Efficiency of potato breeding using FDR 2n gametes for multitrait selection and progeny testing

Summary The objective of this research was to compare the efficiency of the 4x×2x breeding scheme with the traditional 4x×4x method with respect to potato improvement. The basis for such a comparison was the parental value of four 2x and four 4x male parents from the International Potato Center (CIP) as measured by multitrait selection and progeny testing. The 2x parents produced 2n pollen by parallel spindles at anaphase II, which is genetically equivalent to a first division restitution (FDR) mechanism. Both 2x and 4x parents were crossed with four common 4x female parents. Thus, 32 families were evaluated over 2 years at four Peruvian locations. A selection index which considered tuber yield, tuber number, average tuber weight and specific gravity was used for multitrait selection. Three FDR 2x parents had better selection index scores than the 4x parents over the four locations. Estimates of broad-sense heritability for total yield using different number of replications and locations were calculated by using the variance components. The 4x × 2x breeding scheme was found to be better than the traditional 4x × 4x method since fewer replications and locations are required to evaluate tuber yield in 4x × 2x progenies than in 4x × 4x progenies. The FDR 2x parents were also better material than the 4x parents for testing combining ability for tuber yield of the 4x progenitors. This could be the result of the mode of FDR 2n pollen formation. The pollen of FDR 2x parents is more heterozygous, but more homogenous than n pollen from 4x parents.Paper from the Laboratory of Genetics. Research supported by the College of Agricultural and Life Sciences; International Potato Center; USDA-CRGO-88-37234 3619, and Frito-Lay, Inc., USA     

Effects of 2n-pollen formation by first meiotic division restitution with and without crossover on eight quantitative traits in 4x-2x potato progenies

 The 2n-pollen grains formed by first-division restitution without crossover (FDR-NCO) are unique breeding tools, since they can transmit almost 100% of non-additive genetic effects from the parent to the progeny. FDR-NCO gametes are considered superior to those formed by FDR with crossing over (FDR-CO), which can pass on to the progenies approximately 80% of the heterozygosity and a large fraction of the epistasis. However, 2n-pollen formation by FDR-NCO mechanism requires the incorporation (in homozygous condition) of at least two recessive alleles. In the present work, 40 tetraploid families derived from complete 4x-2x factorial crosses were evaluated under short-day conditions to verify whether or not the postulated genetic superiority of FDR-NCO over FDR-CO gametes holds true for eight quantitative traits in potato. Families were derived from crosses between four 4x commercial cultivars, and a random sample of ten diploid Solanum phureja-haploid S. tuberosum hybrids producing 2n-pollen by either FDR-CO or FDR-NCO. The results indicated no significant superiority of FDR-NCO over FDR-CO families for total tuber yield (TTY) and six other traits (haulm maturity – HM; plant vigor – PV; plant uniformity – PU; eye depth – ED; number of tubers per hill – NTH; and commercial over total yield index – CTI). Based upon cytological observations, the FDR-CO and FDR-NCO gametes are expected to be genetically equivalent for all loci between the centromeres and the chromosomal site of maximum recombination. In our experiment, differences between FDR-CO- and FDR-NCO-derived progenies were not observed for TTY. Therefore, our results can be interpreted as additional evidence for the hypothesis that genes with major effect on TTY expression might have a physical location between centromeres and proximal crossovers in the potato chromosomes. In addition, a similar trend was observed for HM, PV, and ED but apparently not for commercial yield – CY (i.e., tubers with more than 33 mm in diameter). Received: 21 September 1998 / Accepted: 26 October 1998     

