrialTable I

No. 2, 1982-1983,

36 locations)

Cultivars(F1) 1 2 4 6 I I 10 l1 t2 13

1.H9P12.}I9P23. G 9-76

4. G t9-77

5. G 24-77

6. HNK-817. HNK-84B. Koflor 1

9. Citosol 3

r0. Ro-2511. RO-36

12. RO-?0

13. NS-H-314. NS-H-415. NS-H-5

100 79

100 78

100

?0

81

ot

100

70

lc

B3

BB

100

7B

B5

86

85

79

100

79

75

85

B3

7B

8B

100

79

87

78

82

76

89

76

100

BO

B1

79

78

/b

81

74

BB

100

B1

tt

75

73

68

79

87aÀ

62

100

67

81

B3

92

92

79

75

7B

76

72

100

78175711 B0

761 81

rol 7s

?41 B0

rrl B0

trl 7e

761 88

,nlB76el ??

7sl B0

100 | 5e

| 100

"lBrrtl 85

rnl 75

81l7e80l 83

uul 87

B3l 83

871 86

url B3

rnl B1

821 81

661 70

85l 88

100 |

B6

| 100

39 locations)

CuItivârs(Fr) 1 , 3 4 7 B o l0 ll L2 l4 IO r7ll8

1. HB-783

2.Il9P 4

3. IH-564. IH-155

5. Gahib 7

6. Citosol 2

7. RO-448. RO-131

9. RO-134

10. RO-141

11. RO-150

12. Sunbred 254

13. Cargill 205

14. Seedtec 5-315

15. Stauffer 3101

16. NS-H-4017. NS-H-4218. NS-H-43

100 BO

100

B2

94

100

70

79

79

100

92

BO

91

67

100

75

91

90

74

76

00

62

B3

B5

d^

63

B1

100

BO

82

79

73

87

B3

BO

100

BO

B4

90

B3

90

B1

86

B5

100

7418083152801816e16776

l8777 l7B80176e3l8382

174100

| 77

| 100

82

s4

85

BO

B6

91

BO

97

B4

93

82

00

B5

90

B8

80

9B

94

B3

76

93

B2

80

86

100

92

B1

66

8B

90

74

B3

68

77

69

73

7S

90

100

73

B3

82

73

IL

93

BO

to

77

94

7l93

90

78

00

90

81

70

80

94

80

B3

81

91

7B

93

94

BO

92

00

7L 17787l83

?71867Ll7e7r

177B2lB07r

174Blls+7718383lB27017e7118078l6B?31?078174e2

l7s00167| 100

CONCLUSIONS

Although some groups of sunflower cultivarsdiffering genetically could be established fromthis study, the genetic diversity of the presentcultivated hybrids seems to be quite limitedto fit all environmental variations ànd to mini-mize the genetic vulnerability of sunflower

crops. The limiting factors are connected pri-marily with the utilization of the same type-of cytoplasmic male sterility, the genetic simi-larity of many of the female pàrents ôrigi-nating from high oil varieties and the reducédnumber of pollen fertility restorer lines.

In- order _to, enlarge the genetic diversity ofsunflower hybrids, breeders have to dev-elop

15

sôurce-populations with various genetic back-ground:which would serve as selfing genepoolsand to'identify new c?zr.s and Rf sources frominterspecific crosses.

The ipignificant variability existing amongvarious pairs of hybrids from the standpointof the magnitude of the coefficients of deter-mination allows to choose those couples ofhybrids with the best reciprocal compensationof the yield variations and so to achieve anappropriate structure of hybrids which mightassure a better stability of high average yield1evels.

As it becomes apparent, beside high yieldingcapacity, sunflower hybrids should also be im-proved for their different reaction to the en-vironmental factors.

REFERENCES

H a b g o o d R. M., L977, Estimation of genetic d'ioer'sita of sel!-fertilizing cereal cultitsars based' onggnotgpe-èntsironment interactions, Euphytica,26,2, 485*489.

J o h n s o n G. R., 19?7, Analgsis of genotgpic simi-laritg in terms ol mean gield and stabilitg otenuironmental response in a set of maize hg-brids, Crop Sci., 17, 6, 837-842.

