Drought susceptibility indexes (DSIGY) of maize and triticale genotypes
According to our earlier research (Grzesiak et al. 2012), drought susceptibility indexes (DSIGY) for maize and triticale genotypes were calculated by determining the changes in grain yield (GY) under two soil moisture levels (irrigated and drought). Variation of DSIGY for maize ranges from 0.381 to 0.650 and for triticale from 0.354 to 0.578. The values of DSIGY made it possible to rank the examined maize and triticale genotypes according to their drought tolerance. In the maize hybrids the drought resistant group (0.381 < DSIGY > 0.439) comprises the hybrids Tina, Garst 8344, Pioneer 3925 and Pioneer 3957, while the drought sensitive group (0.607 < DSIGY > 0.650) comprises the hybrids Ankora, Garst 8702, Garst 8388 and Garst R5515. For triticale the drought resistant group (0.354 < DSIGY > 0.419) comprises the lines CHD 247 and CHD 220, and cultivars Migo and Wanad, while the drought sensitive group (0.544 < DSIGY > 0.578) comprises the strains CHD 147 and CHD 12, and cultivars Mieszko and Maja (Fig. 1).
Germination under simulated drought conditions (FG, PI) and seedling survival (SS)
In the control treatment (0.0 MPa), maize and triticale genotypes did not differ in the final germination index (FG) and the correlation coefficients between FG and DSIGY were insignificant. The imposed drought from −0.43 to 1.30 MPa caused a decrease in the values of stress indexes of FG and the correlation coefficients between FG and DSIGY were statistically significant for maize in −0.43 MPa treatment and for triticale in −0.85 and −1.30 MPa treatments (Table 1). Differences between maize and triticale genotypes in promptness index (PI) values in the control were statistically significant only for maize, whilst the correlation coefficient with DSIGY was statistically significant for both species. As with drought FG, osmotic stress caused a decrease in the values of PI and the correlation coefficients between PI and DSIGY were statistically significant for maize in −0.43 and −0.85 MPa treatments, but for triticale only in −1.30 MPa treatment (Table 2). The results of seedling survival (SS) after grain germination in different water potential of mannitol solutions are presented in Table 3. Germination under osmotic stress conditions (−0.43, −0.85. −1.30 MPa) caused a decrease in the number of live seedlings in maize to 86, 79 and 76 %, respectively, and for triticale to 97, 86 and 84 %, respectively. Correlation coefficients between DSIGY and the values of stress index of seedling survival were statistically significant in −0.43, −0.85 and −1.30 MPa treatments of maize genotypes, but for triticale only −0.85 and −1.30 MPa treatments were affected.
Germination is strongly influenced by plant species, grain age and storage conditions and is also highly sensitive to soil water quality and temperature (Ashraf and Mehmood 1990; Ahmad et al. 2009). Grain germination indexes were used with different results in selecting procedures for identifying drought resistant genotypes (Jajarmini 2009). Opinions on the efficiency of plant drought tolerance assessments on the basis of different parameters of germination are divergent and indicate their limited usefulness. It is believed that germination traits are affected by drought stress less than other physiological and biochemical traits (Winter et al.1988; Morgan 1992). Genotype tolerance to drought stress is process-specific with regard to grain water imbibition, endosperm utilization, activation of the dormant enzyme system and seedling emergence, growth and survival after stress. The tolerance to drought in any of those stages will affect the results and thus screening for drought tolerance reaction should also be process-specific, and the prediction of genotype performance from one process to another would not necessarily be possible (Blum et al. 1980; Ahmad et al. 2009). In this study the tolerance to simulated drought stress proved a relatively good indicator in the screening of maize and triticale genotypes. Our results showed that the efficiency of FG, PI and SS is related to how strongly a trait is expressed and the measurement must be performed at the right moment in order to maximize the expression of genetic variations of the particular trait.
