Potential to select wheat genotypes with improved P utilisation characters

Abstract

Developing new wheat genotypes resistant to phosphate (P) deficiency and responsive to fertilizer P input will be beneficial to increase P use crop efficiency, especially on P deficient calcareous soils. With this objective, a pot experiment, based on a completely randomised design with three replications, was conducted using a P deficient calcareous soil. Twenty different bread and durum wheat genotypes were used, and phosphorus fertilizer as H₃PO₃ at levels of 0, 40, 80 mg P kg⁻¹ was applied to the pots. After harvest, plant dry matter yield was recorded, and total P concentration in the tops of wheat plants was determined. Changes in the parameter of P efficiency index (EI) were related to plant genotypes and P levels. Differences were observed among the genotypes with regard to their effectiveness in P use under the experimental calcareous soil. Dry matter yields and EI were correlated for the individual wheat genotypes. Four classification terms, ER (efficient-responsive), ENR (efficient non-responsive), IR (inefficient responsive) and INR (inefficient non-responsive) were used to characterize the wheat genotypes both for their resistance to P deficiency and also for their responsiveness to fertilizer P input. Most durum wheat genotypes were charecterized as ENR, whereas most bread wheat genotypes were charecterized as INR and IR for this study. We conclude that Dağdaş-94 (T. aestivum), characterized as ER, and Kızıltan-91 (T. durum), Yılmaz-98 (T. durum), Ankara-98 (T. durum), Selçuklu-97 (T. durum), BDMM-98/11S (T. durum), characterized as ENR, are valuable in future breeding programmes.

Keywords:

• Mediterranean agriculture
• P deficiency
• P fertilization
• Phosphate P-use
• P fertilizer P uptake variability
• wheat breeding
• wheat

Introduction

In Mediterranean and West Asian regions, available P deficiency is one of the main nutritional problems in agricultural production (Cooper et al., 1987; Matar et al., 1992). In these regions, many soil conditions such as soil moisture, texture, and calcium carbonate content restrict P availability and uptake by crops. The utilisation of P fertilizers by crops is generally low (10–20%) in the year of application, and fixation of phosphorus significantly affects P accumulation in the soil (Gahoonia & Nielsen, 1999). Much of the applied P remains in soils as a residual portion. Hence there may be considerable variability in the available P content that occurs in field soils (Tisdale et al., 1993).

Phosphorus is one of the essential elements for plant growth, and it is needed for many physiological and biochemical processes in plants. Some studies have indicated that cereal varieties absorb more P than many others (Gahoonia & Nielsen, 1996). Selection of P efficient cultivars has great importance for the efficient use of fertilizer phosphorus in order to obtain the highest yield with reduced P fertilizer. However, P use efficiency and the resistance of plants to P deficiency are affected by many factors. For example, P absorption by plants is an active process, and thus metabolic activity is important (Marschner, 1995). Physiological processes and root growth affect P absorption by plants. Gahoonia & Nielsen (1999) revealed that root hair characteristics could be one of the selection criteria owing to the wide variations in root hairs of cereal cultivars and their substantial role in P uptake. In other studies, phosphorus efficient genotypes have also been selected and developed (Graham, 1984; Çakmak et al., 1997). Many different adaptation mechanisms exist for low P conditions, and thus the development of P efficient varieties is a high priority in regions where soils have P deficiency. This study aimed to characterize P efficient and responsive genotypes using a practical classification method for optimal yield with reduced P input under the experimental P deficient calcareous soil conditions.

Materials and methods

A pot experiment, based on a completely randomised design with three replications, was conducted in 2002 using the available P deficient soil, calcareous usthochrepts. The air dried soils were screened to pass through a 2 mm mesh. Each pot consisted of 4 kg of dry soil. Bread and durum wheat genotypes used for this study, twelve plants per pot, were obtained from The International Winter Wheat Research Centre of Bahri Dağdaş (BD-MIKHAM), Konya-Turkey (Table 1).