The significance of 2N gametes in potato breeding

Summary Phureja-haploid Tuberosum hybrids, which produce 2n gametes in addition to n gametes, were used to obtain diploid progenies in 2x – 2x matings, and tetraploid progeny in 4x – 2x matings. Seven of these diploid clones were intermated in a modified diallel design, and also crossed to seven cultivars, to obtain 21, 16-clone diploid; and 49 (35, 2x X 4x and 14, 4x X 2x) 24-clone tetraploid families, respectively. These progenies were included together with the 14 parental clones, in a tuber yield trial conducted in two locations in Wisconsin. Heterotic responses were striking in 4x X 2x progenies. Three of these entries had mean yields of 5.0, 5.0 and 5.2 lbs/hill, for 24 unselected clones, thus outyielding not only the mid-parent but also the best tetraploid cultivars in the experiment, which are highly selected; Wis 643 (4.8 lbs/hill), and Kennebec (4.8 lbs/hill). As a group, the 14, 4x X 2x families averaged 4.4 lbs/hill, while the mean of all seven cultivars was 4.0 lbs/hill, and the mid-parent was at 3.4 lbs/hill. This unusually high mean performance of 336 unselected clones represented in the 14, 4x X 2x progenies, is interpreted as a manifestation of the capacity of 2n pollen, formed by first meiotic division restitution (FDR), to pass onto the progeny the already heterotic diploid genotype in a largely intact array. Beneficial intra and inter-locus interactions are presumably compounded upon syngamy with an unrelated n egg from the tetraploid parent. -The performance of 2x X 4x progenies was at or below that of the midparent. The failure of these families to perform as well as 4x X 2x families may be a reflection of the incapacity of 2n megasporogenesis to avoid meiotic reassortment as efficiently as FDR does, which would result in inbred 2n gametes. However, the method of 2n megasporogenesis is not known.Hereditary variances were large, both within and among families, and indicated considerable nonadditivity in the action of genes affecting tuber yield at both levels of ploidy. Inbreeding was strongly depressing at both ploidy levels.     

Gene-centromere mapping by 4x-2x matings in potatoes

Summary Two of the four strands of a bivalent are recovered together in tetraploid progeny arising from 4x-2x matings. This provides a method for gene and centromere mapping. The cross pppp x Pp produced 62 nulliplex tetraploid individuals in a total of 951, i.e., 6.5%. The diploid clone was found to produce diplandrous gametes through first division restitution (FDR). The map distance P-centromere was estimated as 13.0 units, the limiting values at the 95% binomial confidence interval being 10.1 and 16.3. The mapping method is explained in detail and a formula is deduced to obtain genotypic series of 2n gametes under particular assumptions. The direction 4x×2x is advantageous, since high seed set diploid clones which give 99% tetraploid progeny, and whose method of diplandroid formation is known are already available. Diploid clones heterozygous for many markers and tetraploids nulliplex for the same markers are needed to fully exploit this method.     

Producing commercially attractive,uniform true potato seed progenies: the influence of breeding scheme and parental genotype

The production of attractive, uniform true potato seed (TPS) progenies was investigated. Four breeding schemes were compared: intercrossing tetraploid cultivars (cv x cv); doubled dihaploids x cultivars (ddh x cv); cultivars x diploid unreduced-gamete producers (cv x FDR) and doubled dihaploids x diploid unreduced-gamete producers (ddh x FDR). Fifty three progenies and five clones were grown in a glasshouse in a randomised complete block design with three replicates of 25 plants per progeny and clone. Each plant's tubers were counted and the colour, shape, quality of skin finish, flesh colour, and commencai attractiveness (which includes yield) recorded. The most uniform progenies were also selected by visual comparison with the clones. For mean attractiveness, differences (P < 0.001) between breeding schemes and between progenies within breeding schemes were detected. The cv x cv and cv x FDR progenies were more attractive than clonal controls. There were significant additive and non-additive effects for attractiveness in all breeding schemes except cv x FDR. There were between-progeny differences (P < 0.001) for uniformity for all characters. Progenies uniform for one character could be variable for other traits. Breeding schemes gave different levels of uniformity (P < 0.001) for all characters except shape and flesh colour, but none gave low levels of variation for all traits. Doubled-dihaploid parents increased the variation in progenies. There were uniformity differences (P < 0.001) between progenies within breeding schemes for all characters. Evidence of additive and nonadditive genetic variation for uniformity in all traits was detected. In each breeding scheme, parents with good general combining ability (GCA) for uniformity in several characters were identified. Visually selected uniform progenies had parents with good GCAs for uniformity in a range of traits and high specific combining abilities (SCAs) for several traits. A desynaptic first-division restitution (FDR) clone and a male-sterile doubled-dihaploid clone had the best GCAs for tuber uniformity in TPS progenies. Achieving multitrait uniformity in TPS is problematic but may be aided by the selection of parents with GCAs for uniformity coupled with progeny testing to allow for non-additive effects.     