Seedling dry matter (DWS, DWR), root length (LS, LSA, LN) and the number of nodal roots (NN)
Soil drought decreased the dry matter of the above ground parts (DWS) in 13 genotypes of maize from 12.0 to 67.0 % and in 11 genotypes of triticale from 5.0 to 17.0 %. Similarly, the dry matter of roots (DWR) decreased in maize genotypes from 3.0 to 30.0 % and in triticale genotypes from 1.0 to 25.0 %. Drought also strongly influenced the ratio of DWS to DWR within maize genotypes from 0.0 to 60.0 % and in triticale genotypes from 0.0 to 25.0 % (Table 4). The variation coefficients (CV) in control (C) seedlings of DWS, DWR and the ratio of DWS to DWR in maize were about 29, 18 and 37 %, respectively, and in triticale 23, 29 and 22 %, respectively. CV calculated for treatment D as a stress index (SI) for those traits in maize were about 34, 13 and 31 %, respectively, and in triticale 15, 9 and 11 %, respectively (Table 4). In the control treatment (C), the statistically significant correlation between DSIGY and DWS, DWR and the ratio of DWS to DWR were found only in triticale for the ratio of DWS to DWR. In the drought treatment, the statistically significant correlation between DSIGY and DWS, DWR and the ratio of DWS to DWR were found in maize for DWS and the ratio of DWS to DWR, and in triticale, only for DWS (Table 4).
The results presented in Table 5 show that in the control (C) treatment, differences within drought resistant and drought sensitive genotypes of maize and triticale in the length of seminal (LS), seminal adventitious (LSA), nodal roots (LN), and number of nodal roots (NN) were in most cases statistically insignificant. Also under control conditions the correlation coefficient between DSIGY and the measured traits was statistically insignificant. LS in drought treatment decreased in the drought resistant genotypes of maize (Tina, Garst 8344) about 10 % and in drought sensitive genotypes (Ancora, Garst 8702) about 23 %. In drought resistant genotypes of triticale (CHD-220, CHD-247) the decrease was about 6 % and in drought sensitive genotypes (CHD-12, CHD-147) about 13 %. In maize the decrease in the total length of two seminal adventitious roots (LSA) in comparison to the control was about 7 % in drought resistant hybrids and in drought sensitive about 25 %. For triticale the decrease in the total length of three seminal adventitious (LSA) roots in comparison to the control was statistically insignificant in drought resistant and sensitive lines. In seedlings exposed to drought, significant differences between drought resistant and sensitive genotypes were observed in the measurements of the total length of nodal roots (LN). In maize the decrease in LN in drought sensitive genotypes was about 22 % and in the drought resistant ones about 10 %, and in triticale about 30 and 13 %, respectively. In C treatments no statistically significant correlation between DSIGY and LS, LSA, NN and LN was found in either maize or triticale genotypes. In seedlings exposed to drought, a statistically significant correlation between DSIGY and SI calculated for the measurements of root traits was found in maize, but in triticale only for LN (Table 5).
As one might expect, a root system characteristic, such as the number and length of particular components of root system structure, their depth and abundance, are known to be associated with performance under drought conditions in many studies of cereal species (Richards 1996; Reynolds 2002). Nevertheless, decreased allocation in roots in the top layer of the soil has been shown to be an effective drought stress adaptive mechanism (Richards 1991). In the studies by Kono et al. (1987) the specific response of cereal species to drought stress was clearly noticeable in root distribution, nodal root number, leaf number and grain yield. Those authors reported that the responses in root growth for maize and rice were different in the downward penetration of the main axis and in the higher order laterals. Species with a “concentrated” type of root system showed less restriction of root and shoot growth compared to species with a “scattered” type. A decrease in size of the root system and an increasing irregularity of root distribution resulted in water and nutrients being transported greater distances to the nearest roots (Tardieu 1991; Lipiec et al. 1996). Drought also modifies root system components, such as the main root axis and lateral roots of different orders in rice and maize (Iijima and Kono 1991). Changes in the morphological structure of the root system in triticale and maize were also observed in our earlier studies in response to waterlogging and soil compaction (Grzesiak et al. 2012).