Table 1. Bread and durum wheat genotypes with respective seed P concentrations

Number	Genotypes	Definition	Seed P concentration g kg^−1
1	Ankara-98	T. durum	2.70
2	BDME-98/3K	T. aestivum	2.70
3	BDME-98/4S	T. aestivum	2.60
4	BDME-98/5S	T. aestivum	2.50
5	BDMM-98/11S	T. durum	2.60
6	BDME-98/33S-CIT	T. aestivum	2.50
7	BDME-00/1K	T. aestivum	2.35
8	BDME-00/2S	T. aestivum	2.35
9	BDME-00/3S	T. aestivum	2.60
10	BDME-00/4S	T. aestivum	2.90
11	Bezostaya-1	T. aestivum	2.95
12	Dağdaş-94	T. aestivum	2.90
13	Gerek-79	T. aestivum	2.65
14	Gün-91	T. aestivum	2.60
15	Kınacı-97	T. aestivum	3.10
16	Kızıltan-91	T. durum	3.65
17	Kunduru-1149	T. durum	3.40
18	Selçuklu-97	T. durum	3.30
19	Sultan-95	T. aestivum	2.70
20	Yılmaz-98	T. durum	2.80

Phosphorus fertilizer as ortho-phosphoric acid (H₃PO₃) at levels of 0, 40, 80 mg P kg⁻¹ was applied to the pots. In addition, a basal dressing of some macro- and micro-nutrients was applied to all pots for normal plant growth. The plants were harvested after 49 days, and dry weights of the tops were recorded. The analysis of P concentration in the tops of the plants was made by spectrophotometry after digestion (Barton, 1948). In the experimental soil, available P analysis was determined by the method of Olsen et al. (1954). Determinations were also made of the textural analysis with a Bouyoucos hydrometer (Gee & Bouder, 1986), organic matter content with the Walkley-Black method (Jackson, 1956), exchangeable potassium and C.E.C.(cation exchange capacity) (Richards, 1954), CaCO₃ (Chapman & Pratt, 1961) and pH values (McLean, 1986). Experimental data were subjected to statistical analysis of variance using MSTAT package programme, and the means were separated by Duncan's multiple range test. Linear and polynomial regression (orders of polynomial: 4 for P-0, 3 for P-40 and 1 for P-80) analysis of some relationships were also calculated using the computer programme StatMost (StatMost, 1995). Dry weight (DM) and total P content of the plants were used to calculate the efficiency index parameter (dm²/total P content) for classification of genotypes (adapted from Siddiqi & Glass, 1981 and Furlani et al., 2001). This classification method also served for the characterization of genotypes as ER (efficient-responsive), ENR (efficient non-responsive), IR (inefficient responsive) and INR (inefficient non-responsive) (Furlani et al., 2001).

The calcareous soil used in this study was clay-loam in texture with 30, 32 and 38% clay, silt and sand, respectively, and the calcium carbonate content was 169 g kg⁻¹. It had the chemical properties of pH (soil:H₂O=1:2.5) of 7.9, and available phosphorus 3.0 mg P kg⁻¹. It also had the following chemical properties: organic matter content 1.2%, cation exchange capacity 36.9 me 100 g⁻¹, exchangeable potassium 1.1 me 100 g⁻¹, DTPA extractable Fe 2.1 μg g⁻¹, Zn 0.11 μg g⁻¹, Cu 1.0 μg g⁻¹, and Mn 3.6 μg g⁻¹.

