genetic control of root weight, root volume and root to ... papers/pert vol. 11 (2) aug....
TRANSCRIPT
Pertanika 11(2), 165-173 (1988)
Genetic Control of Root Weight, Root Volume and Rootto Shoot Weight Ratio in Peas
GHIZAN B. SALEH1 and EARL T. GRITTON2
1 Department of Agronomy and Horticulture,Faculty of Agriculture,
Universiti Pertanian Malaysia,43400 Serdang, Selangor, Malaysia.
2 Department of Agronomy,University of Wisconsin-Madison,
WI53706, U.S.A.
Key words: Pisum sativum L.; inheritance; heritability; gene effects; root characteristics; root system.
ABSTRAK
Pengawalan genetik atas berat akar, isipadu akar dan nisbah berat akar dengan pucuk telah dikaji bagikacang pea (Pisum sativum L) pada peringkat berbunga. Empat populasi yang digunakan telah dibentukdaripada kacukan di antara warisan-warisan dan kultivar-kultivar yang berbeza saiz sistem akarnya. Tum-buhan daripada generasi Fx, F2, kacukan-balik dan induk telah dinilai dalam pasu yang diisi dengancampuran tanah.pasir, dalam rekabentuk berawak lengkap. Analisis min generasi telah digunakan untukmenentukan pewarisan ciri-ciri akar tersebut. Berat dan isipadu akar memberikan nilai kebolehwarisanluas yang sederhana hingga tinggi (0.41 - 0.81 bagi berat akar, dan 0.44 ~ 0.77 bagi isipadu akar). Kesanpenambah dan kesan dominan penting dalam pewarisan berat dan isipadu akar dalam semua populasi,manakala kesan epistasis hanya penting dalam dua populasi, Kepentingan parameter-parameter genetikdalam pewarisan nisbah berat akar dengan pucuk tidak begitu jelas. Dengan wujudnya kesan penambahyang besar dan nilai kebolehwarisan yang tinggi bagi berat dan isipadu akar, pemilihan untuk warisan-warisan tulen unggulyang mempunyai sistem akar yang besar akan memberikan kesan.
ABSTRACT
The genetic control of root weight, root volume and root to shoot weight ratio was studied in peas /Tisumsativum L.) at flowering. The four populations used were developed from crosses between lines andcultivars differing in size of the root systems. Plants of the Flt F2, backcross and parental generationswere evaluated in pots filled with soil: sand mixture, in a completely randomized design. A generationmeans analysis was used to determine the inheritance of the root characteristics. Root weight and volumeexhibited moderate to high broad-sense heritability (0.41 - 0.81 for root weight, and 0.44 - 0.77 forroot volume). Additive and dominance effects were important in the inheritance of root weight andvolume in all populations, while epistatic effects were important only in two populations. The importanceof the genetic parameters in the inheritance of root to shoot weight ratio was unclear. With the presenceof large additive effects and high heritability, estimates for root weight and volume, selection for superiorpure lines with large root systems should be effective.
INTRODUCTION ancj development in crops have been associatedInspite of the vital roles it plays, the plant root with efficiency of water and nutrient uptake andsystem has not been studied as extensively as the tolerance to water stress and root disease con-above-ground parts. Desirable root characteristics ditions (Dereia et al, 1969; Ekanayake et al.}
GHI2AN B. SALLEH AND EARL T. GRITTON
1985; Epstein, 1972; Hurd, 1974; Steponkusetal, 1980; Taylor and Klepper, 1978).
Very few studies have been conducted onthe variability of the pea (Pisum sativum L.)root system, especially at the mature plant stage.Veitenheimer (1981) indicated the presence ofvariability for various root characteristics mea-sured among diverse pea genotypes grown both inhydroponics and soil. It was also noted thatdifferences in the root characteristics were verypronounced when values were compared atflowering. Ali-Khan and Snoad (1977) reporteda considerable amount of genetic variabilityexisting for root characteristics of 10-day old peaseedlings. Broad-sense heritability estimates forroot weight and number of lateral roots wererelatively high, being 67.5% and 53.7%, respective-ly. High heritability estimates for root characteris-tics were also indicated in other crops includingcommon beans (Fawole et al., 1982), kidnevbeans (Lindgren, 1977), oats (Barbour andMurphy, 1984) and rice (Ekanayake et al, 1985).
