Genetic similarity among brassica oleracea l. genotypes as measured by restriction fragment length polymorphisms

J. AMER. Soc. HORT. SCI. 118(2):298-303.
Genetic Similarity Among Brassica oleracea L.
Genotypes as Measured by Restriction Fragment
Department of Horticulture, University of Wisconsin, Madison, WI 53706 Mary K. Slocum2
Department of Biology, University of Utah, Salt Lake City, UT 84112
Dawn A. DeVos3
Harris-Moran Seed Company, 100 Breen Road, San Juan Bautista, CA 95045 Roger Muren4
Sun Seeds, 8850 59th Avenue, NE, Brooks, OR 97305
Additional index words. Brassica oleracea var. italica, B. oleracea var. botrytis, B. oleracea var. capitata, genetic distance, genetic similarity, algorithms, computer software, principal components, multidimensional scaling Genetic similarities were calculated among 89 Brassica oleracea L. genotypes, which included 62 broccolis
(var. italica), 16 cauliflowers (var. botrytis), and 11 cabbages (var. capitata). These entries represented a wide range
of commercially available germplasm, including open-pollinated cultivars, commercial hybrids, the inbred parents of
several hybrid cultivars, and 27 entries that were provided as unknowns. Sixteen random genomic clones were used
as probes in Southern hybridizations to detect restriction fragment length polymorphism (RFLP). From each of the
random probes, an average of four polymorphic bands were classified as to their presence or absence for each
genotype. The genetic similarity between ail pairs of genotypes was calculated. A multidimensional scaling (MDS)
plot indicated that the broccoli, cauliflower, and cabbage groups were clustered with very little overlap. Within
groups, genetic similarity corresponded to relationships based on available pedigree information. Comparison of
banding patterns between hypothetical and actual hybrids was used to correctly identify the parents of several parent-
hybrid combinations. The RFLP pattern of a hybrid and one of the parents (female) were used to predict the genotype
and identity of the other parent (male).
Knowledge of the relative genetic similarity among individuals Martinez et al., 1983; Singh et al., 1991). However environ- or populations is useful in a breeding program because it permits mental effects, as well as genotype by environmental interac- organization of germplasm and provides more efficient sampling tions, can result in the failure of the phenotype to accurately of genotypes. At the inception of a breeding program, knowledge represent the genotype. Moreover, although differences among of the genetic relationships among genotypes can be used to com- genotypes in phenotypic performance would suggest base pair plement phenotypic information in the development of breeding differences, identical phenotypic performance does not preclude populations. Ultimately, knowledge of the genetic similarity be- base pair differences.
tween genotypes can facilitate the choice of individuals to cross Variation for morphological or biochemical markers, such as in hybrid combinations to optimize expression of heterosis (God- isozymes, can be used to measure genetic similarity among shalk et al., 1990; Moll et al., 1965; Smith et al., 1990).
genotypes; however, precise evaluation of these markers may At the molecular level, the genetic similarity between two be confounded with environmental effects (Goodman and Stu- individuals is a function of their base pair sequences. In lieu of ber, 1980; Smith, 1984). RFLPs provide an opportunity to more direct sequence data, indirect measures of genotypic differences precisely measure genetic differences among individuals com- that reflect underlying base pair differences are commonly used.
pared to morphological or biochemical markers because they: Information from pedigrees can be used to calculate the coef- 1) are generally abundant in populations of interest, 2) have no ficient of parentage (r) among genotypes (Malecot, 1948). How- apparent affect on plant fitness per se, and 3) are not affected ever, because of the assumption of no relationship among the by the environment (Soller and Beckmann, 1983).
original parents, r is likely underestimated, to an unknown amount, The taxonomy and origin of Brassica species has been in- relative to the true genetic similarity. Phenotypic information vestigated based on cytogenetic studies (Prakash and Hinata, on quantitative and morphological characteristics can also be 1980), chemotaxonomic studies (Vaughn, 1977), classical tax- used to estimate genetic similarity (Goodman and Stuber, 1980; onomy (Grey, 1982, 1989; Vaughn, 1977), and molecular (RFLP)markers ( Song et al., 1988). In addition, Hu and Quiros (1991)recently identified random amplified polymorphic DNA (RAPD) Received for publication 27 Mar. 1992. Accepted for publication 15 Aug. 1992.
