Buletinul USAMV-CN, 64/2007 (-) ISSN 1454-2382

EFFICIENCY OF NONPARAMETRIC STATISTICS IN EVALUATIN G PHENOTYPIC VARIABILITY AND OF RAPD ANALYSIS IN

EVALUATING VARIABILITY AT THE MOLECULAR LEVEL, OF THE MAIN APPLE VARIETIES GROWN IN TRANSYLVANIA

M.Ardeelean, V.Mitre, L.Lukács, Rodica Pop, Ioana Mitre , Mirela Cordea

University of Agricultural Sciences and Veterinary Medicine 3-5 Mănăştur Street, 400372, Cluj-Napoca, mardelean@usamvcluj.ro

Key words: apple cultivars, nonparametric statistics, phenotypic/genotypic variability, RAPD analysis Abstract. Five apple cultivars (Florina, Golden Delicious, Idared, Jonathan and Starkrimson), were tested in 2004 and 2005 in five commercial orchards located in different counties of Transylvania with quite nonsimilar environments. Fruit yield was registered, the resulting data being evaluated by means of HUHN (1990a and 1990b) nonparametric stability indices: Si

(1) = mean of the absolute rank differences of a genotype over environments and Si

(2) = variance among the ranks over environments. For a genotype with maximum stability Si

(1) = 0 and Si(2) = 0. In June 2004, young shoots were harvested from each cultivar and taken to the laboratory

by means of a refrigerating bag. Young leaves were harvested from the shoots and preserved in refrigerator at –70oC until the DNA extraction was performed. 12 decamer primers were used for DNA amplification out of which ten produced polymorphic bands in each variety/location . Six dendrograms were constructed based on DNA migration in agarose gel and analysis of formed bands and genetic distances, one for revealing the intervarietal polymorphism in all five locations and five for evaluating the polymorphism within each cultivar depending on its location of origin. The highest stability of fruit yield (or the lowest variability), in both experimental years, was found in Florina (Si

(1) = 1.60 and Si(2) = 2.0) which is the only domestic variety tested in

this experiment and, quite expectedly, best adapted to the studied environments. The dendrograms exhibited the existence of obvious differences, at the molecular level, both among cultivars across all the environments and within the same cultivar, depending on location. The last type of differences should be attributed mainly to the lack of uniformity of initial mother plants used in the process of planting material production for the five cultivars. It is concluded that RAPD analysis could successfully be used in checking the authenticity of planting material for apple trees produced in different nurseries, provided there are available standard forms of the interested cultivars from which DNA could be analyzed and used as control.

INTRODUCTION

In apples, the phenotypic stability(or lack of variability) of yield and other quantitative traits which affect the commercial or/and processing qualities of fruits is of utmost importance both for breeding and production. The nonparametric measures for stability, based on ranks, which do not assume a normal distribution of the phenotypic values, are especially fit for apple trials in which, quite often, the possibility of measurement errors is rather high. Therefore, in the presented paper, a nonparametric method , proposed by NASSAR and HUHN (1987), was used for estimating the phenotypic stability or the lack of variability across the tested environments of fruit yield in apple . On the other hand, RAPD analysis has become lately a routine procedure in Romania for fingerprinting plant species and cultivars in order to identity intraspecific (intervarietal) and interspecific diversity at the molecular level (CORDEA et al., 2001; GABOREANU et al., 2001; POP et al., 2003; ARDELEAN et al., 2004). Very seldom has this type of analysis been used in checking the homogeneity/heterogeneity, at the molecular level, of planting materials (grafted trees) produced by different nurseries.

The aim of this research was to reveal, on one hand, the magnitude of phenotypic stability of the five tested apple cultivars across the environments and, on the other hand, the possible differences, at the molecular level, within the planting material of five apple varieties produced in five different nurseries, by means of RAPD analysis. The results of such experiments could be of great help both in apple breeding programs and in the national network of planting material production for apple trees.

MATERIAL AND METHOD

In five intensive (666 trees/ha) commercial orchards of five locations (Cluj-Napoca, Cluj County; Reghin, Mureş County; Mihăieşti, BistriŃa-Năsăud County; Timişoara, Banat County and Seini, Maramureş County), characterized by obviously different environments, five apple varieties, widely grown in Transylvania (Golden Delicious, Jonathan, Starkrimson, Idared, Florina) were studied for two years (2004-2005).

Observations were conducted on the same number of trees (18) from each variety, in all years and locations. Three replications were formed in each location with six trees/plot, resulting an experimental design in randomized blocks. The data on fruit yield were harvested from four trees of each plot, yield/plot being converted to yield/ha. Instead of raw data obtained from field observation, for ranking the studied genotypes over the environments adjusted data were preferred. The ranking was done in each location separately, based on the means of each experimental year and consisted in conferring the rank 5 to the variety which showed the highest adjusted yield and the rank 1 to that with the poorest adjusted yield.