Comparative performance of FDR and SDR progenies from reciprocal 4x-2x crosses in potato

Numerically unreduced (2n) gametes from first division restitution (FDR) are considered to be superior to 2n-gametes from second division restitution (SDR) because they transfer a larger proportion of the total parental heterozygosity and epistasis intact to the tetraploid progeny. This supposed superiority was investigated by comparing 12 sets of reciprocal 4x-2x crosses. Each diploid parent used in a reciprocal set produced 2n-pollen by FDR and 2n-eggs by SDR. Six agronomic characters were investigated. FDR progenies (from 4x.2x) were found to have higher mean yields due to more and bigger tubers. With respect to underwater weight, the overall progeny mean of FDR progenies was significantly higher than that of SDR progenies (from 2x.4x). However, the absolute difference found between both overall progeny means was too small to be of practical significance. No differences between FDR and SDR progeny means were found for vine maturity and chip colour. In addition to the progeny mean, within-progeny variation is important in potato breeding. For vine maturity a higher within-progeny variation was detected in SDR progenies, whereas within-progeny variations for yield, underwater weight and chip colour were not different in FDR and SDR progenies. With regard to vine maturity, we conclude that SDR 2n-gametes are superior to FDR 2n-gametes because, with the same progeny means of FDR and SDR progenies, the within-progeny variation was higher in SDR progenies. Therefore the assumed superiority of FDR 2n-gametes was confirmed for yield but was not observed for vine maturity, underwater weight and chip colour.     

Comparison of FDR- and SDR-derived tetraploid progeny from 2x×4x crosses using haploids of Solanum tuberosum L. that produce mixed modes of 2n eggs

 The relationship between heterozygosity and heterosis in tetraploid potato (Solanum tuberosum subsp. tuberosum L., 2n=4x=48) was examined in a series of first-division restitution (FDR)- and second-division restitution (SDR)-derived tetraploid subpopulations. The subpopulations were constructed using two 2n egg-producing, mixed-mode haploids (2n=2x=24) crossed to three tetraploid (2x= 4x=48) potato clones. Half-tetrad analysis using a co-dominant electrophoretic marker (Pgm-2), which is closely linked to the centromere, discriminated between FDR- and SDR-derived 4x progeny. The FDR:SDR ratio of the 4x progeny observed was dependent upon the haploid parent used in the 2x×4x cross. Field studies were conducted between 1992 and 1996 to compare the yield and specific gravity of the two subpopulations and their parents from three crosses. There was no difference in the total tuber yield or specific gravity between the FDR- and SDR-subpopulations based upon family means, despite the expectation that FDR-derived progeny would transmit a greater portion of the genome’s heterozygosity intact than SDR-derived progeny. The 4x parent in each family had a higher yield than either 4x progeny subpopulation. Inbreeding, as a consequence of the haploidization process and a lack of genetic diversity, may have negated any advantage of the FDR-derived progenies over the SDR-derived progenies. Received: 14 April 1998 / Accepted: 12 May 1998     

Quantitative Genetics of 4x-2x Hybrid Populations with First-Division Restitution and Second-Division Restitution 2n Gametes Produced by Diploid Parents

