Les variétés de blé dur ont constitué jusqu’à la première moitié du siècle dernier, la principale source de production de blé marocain. Ce germoplasme local est encore cultivé dans les zones marginales, en particulier dans les montagnes et les régions sub-sahariennes. Au cours des dernières décennies du 20ème et du début du 21ème siècle, l’amélioration génétique a conduit à la création de nouveaux cultivars de blé dur hautement homogènes et plus productifs. Le présent article étudie l’évolution de la variabilité génétique des populations locales de blé dur aux variétés améliorées en termes de productivité et de qualité, en utilisant une série historique de génotypes marocains de blé dur groupés selon leur période d’inscription au catalogue officiel; “Populations locales/Anciens cultivars”, “Cultivars intermédiaires” et “Variétés améliorées”. Une amélioration significative a été réalisée dans la productivité du blé dur. Les cultivars modernes dépassent leurs prédécesseurs en termes de caractéristiques liées à la productivité. Le gain génétique a été clairement associé à une réduction du cycle végétatif et de la hauteur de la plante, permettant une réduction du rendement en paille et par conséquent, une augmentation du rendement en grains estimé à 15,4 Kg/ha/an. Cependant, les résultats ont révélé une réduction des valeurs de presque tous les caractères de qualité; -0,12 % par an pour la teneur en protéines, -0,30 % par an pour la teneur en gluten, -0,31 % par an pour la teneur en pigments jaunes et -0,19 % par an pour le taux de vitrosité. Par ailleurs, les résultats, soulignent la variabilité importante des attributs de qualité des grains chez les génotypes de populations locales. Ce germoplasme local peut être utilisé comme source de gènes de qualité dans le programme de sélection de blé dur pour développer de nouvelles variétés combinant à la fois une productivité élevée et une meilleure qualité du grain.

Mots-clés: Blé dur, Populations locales, Variétés améliorées, Productivité, Qualité.


Durum wheat (Triticum turgidum L.var. durum) is an important species of the tribe Triticeae, and is one of the most important food crops in the Mediterranean region, because of its adaptation to semi-arid environments and its unique end products like pasta, couscous, burghul, traditional bread, and others artisanal products. In Morocco, durum wheat, is one of the oldest cultivated cereal species. It is highly appreciated by Moroccan consumers, mainly for the preparation of bread, couscous and traditional products. Durum wheat is sown annually over 1.0 million hectares, with an average consumption of around 90 Kg/person/year. Morocco is ranked third in the Mediterranean region and first in the North Africa and Middle East region in term of durum wheat acreage (Nsarellah et al., 2011).

A large number of durum wheat landraces are still growing in Morocco especially in mountains and sub Saharan regions. These ecotypes are a mixture of traditional crop varieties developed by farmers through years of natural and human selection. Although their yields are often low (Ehdaie et al., 2008; Blum et al., 1989), landraces are characterized by their adaptability to local environmental conditions (Ganeva et al., 2010) and their particular qualitative traits (Taghouti et al., 2006).

In Morocco, durum wheat breeding program has been undertaken since the beginning of the twentieth century. The genetic improvement has led to modern durum wheat cultivars highly uniform and productive. Three fundamental steps characterize the long term history of Moroccan durum breeding (Nsarellah et al., 2005). Before 1956, the efforts of the genetic improvement of durum focused on increasing productivity through a better response to improved agronomic packages. Several varieties were extracted from landraces. These varieties were more productive but they were still characterized by a long growing cycle and a high straw, predisposing them to lodging. By 1970s, collaboration with CIMMYT (International Wheat Improvement Centre and Maize) and later with ICARDA (International Center for Agricultural Research in Dry Areas) allowed the introduction of germplasm characterized by earliness and containing the semi-dwarfing Rh1 genes, responsible for height reduction. Many newly released cultivars were distinguished by their earliness and their semi-dwarf size. These varieties were more productive and more suitable than older varieties but their grain quality was lower (Taghouti et al., 2010). Comparisons of cultivars bred in different periods of time can inform on the changes trend in agronomical and qualitative characteristics of the wheat grown in a given region and provide an estimate of breeding progress.

Evaluation of breeding advances was undertaken and reported in several works in many countries. The aims of these studies were to analyze the effects of breeding on crop development during the last century. Motzo et al. (2007) described the effect of breeding on the phenology of landraces and modern Italian durum wheat. Furthermore, breeding progress in morpho-physiological, agronomical and qualitative traits of durum wheat cultivars (De Vita et al. 2007) and genetic improvement effects on yield stability in durum wheat genotypes grown in Italy (De Vita et al. 2010) were reported.