Research conducted at Native Plants Inc, 417 Wakara Way, Salt Lake City, markers that could discriminate within and between 14 broccoli UT 84102. Use of trade names does not imply endorsement of the products and 12 cauliflower cultivars. The objectives of this research named nor criticism of similar ones not named. The cost of publishing this were 1) to analyze RFLP information to determine the genetic paper was defrayed in part by the payment of page charges. Under postal relationships among a sample of genotypes both within and among regulations, this paper therefore must be hereby marked advertisement solelyto indicate this fact.
three of the cultivated Brassica oleracea L. subspecies, var.
2Research Assistant Professor.
3Research Director.
Abbreviations: MDS, multidimensional scaling; RAPD, random amplified po- 4Research Director.
lymorphic DNA; RFLP, restriction fragment length polymorphisms.
J. Amer. Soc. Hort. Sci. 118(2):298-303.
botytis (cauliflower), var. capitata (cabbage), and var. italica the number of bands for genotypes X and Y, respectively. The (broccoli), and 2) to determine if predicted banding patterns of resulting 89 × 89 triangular matrix of genetic similarity values hybrids could be useful in determining their pedigree (parents).
is too lengthy to be presented here, but the data are availableupon request from the senior author. The genetic similarity ma- Materials and Methods trix was subjected to principal components and MDS analysis Germplasm. Eighty-nine B. oleracea cultivars representing a (Wilkinson, 1989).
wide range of commercially available germplasm were selected Crossing algorithm. The following algorithm was used to for use in this study, including 62 broccolis, 16 cauliflowers, predict the inheritance of RFLP fragments in hybrids: 1 × 1 and 11 cabbages (Table 1). The germplasm was obtained by = 1, 1 × 0 = 1, and 0 × 0 = 0. The algorithm is interpreted requesting seeds or lyophilized leaf material from public and as follows: if either parent has a fragment then that fragment private institutions involved in breeding of cole crops. Among will be observed in the Southern blot of their progeny, and only the 89 entries, 27 were obtained from two seed companies as if both parents lacked a fragment will it be absent in their prog- "unknowns" (entry numbers 10 through 36). For these "un- eny (Quinn et al., 1987). The algorithm is predictive if the knowns" no information was provided a priori regarding either observed fragment represents a homozygous parent. If the frag- their classification, e.g., broccoli, cauliflower, or cabbage, or ment represents an allele of a heterozygous parent, then the of the genetic relationships among them.
probability of that allele being transmitted to any one progeny Many of the entries were commercial hybrid cultivars. We is 1/2; however, if the Southern analysis is based on a pooled grew the commercial seed lots of several hybrids, including sample of DNA from several progeny, then the probability of entries 7, 37, 39, 41, and 43, in an unreplicated trial in Brooks, that allele being observed is increased (1-1/2n, where n = the Ore., in 1987, with uniform spacing of 50 cm between plants.
number of progeny sampled). Thus, although heterozygosity in By identifying smaller "off-type" individuals within a row of the parents can be a potential source of error in the algorithm, otherwise uniform plants, we attempted to recover the female this error is minimized by sampling pooled leaf material from inbreds of these commercial hybrids. In the case of entry 43, several individuals from each entry.
‘Shogun', we grew several commercial hybrid seed lots anddiscovered two off-types, one of which had anthocyanin pig- mentation. The possible female parent corresponding to a hybrid Genetic similarity. The 89 × 89 matrix of genetic similarity cultivar is indicated by "–P" and "–H" after the names, measures was subjected to several multivariate analyses proce- respectively (Table 1).
dures. The first and second principal components (PC1 and PC2) RFLP procedures. Entries 10 through 36 were supplied as accounted for 64.5% and 6.0% of the variance, respectively. MDS bulk lyophilized leaf material harvested from six to 10 plants.
