Acta agriculturae Slovenica, 117/4, 1–10, Ljubljana 2021 doi:10.14720/aas.2021.117.4.2119 Original research article / izvirni znanstveni članek Inheritance of plant height, straw yield and flag leaf area in MBB x Gavi- ota durum wheat (Triticum durum Desf.) cross Manel SALMI 1, 2 , Zine El Abidine FELLAHI 3 , Abdelkader BENBELKACEM 4 , Amar BENMAHAMMED 5 , Hamenna BOUZERZOUR 5 Received March 02, 2021; accepted November 02, 2021. Delo je prispelo 2. marca 2021, sprejeto 2. novembra 2021 1 University of Mostefa Ben Boulaid, Life and Natural Sciences Faculty, Ecology and Environment Department, Batna, Algeria 2 Corresponding author, e-mail: salmi.mnl@gmail.com 3 University of Mohamed El Bachir El Ibrahimi, Natural, Life and Earth Sciences and the Universe Faculty, Agronomy Department, Bordj Bou Arréridj, Algeria 4 Algerian National Institute of Agronomic Research (INRAA), Research Unit of Constantine, Algeria 5 University of Ferhat Abbas, Life and Natural Sciences Faculty, Ecology and Plant Biology Department, Setif, Algeria Inheritance of plant height, straw yield and flag leaf area in MBB x Gaviota durum wheat (Triticum durum Desf.) cross Abstract: Plant height, straw mass and flag leaf area are recognized by physiologists as morphological markers of drought stress tolerance. Developing varieties intended for arid and semi-arid zones need to select for these traits. Understand- ing the genetic control of a given trait helps breeder to handle the segregating populations under study in a more efficient and consistent manner by choosing the best breeding method avail- able to realize significant genetic advance. For this purpose, six generations: parents, F1, F2, BC1, BC2, derived from MBB x ‘Gaviota’ durum wheat (Triticum durum Desf.) cross were grown to investigate the nature of gene action involved in the inheritance pattern of the three traits. The results indicated that the six-parameter model fitted the best the data related to the variability present in the generation means of the studied traits. Generation mean analysis indicated that non-allelic interac- tions were important factors controlling the expression of these characters with complementary type of gene action governing FLA and STW inheritance. High heritability estimates, moder- ate to high expected responses to selection, significant genetic correlations with grain yield and greater role of non-additive effects in controlling the inheritance of the three studied traits suggested that breeding methods exploiting both fixable and non-fixable components be applied to break unfavorable link- age and to accumulate useful genes in the base population, fol- lowed by mono-trait or index based selection in late advanced generations. Key words: gene effects; non-allelic interaction; durum wheat; plant height; straw yield; flag leaf area; heritability Višina rastlin, masa slame in površina lista zastavičarja so od fiziologov prepoznane morfološke lastnosti, ki nakazujejo to- leranco na sušni stres Izvleček: Višina rastlin, masa slame in površina lista za- stavičarja so od fiziologov prepoznane morfološke lastnosti, ki nakazujejo toleranco na sušni stres. Sorte, vzgojene za sušna območja morajo biti izbrane glede na lastnosti, ki omogočajo prenašanje suše. Razumevanje genetske kontrole za določeno lastnost pomaga žlahtniteljem uravnavati različne populacije v raziskavi na bolj učinkovit in verodostojen način pri izbiri najboljše metode žlahtnenja za dosego pomembne genetske prednosti. V ta namen je bilo gojeno šest generacij rastlin iz križanja starševske generacije, F1, F2, BC1, BC2 z MBB x ‘Ga- viota’ trde pšenice (Triticum durum Desf.) kot osnova za pre- učevanje delovanja genov, ki so vključeni v vzorec dedovanja treh lastnosti. Rezulatati so pokazali, da se je model šestih para- metrov najbolje prilegal podatkom, povezanih z variabilnostjo preučevanih lastnosti kot povprečje v generaciji. Analiza gene- racijskih povprečij je pokazala, da so bile nealelske interakcije pomemben dejavnik za nadzor izražanja tistih lastnosti, ki so komplementarne delovanju genov in ki vodijo FLA in STW dedovanje. Velike vrednosti dednosti, zmerne glede na priča- kovan odziv selekcije, značilna genetska korelacija s pridelkom zrnja in večja vloga neaditivnih učinkov pri kontroli dednosti treh preučevanih lastnosti nakazujejo, da bi žlahtniteljske me- tode, ki uporabljajo vezane in nevezane komponente lahko bile uporabljene za prekinitev nezaželjenih povezav genov in pospe- šitev uporabnih genov v osnovnih populacijah, ki bi jim sledil izbor posameznih ali indeksiranih lastnosti v kasnejših izbolj- šanih generacijah. Ključne besede: učinki genov; nealelske interakcije; trda pšenica; višina rastlin; pridelek slame; površina lista zastavičar- ja; dednost Acta agriculturae Slovenica, 117/4 – 2021 2 M. SALMI et al. 1 INTRODUCTION In the arid and semi-arid regions, rainfed grown du- rum wheat (Triticum durum Desf.) crop suffers from the combined effects of drought and heat stresses, undergo- ing substantial grain yield losses (Royo et al., 2014; Lui et al., 2015). To minimize yield decline, drought toler- ance improvement is then seen as a key breeding com- ponent for the development of cultivars devoted to these environments, where reducing wind erosion and meet- ing livestock feeding requirement have heightened the importance of increasing straw production along with grain yield. In fact, crop residues are usually not burned but either maintained to protect soil from wind erosion or grazed and straw is balled, stored and fed to livestock during the winter months (Annichiarico et al., 2005; Chenaffi et al., 2011). In this context plant height, a straw yield correlated trait, is seen as an important character- istic