Grain yield and its components

In physiological maturity (Z89) (^{Zadoks et al., 1974}) the 0.4 m² plants were extracted in each plot. The dry weight of leaves, stems and ears was recorded after drying the samples in a forced air oven for 72 h at 70 °C, until reaching constant weight. The grain yield per m² (RG) was determined by weighing the grain obtained in the sample and the individual grain weight (PIG) ​​when counting and weighing 1 000 grains.

Laboratory analysis

The CHS content was determined in anthesis (CHSAnt) and physiological maturity (CHSMf), randomly selecting a subsample of 20 fertile main stems (without spike or leaf blade) from the biomass sample of each plot in each treatment. The CHS were extracted and measured using the anthrone method of ^{Yemm and Willis (1954)} described by ^{Galicia et al. (2009)}. The data was obtained as a percentage of CHS in dry matter. These were used to calculate the CHS content per stem (g stem^-1) or per unit area (g m^-2), according to the procedure described by ^{Pask et al. (2013)}.

The amount of CHS mobilized from the stem to the spike, the remobilization efficiency and the contribution to the grain yield were calculated using the following formulas (^{Shakiba et al., 1996}): 1. Total amount of CHS remobilized or translocated (CHSTransl)= maximum quantity of CHS-quantity of CHS at maturity; 2. Total remobilization efficiency ERCHS= total amount of CHS remobilized x 100/maximum amount of CHS; 3. Contribution to grain yield (CRG)= total amount of CHS remobilized x 100/grain yield; and 4. Water soluble carbohydrates present in the grain (CHSG).

Statistical analysis

Analysis of variance was carried out according to the models mentioned above (^{Littell et al., 1996}) to evaluate the effect of treatments (environments) on each of the variables studied. When analysis of variance revealed significant differences, the mean values for each treatment were compared using Tukey's 5% honest minimum difference (DMSH) (^{Palaniswamy and Palaniswamy, 2006}) using SAS software (^{SAS, 2002}).

Effect of treatments on the variables studied

In the analysis of variance, important effects of the treatments were observed on the variables evaluated in both experiments (Table 1 and 2). It can be seen that in the exp. 1 The effect of the genotypes was highly significant (p< 0.01) for all the variables analyzed in the same way for environments with the exception of CHSAnt and the genotype-environment interaction (GxA) was not significant for the PIG (Table 1). The coefficients of variation ranged from 0.9 to 19.3% corresponding to soluble carbohydrates stored in the stem in anthesis and soluble carbohydrates stored at physiological maturity (CHSMf) (Table 1).

The variability observed among genotypes in terms of RG represented 62 and 48% in experiment 1 and 2, respectively of the total variation, while for the PIG it represented 50% in both experiments. For the accumulation of carbohydrates soluble in the stem in anthesis (CHSAnt) the genotypic effect observed 95% of the total variability, which coincides with evidence in the literature that mention that there is genetic variability for the accumulation of CHS in the wheat stem (^{Ruuska et al., 2006}; ^{Dreccer et al., 2009}).

On the other hand, the variability between genotypes in terms of CHS remobilization and their contribution to grain yield represented 83 and 80% of the total variability, which indicates that the genetic improvement in wheat towards a greater contribution to the RG of the reserves stored could be possible to stabilize the grain yield in environments where stress due to biotic and abiotic factors is frequent (^{Ehdaie et al., 2008}).

General response of environments

With the exception of CHSAnt, defoliation affected significantly (p< 0.05) all the variables studied. The RG, PIG, CHSMf and CHSG decreased on average 14, 9, 35 and 10%, respectively with respect to the non-defoliated treatment, while, CHSTransl, ERCHS and CRG increased significantly by 30, 31 and 52% respectively (Table 2) with defoliation. Drought on average decreased GR, PIG and CHSMf by 14, 9 and 71%, respectively with respect to irrigation, while in the variables CHSAnt, CHStransl, ERCHS, CRG, CHSG, significant increases were observed (56, 142, 56, 181 and 13%, respectively) with respect to the irrigated treatment (Table 2).

In both experiments there were significant increases in the values of CHTransl, ERCHS and CRG (30, 31 and 52%, respectively in defoliation, 142, 56 and 181%, respectively in drought). The contribution of stem reserves to grain yield increased, when the genotypes were subjected to different environments (defoliation and drought), with the contribution and remobilization of the assimilated during the filling of larger grain under drought conditions. Similar results report that the relative contribution of stem reserves to grain yield varies widely depending on environmental conditions and genotypes and their values range between 6 and 100% (^{Borrell et al., 1993}).

Evidence in the literature mentions that the reduction in assimilation under post-anthesis humidity stress will induce a greater mobilization of stem reserves that will be used by the grain (^{Yang et al., 2000}). Also, several studies have shown that water stress during grain filling promotes senescence of plants and increases the remobilization of stored carbon reserves in pre-anthesis towards the grain (^{Ahmadi et al., 2009}). The major contribution to grain yield under stress conditions (defoliation and post-anthesis drought) was in line with a greater efficiency of remobilization of the CHS from the stem.

