The relationship between the genetic distance of parents and both the heterosis of Fl hybrids and the variance of F5 lines was investigated in 72 crosses of pea ...
255
Euphytica 73 : 2 5 5-264, 1994 . • 1994 Kluwer Academic Publishers . Printed in the Netherlands.
Genetic distance and its association with heterosis in peas • Sarawat l , F.L. Stoddard2
& D .R . Marsha11 2 Department of Plant Science, Waite Agricultural Research Institute, University of Adelaide, PMB 1, Glen Osmond, • 5064, Australia 1 ; present address: Khon Kaen Field Crops Centre, Department of Agriculture, Khon Kaen 40000, Thailand; 2 present address : Department of Crop Sciences, The University of Sydney, NSW 2006, Australia Received 24 September 1993 ; accepted 3 November 1993
Key words : Pisum sativum, genetic diversity, genetic variance, breeding strategies, pea, heterosis, genetic distance
Summary The relationship between the genetic distance of parents and both the heterosis of Fl hybrids and the variance of F5 lines was investigated in 72 crosses of pea (Pisum sativum L .) . The genetic distance between each pair of parents was estimated, using isozyme (GDi), morphological (GD m ) or quantitative (GD q ) markers and finally a combination of isozyme and morphological markers (GDi+,,,,) . GD m was poorly correlated with the other measures of genetic distance, which in turn were strongly correlated with each other . Genetic distance was moderately correlated with the level of heterosis for yield over midparent in the Fl generation, with the highest correlation obtained from GDi +m . GD was not significantly correlated with heterosis for yield over the better or best parent but it was significantly correlated with all three measures of heterosis for pods per plant and hundred seed weight . There was no correlation between genetic distance and the level of heterosis for yield and total dry matter in the F2 generation, but GDi, GDi +m, and GD q were predictive for the level of inbreeding depression in grain yield and total dry matter . When parents were high in genetic distance, crosses produced highly transgressive segregants for basal branches per plant, hundred seed weight, harvest index and onset of flowering . Genetic distance between parents was thus a useful measure for predicting a portion of hybrid performance and also of the variance of derived inbred lines . It was concluded that when choosing parents for a cross, consideration should be given to their genetic distance as well as their overall adaptation and their yield . There is considerable potential for optimising choice of parental combinations in the development of improved pea cultivars .
Introduction A positive correlation between genetic diversity of parents and the level of heterosis has been reported in several inbreeding crops . In cotton (Gossypium spp .), for example, interspecific hybrids showed significantly higher heterosis during early growth than their intraspecific counterparts (Marani & Avieli, 1973) . In tomatoes, intercluster crosses tended to have greater heterosis than intracluster crosses (Khanna & Misra, 1977) . In peanut, heterosis was correlated with diversity in both geographical origin (Parker et al ., 1970) and morphological traits (Isleib & Wynne, 1983) . Also in peanut, the chances of high frequency and magnitude of heterosis were greater in crosses between parents within 1 standard deviation of the mean than in
crosses between parents with divergence outside this limit (Arunachalam & Bandyopadhyay, 1984) . On the other hand, the divergence between parents of each cross, estimated using generalized distance and canonical analysis, did not correspond to the level of heterosis in chickpea (Cicerarietinum L .) (Singh & Ramanujam, 1981) . Since the level of heterosis in Fl hybrids often increases with the degree of genetic differentiation between the parental lines, it may be possible to develop a preliminary screen for highly heterotic combinations by measuring the genetic distance between potential parents . To do this requires a rapid and reliable method of estimating the genetic distance using procedures normally available to plant breeders .
256 Ideally, marker loci should distinguish heterozygotes from homozygotes, should be distinguishable from other loci, should display substitutions at every base and should be sampled at random, regardless of function or level of polymorphism (Brown & Weir, 1983) . Only a complete DNA base sequence fits this description, so plant breeders have generally used quantitative or qualitative traits to measure genetic diversity . Many quantitative traits are available for use, but they are often controlled by many genes and they are really phenotypic rather than genotypic . Qualitative morphological characters are usually controlled by few genes and are less affected by non-genetic factors than quantitative characters, but dominance, epistasis and pleiotropy often obscure the simple gene effects (Brown, 1978) . Isozyme techniques are independent of the functional role or the overall level of variation of the enzyme, allowing many individuals to be screened non-destructively and quickly. Furthermore, allelic expression is usually codominant and allows heterozygotes to be distinguished from homozygotes . Nevertheless, post-translational modification can occur, multiple variants can arise through gene duplication, only some base substitutions result in detectable amino acid replacements and only some proteins are suitable for analysis in this way . Several measures have been developed for quantifying genetic diversity and differ according to the qualitative or quantitative nature of the markers . Although a variety of measures of genetic diversity has been proposed and each measure is based on different biological or mathematical assumptions, in practice, many of the methods give similar values (Hedrick, 1975) . Measures using quantitative traits have been based on multivariate statistical analyses such as Pearson's coefficient of racial likeness (Pearson, 1926), Mahalanobis distance (Mahalanobis, 1936), Sokal distance (Sokal, 1961) and Euclidean distance (Goodman, 1972) . In inbreeding crops, qualitative characters have been quantified using the Diversity index of Hunter & Kannenberg (1971) and the Genetic Distance of Nei (1972) . The relationship between genetic distance between parental lines and the level of heterosis in peas has been unclear. In interspecific crosses of Pisum, heterosis appeared to be associated with genetic diversity among the parents (Singh & Malhotra, 1971) . Gupta et al . (1984) and Rao & Narsinghani (1987), however, found that the most heterotic hybrids of field peas were between parents that were moderately divergent as determined from quantitative characters . This experiment therefore set out to compare different methods
of evaluating genetic distance and then to determine the correlation between genetic distance and heterosis in 72 crosses of pea .