Changes in duration of developmental phases of durum wheat caused by breeding in Spain and Italy and its impact on yield were studied by Isidro et al. (2011). More recently, Subira et al. (2014) revealed the breeding progress in the pasta-making quality of durum wheat cultivars released in Italy and Spain during the 20th Century.

The present study was conducted using an historical series of 47 Moroccan durum wheat genotypes grouped according to their period of release into “Landraces/ Old cultivars”, “Intermediate cultivars” and “Modern cultivars”. The aim of this study is to assess the changes in terms of some productivity and quality related traits resulting from the transition of Moroccan durum wheat from landraces to modern cultivars.



A set of twenty three accessions of durum wheat landraces (Triticum turgidum L. var. durum), collected in mountains regions of Morocco by the National Institute of Agronomical Research, was compared to modern genotypes including twenty four durum wheat cultivars. The genotypes were assigned to three groups representing three periods of time; G1 including landraces and cultivars released, mainly from landraces, before 1956, G2 including cultivars released between 1984 and 2000 representing an intermediate period and G3 containing cultivars developed during the period between 2000 and 2015 (Table 1).

Experimental trials

The material was conducted in trials at INRA’s research station of Merchouch (33°60’N, 6°71’W); following a randomized complete blocks design with two replications. Individual plots were made out of 6 rows that were five meter long and 1.8 m wide. Inter-row spacing was 30 cm. The area of each individual plot was 9 m2. The agronomic management including soil preparation, fertilization and weeding were applied. Fertilizer used was a 19-38-0 (N-P-K) complex applied at a rate of 150 kg/ha and ammonium nitrate (33.5%N) applied at a rate of 100 kg/ha. The trial was planted over a previous fallow using a seeding density of 300 seeds/m2.

Morphological and agronomical traits

The number of days to heading was recorded when about half of the culms showed emerging spikes (Zadoks et al. 1974). Date of physiological maturity was recorded when the peduncle was completely yellow for about half of the plant in the plot. Plant height was measured during the maturation stage, when the maximum height level was achieved. Four lines of the plot representing 6 m2 were harvested mechanically and evaluated for grain yield (q/ha) and quality traits.

Quality traits assessment

Seeds samples from each genotype were harvested in each block and analyzed separately. The samples were cleaned manually in order to remove soil particles, broken and foreign seeds. Whole grain flour samples were obtained with a whole–meal grinder (Udy-Cyclone 0.5 mm screen). Grain vitreousness (%) was determined visually by counting the number of vitreous grains or after cutting grains. Gluten strength was determined by the SDS (Sodium Dodecyl Sulfate) sedimentation test following a Moroccan standard assay (N.M. equivalent to American Association for Cereal Chemistry method (AACC). Carotenoid pigments content (ppm) was assayed by AACC 14-50 modified method (Santra et al., 2003). Grain protein content (% DM) and test weight (Kg/hl) were obtained by a near-infrared spectroscope (Infraneo Choppin France) on whole grain of each genotype.

Statistical analysis

The data were subjected to analysis of variance (ANOVA) using SAS Software version 9.0 and Duncan’s Multiple Range test was used to compare means, whenever F-test was found significant.

In addition to a principal components analysis (PCA), regression analysis with a standard linear model applied to year cultivars means was used to calculate rates of change for productivity and quality traits with year of release. Landraces have been assigned 1910 as an arbitrary year of release.


Variation between groups

The combined ANOVA analysis for productivity and quality traits revealed significant group effects for almost all the traits. Significant differences between genotypes within each group were also recorded (Table 2).

Means values for the traits evaluated from Moroccan durum wheat genotypes separated in the three groups are presented in table 3. In terms of productivity traits, a large variation in (i) heading time was detected between the first group (landraces and old cultivars) and the second and third groups including intermediate and modern cultivars, ranging in means from 106.3 (G1) to 90.4 days (G2 and G3); (ii) physiological maturity differed also between groups varying from 160 days for G1 to 144.5 days in means for intermediate and modern cultivars; (iii) plant height decreased significantly from 116.4 cm in G1 to 89.2 cm for intermediate/modern cultivars; (iv) grain yield exhibited a significant trend to increase through breeding history. Grain yield varies from 12.8 qha-1 for landraces to 27.3 qha-1 for modern cultivars (Table 3A).