Results

Agronomic efficiency of P in wheat genotypes

The analysis of variance showed highly significant F values for dry matter production depending on P levels and genotypes. By increasing the P fertilizer level, dry weights were significantly increased as an average of wheat genotypes (Table 2). Significant differences among wheat genotypes were also found for dry matter. The highest dry matter yield was obtained in BDME-98/3K (T. aestivum), whereas the lowest dry matter was obtained in Bezostaya-1 (T. aestivum). The interaction of G×P was statistically significant, meaning that wheat genotypes responded differently to P treatments. Most wheat genotypes were very highly responsive, whereas Yılmaz-98 (T. durum), Ankara-98 (T. durum), Selçuklu-97 (T. durum) and BDMM-98/11S (T. durum) seemed to be non-responsive to P fertilization at the P-40 level (Table 2). By increasing P fertilizer input, Bezostaya-1 (T. aestivum) slightly responded (no more than 10%), whereas BDME-98/3K (T. aestivum), BDME-00/4S (T. aestivum) and Kunduru 1149 (T. durum) did not respond to the higher level of P-80. In general, agronomic P efficiency decreased with increasing P level. Agronomic P efficiency ranged from 9.02 to 103.47% at the P-40 level, whereas it ranged from 42.71 to 132.25% at the P-80 level. Significant differences among wheat genotypes for agronomic P efficiency were also observed. The lowest agronomic P efficiency was found in BDME-98/3K (T. aestivum), whereas Selçuklu-97 (T. durum) and BDMM-98/11S (T. durum) had the highest agronomic P efficiency at the P-40 level. At the P-80 level, bread wheat Gerek-79 variety had the lowest agronomic P efficiency, whereas durum wheat Selçuklu-97 variety had the highest. In general, durum wheat genotypes had a higher agronomic P efficiency than that of bread wheat genotypes at both P levels.

Table 2. Dry matter yield and agronomic P efficiency parameter of bread and durum wheat genotypes at different P levels

	Dry matter yield, g pot^−1				Agronomic P efficiency, % ^a
Genotypes	P-0	P-40	P-80	Av.	P0/P40	P40/P80
Ankara-98	4.22 ej	4.36 ej	4.26 ej	4.28 BD	96.79	102.35
BDME-98/3K	1.15 n	12.75 a	9.93 b	7.94 A	9.02	128.40
BDME-98/4S	1.22 mn	5.86 cg	8.13 bc	5.07 B	20.82	72.08
BDME-98/5S	1.16 n	5.15 di	6.03 cf	4.11 BD	22.52	85.41
BDMM-98/11S	4.47 ej	4.32 ej	4.58 ej	4.45 BD	103.47	94.32
BDME-98/33S-CIT	1.29 mn	4.92 dj	8.10 bc	4.77 BC	26.22	60.74
BDME-00/1K	1.22 mn	6.13 cf	7.50 cd	4.95 BC	19.90	81.73
BDME-00/2S	1.08 n	4.49 ej	6.11 cf	3.89 BE	24.05	73.49
BDME-00/3S	1.20 mn	3.95 el	6.20 cf	3.78 BE	30.38	63.71
BDME-00/4S	1.45 ln	6.21 cf	6.07 cf	4.58 BC	23.35	102.31
Bezostaya-1	1.00 n	2.52 jn	2.71 in	2.08 F	39.68	92.99
Dağdaş-94	1.61 kn	5.16 di	6.33 cf	4.37 BD	31.20	81.52
Gerek-79	1.05 n	2.46 jn	5.76 ch	3.09 DF	42.68	42.71
Gün-91	1.53 ln	4.97 dj	6.26 cf	4.25 BD	30.78	79.39
Kınacı-97	1.07 n	5.46 dh	6.44 ce	4.32 BD	19.60	84.78
Kızıltan-91	1.23 mn	4.03 ek	5.52 dh	3.59 CE	30.52	73.01
Kunduru-1149	1.16 n	3.73 fm	3.20 hn	2.70 EF	31.01	116.56
Selçuklu-97	4.61 ej	4.47 ej	3.38 gn	4.15 BD	103.13	132.25
Sultan-95	1.14 n	3.73 fm	6.57 ce	3.81 BE	30.56	56.77
Yılmaz-98	4.47 ej	4.44 ej	4.22 ej	4.38 BD	100.68	105.21
Average	1.87 C	4.96 B	5.87 A
F values
Genotypes (G): 11.62**
P treatments (P): 261.40**
G×P: 8.14**
^a Agronomic P efficiency=% value related to the response of a genotype to supplied P level. In P efficient genotype, %P efficiency value is higher, which means that the genotype has lower response or non-response to the supplied P levels.