A model is proposed to describe the genetic value of 4x-2x hybrids resulting from crosses between tetraploid genitors and diploid genitors that produce 2n gametes. The model takes into account the genetic consequences of the First Division Restitution (FDR) and Second Division Restitution (SDR) meiosis, particularly on the homozygosity level that 2n gametes contribute to 4x-2x hybrids. As genes can be identical by descent, numerous parameters are needed in the classical approach to describe the inbreeding effects on the mean and variance of 4x-2x hybrids. Using the concept of test value, the model allows a large decrease in the number of required parameters. The model gives the components of genetic variance and usual covariances between relatives using these synthetic parameters. The model is then used to study the efficiency of a recurrent breeding scheme to improve diploid genitors for their combining ability with tetraploid genitors. It appears that, in presence of dominance, ignoring the meiosis pattern will lead to an overestimation of additive variance and then of genetic advance. Some genetic considerations on the differences between FDR and SDR mechanisms lead us to suggest an experimental comparison of their respective advantages and disadvantages for the type of considered recurrent selection. An experimental crossing design is proposed to obtain estimates of the genetic parameters needed for this comparison.     

Statistical epistasis between candidate gene alleles for complex tuber traits in an association mapping population of tetraploid potato

Association mapping using DNA-based markers is a novel tool in plant genetics for the analysis of complex traits. Potato tuber yield, starch content, starch yield and chip color are complex traits of agronomic relevance, for which carbohydrate metabolism plays an important role. At the functional level, the genes and biochemical pathways involved in carbohydrate metabolism are among the best studied in plants. Quantitative traits such as tuber starch and sugar content are therefore models for association genetics in potato based on candidate genes. In an association mapping experiment conducted with a population of 243 tetraploid potato varieties and breeding clones, we previously identified associations between individual candidate gene alleles and tuber starch content, starch yield and chip quality. In the present paper, we tested 190 DNA markers at 36 loci scored in the same association mapping population for pairwise statistical epistatic interactions. Fifty marker pairs were associated mainly with tuber starch content and/or starch yield, at a cut-off value of q ≤ 0.20 for the experiment-wide false discovery rate (FDR). Thirteen marker pairs had an FDR of q ≤ 0.10. Alleles at loci encoding ribulose-bisphosphate carboxylase/oxygenase activase (Rca), sucrose phosphate synthase (Sps) and vacuolar invertase (Pain1) were most frequently involved in statistical epistatic interactions. The largest effect on tuber starch content and starch yield was observed for the paired alleles Pain1-8c and Rca-1a, explaining 9 and 10% of the total variance, respectively. The combination of these two alleles increased the means of tuber starch content and starch yield. Biological models to explain the observed statistical epistatic interactions are discussed.     

Genetic positioning of centromeres using half-tetrad analysis in a 4x-2x cross population of potato

From biological and genetic standpoints, centromeres play an important role in the delivery of the chromosome complement to the daughter cells at cell division. The positions of the centromeres of potato were determined by half-tetrad analysis in a 4x-2x population where the male parent produced 2n pollen by first-division restitution (FDR). The genetic linkage groups and locations of 95 male parent-derived amplified fragment length polymorphism markers could be determined by comparing their position on a 2x-2x highly saturated linkage map of potato. Ten centromere positions were identified by 100% heterozygosity transmitted from the 2n heterozygous gametes of the paternal parent into the tetraploid offspring. The position of these centromeric marker loci was in accordance with those predicted by the saturated 2x-2x map using the level of marker clustering as a criterion. Two remaining centromere positions could be determined by extrapolation. The frequent observation of transmission of 100% heterozygosity proves that the meiotic restitution mechanism is exclusively based on FDR. Additional investigations on the position of recombination events of three chromosomes with sufficient numbers of markers showed that only one crossover occurred per chromosome arm, proving strong interference of recombination between centromere and telomere.     

Evaluation of BC2 progenies derived from 3x-2x and 3x-4x crosses of Lilium hybrids: a GISH analysis