With respect to quality traits, except test weight, which showed a non-significant trend in response to breeding improvement, a clear trend was evident for the other quality parameters. In fact, the traits showed a decreasing tendency over time of cultivars release. The landrace genotypes and old cultivars exhibited significantly higher SDS sedimentation volumes (48.9 ml) and higher vitreousness (95.7%) than intermediate and modern cultivars (SDS=35.9 ml) and (Vit=87.7%) respectively. Yellow pigment content decreased from landraces (7.49 ppm) to cultivars released (5.80 ppm) over time of breeding. For protein content substantial differences were found with a clear difference between the groups of genotypes; the highest value of protein content was recorded in old cultivars (16.8%) while the protein content of cultivars released thereafter showed a trend to decreasing (Table 3B).

Variation among groups

The distributional characteristics for each category of durum wheat genotypes indicates the degree of dispersion of traits related to the productivity (Figure 1) and to quality (Figure 2). Landraces genotypes in the first group were much more variable in terms of physiological parameters. The distribution is more elongated toward the high values of days to heading and days to maturity; more than 75% of genotypes in this group recorded value ranging from 103 to 119 days to heading and from 160 to 169 days to maturity. Whereas, low variability is noted for intermediate and modern cultivars in group 2 and group 3; all cultivars take less than 92 and 147 days to give spikes and to mature (Figure 1A, Figure 1B).

The illustration of plant height and yield in figure 1C and figure 1D, showed much more variability for old cultivars than intermediate and moderns ones; 50% of genotypes in group 1 were tall recording plant height values ranging from 110 cm to 132 cm, whereas plant height in all intermediate and modern cultivars in G2 and G3 was ranging from 82 to 98 cm (Figure 1C). Furthermore, grain yield was low and more variable for landraces and old varieties. However, modern cultivars (G3) were high yielding, 75% of them produce from 26 to 32 q/ha (Figure 1D).

The dispersion of the quality parameters data for each category of durum wheat genotypes is presented in figure 2. Landraces genotypes in G1 showed small spacing between the different parts of the box plots especially for gluten strength (Figure 2A), vitreousness (Figure 2B), and yellow pigment traits (Figure 2C), suggesting a little variation among accessions. Indeed, genotypes in this group were characterized by high values for the three quality recorded traits reaching respectively 58 ml, 99% and 9 ppm. Whereas cultivars in G2 and G3 revealed longer box plots indicating large variation between varieties for the three parameters. Fifteen percent of the cultivars recorded SDS volumes ranging from 30 to 41 ml, vitreousness varying from 82 to 98% and yellow pigment ranging from 3 to 6 ppm.

Principal component analysis (PCA) explained 68.5% of the total variation (Figure 3). The distribution of accessions demonstrates that landraces and old varieties are grouped apart from intermediate and modern cultivars confirming the previous results concerning low productivity and high quality.

Evolution of traits

On the base of regression analysis, the evolution over time of heading time (Figure 4A), maturity time (Figure 4B) and plant height (Figure 4C) showed respectively highly significant correlation (R2=0.66***), (R2=0.84***) and (R2=0.88***). A progressive reduction of heading, maturity days and plant height was observed from landraces to modern varieties respectively (106 days versus 90 days), (160 to 144 days) and (123 to 90 cm). For grain yield, the figure 4(D) showed clearly a variation among cultivars from different phases of breeding (R2=0.76***). Grain yield ranged from 12.8 q/ha for landraces to 27.3 q/ha for varieties released in the intermediate and modern phases of breeding.

The regression over years of quality measured traits is illustrated for gluten strength in Figure 5A, for yellow pigment content in Figure 5B, for protein content in figure 5C and for vitreousness percent in figure 5D. Highly significant differences were detected for SDS volumes (R2= 0.36***) and yellow pigment content (R2= 0.32***). In fact these parameters showed a decreasing trend from landraces to released varieties, respectively varying from 50 ml to 36 ml for SDS; 7 ppm to 5 ppm for yellow pigment. Protein content (R² = 0.14***) and vitreousness (R² = 0.17***) were also affected over time of breeding suggesting a dropping trend from landraces to modern cultivars. Nevertheless, test weight showed a non-significant trend over time (R2 = 0.021ns).


