Combining ability of physiological traits in forrage maize Capacidade combinatória

The exploitation of the existing genetic variability for the physiological traits related to the yield and quality of maize forage can assist in the development of superior inbred lines and hybrids. The objective of this work was to evaluate the general and the specific combining ability and the nature of gene effects of physiological and chemical traits of forage maize genotypes. Two groups of genotypes and 16 hybrids resulting from a 4x4 partial diallel scheme were evaluated. Group I consisted of two commercial hybrids (AG8025, P30B39) and two elite inbred lines (LEM2 and LEM3). Group II consisted of four experimental inbred lines originated from different populations of forage maize breeding program. In total, 24 treatments were evaluated, formed by the genotypes of both groups and the respective crossings. The traits evaluated were: CO 2 assimilation ( A ), stomatal conductance ( gs ), internal CO 2 concentration ( Ci ), transpiration ( E ), calculated activity of Rubisco ( A / Ci ) and efficiency of water use ( A / E ). Forage acid detergent fiber (ADF), neutral detergent fiber (NDF) and digestibility in situ were obtained. There was a predominance of non-additive gene effects for most of the chemical and physiological traits. Crossings LEM2 x 203-218.3, LEM3 x 201-107.2, LEM2 x 101-7.2 and LEM3 x 101-7.2 stood out regarding CO 2 assimilation, and are indicated for future research considering the physiological traits. Inbred lines 101-7.2 and 203-218.3 presented high concentration of favorable alleles to increase carboxylation efficiency, in which inbred line 101-7.2 stood out for NDF, ADF and DIG. Inbred line 201-107.2 has a high concentration of favorable alleles for efficiency of water use. Physiological parameters can assist the selection of inbred lines and hybrids in maize breeding for forage purpose.


INtroDuCtIoN
Physiological and chemical variables are important to understand the differences in the performance of vegetal species and selection of superior genotypes. These traits are used to verify the culture adaptation to new environments, interspecific competition and effects of handling systems, in addition to the yield potential of different genotypes (Alvarez;Crusciol;Nascente, 2012;Chen et al., 2018).
The interaction of environmental factors such as light, temperature, CO 2 concentration and availability and use of water and nutrients affect several traits substantially, being determining factors in the selection of genotypes with greater photosynthetic efficiency, better water use and, as a result, higher productivity, since more efficient genotypes in physiological and chemical traits tend to present higher yield potential (Alvarez;Crusciol;Nascente, 2012;Lins et al., 2017).
Vegetables absorb carbon dioxide from the atmospheric air together with light and water electrons, performing the photosynthesis, which, in short, is the reduction of this atmospheric CO 2 in organic carbon, giving origin to organic molecules for growth and development (Taiz;Zeiger, 2013).
Plants intercept the necessary light for the photosynthesis by the leaves, and the gaseous exchange takes place through the stomatal pores on the foliar surface. The process of opening and closing of the stomata mainly related to light intensity and the hydration state of the leaf. Propitious conditions to carbon fixation favor the opening of the stoma, while the propitious conditions to water loss promote its closing (Taiz;Zeiger, 2013).This mechanism limits the flow of CO 2 to the internal side of the leaf and may limit the CO 2 diffusion rate to the internal part of the leaf, with direct effects on photosynthesis and growth, while it restricts the water flow of the leaf to the atmosphere, decreasing transpiration and directly influencing plant productivity (Huang et al., 2017;Sun et al., 2014).
Maize is a C4 plant, characterized by a mechanism of CO 2 concentration that involves a coordinated functioning between the mesophile and the bundle sheath cells (Retta et al., 2016). It presents a low point of CO 2 compensation, a high photosynthetic rate and low water consumption for the formation of fresh matter (Hlatywayo et al., 2016;Huang et al., 2017). This functioning confers a higher photosynthetic net rate if compared to a C3 plant, since it presents low water loss, it can be cultivated in hot environments with high luminous intensity, and it adapts well to the tropical climate due to its low or null photorespiration (Wu et al., 2016).
Studies on gene effects and combining ability on physiological traits in forage maize are not very common. Ali et al. (2015) carried out a genetic analysis of forage maize genotypes for several physiological traits and concluded that the selection of genotypes with a higher photosynthetic rate, better efficiency of gaseous exchanges and water use can be determining in the increase of forage yield and productivity, which justifies more studies on this area.