No study has been reported to identify kindsand amount of gene effects involved in the controlof root characteristics in peas. In cotton, Eissaet al (1983) found large amounts of additive,dominance and additive x additive epistaticeffects for primary root traits of 16-day oldseedlings. In rice, Ekanayake et al (1985), with50-day old hydroponically grown plants reportedthat additive and dominance effects had equallyimportant contributions to the expression of allroot traits measured.
The objectives of this study were (i) todetermine kinds and amount of gene effects andto estimate heterosis, inbreeding depression andheritability for root weight, root volume androot to shoot weight ratio, and (ii) to determinerelationships among root characteristics andbetween root and other plant characteristics,in peas measured at flowering.
MATERIALS AND METHODS
Experimental Materials and ProceduresParents used in this study were two cultivars andsix breeding lines chosen for their diverse rootcharacteristics revealed in the study by Veinten-heimer (1981). 'Sprite1 and 'Dark Skin Perfection'
(DSP) were the two cultivars used, having verysmall and small root systems, respectively.
'Minnesota 108' (Mn 108) and 'Minnesota 494-A1T (Mn 494-A11) were two root-rot resistant linesused, with relatively small and large root systems,respectively. Four other lines used were 'L 1073',*L 1254' and *L 1482' all with large root systems,and 'L 1532', with small root system.
Four populations were generated from cross-es between the parental lines, namely Population1 (Sprite x L 1482). Population 2 (DSP x L 1973),Population 3 (Mn 108 x L 1254) and Population 4(Mn 494-A11 x L 1532). Initial parental crosseswere made in the greenhouse in February andMarch of 1983. Seeds of the P1? P2) Flt F2 ,BCiPj and B C ^ generations were producedeither in the field or in the greenhouse later inthe same year. Only seeds produced in a commonenvironment were used for evaluation of a popu-lation.
Due to space limitations in the greenhouseand the amount of labour involved in root wash-ing, only two populations were evaluated at atime: Populations 1 and 2 in June to September,while Populations 3 and 4 in September to Decem-ber of 1984. Generations in each population wererepresented by the following number of plants:14 plants for each of the parents and the F^s,30 plants for each of the backcrosses, and 58plants for the F2 's. To minimize reciprocal effect,an equal number of individuals from a cross andits reciprocal was used. The evaluation withineach population was conducted in a completelyrandomized design, in which a plant was a repli-cate for the generation it represented.
Seeds were first germinated in petri disheswith silica sand medium, and then planted inindividual clay pots filled with autoclaved 1:1soil-sand mixture. Each pot contained 3800 cm3
of the medium, with an average bulk density of1.25 g cm"3. Only one plant was grown in eachpot. The pots were placed on capillary wateringmats laid on benches in a temperature-controlledgreenhouse. Water was supplied by wetting themat at regular intervals through dew hoses andcontrolled by an automatic timer. Water statusin the pot was monitored by tensiometers placedin pots on each bench.
Every plant was harvested individually whenthe first flower appeared. The roots were carefullywashed in a screen-lined box placed in a sink filledup with water. Root volume was taken on fresh
166 PERTANIKA VOL. 11 NO. 2, 1988
GENETIC CONTROL OF ROOT WEIGHT, ROOT VOLUME AND ROOT TO SHOOT WEIGHT RATIO IN PEAS
roots of each plant by the water displacementmethod. Root and shoot weights of each plantwere taken on oven dried samples. Leaf area,number of nodes on the main stem, number ofbranches and number of days to flowering werealso recorded for each plant.