Materials and methods The genetic distances between parental lines were estimated using isozyme markers, morphological characters and quantitative traits separately and in combination . Morphological and quantitative characteristics were scored from material grown in an earlier experiment (Sarawat et al ., 1994) . The relationship between genetic distance between parents and the level of heterosis in their F1 progenies was assessed by simple correlation analyses . Evaluation of parents
The four female and 18 male parental lines were characterised for 17 morphological traits and gene symbols were obtained (Blixt, 1977 ; M . Ali, Department of Agriculture South Australia, pers . comm .) . Electrophoretic analyses were carried out on all parents . The enzymes assayed were ADH, alcohol dehydrogenase ; ADL, aldolase ; AMP, aminopeptidase; GOT, glutamate oxaloacetate transaminase ; G6PD, glucose 6 phosphate dehydrogenase ; IDH, isocitrate dehydrogenase ; MDH, malate dehydrogenase ; MRD, menadione reductase ; 6-PGD, 6-phosphogluconate dehydrogenase ; PGI, phosphoglucoisomerase ; PGM, phosphoglucomutase and SDH, shikimate dehydrogenase. Leaves were used for G6PD and IDH and seeds for the other 10 enzymes . Seeds were soaked in water overnight and a small sample from a cotyledon was then crushed in 1-3 drops of extraction buffer . The extract was centrifuged at 2000 rpm for 1-2 min, then the supernatant was absorbed on chromatography paper wicks (6 mm x 3 mm) that were inserted into a vertical slit in the starch gel . Leaflets from the first fully expanded leaves were cut into 1 cm square segments and crushed in 1-3 drops of extraction buffer . The rest of the protocol was identical to that for seeds . The extraction buffer was phosphate buffer (0 .05 M phosphate, pH 7 .0) with DTT (DL Dithiothreitol) 10 mg/ml solution for AMP, GOT, G6PD and MRD, or Carlsons' buffer (12 .1 g tris, 7 .44 g KCI, 1 .86 g EDTA and 37 .58 g sucrose, in 1000 ml H2O) also with DTT (10 mg/ml solution) for the remaining enzymes . The starch gel preparation and electrophoresis techniques were adapted from the procedures described by
257 Table
Cultivar Morphological characters Females Alma bt, G
1 . Qualitative characteristics of pea parental lines
Isozyme system ADH-1 ADL AMP GOT G6PD IDH-1 IDH-2 MDH-1 MDH-2 MRD 6PGD-1 6PGD-2 PGI PGM SDH
1
1
3
1
1
1
2
1
1
2
1
Derrimut bt, G 1 Dundale bt, G 1
1 1
3 3
1 1
1 1
1 1
2
1
2 2
1 1
Wirrega a, bt, Cr, r, ram 1 Males
1
3
1
1
1
2 2
1 1
Garfield a, bt, GB, i, r bt, fn, M, pl SA 15
1
3
1
1
1
i, le, obs, r 2 2 SA 35 cr, fn, k, obs, r, st, Z SA 51 bt, cr, dt, fa, gp, M, r, wlo 2 SA 54 bt, cr, fn, M, pl, r, rms 2
1 1
3 3
1 1
1 1
1 1
1 1 2
3 3 1
1 1 2
1 1 1
SA 123 bt, fr, le, M, r SA 129 bt, cr, fr, le, M, pl, r
2 2
2 2
2 2
2
1
SA 157 bt, Cr, r, wa
1 2
1
3
2 1
1 1
1
1 1
3 3
1 1
a, bt, Cr, r 2 a, GB, le, r 2
2 1
3 3
SA 483 bt, fr, le, M, pl, r 2 SA 688 bt, M, r, wa 2
2 2
SA 828 bt, M, r, ram Solara a, af, bt, GB, Ie, r, SL
2 2
SA 24
SA 236 a, bt, Cr, pl, r SA 247 a, bt, Cr, r, ram SA 248 SA 465
Whero
bt, M, r
2 2 2
1 1
2 2
2 2
1 2
2 2
2 2
1 2
2 2
2 2
1
1 1
2
1
1 1
2
2
1
2
2
1
1
2 2
2 2
1 1
2 2
1 1
2 1
2 2
1 1 2
2 1 2
2 2 2
1 1
2 2 2
1 2 2
1 1
2 1
1 2 1
1 1 1
2 2
2
2 2
1 1
2 2
1 2
1 1
1 1
2 2
1
2 2
2 2
1 1
1 1
2 1
2 2
2 2
1 1
3 3
1
3
1 1
2
1 1
2
1
1 1
2 2
2 2
2 1
2 2
1 2
1 1
2 2
1 1
1 2
2 2
2
2 1
1 1
2 2
2 2
1 1
2 2
1 1
1 1
2 2
2 1
1 2
2 2
2 2
1
2 2
2 1
1 2
1 1
1 1
2 2
1
1
1 1 1
1 2 1
2 2 1
1 I 1
2
1 1 1 2
1 1
2
1 2
1 1 1
2 2 2
1 1
2 2
2 2
1 1
1
2
1
2