The evaluation of physiological and chemical parameters can also be used in an attempt to explain differences of a genetic order (Peixoto et al., 2011). The current search for more efficiency in water use can be applied in the production of forage of good quality, since physiological factors influence and condition plant development and forage maturation (Guarda;Campos, 2014).
The aim of this work was to evaluate the general and specific combining ability and the nature of gene effects of physiological and chemical traits of forage maize genotypes.

mAterIAL AND metHoDs
Two groups of genotypes and 16 hybrids were evaluated, resulting from the crossings between the groups, in a 4x4 partial diallel scheme. Group I consisted of two commercial hybrids (AG8025, P30B39), and two inbred lines (LEM2 and LEM3), used as testers of the inbred lines of group II. Group II consisted offour experimental inbred lines originated from different populations of the breeding program of forage maize of the UNICENTRO. In total, 24 treatments were evaluated, formed by genotypes of both groups and the respective crossings.
The parcels were formed by 2 lines of 5 meters, with spacing of 0.80 m, resulting in a final density equivalent to 69.200 plants ha -1 .
In the first and second crop season, 250 kg ha -1 of the NPK 04-20-20 formulation was used in basis fertilization. The nitrogen fertilizations in top dressing were carried out when the plants reached between 3 leaves (V3 stage) and 5 leaves (V5 stage), with an application of 280 kg ha -1 of urea (46% of N) in each stage, totaling 257,6 kg ha -1 of N.
Three evaluations were carried out in the first crop season, at 42, 63 and 84 days after emergence (DAE), and two evaluations in the second crop season, at 21 and 63 DAE, due to the fact that the crop cycle is shorter. The 21 DAE corresponded to phenological stages V3-V4, when the definition of productive potential takes place; 42 DAE corresponded to V11-V12, when the number of grains and spike size are defined; 63 DAE corresponded to the VT stage, when flowering and pollination took place and; 84 DAE corresponded to the beginning of grain filling (Cañas et al., 2017).
Measurements of gaseous exchanges were carried out with the Infra Red Gas Analyser -IRGA, equipment, LI-6400, LI-COR model. The measurements were carried out between 9 and 11 o'clock on sunny days, in leaves fully expanded from the plants of each parcel. The CO 2 concentration and the light used during the evaluations were the ones existing in the environment. The values of CO 2 assimilation rate (A) (µmol CO 2 m -2 s -1 ) and of the transpiration rate (E) (mmol water vapor m -2 s -1 ) were obtained. The internal CO 2 concentration (Ci) (µmol CO 2 mol -1 air) and the stomatal conductance (g s ) (mol m -2 s -1 ) were calculated by the general equation of gaseous exchanges of Von Caemmerer and Farquhar (1981).
The efficiency of water use (A/E) (µmolCO 2 / mmol H 2 O -1 ) was determined through the ratio between the CO 2 assimilation rate and the transpiration rate, and the carboxylation efficiency or the calculated activity of Rubisco (A/Ci) (µmol m -2 s -1 Pa -1 ) was determined through the ratio between the CO 2 assimilation rate and the internal CO 2 concentration in the leaf, according to Pimentel (2011).
The forage was obtained with the plants in the reproductive stages from pasty to chalky grain, at the point of ¾ of the milkline, which corresponds to the silage point. Acid detergent fiber (ADF-%) and neutral detergent fiber (NDF-%) were evaluated according to Van Soest, Robertson and Lewis (1991). The digestibility in situ of the forage dry mass (DIG-%) was evaluated in accordance with Pereira et al. (2004).
After verifying the normality of errors and homogeneity of the variances by the Bartlett test, the individual and joint analyses of variance were carried out. The individual and joint partial diallel analyses were performed according to method 2 and model 1 proposed by Griffing (1956) and adapted for partial diallels by Geraldi and Miranda Filho (1988), aiming at estimating the general combining ability (GCA) of the genitors and the specific combining ability (SCA) from the pq hybrid combinations, where p inbred lines of group I were crossed with q tester genotypes of group II. The statistical software GENES was used (Cruz, 2013).

resuLts AND DIsCussIoN
The mean squares regarding the general combining ability (GCA) and specific combining ability (SCA) (Tables 1 and 2), when significant, indicate that the genitors of both groups differed from each other in the frequency of favorable alleles, and indicate that the traits evaluated can be used as parameters of selection aiming at the obtainment of hybrids with better physiological efficiency (Ali et al., 2014a;Peixoto et al., 2011;Tiwari et al., 2014).
