Location of the experimental area

The research was carried out in the Valles del Yaqui and del Mayo, as representative sites of highly productive regions of irrigated wheat in Mexico, where about 160 and 97 thousand ha, respectively, are sown in the Valle del Yaqui (27° 30' N; 110° 20' W); and Valle del Mayo (26º 41' N; 109º 30' W), state of Sonora.

The main crop in these valleys is wheat with a surface irrigation system whose water comes from the Álvaro Obregón and Adolfo Ruiz Cortinez dams (^{SAGARPA, 2015}). The air temperature in these valleys has minimum and maximum levels of 0 and 45 ° C in winter and summer respectively. The predominant soils are classified as compacted clayly and alluvial (^{Ortiz and del Carmen, 2015}), based on the methodology proposed by Soil Taxonomy (^{Bhattacharyya et al., 2015}).

Genetic stability index

We used the methodology proposed by ^{Annicchiarico (1992)} to determine the genetic stability index of the grain yield components: plant height, spike length, grain mass and grain yield. The statistical data of grain yield components of the CIRNO C2008 variety were collected from 2007 to 2015 in the Valles del Yaqui and del Mayo where wheat is cultivated, especially the CIRNO C2008 variety (^{SAGARPA, 2015}).

Main characteristics of the CIRNO C2008 variety

The CIRNO C2008 variety is classified as crystalline or hard wheat (Triticum durum L.) and originated from the selection in segregating populations of the SOOTY-9 / RASCON-37 / CAMAYO cross, carried out in the Centro Internacional de Mejoramiento de Maíz y Trigo, CIMMYT (International Center for the Improvement of Maize and Wheat). This variety is directly related to the Átil C2000 variety, which originated from the SOOTY-9/RASCON-37 crossing; its release for cultivation took place in 2008 and is used mainly in the state of Sonora.

This variety grows in the spring, ideal to be cultivated during the autumn-winter cycle under irrigation conditions. Heading occurs from 74-89 d, the physiological maturity at 122 d on average and the length of stems is of 78 cm on average and is classified as a short-size variety; with erect stems and very low or null frequency of plants with flag recurved leaf. Before maturing, the sheath of the flag leaf and the peduncle of the spike present high levels of wax. Its grain yield when released for agricultural extensive use in Sonora was 5600 and 6300 kg ha^-1 with two and three auxiliary irrigations, respectively (^{Figueroa et al., 2010}).

Physiological and agronomic study in the 2015-2016 cycle

The CIRNO C2008 variety was planted in 2015-2016 in order to monitor physiological and agronomic characters in an experimental area of 0.48 ha (88.88 x 54.0 m), under open field conditions, in Block 910 of the Instituto Tecnologico de Sonora´s Centro Experimental de Transferencia de Tecnología (Technology Transfer Experimental Center) (CETT), located at 27 ° 22'0.4 '' N and 109 ° 54' 50.6 '' W (UTM coordinates: 607393.24 m E; 3027508.34 m N).

Sowing, fertilization and maintenance of the crop in the 2015-2016 cycle

Sowing was carried out with a planter (ST 16) on December 8, 2015, in three rows on the quarry with a planting density of 138 kg ha^-1. Fertilization was based on 300 kg ha^-1 of urea + 100 kg ha^-1 of monoammonium phosphate (MAP) 11-52-00. In the first auxiliary irrigation, 150 kg ha^-1 of urea were applied and the other part was divided in the second and third irrigations at a rate of 75 kg ha^-1 each.

The irrigations were at pre-sowing, tillering, heading and grain filling with a mean water depth of 16 cm, and a gross partial standard of Nb = 160 m³ ha^-1 in each irrigation until reaching field capacity. The average watering interval was 25 d. For these tasks we followed the technical instruction for this variety (^{Figueroa et al., 2010}).

Control of pests and diseases

The foliage aphid (Schizaphis graminum) was controlled with the pesticide Muralla Max (ia Imidacloprid + Betaciflutrin) (0.20 L ha^-1 in the periphery of the plot up to 2.0-3.0 m inwards (edge surface), at the beginning of heading, when the pest was present. The broadleaf weeds were controlled with manual weeding before applying auxiliary irrigations.

Agroclimatic variables during the 2015-2016 crop cycle

During the biological cycle of the crop, the average monthly temperature remained between 17 and 24 ° C (average 18.6 ° C). Monthly precipitation was less than 0.2 mm and relative humidity varied from 50 to 78 % (Figure 1).