Physiological efficiency of P in wheat genotypes

The analysis of variance also showed highly significant F values for P concentration and P content depending on P levels and genotypes. In general, total P concentration in the tops of wheat plants varied with P treatments (Table 3). The P concentration of wheat genotypes was significantly increased with increasing P level, but significant differences were apparent. The highest P concentration was detected in BDMM-98/11S (T. durum). The durum variety Kızıltan-91 had the lowest P concentration. The interaction of G x P for P concentration was also statistically significant. By increasing P fertilizer input to the P-40 level, P concentration was increased in most wheat genotypes, whereas there was only a slight increase for P concentration in Gün-91 (T. aestivum) and Selçuklu-97 (T. durum) genotypes. In general, P concentrations in the wheat genotypes were increased at the highest level of P-80. The highest P concentration were detected in BDME-00/4S (T. aestivum) and BDME-98/4S (T. aestivum) at the P-80 level.

Table 3. Phosphorus concentration of bread and durum wheat genotypes at different P levels

	P concentration, g kg^−1
Genotypes	P-0	P-40	P-80	Av.
Ankara-98	1.33 nu	1.75 ip	2.50 ae	1.86 AC
BDME-98/3K	1.10 rv	1.92 fm	2.53 ae	1.85 AD
BDME-98/4S	1.15 rv	1.15 rv	2.98 a	1.76 BF
BDME-98/5S	1.17 qv	1.47 mt	2.68 ac	1.77 BF
BDMM-98/11S	1.62 kr	2.37 bf	2.25 bi	2.08 A
BDME-98/33S-CIT	1.10 rv	1.88 fm	2.25 bi	1.74 BF
BDME-00/1K	0.78 uv	1.73 ip	2.68 ac	1.73 BF
BDME-00/2S	0. 97 tv	1.73 ip	2.33 bg	1.68 BH
BDME-00/3S	0. 97 tv	1.82 gn	2.60 ad	1.79 AE
BDME-00/4S	0. 97 tv	2.03 el	2.95 a	1.98 AB
Bezostaya-1	1.17 qv	1.70 jq	2.48 ae	1.78 BF
Dağdaş-94	0.97 tv	1.78 ho	2.32 bh	1.69 BG
Gerek-79	1.15 rv	1.03 sv	2.10 dk	1.43 GH
Gün-91	1.15 rv	1.25 ou	2.23 bj	1.54 DH
Kınacı-97	0.83 uv	1.32 nu	2.72 ab	1.62 CH
Kızıltan-91	0.67 v	1.45 mt	2.02 el	1.38 H
Kunduru-1149	1.02 sv	1.25 ou	2.17 cj	1.48 FH
Selçuklu-97	1.75 ip	1.77 ip	1.23 pu	1.58 CH
Sultan-95	1.05 sv	1.25 ou	2.33 bg	1.54 DH
Yılmaz-98	1.13 rv	1.55 ls	1.83 fn	1.51 EH
Average	1.10 C	1.61 B	2.36 A
F values
Genotypes (G): 6.57**
P treatments (P): 527.94**
G×P: 7.73**

Total P content was significantly increased with P fertilizer, which can be attributed to the increased dry matter yield (Table 4). Total amounts of P taken up by plants followed a similar pattern. Significant differences among wheat genotypes were found for total P content. The highest P content was found in BDME-98/3K (T. aestivum), whereas the lowest was found in the bread variety Bezostaya-1. The interaction of G×P was statistically significant for total P content. The highest total P content was found in BDME-98/3K (T. aestivum) at the P-40 and P-80 levels, but in BDME-98/4S (T. aestivum) at the P-80 level. Efficiency index (physiological P efficiency) also varied among the genotypes depending on their dry matter yield and total P content. The highest average efficiency index was obtained for BDME-98/3K (T. aestivum), and Bezostaya-1 (T.aestivum) had the lowest efficiency index. Efficiency index (EI) for each P level is also presented in Fig. 1.