Grain yield was marginally reduced (14% on average) both by defoliation and by drought, which can be attributed to the fact that the extent of retranslocation of assimilates was much higher (59%), under that condition, acting as an element of important compensation to ensure grain filling (^{Singh and Singh, 2002}; ^{Ehdaie et al., 2008}).

General effect of genotypes

A wide variability was observed among the 60 genotypes evaluated, in the variables RG, PIG, CHS in anthesis, translocated CHS and contribution of CHS to the grain yield (Figure 1 and 2). The RG ranged from 302 to 585 g m^-2, the PIG was between 31 and 57 mg grain^-1 (Figure 1).

Figure 1 Grain yield (a) and grain weight; and (b) for 60 wheat lines grown in Toluca, Mexico. Average data of environments (defoliated and unfrozen).  

Figure 2 Carbohydrates soluble in anthesis (a) translocated soluble carbohydrates; (b) contribution to grain yield; and (c) for 60 wheat lines. Average data of environments (defoliated and defoliate).  

While the CHs in anthesis were in the range of 125.6 to 545.1 g m^-2, the CHS translocated between 11.2 to 464.9 g m^-2 and the contribution of CHS to the yield ranged between 3 and 155.5% (Figure 2). A greater amount of CHS remobilized to the grains was in line with a greater contribution of these to the RG. In this sense, genotypes 8, 12, 13, 17, 19, 26 and 40, which translocated between 76 and 469.4 g m^-2, in percentage terms also contributed more to the grain yield (76.4 to 155.5%) (Figure 2).

The great variability observed for grain yield, CHSAnt accumulation and contribution to grain yield shows that within elite wheat germplasm (CIMCOG) there is potential to select progenitors for breeding programs with a view to incorporating promising physiological characteristics (CHS) (^{Rebetzke et al., 2008}) that allow to increase the yield under conditions of biotic stress (foliar diseases) (^{Serrago et al., 2011}) or abiotic (drought) (^{Foulkes et al., 2007}). A greater contribution of CHS to the grain yield was related to a greater efficiency of remobilization (Figure 3). In the defoliated plants there was a greater contribution of the CHS to the grain yield (range between 2.4 and 138.4%) with respect to the plants without defoliate (between 5.2 and 100.6%). In the same way, the efficiency of remobilization presented higher average values in defoliated plants (range between 5.4 to 94. 9%) in relation to non-defoliated plants (5.2 to 87.1%) (Figure 3).

Figure 3 Relationship between the contribution of soluble carbohydrates to yield and efficiency of remobilization, in 60 wheat lines grown under defoliated and unfrozen conditions.  

Effect of water stress on CHS and grain yield

Water stress post-anthesis affected the RG, the PIG and the amount of CHS translocated from the stem to the grain in the different cultivars (Figure 4). The average RG in drought was 371.8 g m^-2 and under irrigation 431.5 g m^-2, which represents a 14% reduction with respect to the irrigation treatment. In this sense, the translocation of CHS under drought conditions was higher (318.2 g m^-2) with respect to irrigation (131.5 g m^-2), representing an increase of 141% with respect to the irrigation condition. Similar results of CHS remobilization were found by ^{Zhang et al. (2012)} when water availability represented only 50% of a favorable condition.

Figure 4 Relationship between grain yield (a) individual grain weight; and (b) soluble carbohydrates translocated from the stem to the spike, in 20 wheat lines grown in two environments. Vertical and horizontal lines within the figures represent the average of each environment.  

In regions where the availability of water for grain filling is limiting, CHS play an important role in the definition of RG (^{Ehdaie et al., 2008}), therefore, the identification of genotypes with the ability to accumulate and remobilize CHS from the stem is a priority in breeding programs (^{Dreccer et al., 2009}). Based on the great variability observed in both environments, it was possible to identify genotypes (12 and 16) that did not show significant changes in the RG and in the PIG in both environments, mainly due to their ability to remobilize assimilated from the stem to the spike (Figure 4a, 4b).

The average PIG was marginally affected (9.5%) by the post-anthesis water deficit, while the CHS translocation under this condition represented 59% with respect to the control treatment (Figure 4b). The contribution of the CHS to the yield of grain in drought was higher (range between 31.0 to 155.6%) with respect to irrigation, thus also greater efficiency of remobilization was observed (82.9 to 95.7%) with respect to the cultivars grown under irrigated conditions, with values of 10.8 to 71.7% for the contribution to yield and from 23.3 to 80% in efficiency of remobilization (Figure 5a).

Figure 5 Relationship between CHS contribution to yield and efficiency of CHS (a) CHS remobilization in physiological maturity and stem dry weight (b), in 20 wheat lines grown under irrigated and drought conditions.  

In drought it could be seen that for similar values of remobilization efficiency there was a wide variability in the contribution of the CHS to grain yield (Figure 5a). A higher CHS content per stem at physiological maturity was associated with a higher dry weight of them at that stage (Figure 5b). The content of carbohydrates stored in the stem for the genotypes grown under irrigation conditions was 0.27 to 1.23 g stem^-1, while in the drought situation it was 0.06 to 0.28 g stem^-1 (Figure 5b).