is a multivariate procedure that is related to principal components For the remaining samples, about equivalent amounts of leaf and factor analysis; however, MDS can usually fit an appropriate material were collected from 10 to 12 plants. Plant DNA was model in fewer dimensions than these other procedures (Wilkin- isolated from lyophilized leaf tissue and a restriction endonu- son, 1989). The principal coordinate scatter plots for principal clease (Eco RI) was used to digest the crude DNA samples. The components and MDS analysis were similar in appearance; how- procedures for DNA isolation, restriction endonuclease diges- ever, because it produced a clearer separation of entries in two tion, electrophoresis, Southern blotting, hybridization, and au- dimensions, the MDS plot is presented (Fig. 1).
toradiography have been previously described (Slocum et al., Inspection of the MDS plot indicated distinct clusters, with 1990; Song et al., 1988).
almost no overlap, among the three subspecies, italica (broc- A total of 16 random Brassica genomic clones containing coli), botrytis (cauliflower), and capitata (cabbage) (Fig. 1).
low copy number inserts were individually hybridized to Eco The mean genetic similarity between the broccolis and cauli- RI digested total genomic DNA samples. The set of random flowers was greater, (0.60 ± 0.06), than the mean genetic genomic clones used as probes in this study included EW1G03, similarity between either the broccolis or the cauliflowers com- EW1D02, EW2D03, EW2A06, EW2A07, EW1D03, EW1F08, pared to the cabbages, 0.57 ± 0.07 and 0.59 ± 0.05, respec- EW4A05, EW3D07, EW5F07, EW4G11, EW2B12, EW4G08, tively. The unknown entries 34, 35, and 36 were grouped with EW3C10, EW2E07, and EW1E04 (Pioneer Hi-Bred Interna- the var. capitata group and were therefore predicted to be cab- tional, Johnston, Iowa). Frequently the probes hybridized to bages. Entries 32 and 33, also representing unknown genotypes, multiple restriction fragments within individual samples, which were grouped with the var. botrytis group and were predicted resulted in complex banding patterns. For each hybridization to be cauliflowers. All other unknowns (entries 10 through entry combination, different size restriction fragments across all geno- 31) were grouped with the var. italica group, and were predicted types were assigned numbers (1, 2, 3, .n) according to de- to be broccolis. Similar predictions of subspecies classifications creasing molecular weights. From one to seven bands, which were obtained using discriminate analysis (not shown). All pre- were polymorphic among this sample of genotypes, were scored dictions of subspecies classification of unknowns were con- for each of the 16 probes resulting in a total of 61 scored frag- firmed as accurate by the providers of the source material.
ments. Each polymorphic fragment was treated as a unit char- The only genotype known to be an intermediate between sub- acter, and the genotype was scored for the presence or absence species is ‘Floccoli', a home garden cultivar, sold as a curiosity, of a fragment.
which is a hybrid between broccoli and cauliflower. The mean Calculatinggenetic relationship. Genotypes were coded 1 for genetic similarities between ‘Floccoli' and the broccoli entries presence or 0 for absence of a band (Hill, 1987). Genetic sim- and between ‘Floccoli' and the cauliflower entries were 0.62 ilarities were calculated between pairs of genotypes based on ± 0.06 and 0.71 ± 0.07, respectively. The hybrid origin of the method developed by Nei and Li (1979), i.e., D(XY) = ‘Floccoli' (entry 68) is further indicated by it's position in the 2N(XY)/[N(X) + N(Y)] where D(XY) is the measure of genetic scatter plot near the var. botrytis group and var. italica groups similarity between a pair of genotypes, N(XY) is the number (Fig. 1). Similarly, ‘B-21' is a disease resistant cabbage derived of bands common to genotypes X and Y, and N(X), N(Y) is from an interspecific hybrid between B. napus L. and B. oler- J. Amer. Soc. Hort. Sci. 118(2):298-303.
Cultivars included in the analysis of RFLP data.
Native Plants Inc.
Native Plants Inc.
Native Plants Inc.
Selection from OSU-112 Native Plants Inc.
Possible male parent of entry 43 N.S.S.L. no. 87-151 From National Seed Storage Lab Native Plants Inc.
Native Plants Inc.
Native Plants Inc.
Possible female parent of entry 7 Native Plants Inc.
Cross of entry 38 × entry 7 Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Harris-Moran Seed Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Sun Seeds Co.
Northrup-Ring Seed Co.
Native Plants Inc.
Possible female parent of entry 37 Quali-sel Seed Co.
Native Plants Inc.
Possible female parent of entry 39 Native Plants Inc.