Leaf water potential (ψ) and water loss of excised leaves (WL)
Moderate (D35) and severe (D20) soil drought decreased ψ in maize and triticale (Table 6). Differences between resistant genotypes of maize (Garst 8344, Tina) and triticale (CHD-247, CHD-220) in terms of a decrease of ψ were statistically significant in comparison with sensitive genotypes of maize (Ankora, Garst 8702, Garst 8388) and triticale (Maja, CHD-247).
In seedlings exposed to drought, a statistically significant correlation between DSIGY and SI calculated for the measurement of ψ in maize was found for both treatments but in triticale only for treatment D20. The differences in WL between drought resistant and sensitive genotypes of maize and triticale genotypes were statistically significant. A statistically significant correlation between DSIGY and WL in maize was found only in the measurements taken after 12 h of leaf desiccation but in triticale in the measurements taken after 12 and 24 h (Table 6).
Decreases in leaf water potential initially induced stomatal closure, resulting in a decrease in the supply of CO2 to the mesophyll cells and subsequently in a decrease in the rate of leaf photosynthesis (Williams et al. 1999; Lawlor and Cornic 2002). According to Hura et al. (2007), a statistically significant correlation between water potential and photosynthetic rate and stomatal conductance was found in maize and triticale during various stages of plant development. Dehydration in C3 and C4 plants impairs various physiological processes, especially the changes in leaf water content, water potential and photosynthesis. There is a controversy as to whether drought limits photosynthesis due to leaf water status, stomatal closure, metabolic impairment or injuries to photosynthetic apparatus (Flexas et al. 2006). The first response to leaf water deficit is stomata closure, which limits CO2 diffusion to chloroplasts (Berkowitz et al. 1983; Cornic and Masacci 1996; Muller and Whitsitt 1996). Non-stomatal mechanisms under prolonged or severe soil drought include changes in chlorophyll synthesis, functional and structural changes in chloroplasts and also disturbances in accumulation and distribution of assimilation products (Medrano et al. 2002). However, it is known that during drought stress, plants with C4 photosynthesis increase water use efficiency and suppress photorespiration. Thus, C4 plants are often more competitive than C3 plants in drought-prone areas (Edwards and Ku 1987; King 2011).
Relative loss of intracellular electrolytes from leaf tissues (LI)
In drought sensitive genotypes of maize and triticale, the values of indexes of leaf injury by soil drought (D35, D20), osmotic stress (−0.47, −0.85, −1.30 MPa), and heat temperature (25, 35, 45 °C) were in general higher than in drought resistant genotypes and in most cases the differences between resistant and sensitive genotypes were statistically significant (Table 7). In treatment D35, the values of coefficients of variation (CV) in maize and triticale were higher than in D20 treatment. Under osmotic stress, higher values of CV were found in −0.85 MPa treatment and under high temperature stress in 35 °C treatment.
The ability to maintain the structure and function of cytoplasmatic membranes under water deficit is one of the most important physiological traits. Conductometric measurements of LI are applied as a screening test for the estimation of tolerance to various stresses (Vietor et al. 1977; Richards 1978; Blum and Ebercon 1981; Poljakoff-Mayber 1981; Martinielio and Lorenzoni 1985; Palta 1990). Differences between sensitive and resistant genotypes might stem from the fact that drought resistant genotypes possess more efficient mechanisms protecting membrane functions and structure. Drought stress causes a loosening of lamellar membranes in chloroplasts, loss of a certain amount of grana, and increase in the level of coarse-grain matrix (Haupt-Harting and Fock 2002; Lawlor and Cornic 2002; Tang et al. 2002). Some authors suggest that drought resistant plant species show stronger binding of chlorophyll molecules to the lipid-protein complex of chloroplast membranes (Smirnoff and Colombe 1988; Bukhov et al. 1990). Our earlier results indicate that leaf age is very important because differences in LI between drought resistant and sensitive genotypes were smallest in the oldest and youngest leaves, though the greatest differences were observed in the leaves where cellular divisions had taken place and which had reached the maximal area (Grzesiak et al. 2006).