Genetical and Statistical ProceduresAnalysis of variance was employed separately foreach population to determine the significance ofthe generation effects on root weight and volumeand root to shoot weight ratio. Mid-parentalheterosis was determined by the differencebetween the ¥x mean and the mid-parental value,expressed as a percentage of the mid-parentalvalue. Inbreeding depression was calculated asthe difference between the ¥x mean and theF 2 mean expressed as a percentage of the Fx
mean.Broad-sense heritability was estimated fol-
lowing the equation:
B
while narrow-sense heritability was estimatedfollowing the equation:
2 2
where hg and h N are the broad-sense and
narrow-sense heritabilities, respectively, andVF2> V P ^ V P 2 ' V F r yBCx?x
a n d VBCX?2 a r e
variances associated with the F 2 , Pi , P2 , F j ,BCiPi and BC xP2 generations, respectively.
Where generation effects were significant,the generation means analysis procedure was usedto estimate gene effects for each of the rootcharacteristics, following the model and assump-tions outlined by Mather and Jinks (1971). Meanvalues of each of the six generations were used forthe analysis. The expected relationships amonggenetic parameters were tested by the joint-scalingtest procedure as described by Rowe andAlexander (1980). A weighted least square proce-dure (Mather and Jinks, 1971) was used in thegeneration means analysis to compensate for thedifference in precision involved in the estimationof mean values among the six generations. Eachcharacteristic in each population was tested for
fit to the additive-dominance model by the Chi-square procedure. Where the model did not fit,the analysis was further carried out with com-parisons to fit models that include digenic epistaticparameters in the expectations. Values for indi-vidual genetic parameters were considered sig-nificant if they exceeded twice their standarderrors.
Simple correlation coefficients were cal-culated among root characteristics and betweenroot and other plant characteristics using datafrom the F 2 generation.
RESULTS AND DISCUSSION
Generation mean squares in the analyses ofvariance were significant for all root characteris-tics, root weight, root volume and root to shootweight ratio in all four populations, indicatingthat phenotypic variation exists among gene-rations. Means and variances of these chararacteris-tics are shown in Table 1. The more favourablesummer growing conditions at the time Popu-lations 1 and 2 were evaluated led to larger meanvalues compared to the less favourable fall growingconditions at the time Populations 3 and 4 wereevaluated.
Estimates of mid-parental heterosis weregenerally high for root weight and volume (rangedfrom 10.6 to 126,2 percent) in all four popu-lations (Table 2), indicating a preponderance ofdominance towards the larger parents was presentfor these traits. The two root traits showed almostsimilar estimates of heterosis. Population 4 exhibi-ted the highest degree of heterosis, being 115.9and 126.2 percent for root weight and volume,respectively. The results showed good agreementwith high heterosis and inbreeding depressionfor yield and yield components in peas, as re-ported by Gritton (1975) and Krarup and Davis(1970). Heterosis for root to shoot weight ratiowas not consistent; positive in Populations 1 and4, but negative in Populations 2 and 3. Estimatesof inbreeding depression for root weight andvolume were high in Populations 3 and 4 (rangedfrom 28.8 to 30.1 percent), but low in Population1 (6.1 and 7.9 percent for root weight and volume,respectively). Estimates of inbreeding depressionfor root to shoot weight ratio were generally lowor negative. Negative estimates of inbreedingdepression for all traits in Population 2 were due
PERTANIKA VOL. 11 NO. 2, 1988 167
GHIZAN B. SALLEH AND EARL T. GRITTON
TABLE 1Generation means and variances for root characteristics measured at flowering in four pea populations.