Native Plants Inc.
Possible female parent of entry 41 Northrup-King Co.
Native Plants Inc.
Possible female parent of entry 43 Native Plants Inc.
Phenotypically similar to Green Duke Sun Seeds Co.
Sun Seeds Co.
Early Purple Sprouting Thompsom and Morgan Sprouting broccoli Thompson and Morgan Sprouting broccoli Thompson and Morgan Xmas Purple Sprouting Thompson and Morgan Sprouting broccoli Thompson and Morgan Asgrow Seed Co.
Bejo Seed Co.
Bejo Seed Co.
Oregon State Univ.
Oregon State Univ.
Oregon State Univ.
Oregon State Univ.
Oregon State Univ.
Oregon State Univ.
Oregon State Univ.
Oregon State Univ.
Oregon State Univ.
J. Amer. Soc. Hort. Sci. 118(2):298-303.
Table 1. continued. Kyowa Seed Co.
Kyowa Seed Co.
Kyowa Seed Co.
Thompson and Morgan Sakata Seed Co.
Australian type (?) Snowball Y Improved Peto Seed Co.
Asgrow Seed Co.
Burpee Seed Co.
Burpee Seed Co.
Piracicaba Precose Bioplanta (Brazil) Tetezopolis Precose Bioplanta (Brazil) Terezopolis Gigante Bioplanta (Brazil) Summer type (large curd) Royal Sluis Seed Co.
Bejo Seed Co.
Agriculture Canada Club Root Resistance native Plants 1%.
Male parent of entry 85 Native Plants Inc.
Female parent of entry 85 Native Plants Inc.
Hybrid of entry 83 × entry 84 Native Plants Inc.
Female parent of entry 88 Native Plants Inc.
Male parent of entry 88 Native Plants Inc.
Hybrid of entry 86 × entry 87 Burpee Seed Co.
zi = Brasscia oleracea var. italica (broccoli), b = var. botrylis (cauliflower), c = var. capitata (cabbage), and ? = classification was unknownat the beginning of the study.
yO.P. = open pollinated.
xUnknown: no information provided by supplier about classification or the genetic relationships.
ample, entries 48 and 51 are purple, and 49 and 52 white, cultivarsof biennial sprouting broccoli, which is a type adapted to over- winter under Northern European climatic conditions and flower in the spring (Grey, 1989). These four sprouting broccolis were lo-cated in close proximity on the MDS plot and were apart from theother broccolis (Fig. 1). ‘Waltham 29', ‘DiCicco', and ‘Atlantic'(entries 46, 47, and 53, respectively) are vintage open-pollinatedcultivars of broccoli. These vintage cultivars were grouped in closeproximity to one another and to ‘Calabrese' (entry 50). ‘Calabrese'was first cultivar introduced. into the United States by Italian im-migrants and has been developed by breeding to have a moreannual habit and heading form (Grey, 1989). The relationships onthe MDS plot would suggest that the vintage broccoli cultivars areclosely related to ‘Calabrese' and may have been selected from it.
In addition, a modem hybrid cultivar ‘Green Valient' (entry 37)is located in close proximity to both the the vintage broccoli cul-tivars and ‘Calabrese'. In contrast, the modem hybrid cultivar‘Shogun' (entry 43), its possible female and male parents (entries44 and 4, respectively), hybrid cultivar ‘Cruiser', (entry 7), andthe cultivar ‘Hanemori' (entry 65) appear to represent a distinct Plot of first and second dimensions of MDS analysis of ge- grouping of germplasm compared to the group that includes the netic similarity matrix of Brassica oleracea genotypes. Genotypesare listed in Table 1.
vintage broccoli cultivars, ‘Calabrese' and ‘Green Valient'. Geno- types provided by Oregon State Univ. (entries 56 through 64) areall grouped in close proximity, except for entries 60 and 63, which acea var. capitata (Chaing et al., 1977). The introgression of were intermediate between the two germplasm groups identified exotic genes into ‘B-21' (entry 82) is suggested by its position on the scatter plot as an outlier in the var. capitata cluster, Two unknown entries, 23 and 24, and entry 38, the possible intermediate between that and the var. botrytis cluster.