An allotriploid (ALA, 2n=3 x=36) BC(1) plant was obtained by backcrossing a diploid F(1) interspecific hybrid (LA, 2n=2 x=24), derived from a Lilium longiflorum (L genome) and an Asiatic hybrid (A genome), to the latter parent. This allotriploid was backcrossed to a diploid Asiatic hybrid (2n=2 x=24) and to an allotetraploid (LLAA, 2n=4 x=48) LA hybrid. A total of 25 plants of these crosses were examined for ploidy level, and 12 individuals were analyzed for their genome constitution through genomic in situ hybridization (GISH). In most cases the progenies from the triploid-diploid (3 x-2 x) crosses consisted of aneuploids. Further more, there was evidence for the formation of near-haploid (x=12+2) to triploid (3 x=36) gametes in the allotriploid BC(1) plant. The progenies of triploid-tetraploid (3 x-4 x) cross also consisted of mostly aneuploids but in this case the triploid female parent had contributed predominantly near-triploid (2n) gametes for the origin of BC(2) progenies. The different ploidy levels observed between 3 x-2 x and 3 x-4 x crosses are possibly caused by preferential fertilization or survival resulting in a different ratio of chromosome numbers between the embryo and endosperm. Though Lilium has a tetrasporic, eight-nucleate type of embryo sac formation (Fritillaria type), the observed difference between the progeny types in 3 x-2 x and 3 x-4 x crosses is comparable to that of observed in monosporic eight nucleate types (Polygonum type) that predominate in most genera of Angiosperms. An important feature of the genome constitution of the progenies was that the homoeologous recombinant chromosomes were transmitted intact from BC(1) to BC(2) progenies in variable numbers. In addition, there was evidence for the occurrence of new homoeologous recombinations in the triploid BC(1). Of the two euploid BC(2) plants one had originated through the parthenogenetic development of a 2n egg and the other had originated through indeterminate meiotic restitution (IMR).     

Heterozygosity in 2n gametes of potato evaluated by RFLP markers

The heterozygosity transmitted through 2n gametes in potato was evaluated by following the segregation of RFLP markers in tetraploid progeny from bilateral sexual polyploidization in a cross between two diploid (2x) interspecific hybrids which produce 2n SDR eggs or 2n FDR pollen. Out of 84 probe/enzyme combinations tested, 23 revealed polymorphism between the parents and were heterozygous in at least one of the parents. These probes characterized 13 loci distributed on five different chromosomes of the potato RFLP map. The heterozygosity transmitted through SDR and FDR gametes was estimated to be 31.8% and 71.4%, respectively. Two different indices (LH and RHI) were used to select plants showing a high level of heterozygosity in the tetraploid progeny. The recombination events and the centromere positions were estimated for chromosomes I, VI and VII, following the segregation ratios of SDR or FDR gametes produced by the parents. A different recombination rate was observed between the two interspecific hybrids.Contribution no. 119 from Research Centre for Vegetable Breeding, C.N.R., Portici, Italy. Research supported by National Research Council of Italy, special Project RAISA, Sub-project No. 2, Paper No. 2090     

Single cross alfalfa (Medicago sativa L.) hybrids produced via 2n gametes and somatic chromosome doubling: experimental and theoretical comparisons

Summary The potential breeding value of 2n gametes from diploid alfalfa (2n = 2x = 16) was tested by comparing single cross alfalfa hybrids produced via 2n = 2x gametes from diploids versus n = 2x gametes from somatic-chromosome-doubled, tetraploid counterparts. Three diploid clones, designated 2x-(rprp), homozygous for the gene rp (conditions 2n gamete formation by a first division restitution mechanism) were colchicine-doubled to produce their tetraploid counterparts, designated 4x-(SCD). These six clones were crossed as males to the same cytoplasmic male sterile clone. Yield comparisons of progeny from the six clones demonstrated a significant yield increase of the hybrid progeny from 2n = 2x gametes from the diploids over the hybrid progeny from n = 2x gametes from the chromosome doubled tetraploid counterparts. The yield gain ranged from a 12% increase to a 32% increase. Theoretical comparisons indicated the 2n = 2x gametes from diploids would have 12.5 to 50% more heterozygous loci, on average, than the n = 2x gametes derived from somatic doubling. These results confirm the importance of heterozygosity on alfalfa yield, and the results demonstrate that 2n gametes formed by first division restitution offer a unique method for producing highly heterotic alfalfa hybrids.     