Two rank stability indices , proposed by NASSAR and HÜHN (1987), were computed, )1(

iS being the statistic measure of the mean absolute rank difference of a genotype over environments and )2(

iS being the variance of ranks over the environments. )1(iZ and )2(

iZ are measures of stability computed on the basis )1(

iS and )2(iS , which are supposed to show a

normal distribution allowing the computation of χ2( )1(iZ )2(

iZ ) and χ2sum )1(iZ , )2(

iZ .The specific formula for computation of )1(

iS , )2(iS , )1(

iZ , )2(iZ as well as for correcting the actual

field data are given by NASSAR and HÜHN (1987), HUEHN (1990 a; 1990 b). In June 2004, young shoots were harvested from each cultivar and taken to the

laboratory by means of a refrigerating bag. Young leaves were harvested from the shoots and preserved in refrigerator at – 70oC until the DNA extraction was performed. Genomic DNA was extracted according to LODHI et al. (1990) protocol , modified by POP et al. (2003). The basic method for RAPD was as described by OTONI et al. (1995). There were used 12 primers (Mycroshinth) out of which only ten produced polymorphic bands (table 1).The amplified DNA segments were separated in 1.4% agarose gels prepared in 0.5 x TBE buffer. Electrophoresis was performed at 0.57 V cm-1 for 150 min. and visualization of amplification products was achieved in UV light, the image of gels being taken over by means of a video camera Alpha Innotech. RAPDistance 1.04 software was used for computing the genetic distances and, on this basis, Jaccard`s coefficient was calculated for each species. Cluster analysis was performed using the Neighbor-Joining method and the results were presented as dendrograms generated by RAPDistance 1.04 pack of soft.

RESULTS AND DISCUSSION

The statistics of stability for the mean yield of fruit, in 2004 and 2005, are shown in table 1. It is quite obvious from these data that 2004 was an unfavorable year for apples while 2005 could be considered as a normal one. In spite of the great differences in mean yield of the two years, the ranking of mean yields of the five tested cultivars was rather similar, i.e. in both years the poorest stability of mean fruit yield (or the highest variability) was found in

cultivars belonging to Delicious group , while the highest yield stability was found in Florina and Idared (2004) and Golden Delicious (2005).

From the second part of the table one may look for the stability differences among

genotypes. It is seen that, in 2004, ∑ =

5

1

)1(

i iZ = 20.11 which is higher than 11.07 (the critical

value for χ20.05;5) while ∑ =

5

1

)2(

i iZ = 5.146, less than 11.07. This means that there are

significant differences in rank stability among the five genotypes grown in five environments

in 2004. On the contrary, in 2005, ∑ =

5

1

)1(

i iZ = 3.750 and ∑ =

5

1

)2(

i iZ = 5.146 both values being

well beyond 11.07 meaning that in a favorable year no significant differences in rank stability were found among the five tested cultivars in the five environments.

Estimation and nonparametric test of stability of fruit yield (t/ha)

for five apple cultivars across five environments. 2004 2005 Table 1

Genotype

Mean fruit yield, t/ha

Mean rank

)1(iS

)1(iZ

)2(iS

)2(iZ

Mean fruit yield, t/ha

Mean rank

)1(iS

)1(iZ

)2(iS

)2(iZ

Florina 10.56 2.167 1.60 1.957 2.035 0.001 18.21 2.667 1.45 0.703 1.813 0.106

Golden Del. 11.00 2.500 1.90 2.143 2.313 0.055 14.42 2.833 1.70 0.312 2.639 0.998

Idared 9.99 2.167 1.70 0.238 1.035 0.526 19.10 2.500 1.65 0.078 2.868 0.486

Jonathan 12.41 2.667 2.20 8.571 4.556 3.664 22.77 2.167 1.85 1.953 2.534 3.176

Starkrimson 14.10 3.000 2.15 7.202 3.250 0.881 26.27 2.333 1.75 0.703 5.250 0.096

Sum 20.111 5.146 3.750 4.863

Test statistics

E(S1) E(S2) Var(S1) Var(S2) χ2Z1Z2

χ2sum Z1.Z2 E(S1) E(S2) Var(S1) Var(S2)

χ2Z1Z2

χ2sum Z1.Z2

1.6 2.0 0.042 1.773 13.32 22.41 1.6 2.0 0.022 2.886 5.25 4.48

Grand mean total = 11.61 Grand mean total = 20.15

On inspecting the individual Z values it vas found that , in both years, only Jonathan was

relatively unstable to the others since it showed high Z values compared with the critical value χ2

0,01;1 = 6.62. Data as those analyzed above are of real value since they could offer the opportunity

to choose genitors with high levels of stability for fruit yield. It is assumed that such genitors might transmit to offsprings the genetic basis of this stability.