Figure 1 Climatic variables (temperature, precipitation and relative humidity) during the cultivation cycle December 2015 - May 2016 at the Technological Transfer Experimental Center of the Technological Institute of Sonora (CETT). 

Variables evaluated in the CETT experiment in the 2015-16 cycle

Experimental design and statistical analysis

Genetic stability index of the CIRNO C2008 variety

Wheat, like other crops, fluctuates within a single crop between years and different sites due to the polygenic nature of grain yield components, and these variations depend on the ability of the genotype to adapt to changing conditions such as availability of water and nutrients, solar radiation, temperature, and other factors (^{Shirdelmoghanloo et al., 2016}).

The results showed a significant interaction between the years (cultivation seasons) and the valleys, with the exception of the GLL variable. The variables AP and REND showed low values, and moderate LP and GLL, which denotes the little dispersion and considerable homogeneity of the data obtained for these characteristics, as an indirect indicator that the variety CIRNO C2008 did not have significant variations in its agronomic performance during the years and the valleys where it was evaluated (Table 1).

The four characteristics evaluated had low correlations between them and only r (REND - GILL) = 0.35 was statistically different (p ≤ 0.01). The highest values of heritability were found in the variables plant height and spike length, indicating the lower negative incidence of environmental factors. The number of filled grains and grain yield showed heritability values close to 50%, which indicates greater sensitivity to changes in the environment.

Obtaining a coefficient of low genetic variation for AP, REND and LP in different localities and eight years denotes the presence of monogenic dominant characters (^{Gouache et al., 2016}) or the superdominance of minor genes in the groups of genes that affect positively the grain yield (^{Mirabella et al., 2016}).

The polygenic character of the grain yield of cereals in general, and of wheat in particular, is the main limitation for genetic improvement programs, but numerous genes have been studied that overexpress and show superdominance, although they have been genotypes exposed to the environment for a long time. Many of these genes or groups of genes were characterized and used for the differential selection of genotypes among diverse germplasms or between parents in genetic crosses (^{Vázquez et al., 2016}).

The genetic stability of morphological (^{Liu et al., 2016}) and agronomic characters is important because they can be useful in any breeding program or for the use of genotypes as a model to simulate or predict possible responses to stress conditions (^{Vázquez et al., 2016}).

The negative indices (Table 2) represent unfavorable times for the evaluated variables. The variables GLL and REND presented similar classification as favorable or unfavorable at the same times. The classification result was similar for the other two variables, with the exception of the first period (2007/08).

The 2011/12 season maintained the best environment for all the performance components evaluated in both valleys, classified as favorable, while 2013/14 had an opposite effect.

According to the indices described by ^{Annichiarico (1992)} (Table 3), (Wi) the Valle del Mayo showed better results than the Valle del Yaqui regarding the variables evaluated, with the exception of REND whose higher values occurred in the Valle del Yaqui. Similarly, in this valley the indices for favorable environments (Wi_F) were higher for AP and GLL.

In the present analysis (Tables 2 and 3) the classification is more associated with the variation of the valleys than with the different seasons. The indices obtained, with the exception of the spike length (LP) that had a greater difference between the values of Wi, WiF and WiD, did not register statistical differences in the other three variables, which indicates little variation or high stability of the morphological and agronomic characters of the cultivar in both valleys.

The Wi index was higher in the Valle del Mayo, except for the REND characteristic that had the highest values in the Valle del Yaqui. This type of genotype stability analysis is used in the selection of wheat cultivars (^{Benin et al., 2014}), to identify genotypes of predictable behavior that respond to environmental variations under specific (Wi_F and Wi_D) or general (Wi) conditions.

Physiological and agronomic characteristics of the variety CIRNO C2008 in the CETT 910 in the 2015-2016 crop cycle

The phenology of the variety CIRNO C2008 had no significant changes due to the effect of the cultivation area and there was only a significant difference with the values of the varietal descriptor (^{Figueroa et al., 2010}) in the phenophase of heading. This variation did not influence the subsequent phenophases (Table 4).

Commercial varieties can shorten or lengthen the time of a phenophase due to the edaphoclimatic conditions in the region where they are established (^{Kurepin et al., 2015}). The most frequent in studies of exposure and tolerance to stress conditions is a reduction of phenophase time (^{Guzmán et al., 2016}), which accelerates senescence and abscission, especially during physiological drought associated with drought or salinity stress (^{Mendes et al., 2016}).