In the several evaluations, for most of the traits there was the predominance of non-additive effects, since the quadratic components of the SCA were higher than those of the GCA (Tables 1 and 2), agreeing with the results by Ali et al. (2014b), which favors the high complementarity of parentals considering physiological efficiency. The SCA is expressed because of the effects of dominance, overdominance and epistasis and of the differences in the allelic frequencies of the genitors for the loci involved in the control of a certain trait (Hallauer;Carena;Miranda Filho, 2010;Wang et al., 2018). The predominance of genes of non-additive effects was also verified for forage traits, such as acid detergent fiber (ADF), neutral detergent fiber (NDF) and digestibility by Rosa et al. (2020), so that the selection based on genes of non-additive effects is efficient for both physiological traits and traits related to forage quality.
For the the CO 2 assimilation rate (A) there were significant effects of the genotypes and SCA, at 42 days after emergence in the first crop season (42 DAE-1). There were significant effects for all sources of variation in the evaluation carried out at 63 DAE-1 (Table 1). At 63 days after emergence, in the second crop season (63 DAE-2), there were significant effects for GCA-I and GCA-II (Table 2).
For stomatal conductance (gs), in the evaluation carried out at 42 DAE-1, there was no significant effect only for GCA of group II (Table 1). For internal CO 2 concentration (Ci), at 42 DAE-1, there was a significant effect of groups and of the GCA -I. In the evaluation carried out at 63 DAE-1 there was no significant effect only for the SCA and, in contrast, at 84 DAE-1 only SCA was significant (Table 1). At 21 DAE-2 there was a significant effect for the groups, GCA-I and SCA. At 63 DAE-2 there was a significant effect only for GCA-I (Table 2).
For transpiration (E), at 42 DAE-1, there was a significant effect of genotypes GCA-I and SCA. At 63 DAE-1 there was a significant effect of genotypes GCA-II and SCA. At 84 DAE-1 there was significance only for GCA-II (Table 1). In the evaluation, at 21 DAE-2, the effect was significant for genotypes GCA-I, GCA-II and SCA, and at 63 DAE-2 of the GCA-I (Table 2). For the (A/Ci) ratio there was a significant effect of GCA-II at 63 DAE-1 and of GCA-I at 84 DAE-1, (Table 1). At 21 DAE-2, there was a significant effect for genotypes, groups, GCA-I and GCA-II, and at 63 DAE-2, it was significant for groups and SCA (  For the ratio between CO 2 assimilation and transpiration (A/E), at 42 DAE-1, there was a significant effect of genotypes, groups, of GCA II and SCA, as well as of the SCA at 63 DAE-1. At 84 DAE-1 there was no significant effect only for GCA I (Table 1). There was a significant effect for GCA of both groups at 63 DAE-2 ( Table 2).
The predominance of additive effects in the control of (A) verified in the evaluations at 42 DAE-1 and at 63 DAE-2 (Table 1) implies that in the present work there are potentially favorable genitors to the increase of CO 2 assimilation (Coelho et al., 2014). It is worth pointing out that it is important to improve CO 2 assimilation, since about 90% of the plant dry mass originates from CO 2 fixation, through the photosynthesis process and the regulation of photosynthetic efficiency (Ali et al., 2014a;Magalhaes;Souza, 2009;Minow et al., 2018).  Estimates of general combining ability (ĝ i and ĝ j ) and specific combining ability (ŝ ij ) of CO 2 assimilation (A) (µmol CO 2 m -2 s -1 ), of 24 maize genotypes evaluated in a 4x4 partial diallel at 42 and 63 days after emergence in the first crop season (DAE-1) and at 63 days after emergence in the second crop season (DAE-2), in Guarapuava-PR. The highest estimates of the SCA of stomatal conductance (gs) at 42 DAE-1 occurred in crossings LEM2 x 202-160.1 (0.21 mol m -2 s -1 ) and LEM2 x 203-218.3 (0.19 mol m -2 s -1 ), but they originate from genitors with GCA estimates negative or close to zero (Table 4), with moderate means (Supplement 2), indicating that the performnce of these genitors was not different from the general mean of the other genitors involved in the crossings (Hallauer;Carena;Miranda Filho, 2010;Hlatywayo et al., 2016).