The dendrograms constructed on the basis of genetic distances among cultivars in all five location and within the same cultivar in the five locations are presented in fig.1. At the first sight it can be noticed that the five apple cultivars are assembled in two distinct groups which rather logically represent phylogenetic groups (fig. 1, f). The first group comprises two cultivars closely related to each other (Starkrimson and Idared), both of them belonging directly or indirectly to Delicious group of apples. The second group comprises three cultivars, Jonathan Florina and Golden Delicious which are also phylogenetically tied to each other. Concretely, Golden Delicious is one of the genitors of Florina while Jonathan is a remote ancestor of many recent Romanian apple varieties, including Florina. On the basis of the above considerations it can be stated that RAPD analysis fairly illustrated the genetic differences among the five apple cultivars under study emphasizing, at the same time, the phylogenetic relationship existent among them. There can be admitted that

e. Starkrimson f. Dendrogram generated on the basis of genetic distances among cultivars in all five locations

Fig. 1. Dendrograms generated on the basis of genetic distances among

cultivars and among proveniences of the same apple cultivar

each of the five cultivars represents a distinct genetic entity, easily recognizable at the molecular level. These assumptions would allow one to expect that each tested cultivar show the same DNA pattern at the molecular level, no matter which nursery the trees were originated in, but the dendrograms presented in fig.1 (a;b;c;d;e) obviously contradict such expectations.

As it can be easily noticed, the standard is different from the five dendrograms presented in fig. 1 (a;b;c;d;e), within each of the five cultivars there had been revealed a significant variability, at the molecular level, which can only satisfactorily be explained by an improper management of the planting material production. This explanation stands valid since, for fruit trees, each Fruit Experimental Station had its own mother plant orchards for cultivars to be multiplied and, most probably, these mother plants were rather local clones of respective varieties which might have accumulated a significant proportion of somatic mutations.

CONCLUSION

1. The nonparametric stability tests proposed by NASSAR and HÜHN (1987) proved to be an accurate tool in revealing the variability/stability of fruit yield of tested cultivars across environments 2. Data as those analyzed above are of real value since they could offer the opportunity to choose genitors with high levels of stability both for fruit yield and other quantitative characters of interest in apple breeding programs. It is assumed that such genitors might transmit to offsprings the genetic basis of this stability. 3. RAPD analysis fairly illustrated the genetic differences among the five apple cultivars under study emphasizing the phylogenetic relationship existent among them. 4. Dendrograms constructed on the basis of genetic distances among proveniences within the same cultivar revealed a significant variability, at the molecular level, which can only satisfactorily be explained by an improper management of the planting material production. 5. RAPD analysis proved to be a reliable tool for evaluating the authenticity of planting material for apple trees produced in different nurseries, provided there are available standard forms of the interested cultivars from which DNA could be analyzed and used as control.

BIBLIOGRAPHY 1. Ardelean,M., Mirela Cordea, D.Pamfil, .W.Blackhall, S.C.Andras, M.R.Davey, 2004, Revealing Genetic Diversity of Three Sections (Pharbitis, Quamoclit and Batatas) of Ipomoea Genus by Means of RAPD Analysis, Acta Horticult., 651, 149-153. 2. Cordea Mirela, N.W.Blackhall, S.C.Andras, M.Ardelean, J.B.Power, D.Pamfil, M.R.Davey, 2001, Fingerprinting Lines and Species of Ipomoea Genus, Bul.USAMV-CN, Seria Hort., vol.55-56, 63-68. 3. Gaboreanu Ioana, N.W. Blackhall, C. Andras, M.R. Davey, D. Pamfil, C. Panfil, 2001, Gnetic relationships of Mentha species evaluated by RAPD analysis, Bu. USAMV-CN, 55-56, 149-153. 4. Huehn, M., 1990, Nonparametric measures of phenotypic stability. Part 1: Theory, Euphytica, 47, 180-194. 5. Huehn, M., 1990, Nonparametric measures of phenotypic stability. Part 2: Applications, Euphyt., 47,195- 201. 6. Lodhi, M.A., N.F.Weeden, B.I. Reisch, 1997, Characterization of RAPD markers in Vitis, Vitis 36: 133-140. 7. Nassar, R. and M. Hühn, 1987, Studies on estimating phenotypic stability: tests of significance for nonparametric measures of phenotypic stability, Biometrics, 43, 45-53. 8. Otoni, W.C., N.W. Blackhall, F.B. D`Utra Vaz, V.W. Casali, J.B. Power, M.R. Davey (1995) Somatic hybridization of the Passiflora species, P.edulis f. Flavicarpa degener. and P. incarnata L., Journal Exp. Bot. 46: 777-785. 9. Pop Rodica, M.Ardelean, D.Pamfil, Ioana Gaboreanu, 2003, The efficiency of different DNA isolation and purification protocols in ten cultivars of Vitis vinifera, Bul. USAMV-CN, 59, ser. ZB, 259-261.