Sàulescu N. N., Tapu C., Ittu Gh., 1981, Ge-netic ilioersitg ol utinter wheat oarietal assort'ment cultitsoted at present in Rom'ania (in Ro-manian), Probleme de geneticà teoreticà çi apli-catà, XIII (2), 65-80.

LA SIMILITUDE GÉNÉTIQUE DES CULTIVARS DETOURNESOL EXPÉRIMENTES DANS LES ESSAIS

INTERNATIONAUX

Résumé

La similitude génétique de 101 hybrides et 13 cul-tivars de tournesol a été étudiée, ceux-ci étant testésdans le cadrg du Réseau de recherches de Ia F.A.O.pour le tqurnesol pendant 1976-1983, dans un grandnombre de tqcalités d'Europe et hors Europe. La si-militude génétique a été estimée en utilisant le carrédes coefffpients de corrélation entre les rendementsen graind$ de chaque paire de cultivars, c'est-à-direle coefficient de détermination (12.100). En considé-rant la similitude des rendements de chaque pairede cultivars en conditions très variées de milieu (16-41 localités), une large amplitude des coefficients dedétermination a été observée, allant de B à 9B%.

Bien que certaines groupes de cuitivars différantgénotypiquement peuvent être établies, la diversitégénétique des hybrides actuels de tournesol est assezlimitée, pour correspondre à toutes les variations dumilieu et pour réduire au minimum la vulnerabilitégénétique des .cultures de tournesol. Les facteurs limi-tatifs sont liés à l'utilisation du même type de sté-rilité mâle cytoplasmique, ainsi qu'à la similitudegénétique de beaucoup de formes parentales femelles,provenant des cultivars à teneur élevée d'huile et aunombre réduit de Iignées restauratrices de la ferti-lité du pollen.

Afin d'élargir la diversité génétique des hybridesde tournesol, il est nécessaire de créer des sources-populations à base génétique différente, pour servircomme matériel initiel pour l'autofécondation, et d'i-dentifier des nouvelles sources de stérilité mâle cyto-plasmique et de restauration de la fertilité du pollen,lors des croisements interspécifiques.

SEMEJANZA GENÉTICA DE LOS CULTIVARESDE GIRASOL TESTADOS EN LOS CULTIVOS

COMPARATIVOS INTERNACIONALES

Res{tmen

Se ha estudiado la semejanza genética de un nû-merp de 101 hibridos y 13 variedades de girasol testa-dos en la Red de investigaciones F.A.O. para elgirasol en el periodo 1976-1983 en gran nûmero delocalidades de Europa y fuera de Europa. La seme-janzâ genética fue estimada empleândose el cuadradode los coeficientes de correlaci6n entre las produccio-nes de semillas de cada pareja de cultivares, esto esel coeficiente de determinaci6n (r2.100). Tomando enconsideraciôn la similitud de las producciones de cadapareja de cultivares en condiciones variadas de me-dio, se nota una gran amplitud de los coeficientes dedeterminaci6n comprendida entre B y 98 por ciento.

A pesar de que se pueden establecer algunos gruposde cultivares que difieren genotipicamente, la diver-sidad genética de los hibridos actuales de girasol esbastante limitada para corresponder a todas la varia-ciones de medio y para minimalizar la vulnerabilidadgenétlca de los cultivos de girasol. Los factores limi-tativos estân relacionados al êmpleo de un mismotipo de androesterilidad citoplasmâtica, asi como a Iasimilitud genética de muchas formas parentales feme-las provinientes de las variedades con contenido ele-vado de aceite y del nrimero reducido de lineas res-tauradoras de la fertilidad del poleno.

Para ampliar la diversidad genética de los hibridosde girasol hace falta crearse fuentes-poblaciones conbase genética diferente que sirvan como. material ini-cial para autofecundaciôn y que identifiquen nuevasfuentas de androesterilidad citoplasmâtica y restaura-ci6n de la fertilidad del poleno dentro de los crucesinterspecificos.