Evaluation of breeding advances on durum wheat was undertaken and reported in several works in many countries such as Canada (McCaig et al., 1995), Italy (De Vita et al., 2007) and Spain (Subira et al., 2014). The changes occurred in the present study, in yield potential and its morpho-phenological traits due to genetic breeding are similar to those reported in previous studies.

Landraces and old cultivars (G1) are characterized by late heading and maturity time, high plant height, low yield potential and high quality parameters. While, intermediate (G2) and modern varieties (G3) are earlier, shorter, more productive with lower grain quality than G1.

Indeed, the genetic progress in earliness and yielding through novel varieties with short growing season is a consequence of farmers’ requirement in order to overcome the increase of drought and heat incidence at the end of the growing season. Annual yield increase in the present study (+ 15.42 Kg/ha/year) is similar to the gain in grain yield estimated in previous studies conducted on Italian durum wheat cultivars reported by Pecetti al. (1998) (17 Kg/ha/year) and De Vita et al. (2007) (19.9 Kg/ ha/year), on Canadian and Brazilian durum wheat cultivars respectively 22.6 Kg/ha/year and 29 Kg/ ha/year reported respectively by McCaig et al. (1995) and Beche et al. (2014). Furthermore, the reduction of plant height (-0.40 cm/year) indicates the introgression of the semi dwarfing Rh gene responsible for height reduction. In general, most of durum wheat breeding program showed very similar tendency in terms of grain yield genetic gain over the last century in order to satisfy food security despite environmental, economic and political differences among countries.

However, this gain was not without negative impact on Moroccan modern varieties quality, so a decreasing trend over time of quality traits was observed especially grain protein (-0.12% per year) in accordance with (Subira et al., 2014; Amallah et al., 2014; Amallah et al., 2016) studies; yellow pigment (-0.31% per year); gluten strength (-0.30 % per year) not in accordance with some studies that revealed rather an increase of 0.33% per year of gluten strength (De Vita et al., 2007; Subira et al. ,2014). Many studies revealed that the increase of grain yield is accompanied by a decrease of protein concentration suggesting genetic incompatibility and competition for photosynthetic energy between nitrogen and carbon-1, 2, 3 (Dotlacil et al., 2010; Nazco et al., 2012).

Quality traits are more required by consumers and industrials. In this context, improving durum wheat grain quality has become the main objectives in Moroccan breeding program only in the last decade of the 20th century. The main quality characteristics of durum wheat required for high-quality end-products such as bread, pasta and couscous concern protein content, vitreousness, gluten strength, yellow pigment content, and test weight. Therefore, the first varieties combining good productivity and quality “Faraj” and “Louiza” cultivars were registered since 2007 and 2011 respectively.


The results of this study showed that a significant improvement was achieved in durum wheat productivity during the 20th and the beginning of 21st centuries. Firstly, the released varieties in 20th century exceed their predecessors in terms of earliness (-0.20 days/year), plant height reduction (-0.40 cm/year) and productivity (15.42 Kg/ha/year). This improvement has impacted negatively the grain quality parameters; the landraces still exceed released varieties in terms of grain quality. Since the last decade, grain quality become among priorities in Moroccan durum wheat breeding program.

These results underline the important variability in grain quality attributes among landraces genotypes. Therefore local landraces may be used as sources of quality–improving attributes in breeding program to develop new varieties combining both high productivity and grain quality.


Amallah L., Taghouti M., Rhrib K., Gaboun F., Arahou M., Hassikou R., Diria G. (2016). Validation of simple sequence repeats associated with quality traits in durum wheat. Journal of Crop Science and Biotechnology 19: 137-150.

Amallah L., Hassikou R., Rhrib K., Gaboun F., Ennadir J., Ouazza F., Rochdi A., Arahou M., Diria G., Taghouti M. (2016). Analyse de la diversité génétique d’une collection de blé dur par les marqueurs agro-morphologiques et biochimiques. J. Mater Environ. Sci. 7: 2435-2444.

Amallah L., Taghouti M., Rhrib K., Gaboun F., Hassikou R. (2014). Genetic variability in agro-morphological and quality traits of a Mediterranean durum wheat landraces. Cereal Research Communication 43: 123–132.

Beche E., Benin G., Lemes da Silva C., Berger Munaro L., Marchese J. A. (2014). Genetic gain in yield and changes associated with physiological traits in Brazilian wheat during the 20th century. European Journal of Agronomy 61: 49–59.