influencing cultivars adoption under such growing conditions (Rabti et al., 2020; Jatayev et al., 2020; Had- dad et al., 2021). Because of their straw yield advantage, tall varieties, derived from land-races, are still cultivated and not replaced by newly released reduced height cul- tivars (Rabti et al., 2020). For a full expression of their potential, dwarf wheat varieties need to be grown in fa- vorable, well-watered conditions which permit applica- tion of relatively high levels of nitrogen fertilizer which is usually not the case in arid and semi-arid environments. This type of plant material becomes, under drought stress conditions, too short, yielding less than tall varieties, showing reduced flag leaf area, kernel size, kernel Zn, Fe, Mg and Mn concentration, coleoptiles, and roots length (Aziz et al., 2017; Velu et al., 2017; Jatayev et al., 2020; Rabti et al., 2020). Positive relationship between plant height and grain yield, under drought stress growing conditions, implied that there is a minimum height be- low which grain yield limitation becomes evident (Slafer et al., 2005; Royo et al., 2014). in this context, Yani and Rashidi (2012) reported that straw yield and plant height had positive and significant role in the expression of grain yield under drought conditions. Asadi et al. (2019) noted that straw yield explained 65 % of grain yield variation and exerted substantial direct effect on grain yield under water deficit conditions. Belagrouz et al. (2018) reported significant correlation between grain yield water use effi- ciency, harvest index and plant height, suggesting that se- lection for plant height and harvest index could improve both water use efficiency and grain yield under drought prone environments. Flag leaf is the main photosynthetic organ providing the major assimilate required for spike growth. It senses environmental signals and consequent- ly adapts to surrounding environment by minimizing area reduction and delaying senescence, caused by termi- nal drought stress (Farook et al., 2009; Belkherchouche et al., 2015). Joshi et al. (1984) found that flag leaf size was positively correlated with grain yield, suggesting that optimal flag leaf dimensions could be an important breeding target under drought prone environments. In- formation on straw yield, plant height and flag leaf area inheritance can assist developing adapted cultivars for areas practicing cereal-livestock farming systems and conservation agriculture (Chenaffi et al., 2011; Jatayev et al., 2020). Genetic variation for quantitative characters in segregating population is of prime concern to breeders. It determines selection efficiency which depends upon the nature and magnitude of genetic variability available. Genetic models devoted to the estimation of different genetic effects have been developed. Among these mod- els, generation means analysis provides information on the relative importance of gene effects due to additive, dominance deviations and non-allelic interactions, in determining generation means (Mather and Jinks, 1982; Shayan et al., 2018; Salmi et al., 2019). Plant height was found to be controlled mainly by over dominance while flag leaf area was reported to be mostly under additive control combined with partial dominance and epistatic gene actions (Saleem et al., 2005; Shabbir et al., 2012; Yang et al., 2016). Salmi et al. (2019) found that domi- nance acted in the direction of increased plant height. Joshi and Sharma (1984) reported that genes affecting smaller leaves are partially dominant over genes affecting larger leaves. Evidence for non-allelic interactions was reported and linkage among loci appeared to be an im- portant component of flag leaf dimensions heredity (De Pace et al., 2001). The present study aimed to investigate the inheritance pattern of straw yield, plant height and flag leaf area in a durum wheat cross involving a tall and a semi-dwarf varieties. 2 MATERIALS AND METHODS 2.1 SITE, PLANT MATERIAL, AND EXPERIMEN- TAL DESIGN Two durum wheat (Triticum durum Desf.) varieties, namely Gaviota (GTA), a semi-dwarf cultivar derived from ‘Crane’ /4/ ‘Polonicum PI 185309’//Triticum gluti- nosum enano/2 * ‘Tehuacan 60’/3/’Grulla’ cross (http:// wgb.cimmyt.org/gringlobal/AccessionDetail.aspx?id= 1783 ), and a tall cultivar, Mohammed Ben Bachir (MBB), a head-row selection derived from a land race adapted to Setif’s high plateaus region (Algeria), were hybrid- ized during the 2015-2016 cropping season at the Field Crop Institute, Agricultural Experimental Station of Setif (ITGC, AES, 36°12’N 05°24’E, 1080 m above sea level, Acta agriculturae Slovenica, 117/4 – 2021 3 Inheritance of plant height, straw yield and flag leaf area in MBB x Gaviota durum wheat (Triticum durum Desf.) cross Setif, Algeria). During the 2016-2017 cropping season, F 1 was selfed to produce F 2 , and crossed to the parents to obtain back cross generations (BC 1 and BC 2 ). The parents were crossed again to get the first filial generation (F 1 ). The following cropping season (2017-2018), the six ba- sic generations were grown in a randomized completed block design, with five replications. Parents, F 1 , and BC’s generations were sown in one row, 2 m long, 20 cm inter- row spacing and 10 cm plant-plant spacing in the row. F 2 generation was sown in thirty rows 2 m long. Recom- mended cultural practices for the area were followed as reported in Salmi et al. (2019). 