Generation
Population 1
L1482 (Pj)Sprite (P2)
p
B C l P 2
Population 2
I 1073 (Pj)DSP (P2)F,
BCiPj
Population 3
L1254(P 1 )Mn 108 (P2)Pj
^2
BCX?2
Population 4
Mn494-All(P1)L1532(P2)Fj
^2
BC! P2
Root Weight
Mean
g
1.68afl
0.55c1.38ab1.29b1.67a0.84c
2,77ab1.27d2.23bc2.79a2,78ab2.04c
0.99b0.47e1.13ab0.80c1.15a0.67d
0.67d0.60d1.38a0.98c1.14b0,93c
Variance
0.1300.0160.2630.3210.1860.165
1.3080.0860.1211.2400.5620.411
0.0580.0150.1140.0790.0590.028
0.0380.0340.0540.0880.0690.091
Root Volume
Mean
cm3
32.2a10.4c26. lab24.1b29.3a16.9c
42.9a23.2c37.6ab41.5a42.5a33.2b
16.3a7.1c
17.6a12.3b18.0a10.5b
11.Od9.1d
22.8a15.9c18.8b14.7c
Variance
46.206.43
90.2093.7434.1948.34
166.1316.8517.82
161.4886.1397.98
16.042.50
36.6520.4920.42
8.48
7.307.91
20.1623.6919.5224.69
Root to shootweight ratio
Mean
0.95a0.48d0.66bc0.65bc0.73b0.53cd
0.55a0.50abc0.42c0.48abc0.5 lab0.46bc
0.30a0.27ab0.26ab0.25b0.26ab0.26ab
0.32a0.15d0.30ab0.26b0.31a0.23c
Variance
0.0400.0150.0580.0470.0300.045
0.0190.0180.0050.0160.0110.031
0.0010.0040.0070.0050.0040.004
0.0030.0020.0030.0030.0060.004
aMeans followed by the same letters in the same column and population are not significantly different at P = 0.05 level.
to values of F 2 means larger than those of theF^s(Table:) . *
Broad-sense heritability estimates for rootweight and volume were moderate to high (0.75,0.81. 0.41 and 0.53 in Populations 1, 2, 3 and 4,respectively, for root weight; and 0.68, 0,77,0.44 and 0.5b in Populations 1, 2, 3 and 4, res-
\ely, for root volume), indicating the pre-seiue of relatively high genetic variance for thesetraits (Table 2). Broad-sense heritabilities for rootto shoot weight ratio were low to moderate
If>s PERTANikA \ o i
(0.31, 0.25, 0.24 and 0.19 in Populations 1, 2,3 and 4, respectively).
Narrow-sense heritability estimates werehighly variable for all characteristics measured(Table 2). Estimates ranged from 0.18 to 0.91 forroot weight, 0.13 to 1.00 (calculated value is 1.12)for root volume and zero (calculated value is-1.23) to 0.40 for root to shoot weight ratio.The wide range in values (including negativeestimates and estimates exceeding 1.00) suggeststhat the estimates were very imprecise. Con-n NO .\ 1988
GENETIC CONTROL OF ROOT WEIGHT, ROOT VOLUME AND ROOT TO SHOOT WEIGHT RATIO IN PEAS
TABLE 2Estimates of mid-parental heterosis, inbreeding depression and heritability for root characteristics measured at
flowering in four pea populations.
PopulationMid-parental
heterosis
Inbreeding
depression hB
Root weight
Pop. 1 (Sprite x L 1482)
Pop. 2 (DSP xL 1073)
Pop. 3(Mnl08xL1254)
Pop. 4(Mn494-AUxL1532)
Root volume
Pop. 1
Pop. 2
Pop. 3
Pop. 4
Root to shoot weight ratio
Pop. 1
Pop. 2
Pop. 3
Pop. 4
23.2
10.6
55.2
115.9
22.8
13.7
49.8
126.2
7.7-20.0-6.827.7
6.124.8
29.528.8
7.7
-10.4
30.1
30.1
1.5-14.3
3.8
13.3
0.750.81
0.41
0.53
0.68
0.770.44
0.56
0.310.250.24
0.19
0.911.22
0.900.18
1.12
0.86
0.59
0.13
0.40-0.63
0.33-1.23
: broad-sense heritability, hXT - narrow-sense heritability.
sequently, we make no inference regarding thenarrow-sense heritabilities.
From the generation means analysis, it wasfound that an additive dominance model wasadequate in explaining the variation in generationmeans for root weight and volume in Populations1 and 4, as indicated by the non-significant Chi-square values in the joint-scaling tests (Table 3).Both additive [d] and dominance [h] parameterswere significant for these characteristics in bothpopulations, as shown by their values larger thantwice their standard errors. This indicates thatsignificant additive and dominance gene effects,in the direction of the larger parents were presentfor root weight and volume in Populations 1 and4. Magnitude of additivity was larger than thatof dominance in Population 1, but was the oppo-site in Population 4.