female parent of the modern hybrid cultivar ‘Green Valient' The var. italica (broccoli) entries did not reveal any independent (entry 37), are located in close proximity on the MDS plot (Fig.
clusters. Nevertheless, entries known by pedigree to be closely 1). The genetic similarities between entry 38 and entries 23 and related were located in close proximity to one another. For ex- 24 are 0.918 and 0.950, respectively. We subsequently learned J. Amer. Soc. Hort. Sci. 118(2):298-303.
that entries 23 and 24 were developed by pedigree selection Moran Seed Co. (comparisons 7, 8, and 9, in Table 2), the from an individual plant identified as the possible female parent genetic similarity between actual and predicted hybrids was low, of hybrid ‘7181' (entry 22). The results of these analyses would ranging from 0.717 to 0.850.
suggest that the hybrid cultivars ‘Green Valient' and ‘7181' may Determination of male parent, given the hybrid and the fe- share a closely related female parent. In addition, ‘7184' (entry male parent. Entries 7, 37, 39, 41, and 43 are commercial F1 25) is a doubled haploid (DH) line derived from anther culture hybrid broccoli cultivars (Table 1). The likely female parents of hybrid ‘7181'. The genetic similarity between ‘7184' and of these hybrids were identified as off-types in uniform trials.
‘7181' is 0.720. An additional DH line is ‘7186' (entry 27) that The same algorithm used to predict hybrids from crosses be- was derived from a broccoli hybrid ‘7187' (entry 28); the ge- tween parents was used to predict the likely male parent of a netic similarity between these genotypes is 0.765.
hybrid combination, given the hybrid and the female parent.
The var. botrytis (cauliflower) entries alone did not reveal Entries 43 and 44 represent the commercial hybrid ‘Shogun' any distinct groups (Fig. 1). Entries 77, 78, and 79, all of which and it's likely female parent, respectively. The predicted male are Brazilian cultivars adapted to tropical environments, did not parent of ‘Shogun' is entry 4 (Table 3). Entry 4 is an off-type form a distinct group. The only hybrid cauliflower, ‘Early White' different from the female in that it possesses anthocyanin pig- (entry 75), was located near the center of the cauliflower cluster.
mentation on the stem. This prediction suggests that both the Determination of parents of F1 hybrids. At the onset of this male and female parents of ‘Shogun' are self incompatible, and study, the only known parental-F1 hybrid relationships were one that hybrid seed is harvested off both parents. Similarly, entry broccoli and two cabbages (comparisons 1, 2, and 3, respectively, 37 and 38 represent the commercial broccoli hybrid ‘Green Val- in Table 2). The inbred parents of the broccoli hybrid, entries 38 ient' and its likely female parent, respectively. The predicted and 5, were crossed using the crossing algorithm (described in the male parent is entry 73, which is ‘Snowball Y', a cauliflower.
materials and methods section), and the genetic similarity between ‘Green Valient' is not a broccoli × cauliflower hybrid; however, the actual and the predicted hybrid banding pattern was 0.918. In it is a modern hybrid characterized by small bead size. Cauli- contrast, the genetic similarity between the actual and predicted flower has been used by plant breeders as a source of genetic hybrid banding patterns in both cabbages was 0.836. In the cab- variance for bead size in broccoli (Grey, 1982); thus, the results bage hybrids, the genetic similarity between one of the parents here suggest that the male parent of ‘Green Valient' has some (parent ‘A') and the actual hybrid was greater (0.935) than the cauliflower in its parentage.
genetic similarity between the parents (0.885 and 0.820, respec-tively). A similar relationship was observed among the parents and F1 hybrid of the cabbage entries originally provided as unknownsby Sun Seeds (comparison 4). The observed genetic relationships In these analyses and in the application of the crossing al- among the parents and their F1 hybrid progeny suggests that the gorithm, the allelic relationships between bands was unknown; hybrid cabbages used in this study were contaminated with selfed thus, we could not take advantage of the co-dominant infor- seed of their female parents.
mation often available for RFLP data. Therefore, the individual Random hybrid combinations were tested from among the bands are considered as dominant rather than co-dominant mark- entries to identify possible combinations of parents and hybrids.