2.2 DATA COLLECTION AND ANALYSIS Data were collected from 5, 5, 10, and 30 plants per replication for the parents, F 1 , BC’s, and F 2 generations, respectively. Prior to harvest, length of the main stem of each plant was measured from the ground level to the base of the spike and recorded as plant height estimate in cm (PHT, cm). Straw mass (STM, g plant -1 ) was deter- mined as the difference between above ground plant bio- mass (BIO, g plant -1 ) and plant grain yield (GY , g plant -1 ). Flag leaf area (FLA, cm²) was estimated by the product of leaf length x leaf width x 0.749 (Spagnoletti-Zeuli and Qualset, 1990). Collected data were subjected to an analysis of variance using Cropstat software (2007) to test generation effect. Whenever this effect, tested against the residual mean square, was significant, genetic analysis for the specific trait was undertaken. To test the presence of additive vs. dominance genes effects, contrast method (Steel and Torrie, 1982) was applied to check the signifi- cance of the following comparisons: F 1 vs. mid-parent, P 1 vs. P 2 , F 2 vs. average BC’s, and BC 1 vs. BC 2 . The notations adopted for gene effects were [m], [d], [h], [i], [j], and [l] representing main, additive and dominance gene effects, and additive × additive, additive × dominance, and dom- inance × dominance epistatic interactions, respectively. The appropriate genetic model (three vs. six parameters) was also determined using both ABCD and joint scal- ing tests. These tests provide information regarding the absence or the presence of gene interactions (Mather and Jinks, 1982). Significance of any one or both scaling tests implies inadequacy of the additive–dominance model. The C and D scaling tests provide check for dominance x dominance (l) and additive x additive (i) types of epis- tasis, respectively. The genetic parameters [m], [d], [h], [i], [j], and [l] were estimated by weighted least square method. The purpose of using weights was to account for differential precision with which means of different generations were estimated based on varying sample size. Gene effects were tested for significance using the t-test (Kearsey and Pooni, 1996). Three vs six-parameter models testing ware performed using GENMEANS sub- routine implemented in Tnaustat software (Manivannan, 2014). Genotypic and environmental variance compo- nents, of the measured traits, were estimated by equating the observed values of the different generations, accord- ing to Mather and Jinks, (1982) as follows: σ² E = ¼ (σ² P1 + σ² P2 + 2σ² F1 ), σ² D = (2σ² F2 – σ² BC1 – σ² BC2 ) and σ² H = 4 (σ² F2 -½σ² D - σ² E ). The significance of the mean value of a particular parameter was tested against its correspond- ing standard error, via a Student’s t-test, as suggested by Mather and Jinks (1982). Broad-sense heritability (H² bs) was calculated according to Kearsey and Pooni, (1996), as follow: H²bs = (σ 2 D + σ 2 H ) / (σ 2 D + σ 2 H + σ 2 E ) = (σ 2 G ) / (σ 2 P ), where σ 2 D , σ 2 H , σ 2 E , σ 2 G , and σ 2 P stand for the ad- ditive, dominance, environmental variance components, genetic, and phenotypic variances, respectively. Narrow sense heritability (h²ns) was estimated according to Hal- lauer and Miranda Filho (1989) as follow: h² ns = σ 2 D / (σ 2 D + σ 2 H +σ 2 E ) = (σ 2 D ) / (σ 2 P ), Standard errors (SE) of these estimates were calculated as: SE (h2bs) = [SE (σ 2 G )] / (σ 2 P ) and SE (h2ns) = [SE (σ 2 D )] / (σ 2 P ). Significance of these pa- rameters, h² bs and h² ns was tested using a t-test equals to the ratio of heritability over its standard error (Halloran et al., 1979). The expected response to selection (ERS) was derived according to Sing and Chaudhary (1999) as follows: ERS = 2.06 * h² bs * √σ² F2, and expressed as per- cent of the over mean (X bar ) of the given trait: ERS (%) = (100 * ERS) / X bar . Relationship between studied traits and grain yield was inspected through genotypic corre- lation coefficient (rg), which was derived as the ratio of covariance to the square root of the product of the cor- responding variance of the two traits considered. Geno- typic covariance was determined using the property of the analysis of variance of the sum of two variables as suggested by Kwon and Torrie (1964) and described in Mansouri et al. (2018), using Past software (Hammer et al., 2001). The standard error of rg was derived using the formulae of Reeves (1955), reported by Koots and Gibson (1996), as follows: SE rg = [(1- rg²) / √2] * [(√SE h²i * SE h²j ) / (√h²i*√h²j)], where h²i and h²j are the traits heritability’s. Student’s t-test was used to determinate the significance of the correlation coefficient. 3 RESULTS AND DISCUSSION 3.1 V ARIATION AND MEAN PERFORMANCES Significant generation effect was revealed for flag leaf area (FLA), plant height (PHT) and straw mass (STM) by the analysis of variance, indicating the pres- ence of substantial genetic variability and allowing to Acta agriculturae Slovenica, 117/4 – 2021 4 M. SALMI et al. carry out in deep bio-metrical analysis (Table 1). A sig- nificant generation effect is a prerequisite to perform the inheritance study of the targeted traits applying genera- tion means analysis model. Among generations traits mean estimates, varied from 19.67 to 29.70 cm² for FLA, from 82.00 to 136.25 cm for PHT , and from 12.32 to 31.98 g plant -1 for STM (Figures 1, 2 and 3). Contrast analysis indicated that deviations between parental mean values for the studied traits were significant (Table 1). ‘Gavi- ota’ had significantly greater FLA (29.70 cm²) than MBB (26.08 cm²), while MBB showed significantly greater PHT (136.25 cm) and STM (31.55 g plant -1 ) than ‘Gavi- ota’ whose mean values for both traits were 86.87 cm and 15.24 g/plant, respectively (Figure 1, 2 and 3). These results suggested that the crossed parents carry different allelic combinations involved in the ge- netic control of the studied traits. On average, F1 ex- hibited equal FLA (29.35 cm²) to the best parent GTA (29.70 