The five-parameter model fits in explainingthe genetic control of root weight and volume in
Populations 2 and 3 (Table 3). Additive effects[d] were significant for both characteristics inboth populations. Dominance effects [h] weresignificant and in the direction of the largerparents for these two characteristics only inPopulation 3. Dominance was towards the smallerparent in Population 2 (negative values of [h]).These negative estimates of dominance were notexpected and were contrary to the presence ofheterosis for these characteristics in Population 2,as shown in Table 2. These estimates of dominanceeffects, was thus, felt unreliable. The negativeestimates might have been caused by mean valuesof the F 2 generation higher than the Fx 's (Table1). It might have also been possibly caused by thepresence of linkage or generation x environmentinteraction. Linkage between the root characteris-tics and other traits in any of the two parents in apopulation could have led to the weakness of thegeneration means analysis (Jinks, 1979). Attempts
PERTANIKA VOL. 11 NO. 2, 1988 169
TABLE 3Estimates of gene effects in three-parameter and five-parameter genetic models for root characteristics measured at flowering in four pea populations.
szo
o
Population
Root weight
Pop. 1 (Sprite x L 1482)
Pop. 2 (DSP xL 1073)
Pop. 3(Mnl08xL1254)
Pop. 4(Mn494-AllxL1532)
Roof volume
Pop. 1
Pop. 2
Pop. 3Pop. 4
Root to shoot weight ratio
Pop. 1
Pop. 2
Pop. 3
Pop. 4
1.16 ±0.05*
3.28 ±0.26
0.45 ±0.12
0.64 ±0.03
21.50 ±0.87
44.89 ±3.18
6.52 ±1.96
10.07 ±0.51
0.66 ±0.04
0.54 ±0.02
0.28 ±0.01
0.23 ±0.01
0.62 ±0.05
0.74 ±0.12
0.27 ±0.04
0.07 ±0.03
11.27 ±0.83
9.52 ±1.46
4.68 ±0.59
1.48 ±0.50
0.19 ±0.04
0.04 ±0.02
0.01 ±0.01
0.08 ±0.01
0.21 ±0.10
-1.05 ±0.32
0.75 ±0.18
0.74 ±0.07
4.11 ±2.01
-7.40 ±3.61
12.06 ±3.18
12.77 ±1.10
-0 .04 ±0.08
- 0.11 ±0.03
- 0.04 ±0.02
-0 .06 ±0.02
l~ 1.27 ±0.31
0.30 ±0.12
-12.17 ±3.74
5.34 ±2.02
•
-0.02 ±0.51
0.44 ±0.19
-1.13 ±5.41
5.89 ±2.42
x 2 '
5.25
0.17
1.35
5.66
0.97
0.12
1.23
6.82
0.89
0.77
2.03
0.92
P
0.10-
0 .90-
0 .10-
0.10-
0 .75-
0 .90-
0.25 -
0 .05-
0 .75-
0.75-
0.50-
0.75 -
c
0.25
0.95
0.25
0.25
0.900.95
0.50
0.10
0.90
0.90
0.75
0.90
^Significance of each gene effect estimate can be determined by t-test procedure, where value of t is found by dividing each estimate by its SE.
[m] = mean effect, [d] = additive gene effect, [hj = dominance gene effect, [i] = additive x additive gene effect, [j] • additive x dominance gene effect.Dominance x dominance gene effect [ 1 ] is not included because the estimate is not significant in all cases.
c X denotes Chi-square value for testing goodness of fit of model; P denotes probability leveL &
Om
8
S5
I5
cm
3
a
i6
GHIZAN B. SALLEH AND EART T. GRITTON
were made to adjust the values using a commonweighting factor for all generations, but was foundnot to have changed the estimates significantly.Likewise, due to the same reason, estimates ofepistasis for these two traits in Population 2 werenegative. However, only additive x additive epista-tic parameter [i] was significant for both traits.In Population 3, both additive x additive [i] andadditive x dominance [i] epistatic effects weresignificant for root weight and volume.