Blum A., Golan, G., Mayer, J., Sinmena, B., Shpiler L., Burra J. (1989). The drought response of landraces of wheat from the northern Negev desert in Israel. Euphytica 43: 87-96.

De Vita P., Li Destri Nicosia O., NigroF., Platani C., Riefolo C., Di Fonzo N., Cattivelli L. (2007). Breeding progress in morpho-physiological, agronomical and qualitative traits of durum wheat cultivars released in Italy during the 20th century. European Journal of Agronomy 26: 39–53.

De Vita P., Mastrangelo A.M., Matteua L., Mazzucotelli E., Virzì N., Palumbo M., Lo Stortod M., Rizza F., Cattivelli L. (2010). Genetic improvement effects on yield stability in durum wheat genotypes grown in Italy. Field Crops Research 119: 68–77.

Dotlacil L., Hermuth J., Stenho Z., Dvoracek V., Bradova J., Leisova L. (2010). How can landraces contribute to present breeding? Czech Journal of Genetics and Plant Breeding 46: 70-74.

Ehdaie B., Waines J. G., Hall A. E. (1988). Differential responses of landrace and improved spring wheat genotypes to stress environments. Crop Sciences 28: 838-842.

Ganeva G., Korzum V., Landjeva S., Popova Z. (2010). Genetic diversity assessment of Bulgarian durum wheat (Triticum durum Desf.) landraces and modern cultivars using microsatellites markers. Gen. Res. Crop E. 57: 273-285.

Isidro J., Álvaro F., Royo C., Villegas D., Miralles D.J., García del Moral L.F. (2011). Changes in duration of developmental phases of durum wheat caused by breeding in Spain and Italy during the 20th century and its impact on yield. Annals of Botany 107: 1355–1366.

McCaig T.N. Clarke J.M. (1995). Breeding durum wheat in western Canada: historical trends in yield and related variables. Can. J. Plant Sci. 75: 55–60.

Motzo R., Giunta, F. 2007. The effect of breeding on the phenology of Italian durum wheats: From landraces to modern cultivars. Europ. J. Agronomy 26: 462-470.

Motzo R., Fois S., Giunta F. (2004). Relationship between grain yield and quality of durum wheats from different eras of breeding. Euphytica 140: 147-157.

Nazco R., Villegas D., Ammar K., Pena R. J., Moragues M., Royo C. (2012). Can Mediterranean durum wheat landraces contribute to improved grain quality attributes in modern cultivars? Euphytica 185: 1–17.

N.M.08.1.217.(1999). Norme Marocaine homologuée. Blé dur Indice de sédimentation au dodecyl sulfate de sodium.

Nsarellah N., Amamou A., Taghouti M., Annicchiarico P. (2011). Adaptation of Moroccan durum wheat varieties from different breeding era. Journal of Plant Breeding and Crop Science 3: 34-40.

Nsarellah N., Amri A., Nachit M. (2005). Amélioration génétique du blé dur. In La création variétale à l’INRA Méthodologie, acquis et perspectives. F. Abbad Andaloussi A. Chahbar (Eds), Maroc.

Pecetti L., Annicchiarico P. (1998). Agronomic value and plant type of Italian durum wheat cultivars from different eras of breeding. Euphytica 99: 9–15.

Santra M., Rao V.S., Tamhankar S.A. (2003). Modification of AACC procedure for measuring B-carotene in early generation durum wheat. Am. Assoc. of cereal Chemists, Inc. Cereal Chem. 80: 130-131.

Subira J., Pena RJ., Alvaro F., Ammar K., Ramdani A., Royo C. (2014). Breeding progress in the pasta-making quality of durum wheat cultivars released in Italy and Spain during the 20th Century. Crop et Pasture Science 65: 16–26.

Taghouti M., Gaboun F., Nsarellah N., Rhrib K., El-Haila M., Kamar M., Abbad-Andaloussi F., Udupa S.M. (2010).Genotype x Environment interaction for quality traits in durum wheat cultivars adapted to different environments. African Journal of Biotechnology 9: 3054-306.

Taghouti M., Gaboun F., Nsarellah N., Rhrib K., Nachit M., Saidi S., Boujnah M., Maataoui, F. (2006). Analyse de la diversité génétique des populations locales de blé dur et potentialités d’utilisation en amélioration génétique. Al Awamia 120.

Zadoks J.C., Chang T.T., Konzak C.F. (1974). A decimal code for the growth stage of cereals. Weed Res. 14 : 415-421.