cm²), equal PHT (82.00 cm) to the semi-dwarf parent GTA (86.87 cm) and equal STM (31.98 g plant -1 ) to the tall parent MBB (31.55 g plant -1 ), within the lim- its of their standard errors (Figures 1, 2 and 3). Further- more, F1 means differed significantly from mid-parent average for PHT and STM but not for FLA, suggesting that dominance was predominantly involved in the ge- netic control of PHT and STM; while additive genetic control was predominantly expressed for FLA (Table 1, Figures 1, 2 and 3). Dominance acted in the direction of Sources of variation DF FLA (cm²) PHT (cm) STM (g/plant) Generations 5 89.56 ** 2091.65 ** 398.08 ** Replications 4 9.56 8.81 25.00 Homogeneous (Homo) 2 21.56 ** 4504.12 ** 455.19 ** P1 vs P2 1 8.10 ** 1190.50 ** 18.10 ** F1 vs ½(P1+P2) 1 2.55 ns 568.92 ** 6.69 * Heterogeneous (Het) 2 20.38 ns 579.97 ** 164.99 ns F2 vs ½(BC1+BC2) 1 4.95 * 177.96 ** 8.88 ** BC1 vs BC2 1 16.57 ** 9.17 ** 1.11 ns Homo vs Het 1 89.86 ** 56.58 ** 20.43 ** Residual 20 4.05 5.12 36.72 Table 1: Mean square deviations of the analysis of variance for flag leaf area, plant height and straw mass in Gaviota x MBB durum wheat cross Ns, *, ** : non-significant and significant effects at 5 % and 1 % probability level, respectively. FLA: Flag leaf area, PHT: Plant height; STM : Straw mass Figure 1: Mean of flag leaf area of the basic generations Acta agriculturae Slovenica, 117/4 – 2021 5 Inheritance of plant height, straw yield and flag leaf area in MBB x Gaviota durum wheat (Triticum durum Desf.) cross reduced PHT and increased STM, suggesting that GTA carries more dominant genes controlling PHT, while MBB carries more dominant genes controlling STM. Based F1 data analysis, these results agreed with find- ings of several authors who reported that non-additive genetic effects appeared as an important component of the genetic architecture of PHT and STM, while additive gene effects were prevalent for FLA (Saleem et al., 2005; Shabbir et al., 2012; Y ang et al., 2016). Mean values of the F2 generation deviated significantly from the average of BCs generations for the three studied traits, being sig- nificantly higher and laying within the parental range for PHT and STM and outside of this range for FLA (Table 1, Figures 1, 2 and 3). 3.2 GENE EFFECTS ABCD and joint scaling tests, applied to appraise presence of epistasis, were found significant, invaliding the additive–dominance model adequacy for explain- ing the inheritance pattern of PHT, STM and FLA, and suggesting the adoption of higher than three-parameter model (Table 2). These results indicate that higher order Figure 2: Mean of straw yield of basic generations Figure 3: Mean of plant height of the basic generations Acta agriculturae Slovenica, 117/4 – 2021 6 M. SALMI et al. inter-allelic interactions played an important role in the expression of the measured traits, and additive–domi- nance model alone will not be sufficient to deal with the inheritance pattern of such traits. Estimates of the gene effects derived from this six-parameter model are given in Table 2. Gene main effect [m] was significant for all three analyzed traits, indicating that these traits are con- trolled by minor genes and quantitatively inherited. For PHT, additive [d] and dominance [h] gene affects, and additive * additive [i] and additive * dominance [j] non- allelic interactions were significantly involved in the in- heritance of this trait. Dominance [h] gene effects and additive * additive [i] non-allelic interactions came out as the salient features of the genetic control of this char- acter as this is indicated by the high absolute values of the genetic parameters. The negative sign of the additive * dominance [j] component indicated that genes involved in the control of this trait were dispersed the parents. These results agreed with Novoselović et al. (2004), Ojaghi and Akhundova (2010), Mohamed et al. (2013), Dorri et al. (2014) and Fellahi et al. (2016) who re- ported that non-additive gene effects played an impor- tant role in the inheritance of PHT. Salmi et al. (2019) found that dominance acted in the direction of in- creased plant height But Akhtar and Chowdhry (2006) as well as Hannachi et al. (2013) reported that additive gene effects were predominant in the genetic control of this character. For STM, the additive [d] gene effects were not significant while dominance [h] gene effects, additive * additive [i], additive * dominance [j] and dominance * dominance [l] allelic interactions were sig- nificant. The gene effects [h], additive * additive [i], and dominance * dominance [l] allelic interactions exhibited the largest effects. Being significant and of the same sign, dominance [h] gene effects and dominance * dominance [l] non-allelic interactions suggested the implication of complementary type of epistasis in the genetic control of this trait. A greater magnitude of dominance [h] com- pared to additive [d] gene effects, as this is the case in the present study, for this trait, arises, according to Ke- arsey and Pooni (1996), when genes are dispersed in the parents. For FLA, the additive * dominance [j] epistatic component was not significant while the additive [d] and dominance [h] gene effects, the additive * additive [i], and dominance * dominance [l] allelic interactions were significant. The dominance * dominance [l] com- ponent exhibited the largest effect. The implication of complementary type of epistasis in the genetic control Traits PHT STM FLA Scaling test A 9.65 ± 4.96 ns 37.14 ±1.11 ** 17.72 ± 1.71 **   B 31.25 ± 2.56 ** 20.44 ± 2.18 ** 15.86 ± 0.66 **   C 84.03 ± 4.41 ** 26.38 ±2.65 ** 17.04 ± 1.35 **   D 21.55 ± 3.54 ** 41.99 ± 1.69 ** 8.12 ± 1.11 ** Joint scaling test X² 860.27 ** 569.90 ** 745.72 ** Genetic parameters m 117.79 ± 1.09 ** 34.25 ± 0.62 ** 24.51 ± 0.32 **   [d] 13.87 ± 2.78 ** 0.19 ± 1.14 ns 2.89 ± 0.90 **   [h] - 72.67 ± 7.09 ** 75.39 ± 3.42 ** 14.49 ± 2.23 **   [i] - 43.11 ± 7.09 ** 83.98 ± 3.39 ** 16.24 ± 2.22 **   [j] - 10.84 ± 2.79 * 8.35 ± 1.21 ** 1.08 ± 0.91 ns   [l] 2.18 ± 11.99 ns 141 ± 5.27 ** 49.54 ± 3.85 ** Type of gene actions ----- Complementary Complementary H² bs ±SE 0,98 ± 0.40* 0,99 ± 0.41* 0,98 ± 0.40* h² ns ±SE 0,57 ± 0.23* 0.76 ± 0.31* 0,39 ± 0.16* rg GY/ -0.57 ± 0.15* 1.21 ± 0.17* 0.91 ± 0.07* ERS 8.25 22.72 7.03 ERS% 6.99 99.44 29.69 Table 2: Scaling tests, gene action types, heritability (H²bs, h²ns), expected response to selection (ERS) and genotypic correlation (rg GY /…) estimates for plant height, straw yield and flag leaf area in MBB x GTA durum wheat cross PHT = Plant height, STM = straw mass, FLA = Flag leaf area. m = mean main effect, [d] = assistive effect, [h] = dominance effect, [i] = additive x additive effect, [j] = additive x dominance effect, [1] = dominance x dominance effect. Ns, * and ** = non-significant and significant effects at 5 % and 1 % probability levels, respectively Acta agriculturae Slovenica, 117/4 – 2021 7 Inheritance of plant height, straw yield and flag leaf area in MBB x Gaviota durum wheat (Triticum durum Desf.) cross of this trait is suggested by the dominance gene effects [h] and the dominance * dominance [l] allelic interaction which were significant and of the same sign. Dominance [h] component was greater than additive [d] gene effects suggesting that genes controlling FLA are dispersed in the parents. In this context Saleem et al. (2005); Inamul- lah et al. (2006); Munir et al. (2007); Ijaz et al. (2013) and Y ang et al. (2016) found that FLA was mostly under addi- tive genetic control combined to partial dominance and epistasis type of gene actions. Joshi and Sharma (1984) mentioned that dominance acted in the direction of re- duced FLA. Shayan et al. (2019) reported that the additive-dom- inance model fitted best the variation present among generation means for FLA, while for PHT and STM, the six-parameter model was adequate implying the pres- ence of non-allelic interactions in the inheritance of these two traits. Divergence in the result among various studies seems to indicate that genetic model adequacy as well as the preponderance of significant gene effects and non-allelic interactions are dependent upon the cross combination genetic background and the experimen- tal growth conditions experienced. The complementary epistasis type implicated in the inheritance of STM and FLA suggested the possibility of heterosis expression for these two traits. In fact, Punia et al. (2011), referring to Jinks and Jones (1958), mentioned that heterosis is likely to be expressed with greater magnitude in crosses where complementary type of interaction is expressed. The fact that generation means variation fitted a digenic epistatic model suggested that improvement of PHT, STM and FLA would be fairly difficulty compared to the situation where the additive-dominance model was the most ad- equate. Furthermore, Sirohi and Gupta (1993) suggested that traits showing high magnitude of dominance [h] than additive [d] gene effects, as this is the case for PHT, STM and FLA, in the present study, can be improved through conventional breeding approaches. But selec- tion need to be delayed until later generations when the dominance effects would have diminished and desirable segregants become available. 3.3 HERITABILITY, EXPECTED RESPONSE TO SE- LECTION, AND GENOTYPIC CORRELATIONS Being significant, broad sense heritability estimates for FLA, PHT and STM were appreciably high, taking values of 98.00, 98.00 and 99.00 %, respectively. Their corresponding narrow sense counterparts were also significant, but of lower magnitude, being still high for STM (76.00 %), moderate for PHT (57.00 %) and low for FLA (39.00 %). High h²ns values indicate that the environment influences less the expression of the given character. In fact, estimates of heritability are useful for a breeder to weigh the proportion of variation which is inheritable from that which is non-inheritable. Heritabil- ity values observed in the present study were in the range of those reported in similar studies. Fellahi et al. (2020) reported h²bs values of 86.50 % for PHT and 77.40 % for FLA. Novoselovic et al. (2004) reported h²bs values rang- ing from 0.54 to 0.81 for PHT of several crosses. These high H²bs values suggested that these traits are less im- pacted by environmental variation, and then are easily amenable to improvement. In this context, Johnson et al. (1966) mentioned that since plant height and straw mass heritability were sufficiently high then selection in the F 2 for these traits could be effective. So, based on the herita- bility estimates observed in the present study, STM, FLA and PHT appeared amenable to significant improvement applying early selection. Genotypic correlation coefficients, relating grain yield (GY) to FLA, PHT and STM, found in this study, were significant, taking values of 0.91, -0.57 and 1.21, re- spectively (Table 2). while PHT was posirivelly correlated with STM (r = 0.315**), but negatively correlated with FLA (r = -0.153*), and STM was positivelly correlated with FLA (r = 0.269**). These correlation coefficients in- dicated that selection of high values for FLA and STM will be accompanied by increased GY, but selection to increase PHT had a negative impact on GY and on FLA. The negative correlation relating PHT to GY , observed in the present study, contradicted Ataei et al. (2017) results which showed that GY was positively and highly corre- lated with PHT