An additive-dominance model was adequatein explaining the variability in root to shootweight ratio in all four populations (Table 3).However, the importance of the genetic para-meters in the control of this trait was not consis-tent with populations, and therefore, inconclusive.
Simple correlation coefficients among theroot characteristics, and between the root andother plant characteristics are presented in Table4. In all populations, root weight and volume werehighly correlated (r values ranged from 0.832 to0.963). Similarly, high correlations were alsofound by Veitenheimer (1981) who utilizeddiverse pea genotypes in the study. No consistentcorrelation was found between root to shootweight ratio and size of the root system. Root toshoot weight ratio was positively correlated withroot weight in Populations 3 and 4 (r = 0.345 andr = 0.424, respectively), but was not correlated inPopulations 1 and 2 (r = 0.076 and r = 0.248, res-pectively). Similarly, root to shoot weight ratiowas correlated with root volume in Populations 2and 4 (r = 0.305 and r = 0.360, respectively), butwas not correlated in Populations 1 and 3(r = -0.011 and r = 0.006, respectively).
In all populations, in general, correlationsof root weight and volume with all shoot charac-teristics measured were positive, except for thecorrelations between root weight and number ofbranches in Population 3, and between rootvolume and number of branches in Population 4,which in both cases were not significant (Table 4).This indicates that, in general, plants with largerand heavier root systems had larger and heaviershoots and larger total leaf area than those withsmaller and lighter root systems, althoughexceptions were found. As expected, root to shootweight ratio was negatively correlated with shootweight in all populations. Number of days toflowering in Populations 1 and 2 was highly,
positively correlated with root weight and volume,indicating that, in these populations, plants thatflowered earlier had smaller root systems thanthose that flowered later. In contrast, however, inPopulation 4, days to flowering was highly, nega-tively correlated with root weight and volume. InPopulation 3, days to flowering was positivelycorrelated only with root volume, but not corre-lated with root weight. Days to flowering wasnegatively correlated with root to shoot weightratio in Populations 3 and 4, but was not corre-lated in Populations 1 and 2.
CONCLUSIONS
The significant additive and dominanceeffects for root weight and volume showed theimportance of these parameters in the geneticcontrol of these two characteristics, althoughepistatic effects were also important in somepopulations. Large additive effects and highheritability estimates indicated that selection forsuperior lines with large root systems should beeffective. Maximum utilization of additive effectscan be obtained by selection after each generationof selfing. Selection procedures using any kindof non-destructive methods like the use of hydro-ponically grown plants or partial root harvest,and simultaneously selecting for other desirabletraits could be employed. Selections may bedelayed until later generations in populationswhere additive x additive gene effect is important.The tendency of the larger root systems to beassociated with later flowering plants was notabsolute, thus, effective selection for plants withlarge root systems but without concommitantlateness should be possible.
REFERENCES
ALI-KHAN, S.T., and B. SNOAD. (1977): Root andshoot development in peas. L Variability in sevenroot and shoot characters of seedlings. Ann. Appl.Bioi 85: 131-136.
BARBOUR, N.W., and C.F. MURPHY. (1984): Fieldevaluation of seedling root length selections in oats.CropSci. 24: 165-169.
DERERA, N.F., D.R. MARSHALL, and L.N. BALAAM.(1969): Genetic variability in root development inrelation to drought tolerance in spring wheats.ExplAgric. 5:327-337.
EISSA, A.M., J.N. JENKINS, and C.E. VAUGHN.(1983): Inheritance of seedling root length and
PERTANIKA VOL. 1! NO. 2, 1988 171
TABLE 4Simple correlation coefficients between pairs of root characteristics and between root
and other plant characteristics measured at flowering in four pea populations
RV RSWR SW LA NN NB
PE
RI
>
1§
i00
Population 1
(Sprite x L 1482)RW
RV
RSWR
Population 2
(DSPx L 1073)
RW
RV
RSWR
Population 3
(Mn 108 x L 1254)
RW
RV
RSWR
Population 4
(Mn494 All x L 1532)
RW
RV
RSWK
0.963**c
O.93O*<
0.832**
0.936* 0.424**
0.360**
0.735**
0.715**
-0.275*
0.621**
0.664**
-0.260
0.523**
0.467**
-0.101
Correlation coefficient differs significantly from zero based on the 0.05 or 0.01 level of probability, respectively.a RW = root weight, RV = root volume, RSWR • root to shoot weight ratio, SW = shoot weight, LA • leaf area, NN" stem, NB = number of branches, DF = days to flowering.b All correlation coefficients are based on 58 F 2 individuals.