For thermal stress conditions, the time of phenophases decreases due to the increase in the concentration of gibberellic acid (^{Argentel et al., 2016}). The time of the crop phenophases is an efficient indicator of the good vegetative and nutritional state of the plants, and has a positive correlation with the yield, but it is necessary to carry out complementary studies such as the vegetable biomass index in each phenophase (^{Ballesteros et al., 2016}).

Normalized difference vegetation index

The NDVI evaluated in the phenophases of the CIRNO C2008 crop had a significant increase since the tillering phenophase and the maximum point was obtained in the heading (Figure 2). This result shows the good physiological and nutritive state of the plant. The decrease in the NDVI values since heading is probably due to the translocation of nitrogen compounds, which favors the filling of the grains and is an important sign of the senescence of the plants (^{Gaju et al., 2016}).

Figure 2 Normalized difference vegetation index in the main phenophases of the CIRNO C2008 wheat variety in the CETT during the 2015-16 crop cycle. The bars represent the standard deviations of the measurements in each phenophase. 

Measurements of the nutrient status of plants in the different phenophases of the crop through the Greenseeker has been very successful in the intensive and extensive wheat systems and allows to quantify the efficiency of the use of nitrogen fertilizer, as well as to predict the grain yield (^{Stefen et al., 2016}).

Water use efficiency (WUE) based on the photosynthesis-transpiration ratio

The WUE was similar in the leaf in the three phenophases evaluated (Figure 3); however, in the spikes, although it was higher in the phenophases of heading and flowering than in the leaf, there was a significant decrease in their values after the phenology, which could be 99 % due to the effect of the phenophases.

Figure 3 Water use efficiency in leaves and spikes of the CIRNO C2008 variety during the phenophases of heading, flowering and filling of the grains in the CETT in the 2015-16 cycle. R²: coefficient of determination, without adjustment, of the WUE variation in the different phenophases. 

The decrease in the photosynthetic activity of the spikes was the factor that most influenced the WUE variations (Table 5), which is attributed to the loss of the chlorophyll pigments as the spike grows (^{Jamil et al., 2016}), the rupture of chloroplasts due to the concentration of ethylene during maturation, and the increase in the activity of the chlorophyllase enzyme that accelerates chlorophyll catabolism (^{Sánchez et al., 2016}).

The highest value of WUE in the spikes was obtained during the heading, which is due to the fact that the spikes have, in their green state, considerable advantage for capturing the light with respect to the flag leaf and the remaining leaves, as a result of the little or no interference from shading or high irradiance, allowing it to maximize photosynthesis (^{Blum, 2005}).

In addition, low transpiration also contributed since the spikes are not bodies specialized in the realization of this process. This response was the main contributor to WUE superiority of spikes over the leaf in the present study.

The evaluation of WUE determined by the photosynthesis/transpiration ratio is an important indicator to select genotypes with high photosynthetic capacity in conditions of low water availability, and tolerance to drought stress is attributed to it (^{Messina et al., 2015}).

These genotypes might be tolerant to thermal stress, although it is assumed that this type of stress considerably increases transpiration to achieve thermoregulation of the leaf mesophyll (^{Cowie et al., 2016}).

Grain yield components

Decreases in the spike mass (5.35 %) and the number of full kernels (3.3 %) were significant with respect to the varietal descriptor (Table 6), although there was no negative correlation between these variables and agricultural yield (r _{(MP- REND)} = 0.012; r _(GLLP-REND) = 0.018). On the contrary, the dry mass of the grain and the grain yield were statistically higher.

Obtaining a superior yield in relation to the varietal descriptor shows that the variety still maintains its genetic-productive potential, perhaps due to the ability of the plant to mobilize the photoassimilates from the foliage to the spike and propitiate an adequate filling of the grains. (^{Vázquez et al., 2016}). This can be corroborated by the significant difference found between the straw and grain dry mass values. The dry mass relation in terms of grain/straw can vary because the edaphoclimatic conditions and of cultivation affect the translocation and distribution of photoassimilates from the sources and reservoirs up to the sinks (^{Hortelano et al., 2013}).

The variables plant height and spike length did not present significant differences with respect to the varietal descriptor (Table 6), which shows the genetic stability of the variety during its exposure to field conditions. This is notable because most of the yield components have a polygenic nature but perhaps there are minor genes that still maintain their dominant character before a considerable variability of edaphoclimatic conditions (^{Asseng et al., 2015}). In wheat, plant height and in particular the NDVI are efficient variables for the estimation of yield, and are reported of being precise indicators for differential selection programs (^{Pantazi et al., 2016}).