ers and provide information similar to that provided by RAPD All entries provided as unknowns were crossed, and the result- markers. Thus, at a given locus, heterozygotes will have a band- ing hybrid patterns were compared to the existing database en- ing pattern identical to one of the parents. Moreover, hetero- tries. Among the unknowns provided by Sun Seeds, entry 26 zygous individuals, either F1 hybrid or open pollinated, will was predicted to be a hybrid (0.967), with entries 25 and 30 as likely have more bands coded as present than inbred genotypes.
parents (comparison 5, Table 2). Similarly, entry 29 was pre- This difference can complicate the observed relationships be- dicted to be a hybrid (0.934), with entries 28 and 31 as parents tween individual parents and their hybrid progeny, as hetero- (comparison 6, Table 2). These predictions were later confirmed zygous parents may appear as more closely related to their F1 as accurate by Sun Seeds. In contrast, although several parent- hybrid progeny than the other more inbred parent. This devia- hybrid combinations were among the unknowns sent by Harris- tion was observed in this experiment (Table 2); however, it did Genetic similarity between actual hybrids and hybrids predicted based on the inheritance of banding patterns between parents.
Genetic similarityz between zGenetic similarity calculated based on the presence or absence of RFLP bands.
ySee Table 1 for cultivar names and classification.
xArbitrary designation of the parents of a hybrid.
J. Amer. Soc. Hort. Sci. 118(2):298-303. 1993.
Table 3. The predicted male parent of a hybrid cultivar based on RFLP genotypes of the hybrid and the likely female parent.z Genetic similarity between Actual hybrid and hybrid made using predicted male parent zThe likely female parent was identified as an "off-type" in a row of plants from acommercial F1 hybrid seed lot.
ySee Table 1 for cultivar names and classification.
not preclude the ability to identity the parents of most of the single cross grain yield performance (Godshalk et al., 1990; hybrid cultivars.
Smith et al., 1990) suggest that molecular marker information The reasons for the contrasting results in prediction of parental can be used to identify parents with high heterotic potential.
F1 hybrid combinations between the unknowns provided by Sun Seeds and those provided by Harris-Moran were unknown. Welater learned that some of the parents used in the hybrids provided Arus, P., S.D. Tanskley, T.J. Orton, and K.A. Jones. 1982. Electrophoretic variation as a tool for determining seed purity and for breeding hybrid varieties of Brassica by Sun Seeds were highly inbred and, in fact, entries 28 and 25 oleracea. Euphytica 31:417-428.
were DH lines. In contrast, the parents of the hybrids provided by Chaing, M.S., B.Y. Chaing, and W.F. Grant. 1977. Transfer of resistance to race 2 of Harris-Moran Seed Co. were early generation (F3 and F4 ) de- Plasmodiophora brassicae from Brassica napus to cabbage (B. oleraea var. capi- tata). I. Interspecific hybridization between B. napus and B. oleracea var. capitata. veloped by pedigree selection. Thus, the higher levels of heter- ozygosty in the parents of the hybrids provided by Harris-Moran Godshalk, E.B., M. Lee, and K.R. Lampkey. 1990. Relationship of restriction frag- Seed compared to those provided by Sun Seeds might account for ment length polymorphisms to single-cross hybrid performance on maize. Theor.
the differences observed in prediction of parental F1 hybrid com- Applied Genet. 80:273-280.
Goodman, M.M. and C.W. Stuber. 1980. Genetic identification of lines and crosses binations. Determination of the level of heterozygosity in an inbred using isoenzyme electrophoresis. Proc. 35th Annu. Corn, Sorghum Res. Conf.
parent population is possible using RFLPs, and such data would help to resolve the sometimes unexpected relationships observed Grey, A.R. 1982. Taxonomy and evolution of broccoli (Brassica oleracea var. ital- ica). Econ. Bot. 36:397-410.
in this study between parents and hybrids. In addition, analysis of Grey, A.R. 1989. Taxonomy and evolution of broccolis and cauliflowers. Baileya individuals rather than a bulk would permit determination of the percentage of contamination of selfed seed in the cabbage hybrids.