DF*
0.076
-0 .011
0.248
0.305*
0.345**
0.006
0.773**
0.798**
-0.492**
0.846**
0.750**
-0.272*
0.610**
0.659**
-0.451**
0.543**
0.584**
-0.545**•','
0.790**
0.694**
-0.321*
0.658**
0.728**
-0.357**
0.821**
0.765**
-0.181
0.813**
0.694**
0.021
0.413**
0.496**
-0.473**
0.425**
0.471**
-0.502**
0.479**
0.549**
-0.182
0.205
0.290*
-0.343*
0.745**
0.653**
-0.139
0.793**
0.638**
0.091
0.147
0.312*
-0.577**
GH
IZA
N B
. ! 3AL
LE
H J
a' EA
RT
T. G
B
i
0.275* -0.592**
0.259 -0.522**
-0.139 -0.484**
• number of nodes on main
GENETIC CONTROL OF ROOT WEIGHT, ROOT VOLUME AND ROOT TO SHOOT WEIGHT RATIO IN PEAS
relative root weight in cotton. Crop Sci 23: 1107 —1111.
EKANAYAKE, I.J., J.C. O'TOOLE, D.P. GARRITY, andT.M. MASAJO. (1985): Inheritance of root charac-ters and their relation to drought resistance in rice.Crop Sci. 25:927-933.
EPSTEIN, E. (1972): Mineral nutrition in plants: Prin-ciples and perspectives. John Wiley & Sons* Inc., NY.
FAWOLE, L, W.H. GABELMAN, and G.C. GERLOFF.(1982): Genetic control of root development inbeans (Phaseolus vulgaris L) grown under phos-phorous stress, i. Am. Soc. Hort Sci 107: 98-100.
GRITTON, E.T. (1975): Heterosis and combining abilityin a diallel cross of peas. Crop Sci 15: 453-457.
HURD, E.A. (1974): Phenotype and drought tolerancein wheat. Agric. Meteorol 14: 39-55.
JINKS, J.L. (1979): The biometrical approach to quanti-tative variation, p. 81-109. In J.N. Thompson, Jr.,and J.M. Thoday (ed.) Quantitative genetic variation.Academic Press, Inc., NY.
KRARUP, A., and D.W. DAVIS. (1970). Inheritance ofseed yield and its components in a six-parent diallelcross in peas. /. Am. Soc. Hort Set 95: 795-797.
LINDGREN, D.T. (1977): Genetic variability of rootmorphology in kidney beans. Annu. Rep. BeanImprovement Cooperative. 20: 23-24.
MATHER, K., and J.L. JINKS. (1971): Biometricalgenetics. Cornell University Press, Ithaca, NY.
ROWE, K.E., and W.L. ALEXANDER. (1980): Compu-tations for estimating the genetic parameters in joint-scaling tests. Crop Sci. 2(h 109-110.
STEPONKUS, P.J., J.M. CUTLER, and J.C. O'TOOLE.(1980): Adaptation to water stress in rice. p. 401 -418. In N.C. Turner and PJ. Kramer (ed.) Adap-tation of plants to water and high temperature stress.John Wiley & Sons Inc., NY.
TAYLOR, H.M., and B. KLEPPER. (1978): The role ofrooting characteristics in the supply of water toplants. Adv. Agron. 30: 99-128.
VEITENHEIMER, E.E. (1981). Pea (Pisum sativum L.)root growth and development as affected by geno-type, media and Aphanomyces euteiches Drechs.M.S. thesis (Unpublished). Univ. Wisconsin Library,Madison, WI.
(Received 28 December 1987)
PERTANIKA VOL. 11 NO. 2, 1988 173