Experimental Agriculture (2022), vol. 58, e23, 1–13
doi:10.1017/S0014479722000151
RESEARCH ARTICLE
Balanced fertilization increases wheat yield response on
different soils and agroecological zones in Ethiopia
Sofonyas Dargie1, Tsegaye Girma2,3,* , Tilahun Chibsa4, Sofia Kassa5, Shiferaw Boke2,
Abate Abera6, Bereket Haileselassie1, Samuel Addisie7, Sosina Amsalu8, Mehretab Haileselassie1,
Shure Soboka9, Wuletawu Abera10 and Sileshi G. Weldesemayat11,12
1Tigray Agricultural Research Institute, Mekelle Soil Research Centre, Mekelle, Ethiopia, 2Southern Agricultural Research
Institute, Hawassa, PO. Box 06, Ethiopia, 3Centre for Dryland Agriculture (CDA), Bayero University, Kano, Nigeria,
4Oromia Agricultural Research Institute, Sinana Agricultural Research Centre, Sinana, Ethiopia, 5Ethiopian Institute of
Agricultural Research, Addis Ababa, PO Box 2003, Ethiopia, 6Southern Agricultural Research Institute, Areka
Agricultural Research Center, Areka, Ethiopia, 7Amhara Regional Agricultural Research Institute, Srinka Agricultural
Research Center, Woldia, Ethiopia, 8Ethiopian Institute of Agricultural Research Institute, Debre Zeit Agricultural
Research Center, Debre Zeit, PO Box 32, Ethiopia, 9Oromia Agricultural Research Institute, Addis Ababa, Ethiopia,
10International Center for Tropical Agriculture (CIAT), Alliance Bioversity, Addis Ababa, Box 1569, Ethiopia,
11Department of Plant Biology and Biodiversity Management, Addis Ababa University, Addis Ababa, Ethiopia and
12School of Agricultural, Earth and Environmental Sciences, University of KwaZulu-Natal, Pietermaritzburg, South Africa
*Corresponding author: Email: elositseg28@gmail.com
(Received 29 September 2021; revised 07 March 2022; accepted 14 March 2022)
Abstract
The response of wheat to the application of different rates of nitrogen (N), phosphorus (P), potassium (K),
and sulfur (S) under balanced fertilization on different soil types and agroecologies has not been well
studied in Ethiopia. Therefore, the objectives of this study were to (1) determine soil-specific responses
of wheat to N, P, K, and S under balanced fertilization; (2) quantify agroecology-specific N, P, K, and
S response of wheat under balanced fertilization; and (3) determine nutrient use efficiency of wheat on
different soil types under balanced fertilization. Trials were conducted on farmers’ fields across 24 locations
covering 4 soil types and 5 agroecological zones (AEZs) from 2013 to 2017. The mean grain yields of wheat
significantly varied with applied N and P fertilizer rates with soil types and AEZs. With balanced applica-
tion of other nutrients, the optimum N rates for wheat were 138 kg N ha−1 on Cambisols and Luvisols,
92 kg N ha−1 on Vertisols, and 176 kg N ha−1 on Nitisols, while the optimum P rate was 20 kg P ha−1 on
Cambisols and Vertisols. The nutrient dose–response curve did not reveal consistent pattern for
K and S applications on all soil types. The agronomic efficiency of wheat decreased with increasing rates
N and P on all soil types. The highest agronomic efficiency of N (15.8 kg grain kg−1 applied N) was
recorded with application of 92 kg N ha−1 on Vertisols, while the highest agronomic efficiency of P
(49 kg grain kg−1 applied P) was achieved with application of 10 kg P ha−1 on Cambisols. We conclude
that applications of 92–138 kg N ha−1, 20 kg P ha−1, 18 kg K ha−1, and 10 kg S ha−1 under balanced
application of zinc and boron could be recommended depending on soil type for wheat production
in the study areas.
Keywords: Agronomic efficiency; Balanced fertilization; Cambisols; Luvisols; Nitisols; Vertisols
Introduction
Wheat (Triticum aestivum L.) is one of the topmost staple cereals in sub-Saharan Africa (SSA)
with over 10 million ha of land under production. In terms of per capita calories consumed, food
© The Author(s), 2022. Published by Cambridge University Press. This is an Open Access article, distributed under the terms of the Creative
Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduc-
tion, provided the original article is properly cited.