Cytological,morphological and molecular analyses of controlled progenies from meiotic mutants of alfalfa producing unreduced gametes

A program of sexual polyploidization was carried out in alfalfa using plants from wild diploid species that produced male or female unreduced gametes. Sixteen progenies from 2x-4x and 2x-2x crosses were examined with a combination of morphological, cytological and molecular analyses. The chromosome counts revealed diploid, tetraploid and aneuploid plants. Plants with B chromosomes were also detected. The leaf area of the plants was a useful characteristic for distinguishing tetraploid from diploid plants obtained by unilateral or bilateral sexual polyploidization. Leaf shape and leaf margin were not correlated with the ploidy levels. Plants with supernumerary chromosomes displayed obovate or elliptic leaves which differed markedly from the range of forms typical of diploid and tetraploid alfalfa plants. RAPD markers were investigated in all progeny plants to determine maternal and paternal amplification products. Three alfalfa-specific primers proved to be effective in revealing the hybrid origin of the plants. A combination of cytological, morphological and molecular analyses is essential for a detailed genetic characterization of progenies in programs of sexual polyploidization.     

QTL mapping of yield, agronomic and quality traits in tetraploid potato (Solanum tuberosum subsp. tuberosum)

Interval mapping of quantitative trait loci (QTL) for 16 yield, agronomic and quality traits in potato was performed on a tetraploid full-sib family comprising 227 clones from a cross between processing clone 12601ab1 and table cultivar Stirling. Thirty-eight AFLP primer combinations and six SSRs provided 514 informative markers which formed a molecular marker map comprising 12 linkage groups (LGs) in 12601ab1 (nine with four homologous chromosomes) which were aligned with 12 in Stirling (11 with four homologous chromosomes), with four partial groups remaining in 12601ab1. Two LGs were identified unequivocally as chromosomes IV and V and eight others were tentatively assigned with chromosomes VII and X unidentified. All of the traits scored had moderately high heritabilities with 54–92% of the variation in clone means over 3 years and two replicates being due to genetic differences. A total of 39 QTLs were identified. A QTL for maturity was identified on chromosome V which explained 56% of the phenotypic variance, whereas the other QTLs individually explained between 5.4 and 16.5%. However, six QTLs were detected for after-cooking blackening and four for each of regularity of tuber shape, fry colour both after storage at 4 and 10°C and sprouting. Just two QTLs were found for each of yield, the two ‘overall’ scores, crop emergence, tuber size and common scab and just one QTL was detected for each of dry matter content, keeping quality, growth cracks and internal condition. The implications for practical potato breeding and for practical linkage and QTL analysis in autotetraploids are discussed.     

Direct estimation of the effects of meiotic recombination on potato traits via analysis of 4x-2x progenies from synaptic mutants with 2n-pollen formation by FDR without crossing-over

Diploid potato clones with 2n-pollen formation by first-division restitution without crossing-over (FDR-NCO) are ideal testers to estimate the breeding value of elite 4x cultivars by virtue of transmitting their genotypes practically intact to their progenies. This characteristic facilitates genetic analysis, since meiotic recombination would take place only in the 4x parent and not in the diploid parent. We evaluated (under short-day conditions) families from complete factorial crosses between four 4x cultivars and five 2x(FDR-NCO) clones. Families were compared with two standard 4x cultivars (’Bintje’ and ’Delta’) for total tuber yield (TTY), commercial yield (CY), haulm maturity (HM), plant vigor (PV), plant-top uniformity (PU), eye depth (ED), number of tubers per hill (NTH), and the CY/TTY index (CTI). For TTY, the contrasts family group (310 g/ hill) vs ’Delta’ (430 g/hill) and the family group vs ’Bintje’ (210 g/hill) were significant. Only 25% of the families were different from ’Delta’ and 20% of them outyielded ’Bintje’. For CY, differences were observed between families (240 g/hill) vs ’Delta’ (340 g/hill) and families vs ’Bintje’ (150 g/hill). The two best families had 53% CY over ’Bintje’. Surprisingly, only one family had a higher NTH than ’Bintje’. No differences were observed for HM. Seventy five and 30% of the families had an ED similar to ’Delta’ (ED = 2) and ’Bintje’ (ED = 1), respectively. A multivariate analysis indicated that 63% of the data variability could be explained by two factors. TTY, CY, and PV had high loading on the first factor, whereas ED, PU and HM had high loading on the second factor; CTI and NTH had equal sizes on both factors. High TTY and PV were associated with high NTH and CTI. Deep eye, PU, and late maturity were associated with high NTH and reduced CTI. The distributions of factor scores of the entries indicated that some 2x parents had strong influences (irrespective of the direction of their effects) on the crosses. Six crosses due to two 2x males were in the ’Bintje’ quarter with negative scores for both factors (implying low TTY, poor vigor, and low NTH). Also three crosses due to another 2x clone were distributed in the quarter of positive factor 1 and negative factor 2. These crosses plus another one were in the same quarter of ’Delta’ (implying high yields, low ED, low PU, and early maturity). The FDR-NCO clones provide a homogeneous sample of heterozygous 2n-gametes allowing the unique opportunity to estimate the relative contribution of the random meiotic products (from the 4x parents) and the ’somatic’ 2x genome for the phenotypic expression of quantitative traits. The interesting result was that measurable effects (favorable or not) on the data variability were mainly determined by the genomic contribution of the haploid-species hybrids. Three out of five 2x-male parents showed rather strong effects on progenies. No such effects were observed on the four 4x-female parents. Received: 3 September 1999 / Accepted: 24 November 1999     