I = fast migrating band ; 2 = slow migrating band ; 3 = medium migrating band . Morphological characters : a = white flower colour ; of = leaflets converted to tendrils ; bt = pod apex pointed ; cr = crimson flower colour; Cr = creamy testa colour ; dt = short peduncle ; fa = fasciated plant type ; fn = increased numbers of flowers per node ; fr = increased numbers of basal branches ; G = green testa colour; GB = greenish blue testa colour; gp = yellow pod colour ; i = green cotyledon colour; k = wing petal appressed to keel petal ; le = short internode ; M = marbled testa colour ; obs = violet pattern testa colour ; pl = black hilum; r = round and smooth seed ; ram = increased numbers of upper branches ; rms = hard stems ; SL = semi- leafless; wa = plant without wax ; wlo = upper surface of leaflet without wax ; Z = zig zag stems .
Shields et al . (1983) . Following electrophoresis, the gel was sliced into two to three pieces, and each slice was immersed in 50 ml of prepared staining solutions . Each gel slice was stained for a particular enzyme system . The staining procedure for MRD was taken from Burdon et al . (1980), those for ADH and AMP were from Shaw & Prasad (1970) and the remainder from Soltis et al . (1983) . After gels were sufficiently stained (30 min to overnight depending on the enzyme), they were washed with water and fixed with 50% ethanol . The genotype of each accession for each enzyme system was scored from the gels .
Eleven quantitative characteristics of parents (basal branches per plant, plant height, podded nodes per plant, pods per plant, seeds per pod, hundred seed weight, grain yield, total dry matter, harvest index, days to onset of flowering and duration of flowering) were evaluated in field trials in 1989 and 1991 . Three measures of the level of heterosis, Hmp (over the midparent), Hbp (over the better parent), Hcm (over the best parent) were calculated for Fl hybrids and F2 populations (Sarawat et al ., 1994) .
25 8 Table 2 . Genetic distance of Nei . Estimates of genetic distance between each pair of parents, based
on 15 isozyme markers (GD ; ), 17 morphological markers (GD,,,) and the combination of isozyme and morphological markers (GD, + ) Male
Garfield
GDS between male and
GDm between male and
GD ;+,,, between male and
Alma,
Derrimut,
Alma,
Alma,
Dundale
Wirrega
Derrimut, Dundale
Wirrega
Derrimut
Wirrega
Dundale
0 .20 0 .16 0 .16 0 .11
0 .13 0 .10
0 .10 0 .22
0.17 0.12
0 .14 0 .10
0 .12 0 .17
0 .11
0 .17 0 .32
0 .22 0 .32
0.14 0.19
0 .17 0 .21
0 .19 0 .21
0 .27 0 .22
0 .32 0 .22
0.26 0.35
0 .29 0 .32
0 .32 0 .32
0 .13 0 .22
0 .17 0 .22
0 .35 0 .32
0 .32 0 .29
0.35 0 .29
0 .10
0 .10
0 .07
0 .08
0.08
0 .13 0 .13
0 .06 0 .00
0 .12 0 .14
0 .10 0 .12
0.07 0.08
0 .10 0 .17
0 .03 0 .13
0 .17 0 .17
0 .14 0 .14
0.10 0.12
0 .46 0 .38
0 .17 0 .10
0 .22 0 .13
0 .32 0 .26
0 .29 0 .24
0.32 0.24
SA 828 0 .03 Solara 0 .07
0 .07 0 .11
0 .10 0 .22
0 .06 0.17
0 .07 0 .14
0 .08 0 .17
0.07 0.14
Whero
0 .16
0 .06
0.10
0 .12
0 .10
0 .12
SA 15 SA 24
0 .16
SA 35 SA 51
0 .07 0 .11 0 .26 0 .31
SA 54 SA 123
0 .55 0 .80
0 .46 0 .66
SA 129
0 .46 0 .03
0 .38 0 .07
SA 157 SA 236 SA 247 SA 248
0 .11 0 .07 0 .16 0 .20 0 .26
SA 465 0 .16 SA 483 0 .55 SA 688 0 .46
0.20
0 .20 0 .11
Genetic distance estimates
Results
Four measures of genetic distance (GD) were calculated between all 72 parental combinations ; GDi, based on isozymes ; GD,,,, on morphological markers, GDi + ,,,,, on isozymes and morphological markers ; and GDq , on quantitative traits . Allelic frequencies from morphological and isozyme markers were first transformed to is and Os, then the Genetic Distance of Nei (1972) was calculated separately for GDi, GDm and GDi+,,,, . Following orthogonal transformation of the quantitative traits to give principal components Euclidean distance was calculated (Goodman, 1972) . Simple correlation coefficients were computed for all possible types of GD with all measures of heterosis .