and peduncle under drought stress con- ditions, emphasizing the importance of PHT as selection criterion to improve drought tolerance. In this context and according to Davidson et al. (1992) and Belkher- chouche et al. (2015) wheat peduncle is a transient source of water-soluble carbohydrates, playing a crucial role in minimizing grain yield decline under drought stress conditions. Similarly, Mohsin et al. (2009) found that grain yield correlated positively with FLA, PHT, biomass (BIO), under drought stress. Under rainfed growing conditions, Mansouri et al. (2018) reported that above ground plant biomass exhibited significant and positive correlation coefficients with GY, STM and PHT. The ex- pected response to selection estimates were low, being less than 10 % for PHT (8.24 cm or 6.99 %, X barF2 = 117.87 cm) and high, above 20 %, for STM (22.74 g plant -1 or 99.4 % X barF2 = 22.85 g plant -1 ), FLA (7.03 cm 2 or 26.6 %, X barF2 = 23.67 cm 2 ) (Table 2). These results indicated that, based on the magnitude of the variability expressed by each trait, moderate to appreciable genetic gain could be made via mono trait selection. Durum wheat production is often impacted by Acta agriculturae Slovenica, 117/4 – 2021 8 M. SALMI et al. drought stress, particularly in the arid and semi-arid re- gions. T o overcome this situation, it is necessary to devel- op improved varieties devoted to these specific environ- ments. Morphological characters like plant height, flag leaf area, and straw yield, had been identified and pro- posed as morphological markers for drought tolerance. Breeding procedure for drought tolerance depends upon the pattern of inheritance and the nature of actions of the genes involved in the genetic control of the drought related traits. A better understanding of the complexi- ties of the genetic control of these traits will be useful for cultivar improvement. Globally, from the results of this study, it can be summarized that the additive–dominance model was inadequate, suggesting the adoption of a six- parameter genetic model. Additive [d] and dominance [h] gene effects, and non-allelic interactions ([i], [j] and [l]) were involved in the inheritance of PHT, with the predominance of [h] and [i] components. Complemen- tary type of epistasis was implicated in the inheritance of STM and FLA. These results were in lines with findings of some studies (Novoselovic et al., 2004; Ojaghi and Ak- hundova, 2010; Mohamed et al., 2013; Dorri et al., 2014; Fellahi et al., 2016) and diverged from those of others studies (Inamullah et al., 2006, Munir et al., 2007; Ijaz et al., 2013; Y ang et al., 2016; Shayan et al., 2019), suggesting that genetic model adequacy as well as the preponderance of significant gene effects and non-allelic interactions, governing the inheritance of a given trait, are dependent upon the cross combination genetic background and the experimental growth conditions experienced. Due to the presence of non-allelic interactions PHT, STM and FLA improvement would be fairly difficulty, requiring the implementation of conventional breeding approaches such as the inclusion of F2’s showing high performances in multiple crosses for further improvement of the stud- ied traits in order to synthesize a dynamic population accumulating most of the favorable genes. This mating procedure seems to be a good technique to disrupt link- age, to generate usefull recombination and to accumulate favorable genes in the base population. Selection need to be delayed until later generations when the dominance effects would have diminished and desirable segregants become available. This strategy is supported by the high heritability estimates, the moderate to high expected ge- netic gains and the significant genotypic correlation coef- fcicients relating the studied traits to grain yield. 4 CONCLUSION The results of the present study indicated that non- allelic interactions, in addition to additive and dominant gene effects are important factors controlling the expres- sion of PHT, STM and FLA. Application of conventional selection procedure may not be rewarding for the im- provement of these characters. But inter mating among the selected segregants followed by few generations of selfing could be useful to break the undesirable linkage and allow accumulation of favorable alleles for improve- ment of these traits. 5 REFERENCES Akhtar, N., & Chowdhry, M. A. (2006). Genetic analysis of yield and some other quantitative traits in bread wheat. Interna- tional Journal of Agriculture and Biology, 4, 523–527. Annicchiarico, P., Abdellaoui, Z., Kelkouli, M., Zerargui. H. (2005). Grain yield, straw yield and economic value of tall and semi-dwarf durum wheat cultivars in Algeria. Journal of Agricultural Science, 143, 57–64. https://doi.org/10.1017/ S0021859605004855 Asadi, A., Valizadeh, M., Mohammadi, S. A., Khodarahmi. M. (2015). Genetic analysis of some physiological traits in wheat by generation means analysis under normal and wa- ter defict conditions. Biological forum, 7, 722-733. Ataei R., Gholamhoseini, M., Kamalizadeh, M. (2017). Genetic analysis for quantitative traits in bread wheat exposed to ir- rigated and drought stress conditions. Phyton, 86, 228-235. https://doi.org/10.32604/phyton.2017.86.228 Aziz, T., Mahmood, Z., Mahmood, K., Shazadi, A., Kazi, M., Rasheed, A. (2018). Genotypic variation and genotype x environment interaction for yield-related traits in syn- thetic hexaploid wheats under a range of optimal and heat- stressed environments. Crop Science, 58, 295-303. https:// doi.org/10.2135/cropsci2017.01.0035 Belagrouz A., Chennafi, H., Bouzerzour, H., Hakimi, M., Ra- zem, R., Hadj Sahraoui, A. (2018). Relationships among water use efficiency