Relationships among genetic distance,forage yield and heterozygosity in isogenic diploid and tetraploid alfalfa populations

Isogenic diploid and tetraploid alfalfa (Medicago sativa L.) was studied with molecular markers to help understand why diploid performance and breeding behavior does not always predict that of tetraploids. In a previous study of partially heterozygous alfalfa genotypes, we detected a low correlation between yields of isogenic diploid (2x) and tetraploid (4x) single-cross progenies, and genetic distances were more highly correlated with yields of tetraploids than diploids. These differences may be related to the level of RFLP heterozygosity expected among progenies derived from heterozygous parents at the two ploidy levels. The objectives of this study were to determine the relationships among genetic distance, forage yield and heterozygosity in isogenic 2 x and 4 x alfalfa populations. Four diploid genotypes were chromosome doubled to produce corresponding isogenic autotetraploids, and these genotypes were mated in 4 × 4 diallels to produce 6 single-cross families at each ploidy level for field evaluation. Allele compositions of parents were determined at 33 RFLP loci by monitoring segregation of homologous restriction fragments among individuals within progenies, and these were used to estimate RFLP heterozygosity levels for all single-cross progenies at both ploidy levels. RFLP heterozygosity rankings were identical between progenies of isogenic diploid and tetraploid parents; but significant associations (P < 0.05) between estimated heterozygosity levels and forage yield were detected only at the tetraploid level. Since tetraploid families were nearly 25% more heterozygous than the corresponding diploid families, inconsistencies in the association between molecular marker diversity and forage yields of isogenic 2 x and 4 x single crosses may be due to recessive alleles that are expressed in diploids but masked in tetraploids. The gene action involved in heterosis may be the same at both ploidy levels; however, tetraploids benefit from greater complementary gene interactions than are possible for equivalent diploids. Present address: AgResearch Grasslands, New Zealand Pastoral Agriculture Research Institute, Palmerston North, New Zealand     

Molecular markers flanking citrus tristeza virus resistance gene from Poncirus trifoliata (L.) Raf.

 Two segregating populations for citrus tristeza virus (CTV) resistance derived from Poncirus trifoliata var ‘Flying Dragon’ by self-pollination and pollination to Citrus medica L. var ethrog ‘Arizona’ were inoculated with a common CTV isolate. The presence of virus was checked by the Double Antibody Sandwich Enzyme-Linked Assay and Direct Tissue Blot Inmunoassay at 3, 6, and 12 months after inoculation. Seven RAPDs were found linked to the CTV resistance gene by bulked segregant analysis. The closest linked RAPDs were cloned to obtain linked codominant RFLPs and to increase the precision of the genetic distance estimation. The CTV resistance gene seems to be located between cW18 and cK16. Differences in genetic distances among progenies are large and can be explained by genome-wide reduction in the recombination of progeny derived from male versus female gametes. Received: 5 June 1996 / Accepted: 26 July 1996     