Fig. 1. Polynomial relationships between EI (Efficiency index=dry matter yield²/total P) and dry matter yield of bread and durum wheat genotypes for P-0 (a), P-40 (b) and P-80 (c) levels.

Table 4. Total P content and efficiency index of bread and durum wheat genotypes for different P levels

	Total P content, mg pot^−1				Average efficiency index, EI^a
Genotypes	P-0	P-40	P-80	Av.
Ankara-98	5.66 kt	7.50 io	10.68 fl	7.95 BE	2.46
BDME-98/3K	1.28 rt	24.35 a	25.20 a	16.94 A	3.87
BDME-98/4S	1.41 rt	6.75 js	24.15 a	10.77 B	2.96
BDME-98/5S	1.35 rt	7.43 io	15.67 bg	8.15 BE	2.29
BDMM-98/11S	7.22 ip	10.24 gm	10.24 gm	9.23 BC	2.21
BDME-98/33S-CIT	1.43 rt	9.28 hn	18.22 bc	9.64 BC	2.46
BDME-00/1K	0.95 t	10.72 fl	20.19 ab	10.62 B	2.62
BDME-00/2S	1.04 st	7.77 io	14.14 ch	7.65 BE	2.12
BDME-00/3S	1.16 rt	7.37 ip	16.10 bf	8.21 BE	1.92
BDME-00/4S	1.40 rt	12.72 di	17.78 bd	10.63 B	2.20
Bezostaya-1	1.17 rt	4.29 nt	6.78 js	4.08 G	1.14
Dağdaş-94	1.57 qt	9.36 hn	14.69 ch	8.54 BD	2.41
Gerek-79	1.20 rt	2.48 ot	12.07 ej	5.25 EG	2.04
Gün-91	1.76 pt	6.21 kt	13.85 ch	7.27 CF	2.71
Kınacı-97	0.89 t	7.17 iq	17.38 be	8.48 BD	2.61
Kızıltan-91	0.80 t	5.87 kt	11.14 fk	5.94 DG	2.47
Kunduru-1149	1.17 rt	4.72 mt	6.86 jr	4.25 FG	1.86
Selçuklu-97	8.03 io	7.91io	4.42 nt	6.78 CG	2.59
Sultan-95	1.19 rt	4.86 mt	15.30 bg	7.12 CG	2.26
Yılmaz-98	5.07 lt	6.88 jr	7.73 io	6.56 CG	2.80
Average	2.29 C	8.20 B	14.13 A
F values
Genotypes (G): 14.79**
P treatments (P): 433.14**
G×P: 9.96**
^a Efficiency index (EI)=dry matter yield^2/total P content, and serves to select wheat genotypes with improved P utilization characters as ER (efficient-responsive), ENR (efficient non-responsive), IR (inefficient responsive) and INR (inefficient non-responsive).

Classification and characterization of wheat genotypes for P use efficiency

The linear regression analysis, conducted between dry matter yield and EI, for average P levels, had a significant degree of association (R ²=0.82, P<0.01), and the regression equation was DM=0.1734+0.0024 * EI. Significant polynomial relationships were also found between dry matter and EI for the three P-levels and all genotypes (Fig. 1). The polynomial regression analyses showed that there were significant relationships for the variants. However, polynomial regression analysis for each P level indicated that the polynomial relationship between DM and EI varied with supplied P level, and the highest correlation coefficient for the genotypes was observed at the P-0 level (Fig. 1). This result also showed that EI was a more characeristic parameter under P stressed conditions than that of P supplied conditions to characterize P use efficiency. Classification of bread and durum wheat genotypes according to the values of EI (at the P-0 level) and DM were made to characterize these genotypes as ER (efficient-responsive), ENR (efficient non-responsive), IR (inefficient responsive) and INR (inefficient non-responsive). As a result of this classsification, Dağdaş-94 (T. aestivum) was characterized as ER; Kızıltan-91 (T. durum), Yılmaz-98 (T. durum), Ankara-98 (T. durum), Selçuklu-97 (T. durum), BDMM-98/11S (T. durum) were characterized as ENR; Gün-91 (T. aestivum), Sultan-95 (T. aestivum), BDME-98/4S (T. aestivum), BDME-98/33S-CIT (T. aestivum), BDME-00/1K (T. aestivum), BDME-00/2S (T. aestivum), BDME-00/3S (T. aestivum), BDME-00/4S (T. aestivum) were characterized as IR, and Bezostaya-1 (T. aestivum), Gerek-79 (T. aestivum), BDME-98/5S (T. aestivum), Kunduru-1149 (T. durum) were characterized as INR (Fig. 2).