Genitor P30B39 had favorable estimates of GCA(Ci) at 21 DAE-2 (ĝ i = 9.18 µmol CO 2 mol -1 air) and at 63 DAE-2 CGC (ĝ i = 8,64 46 µmol CO 2 mol -1 air) and participated in crossing P30B39 x 202-160.1, which had high SCA estimates in the respective evaluations (ŝ ij = 16.55 µmol CO 2 mol -1 air and ŝ ij = 10.78 µmol CO 2 mol -1 air), demonstrating that tester P30B39 allowed to discriminate the most favorable inbred line to improve Ci. Also, at 63 DAE-2, the tester P30B39 participated in the crossing of the highest estimate of SCA (P30B39 x 203-218.3; ŝ ij = 23.46 µmol CO 2 mol -1 air) (Table 5). Ali et al. (2014b) indicated that the selection based on internal CO 2 concentration can be useful in the development of maize hybrids with forage purpose, since the authors verified a direct relationship between the Ci and the accumulation of dry mass.
At 21 DAE-2, the inbred lines LEM 3 (ĝ i = 0.31 mmol water vapor m -2 s -1 ) and 203-218.3 (ĝ i = 0.39 mmol water vapor m -2 s -1 ) had the highest estimates of GCA(E) ( Table 6). The evaluation of transpiration rate of young plants can be effective in the selection of maize genotypes aiming at better photosynthetic efficiency, as higher stomatal conductance causes the increase of E, which increases internal CO 2 concentration and, consequently, causes an increase of the photosynthetic rate, directly reflecting on forage quality (Ali et al., 2013).  Estimates of general combining ability (ĝ i and ĝ j ) and specific combining ability (ŝ ij ) of internal CO 2 concentration (Ci) (µmol CO 2 mol -1 air), of 24 maize genotypes evaluated in a 4x4 partial diallel at 42, 63 and 84 days after emergence in the first season (DAE-1) and at 21 and 63 days after emergence in the second crop season (DAE-2), in Guarapuava-PR.
The A/E ratio reveals the efficiency of water use and reflects the plant's ability to limit water loss and, at the same time, allows CO 2 absorption (Taiz;Zeiger, 2013).
The high efficiency in water use is directly related to the stomatal opening time, since while the plant absorbs CO 2 for the photosynthesis, water is lost by transpiration, with varied intensity, depending on the potential gradient between the foliar surface and the atmosphere, following a chain of water potentials (Muraya et al., 2017;Silva et al., 2015). Thus, it is assumed that more efficient genotypes in water use tend to produce higher amounts of dry mass per gram of transpired water, directly affecting forage performance (Ali et al., 2013).
Given the above, hybrid LEM2 x 101-7.2 presented the highest rate of CO 2 assimilation, in addition to desirable values for neutral detergent fiber (NDF) and acid detergent fiber (ADF), in the evaluation of both crop seasons, as well as good digestibility (DIG) (Figure 1). The contents of fiber and digestibility directly reflect on forage quality and it is essential to study them to obtain promising genotypes, justifying the study of both physiological and chemical components, and their effects on forage quality (Ali et al., 2013).
The inbred lines which most contributed to improve DIG, considering the mean values, were the tester LEM2 (group I) and the experimental inbred line 201-107.2 (group II) (Figure 1). Consequently, they presented high CO 2 assimilation and high efficiency of water use, which reasserts the relationship of physiological traits with qualitative bromatological traits of the forage, (Figure 1), justifying this study.

CoNCLusIoNs
There was a predominance of non-additive gene effects for most of the chemical and physiological traits. Elite inbred lines LEM2 and LEM3 were the best testers to discriminate the relative merit of the experimental inbred lines. Crossings LEM2x203-218.3, LEM3 x 201-107.2, LEM2 x 101-7.2 and LEM3 x 101-7.2 stood out regarding CO 2 assimilation, and are indicated for future research considering the physiological traits. Inbred lines 101-7.2 and 203-218.3 presented the high concentration of favorable alleles to increase carboxylation efficiency, in which inbred line 101-7.2 stood out for NDF, ADF and DIG. Inbred line 201-107.2 has a high concentration of favorable alleles for efficiency of water use. Physiological parameters can assist the selection of inbred lines and hybrids in maize breeding for forage purpose.