Genetic distances A single zone of isozyme activity on the stained gel was found and resolvable for each of the enzymes ADL, G6PD and MRD (Table 1) . Double zones were found for AMP, GOT, IDH, 6-PGD, PGI and SDH but only one of them was resolvable in AMP, GOT, PGI and SDH . Three zones were found for enzymes ADH, MDH and PGM, but again only two zones were readable in MDH and one zone in ADH and PGM . Alma and Dundale were identical at every locus, as were Derrimut and Wirrega . The genetic distance for 72 pairs of parents was in a range of 0 .03 to 0 .80 (Table 2), with the male parent SA 123 most divergent from the female parents and SA 157 the least divergent . The `normal' plant had a full wax cuticle, deep rose-pink flowers, rounded pod apex, green colour of young pods, dimpled seed shape, unpigmented hilum and yellow cotyledon colour. GD,,, between each pair
259 Table 3 . The Euclidean genetic distance (GD q ) between each pair of parents estimated from ten quantitative traits
Female Alma Derrimut
Dundale
Wirrega
Correlation matrix for genetic distance estimated from isozyme markers (GD,), morphological polymorphisms (GD m ), isozyme markers and morphological polymorphisms (GDi+m) and quantitative traits (GDq ) Table 4 .
Female Derrimut Dundale Wirrega
6 .7 4 .6 6 .4
GD, 6 .0 4 .9
7 .4
4 .8
5 .2
13 .0
10 .1
5 .9 16 .1
9 .3
12 .6 6 .2
12 .4 7 .6
Male Garfield SA 15
3 .3 13 .1
SA 24 SA 35
12 .2 5 .0
SA 51
7 .8
6.5 5 .4
8 .3
6 .3
SA 54 SA 123
31 .9 27 .8
28 .0 23 .7
33 .1 29 .1
26 .1 23 .2
SA 129 SA 157
27 .5 4.2 7 .0
23 .2 4 .4
28 .6 6 .1 7 .4
23 .0 3 .2 11 .9
9 .2 7 .6 14 .9
4.9 6.0
SA 236 SA 247 SA 248
7 .1 6 .1
11 .8 6 .1 6 .4
SA 465 SA 483
12 .9 27 .0
10 .0 23 .1
28 .1
11 .2 22 .0
SA 688
15 .4 12 .6
13 .3 12 .2
17 .6 15 .1
11 .0 9 .4
10.8 2 .2
10.0
9 .9 5 .2
12 .9 7 .4
SA 828 Solara Whero
7 .2
of parents estimated from these ranged from zero to 0 .32 . The highest distance was found between SA 35 and all female parents whereas SA 247 and Wirrega were almost identical (Table 2) . Wirrega had a genetic distance of 0 .10 from Alma, Dundale and Derrimut which were identical in all morphological characters . The 15 isozyme and 17 morphological markers were pooled to give a GD based on 32 characteristics (Table 2) . Equal highest distances of 0 .35 were found between the female parents and both SA 54 and SA 123 whereas the lowest distance of 0 .07 was found between both SA 157 and SA 828 and both Alma and Dundale . The Euclidean genetic distance between male and female parents ranged from 2 .2, between Whero and Alma, to 33 .1, between Dundale and SA 54 (Table 3) . The genetic distances among the female parents were in a range of 4 .6 to 7 .4 (Table 3) . Three of the four measures of GD, namely GDi, GDi +m and GD 9 were highly correlated (0 .76 to 0 .89)
GD m
0 .14
GD,+m GD y
0 .89** 0 .78**
GD,,,,
GDi+ m
0.57** 0.23
0 .76**
** :P