Hill, W.G. 1987. DNA fingerprints applied to animal and bird populations. Nature Genetic marker data on many individuals might be more econom- Hu, J. and C.F. Quiros. 1991. Identification of broccoli and cauliflower cultivars with ically obtained by using diagnostic isozymes (Arus et al., 1982) RAPD markers. Plant Cell Rpt. 10:505-511.
rather than RFLPs. However, a limitation of isozymes is the lack Malecot, G. 1948. Les mathematiques de l'heredité. Masson et Cie, Paris.
of observed polymorphism among closely related genotypes com- Martinez, O.J., M.M. Goodman, and D.H. Timothy. 1983. Measuring racial dif- ferentiation in maize using multivariate similarity measures standardized by vari- pared to DNA-based markers, such as RFLPs or RAPDs (Hu and ation in F2 populations. Crop Sci. 23:775-781.
Quiros, 1991).
Moll, R.H., J.H. Lonnquist, J.V. Fortuna, and C.E. Johnson. 1965. The relationship Only limited pedigree information is available for the 89 en- of heterosis and genetic divergence in maize. Genetics 52:139-44.
tries included in this experiment; nevertheless, the available Nei, M. and W. Li. 1979. Mathematical model for studying genetic variation in terms of restriction endonucleases. Proc. Natl. Acad. Sci. 76:5256-73.
pedigree information, including information available a priori Prakash, H. and K. Hinata. 1980. Taxonomy, cytogenetics and origin of crop Bras- and a posteriori, was highly correlated with genetic similarity sica, a review. Opera Bot. 55:1-57.
as measured by RFLP polymorphisms. The information con- Quinn, T.W., J.S. Quinn, F. Cooke, and B.N. White. 1987. DNA marker analysis detect multiple maternity and paternity in single broods of the lesser snow goose.
tributed by 61 RFLP bands provided a level of resolution of Nature 326:392-394.
genetic relationships among cultivars and between parents and Singh, S.P., J.A. Gutierrez, A. Molina, C. Urrea, and P. Gepts. 1991. Genetic hybrids that might previously have been considered protected diversity in cultivated common bean: II. Marker-based analysis of morphological as "trade secrets" by plant breeding companies. Implied in and agronomic traits. Crop. Sci. 31:23-29.
Slocum, M.K., S.S. Figdore, W.E. Kennard, J.Y. Suzuki, and T.C. Osborn. 1990.
these analyses, especially those dealing with the identification Linkage arrangement of restriction fragment length polymorphism loci in B. oleracea. of parents of F1 hybrids, is the ability to dissect hybrid cultivars Theor. Applied Genet. 80:57-64.
and better understand how both the individual parents and their Smith, J.S.C. 1984. Geneticvariability within U.S. hybrid maize: Multivariate analy- sis of isozyme data. Crop Sci. 24:1041-1046.
hybrid are related to known germplasm resources. Moreover, Smith, O.S., J.S.C. Smith, S.L. Bowen, R.A. Tenborg, and S.J. Wall. 1990. Sim- by using RFLP information it is possible to identify one or more ilarities among a group of elite maize inbreds as measured by pedigree, F1 grain of the parents used in a F1 hybrid cultivar.
yield, grain yield, heterosis, and RFLPs. Theor. Applied Genet. 80:833-840.
Plant breeders can use the RFLP information to organize ge- Soller, M. and J.S. Beckmann. 1983. Genetic polymorphism in varietal identification and genetic improvement. Theor. Applied Genet. 67:25-33.
netic resources into related groups to make more informed de- Song, K.M., T.C. Osborne, and P.H. Williams, 1988. Brassica taxonomy based on cisions in the choice of parents. Eventually, the hypothetical nuclear restriction fragment length polymorphisms (RFLPs). 2. Preliminary analy- banding patterns of test hybrids might be correlated with phe- sis of subspecies within R. rapa (syn. campestris) and B. oleracea. Theor. Applied notypic information to predict both performance and adaptation Genet. 76:593-600.
Vaughn, J.G. 1977. A multidisciplinary study of the taxonomy and origin of Brassica of experimental hybrids. Moreover, the recent studies indicating crops. Bioscience 27:35-40.
a high correlation between RFLP-based genetic distance and Wilkinson, L. 1989. SYSTAT: The System for Statistics. SYSTAT, Evanston, Ill.
J. Amer. Soc. Hort. Sci. 118(2):298-303.

Source: http://journal.ashspublications.org/content/118/2/298.full.pdf