Segregation for sexual seed production in Paspalum as directed by male gametes of apomictic triploid plants

BACKGROUND AND AIMS: Gametophytic apomixis is regularly associated with polyploidy. It has been hypothesized that apomixis is not present in diploid plants because of a pleiotropic lethal effect associated with monoploid gametes. Rare apomictic triploid plants for Paspalum notatum and P. simplex, which usually have sexual diploid and apomictic tetraploid races, were acquired. These triploids normally produce male gametes through meiosis with a range of chromosome numbers from monoploid (n = 10) to diploid (n = 20). The patterns of apomixis transmission in Paspalum were investigated in relation to the ploidy levels of gametes. METHODS: Intraspecific crosses were made between sexual diploid, triploid and tetraploid plants as female parents and apomictic triploid plants as male parents. Apomictic progeny were identified by using molecular markers completely linked to apomixis and the analysis of mature embryo sacs. The chromosome number of the male gamete was inferred from chromosome counts of each progeny. KEY RESULTS: The chromosome numbers of the progeny indicated that the chromosome input of male gametes depended on the chromosome number of the female gamete. The apomictic trait was not transmitted through monoploid gametes, at least when the progeny was diploid. Diploid or near-diploid gametes transmitted apomixis at very low rates. CONCLUSIONS: Since male monoploid gametes usually failed to form polyploid progenies, for example triploids after 4x x 3x crosses, it was not possible to determine whether apomixis could segregate in polyploid progenies by means of monoploid gametes.     

Indeterminate meiotic restitution (IMR): a novel type of meiotic nuclear restitution mechanism detected in interspecific lily hybrids by GISH

A detailed analysis of microsporogenesis was carried out in three diploid lily cultivars (2n=2x=24) and three diploid interspecific hybrids (2n=2x=24) using DNA in situ hybridisation methods (GISH and FISH). In cvs. Gelria (Lilium longiflorum; L genome), Connecticut King and Mont Blanc (both Asiatic hybrids; Agenome) meiosis was regular and only haploid gametes were formed while the three interspecific hybrids between L. longiflorum×Asiatic hybrid (LA) showed a variable frequency of meiotic nuclear restitution and stainable 2n-pollen formation ranging from 3% to 30%. An analysis of meiotic chromosome behaviour of the LA hybrids through GISH and FISH revealed that: (1) the parental chromosomes could be clearly discriminated into univalents, half-bivalents and bivalents in the PMCs; (2) in some of the PMCs the entire complement was present either as univalents or half-bivalents which had the potential to divide equationally (following centromere division) during the first division leading to first division restitution (FDR) gametes; (3) more frequently, however, in one and the same PMC the univalents and half-bivalents divided equationally whereas the bivalents disjoined reductionally at the same time giving rise to 2n-gametes that could vary from the well-known FDR or SDR 2n-gametes. We indicate this novel type of restitution mechanism as Indeterminate Meiotic Restitution (IMR). In order to confirm the occurrence of IMR gametes, the chromosome constitutions of eight triploid BC₁ progenies derived from backcrossing the 2n-gamete producing the LAhybrids to the Asiatic hybrid parents were analysed through in situ hybridisation. The results indicated that there were seven BC₁ plants in which FDR 2n-gametes, with or without homoeologous recombinations, were functional, whereas in one case the 2n-gamete resulting from IMR was functional. In the latter, there was evidence for the occurrence of genetic recombination through homoeologous crossing-over as well as through the assortment of homoeologous chromosomes. A singular feature of the IMR 2n-gamete was that although it transmitted a euploid number of 24 chromosomes to the BC₁ progeny, the number of chromosomes transmitted from the two parental species was dissimilar: 9 L-genome chromosomes and 15 A-genome chromosomes instead of 12 of each. Received: 15 May 2000 / Accepted: 4 December 2000