Correlations among stress parameters (DWS, DWR, DWS/DWR) were significant, except for the relationship between DWS and DWR in maize. In this experiment the correlations between DWR and other traits were not significant except for seedling survival (SS). For both species, high and significant correlations were found between measurements of membrane injures due to drought (LID), osmotic stress (LIOS), and high temperature (LIHT). Also for both species, high and significant correlations were found between leaf water potential (ψ) and water loss (WL) and between FG and PI, except for maize (Table 8).
Most of physiological processes are affected by the stresses of soil drought, osmotic and high temperatures (Levitt 1980). The usefulness of methods for studying plant drought tolerance has been discussed in many papers and reviews (Blum et al. 1980; Kpoghomou et al. 1990; Zagdańska 1992; Reynolds 2002). According to some authors, besides the evaluation criteria used in the present research, positive results were also obtained by utilizing the measurements of canopy infra-red temperature, changes of leaf color and responses of root architecture (Clarke and McCaig 1982; Stanley 1990). Partly positive results were obtained in tests of the plant’s ability to reduce leaf area, the development of wax bloom on leaves, leaf rolling and the ability of the leaf to hold water (Passioura et al. 1993). On the other hand, the effectiveness of the evaluation of tolerance based on the measurement of different parameters of leaf gaseous exchange and the content of various metabolites, including proline, has not been definitely confirmed. (Hanson and Nelson 1985; Farquhar et al. 1993; Dubey 1997; Bandurska and Stroiński 2003).
The maize and triticale genotypes used in these experiments show a relatively wide range of drought tolerance. This study contributes to the understanding of responses of different genotypes to drought, though in this research, the method of estimating drought susceptibility index was relatively simple and did not take into account other important factors of soil–water-plant relationship. The correlations between DSIGY and laboratory tests (LI, FG, SS) showed that they are good indicators of plant drought tolerance. The correlation coefficients between LIOS and LIHT tests were, in most of the considered cases, statistically significant, thus indicating that in maize and triticale the mechanisms of membrane injury caused by simulated drought or high temperature were physiologically similar. It can be concluded that an approach to the breeding of maize and triticale for drought tolerance using these tests can be implemented on the basis of separate selection for each trait or for all of them simultaneously. In that case, it would be necessary to determine the importance of the trait in relation to the growth phase, drought timing and level as well as associations with morphological traits contributing to drought tolerance (Kono et al. 1987; Kpoghomou et al. 1990).
The results presented in this paper, our earlier research (Grzesiak 1990; Grzesiak et al. 2012) and the results of other authors (Lorens et al. 1987; Martinielio and Lorenzoni 1985; Kono et al. 1987) confirm the existence of a wide range of genotypic variability of response to drought in cereal plants. In maize and triticale, similar to other crop plants, the physiological reasons for this variability have not as yet been fully recognized and explained. In the literature one can find many contradictory conclusions, as the reduction of crop yield of the tested genotypes depends not only on the drought level and its duration, but also on the phase of plant growth and development and interaction with other environmental factors (Hanson and Nelson 1985; Blum 1988; Naylor and Su 1998).
The grounds for genotypic variation of drought tolerance in cereals has not yet been entirely elucidated and future research is necessary (Zagdańska 1992; Royo et al. 2000; Reynolds 2002). The frequent occurrence of drought stress in many regions of the world and the deteriorating water conditions for plant growth and productivity have raised interest in research into the responses of crops to periodic water deficiency. Progress in the breeding of drought resistant cereal plants requires future study of the physiological mechanisms underlying the responses of plants to water stress. It is a widely held opinion that the breeding of drought resistant crop plants will not be an easy task (Hanson and Nelson 1985; Zagdańska 1992; Reynolds 2002). The complexity of the property of drought tolerance will make it necessary to take into consideration various tolerance evaluation criteria in the breeding programs (Levitt 1980; Turner 1986; Blum 1988; Richards 1991; Jones 1993; King 2011).
MTG, FJ, PW, KH designed the research; MTG, PS, IM and TG conducted the research; MTG, FJ, PS and KH analyzed the data; MTG, FJ, IM and KH wrote the paper; MTG had primary responsibility for the final content. All authors have read and approved the final manuscript.