and the physio-agronomic traits in durum wheat (Triticum durum Desf.) cultivars assessed under rainfed conditions of the eastern high plateaus of Al- geria. The Journal Agriculture and Forestry, 64(3), 159-172. https://doi.org/10.17707/AgricultForest.64.3.14 Belkherchouche, H., Benbelkacem, A., Bouzerzour, H., Benma- hammed. A. (2015). Flag leaf and awns ablation and spike shading effects on spike yield and kernel weight of durum wheat (Triticum turgidum L. var. durum) under rainfed conditions. Advances in Environmental Biology, 9(8), 184- 191. Chennafi, H., Hannachi, A., Touahria, O., Fellahi, Z.E.A., Makhlouf, M., Bouzerzour, H. (2011). Tillage and residue management effect on durum wheat [Triticum turgidum (L.) Thell. ssp. turgidum conv. durum (Desf.) Mackey] growth and yield under semi-arid climate. Advances in En- vironmental Biology, 3231–3241. Cropstat. (2007). CropStat for Windows 7.2. Dapo, Manila IRRI - International Rice Research Institute. Davidson, D. J. and Chevalier, P . M. (1992). Storage and remo- bilization of water-soluble carbohydrates in stems of spring wheat. Crop Science, 32, 186-190. https://doi.org/10.2135/cr opsci1992.0011183X003200010038x Acta agriculturae Slovenica, 117/4 – 2021 9 Inheritance of plant height, straw yield and flag leaf area in MBB x Gaviota durum wheat (Triticum durum Desf.) cross De Pace, C., Snidaro, D., Ciaffi, M., Vittori, D., Ciofo, A., Cenci, A. (2001). Introgression of Dasypyrum villosum chromatin into common wheat improves grain protein quality. Euphyt- ica, 117, 67–75. https://doi.org/10.1023/A:1004095705460 Dorri, P., Khorasani, S. K., Shahrokhi, M. (2014). Generation means analysis. A case study of variance components in KSC 500 generations of maize (Zea mays L.). International Research Journal of Applied and Basic Sciences, 8(2), 194- 200. Farooq, M., Wahid, A., Kobayashi, N., Fujita, D., Basra, S. M. A. (2009). Plant drought stress: effects, mechanisms and man- agement. Agronomy for sustainable development, Spring- er Verlag/EDP Sciences/INRA, 29, 85-212. https://doi. org/10.1007/978-90-481-2666-8_12 Fellahi, Z. E. A., Hannachi, A., Bouzerzour. H. (2020). Expected genetic gains from mono trait and index-based selection in advanced bread wheat (Triticum aestivum L.) populations. Revista Facultad Nacional de Agronomía Medellín, 73, 9131- 9141. https://doi.org/10.15446/rfnam.v73n2.77806 Fellahi, Z. E. A., Hannachi, A., Bouzerzour, H., Dreisigacker, S., Yahyaoui, A., Sehgal, D. (2017). Genetic analysis of morpho-physiological traits and yield components in F2 partial diallel crosses of bread wheat (Triticum aestivum L.). Revista Facultad Nacional de Agronomía Medellín, 70, 8237–8250. https://doi.org/10.15446/rfna.v70n3.61927 Haddad, L., Bachir, A., Yakhlef, N., Benmahammed, A., Bou- zerzour, H. (2021). Durum wheat [Triticum turgidum (L.) Thell ssp turgidum conv. durum (Desf.) Mackey] during the past 70-year in Algeria: Performance assessment of a set of historical varieties under rainfed conditions of the eastern high plateaus. Jordan Journal of Biological Sciences, 14. (in press). Hallauer, A. R., & Mirinda Filho, J. B. (1989). Quantitative ge- netics in maize breeding. Second edition, Ames, IOW A State University Press. 468 p. Halloran, G. M., Knight, R., McWhirter, K. S., Sparrow, D. H. B. (1979). Plant breeding. Knight, R. (Ed.). Poly-Graphics Pty. Ltd. Brisbane, Australia. pp. 61 - 62. Hammer, O., Harper, D. A. T., Ryan, P . D. (2001). Paleontologi- cal statistics software package for education and data analy- sis. Palaeontologia Electronica, 4(9). Hannachi, A., Fellahi, Z. E. A. Bouzerzour, H., Boutakrabt. A. (2013). Diallel-cross analysis of grain yield and stress tolerance-related traits under semi-arid conditions in Durum wheat (Triticum durum Desf.). Electron- ic Journal of Plant Breeding, 4, 1027-1033. https://doi. org/10.1155/2013/201851 Ijaz U., Miullah, S., Kashif. M. (2013). Generation means analy- sis for five physiological traits of bread wheat under rain- fed condition. Universal Journal of Plant Science, 1, 21-26. https://doi.org/10.13189/ujps.2013.010103 Inamullah, A., Hussain, M., Hassan, G. F., Gul, R. (2006). Dial- lel analysis of the inheritance pattern of agronomic traits of bread wheat. Pakistan Journal of Boutany, 38, 1169-1175. Jatayev, S., Sukhikh, I., Vavilova, V., Smolenskaya, S. E., Gon- charov, N. P., Kurishbayev, A., Zotova L. (2020). Green revolution ‘stumbles’ in a dry environment: Dwarf wheat with Rht genes fails to produce higher grain yield than tall- er plants under drought. Plant, Cell and Environment, 43, 2355-2364. https://doi.org/10.1111/pce.13819 Johnson, V . A., Biever, K. J., Haunold, A., Schmidt, J. W . (1966). Inheritance of plant height, yield of grain, and other plant and seed characteristics in a cross of hard red winter wheat, Triticum aestivum L.. Crop Science, 6, 307-312. https://doi. org/10.2135/cropsci1966.0011183X000600040011x Joshi, S. K., Sharma, S. N., Singhania, D. L., Sain. R. S. (2004). Combining ability in the F1 and F2 generations of diallel cross in hexaploid wheat (Triticum aestivum L. em. Thell). Hereditas, 141, 115-121. https://doi.org/10.1111/j.1601- 5223.2004.01730.x Kearsey, M. J. and Pooni, H. S. (1996). The genetic analysis of quantitative traits. Chapman and Hall, London. 