Fig. 2. Classification of bread and durum wheat genotypes according to the EI (efficiency index at the P-0 level) and the maximum dry matter (at the 40 or 80 mg kg⁻¹ P levels). ER; efficient-responsive, ENR: efficient non-responsive, IR: inefficient responsive, INR: inefficient non-responsive.

Discussion

We have shown that a variability within wheat genotypes in P use efficiency exists under the P deficient and non-deficient conditions. The performance of a specific wheat variety for P use efficiency was not similar at different P levels. The causes of differences in P uptake and P use efficiency may be due to genetic differences among the genotypes. Agronomic P efficiency ratios of durum wheat were higher than those of bread wheat genotypes, which means that durum wheats seemed to be more P efficient than bread wheats under P stressed conditions. However, it has been found that durum and wheat genotypes showed intra- and inter-specific differences in agronomic and physiological P efficiency, and agronomic P efficiency of bread wheat was higher than that of durum wheat genotypes. The extent of P deficiency of wheat species varies, depending on wheat genotypes (Graham, 1984; Gahoonia & Nielsen, 1999).

Selection of P efficient cultivars has great importance in the efficient use of fertilizer phosphorus to obtain the maximum yield with reduced P fertilizer inputs, and we suggest that classification and characterization of genotypes with respect to P use should be included in future breeding programmes. Furlani et al. (2001) classified 29 soybean cultivars in relation to response to P levels. In our study, the classification method characterized the genotypes as responsive or non-responsive. This classification method is a basic way of characterizing the varied amount of genotypes for both P efficiency and P use responsiveness under the varied soil conditions (Furlani et al., 2001). Hence, the four classification terms, ER, ENR, IR and INR were confidently used for the different genotypes to characterize both the performance of their resistance to P deficiency and their P use efficiency under P stressed and P supplied conditions. As a result of this classification, bread wheat Dagdas-94 is characterized as ER and durum wheat Kızıltan-91, Yılmaz-98, Ankara-98, Selçuklu-97 and BDMM-98/11S are characterized as ENR, and this appears to be valuable for breeding studies under the experimental calcareous soil. Most durum wheat genotypes apart from Kunduru-1149 were characterized as ENR, whereas most bread wheat genotypes were characterized as INR and IR. In general, the best genotypes showed higher dry matter yield and P efficiency index. In this classification method, variability of P efficient and inefficient wheat genotypes in relation to responsiveness of supplied P level was an important parameter in breeding studies aimed at increasing fertilizer P use efficiency, and decreasing fertilizer P inputs under the even soil conditions. Characterization of P efficient and responsive genotypes will provide valuable genetic resource for sustaining the optimal yields and quality with reduced P supply. Our results showed that adopting this classification method for newly developed genotypes under varied soil conditions is beneficial for rapid testing for P efficiency of these cultivars. Thus economic and environmental benefits will be obtained by including characterization of P deficiency resistance and P use efficiency in future breeding programmes of wheat genotypes by our practical classification method.

Additional information

Notes on contributors

M. Rüştü Karaman*

Karaman, M. R. and Sahin, S. (Department of Soil Science, Agricultural Faculty, Gaziosmanpaşa University, TR-60250 Tokat, Turkey). Potential to select wheat genotypes with improved P utilisation characters.

Notes

Karaman, M. R. and Sahin, S. (Department of Soil Science, Agricultural Faculty, Gaziosmanpaşa University, TR-60250 Tokat, Turkey). Potential to select wheat genotypes with improved P utilisation characters.