396 p. https://doi.org/10.1007/978-1-4899-4441-2 Koots, K. R., Gibson. J. P. (1996). Realized sampling variances of estimates of genetic parameters and the difference be- tween genetic and phenotypic correlations. Genetics, 143, 1409-1416. https://doi.org/10.1093/genetics/143.3.1409 Kwon, S. H. and Torrie, J. H. (1964). Heritability and inter-re- lationship among traits of two soybean populations. Crop Science, 4, 196-198. https://doi.org/10.2135/cropsci1964.00 11183X000400020023x Liu, H., Searle, I. R., Mather, D. E., Able, A. J., Able. J. A. (2015). Morphological, physiological and yield responses of durum wheat to pre-anthesis water-deficit stress are genotype-de- pendent. Crop and Pasture Science, 66, 1024–1038. https:// doi.org/10.1071/CP15013 Manivannan, N. (2014). TNAUSTAT- Statistical package: htt- ps://sites.google.com/site/tnaustat. Mansouri, A., Oudjehih, B., Benbelkacem, A. Fellahi, Z. E. A., Bouzerzour, H. (2018). Variation and relationships among agronomic traits in durum wheat [Triticum turgidum (L.) Thell. ssp. turgidum conv. durum (Desf.) Mackey] under south Mediterranean growth conditions: Stepwise and path analyses. International Journal of Agronomy, 1-11. https:// doi.org/10.1155/2018/8191749 Mather, K., & Jinks, J. L. (1982). Biometrical Genetics. The Study of Continuous Variation. Third edition. Chapman and Hall, London – New York, pp. 279. https://doi.org/10.1007/978- 1-4899-3406-2 Mohsin, T., Khan, N., Naqvi, F. N. (2009). Heritability, pheno- typic correlation and path coefficient studies for some agro- nomic characters in synthetic elite lines of wheat, Journal of Food, Agriculture and Environment, 7, 278–282. Munir, M., Chowdhry, M. A., Ahsan, M. (2007). Generation mean studies in bread wheat. Internaional Journal of Ag- riculure and Biology, 9, 282-286. Novoselović, D., Barić, M., Drenzer, G., Gunjača, J., Lalić, A. (2004). Quantitative inheritance of some wheat plant traits. Genetics and Molecular Biology, 27, 92-98. https://doi. org/10.1590/S1415-47572004000100015 Ojaghi, J., & Akhundova, E. (2010). Genetic analysis for yield and its components in doubled haploid wheat. African Jour- nal of Agricultural Research, 5, 306-315. Punia, S. S., Baldev, R. Koli, N. R. Ranwah, B. R. Rokadia, P., Maloo, S. R. (2011). Genetic architecture of quantitative traits in field pea. Journal of Food Legumes, 24, 299–303. Rabti, A., Mekaoussi, R., Fellahi, Z. E. A., Hannachi, A., Ben- Acta agriculturae Slovenica, 117/4 – 2021 10 M. SALMI et al. belkacem, A., Benmahammed, A., Bouzerzour, H. 2020. Characterization of old and recent durum wheat [Triticum turgidum (L.) Tell. convar. durum (Desf.) Mackey] varieties assessed under south Mediterranean conditions. Egyptian Journal of Agronomy, 42, 307-320. https://doi.org/10.21608/ agro.2020.43329.1230 Rad, M. R. N., Kadir, M. A., Yusop, M. R., Jaafar, H. Z., Danaee, M. (2013). Gene action for physiological parameters and use of relative water content (RWC) for selection of tolerant and high yield genotypes in F2 population of wheat. Austo- ralian Journal of Crop Science, 7, 407-413. Royo, C., Nazco, R., Villegas, D. (2014). The climate of the zone of origin of Mediterranean durum wheat (Triticum durum Desf.) landraces affects their agronomic performance. Ge- netic Resources and Crop Evolution, 61, 1345–1358. https:// doi.org/10.1007/s10722-014-0116-3 Saleem, M., Aslam, M., Muhammad, C., Kashif, K., Khaliq. M. (2005). Inheritance pattern of plant height, grain yield and some leaf characteristics of spring wheat. International Journal of Agriculture and Biology, 7, 1019-1025 Salmi, M., Benmahammed, A., Benderradji, L., Fellahi, Z. E. A., Bouzerzour, H., Oulmi, A., Benbelkacem. A. 2019. Genera- tion means analysis of physiological and agronomical tar- geted traits in durum wheat (Triticum durum Desf.) cross. Revista Facultad Nacional de Agronomía Medellín, 72, 8971- 8981. https://doi.org/10.15446/rfnam.v72n3.77410 Shabbir, G., Kiran, T., Akram, Z., Ijaz, M., Shah. K. N. (2012). Genetics of some biometric traits in bread wheat (Triticum aestivum L.). Journal of Agricultural Research, 50, 457-468. Shayan, S., Moghaddam Vahed, M., Norouzi, M., Mohammadi, A., Tourchi, M., Molaei. B. (2018). Inheritance of agronom- ical and physiological traits in the progeny of Moghan3 and Arg bread wheat varieties cross. Plant Genetic Research, 4, 43-60. https://doi.org/10.29252/pgr.4.2.43 Singh, R. P., Huerta-Espino, J., Rajaram, S., Crossa, J. (2001). Grain yield and other traits of tall and dwarf isolines of modern bread and durum wheats. Euphytica, 119, 241–244. https://doi.org/10.1023/A:1017541805454 Sirohi, A., & Gupta, V . P . (1993). Additive, dominance and epi- static components of variation for seed protein content in pea (Pisum sativum L.). Indian Journal of Genetics, 53, 252- 256. Slafer, G. A., Araus, J. L. Royo, C., Del Moral, L. F. G. (2005). Promising eco-physiological traits for genetic improvement of cereal yields in Mediterranean environments. Annals of Applied Biology, 146, 61–70. https://doi.org/10.1111/j.1744- 7348.2005.04048.x Spagnoletti-Zeuli, P. L., & Qualset, C. O. (1990). Flag leaf variation and the analysis of diversity in du- rum wheat. Plant Breeding, 105, 189–202. https://doi. org/10.1111/j.1439-0523.1990.tb01196.x Steel, R. G. D., & Torrie, J. H. (1960). Principles and procedures of statistics, McGraw-Hill Books, New Y ork. 481 p. Velu, G., Singh, R. P. Huerta, J., Guzman C. (2017). Genetic impact of Rht dwarfing genes on grain micronutrients concentration in wheat. Field Crops Research, 21, 373–377. https://doi.org/10.1016/j.fcr.2017.09.030 Yang, D., Liu, Y., Cheng, H., Chang, L., Chen, J., Chai, S., Li, M. (2016). Genetic dissection of flag leaf morphology in wheat (Triticum aestivum L.) under diverse water regimes. BMC Genetics, 17, 1-15. https://doi.org/10.1186/s12863- 016-0399-9 Yani, S. C., & Rashidi, V. (2012). Selection indices in the im- provement of wheat grain yield on drought stress condi- tions, the success of selection in the cross populations depend on the knowledge of inheritance pattern of the desired traits. African Journal of Agricultural Research, 7, 1177-1183. https://doi.org/10.5897/AJAR11.1616