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
2 Sofonyas Dargie et al.
supply, and value of imports in Africa, wheat is ranked number one among the crops (Sileshi and
Gebeyehu, 2021). Its demand has considerably increased in the past 20 years as a result of growing
population, changing food preferences, and socioeconomic change associated with urbanization
(Macauley and Ramadjita, 2015). Since domestic wheat production in SSA is unable to meet the
demand, about 41 million metric tons (t) of wheat valued at US$ 12 billion are imported annually
(Sileshi and Gebeyehu, 2021). Annual wheat imports account for 25.4% of wheat imports on the
global market (Sileshi and Gebeyehu, 2021), thus making this region the world’s biggest wheat
importer (Mason et al., 2012). Ethiopia is the second largest wheat producer in Africa following
South Africa (FAO 2016). With Ethiopia’s rising population and urbanization, the demand for
wheat has surpassed the national supply making the current production insufficient to meet
domestic needs. This has forced the country to import up to 50% of the domestic demand to fill
the gap (Minot, 2014). Wheat is mainly cultivated as mono-crops and usually involved in crop
rotations (tef-wheat-food legumes) in Ethiopia. It is the major cereal crop grown in the major
AEZs of Ethiopia, with an estimated area of 1.7 million ha of land and production of 4.64 million
tons per year (CSA, 2018). However, the national average yield is 2.74 t ha−1 (CSA, 2018), which is
lower than the experimental yield of over 5 t ha−1 and the world average yield of 4 t ha−1
(FAO, 2016).
The low wheat yield is mainly associated with the depletion of soil fertility and soil acidity
(Agegnehu et al., 2019; Regassa and Agegnehu, 2011), continuous nutrient offtake by crops,
mono-cropping in the major wheat-growing areas, lack of supply of improved seed varieties,
improper use of fertilizer and low fertilizer use efficiency (Tarekegne and Tanner 2001; Yirga
et al., 2002), and occurrence of disease and insect pests (Oerke, 2006; Sileshi and Gebeyehu,
2021). To meet the growing demand, either the area under wheat production and/or productivity
per unit area should be markedly increased (Sharma et al., 2015). Varietal development and adop-
tion of improved agricultural technologies including soil fertility management are among the key
interventions to improve the productivity of wheat.
Current fertilizer recommendations in Ethiopia are based on very general and a blanket rec-
ommendation of 64 N-20 P-0 K kg ha−1 in the form urea and Diammonium Phosphate (DAP).
This blanket recommendation often fails to take into consideration differences in agroecology
and soil type, which are key determinants of nutrient use efficiency and productivity (Sileshi
et al., 2022). It also does not make allowances for dramatic changes in input/output price ratio,
thereby discouraging farmers from fertilizer application. Moreover, the nutrients in the blanket
recommendation are not well balanced, and their continued use will gradually exhaust soil
nutrient reserves. Therefore, neither yields nor profits can be sustained by using unbalanced
fertilizer applications, as the practice results in accelerated deficiencies of other essential
nutrients. Absence of one or more nutrients or imbalances can significantly depress yield.
Deficiencies of N and P, S, B, and Zn are widespread in Ethiopian soils, while some soils
are also deficient in K, Cu, Mn, and Fe (Habte and Boke, 2017). This could explain, in part,
the modest crop yield improvements observed over the decades despite significant increases
in fertilizer use in the country (Zeleke et al., 2010).
There is an urgent need to develop crop-specific nutrient recommendations that are rationally
differentiated according to AEZs, soil type, nutrient uptake, and socioeconomic circumstances
of farmers. Better matching of fertilizers and balanced application of nutrients at rates suitable
to the local climate and soil type can increase the productivity of wheat and can optimize
nutrient use efficiency and productivity (Sileshi et al., 2022). However, wheat yield response
and agronomic efficiency of different rates of N, P, K, and S fertilizers under balanced fertili-
zation have not been studied on different soil types and AEZs in Ethiopia. Therefore, the
objectives of this study were to (1) determine soil-specific responses of wheat to N, P, K,
and S under balanced fertilization, (2) quantify agroecology-specific N, P, K, and S response
of wheat under balanced fertilization, and (3) to determine nutrient use efficiency of wheat on
different soil types under balanced fertilization.
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
Experimental Agriculture 3
Table 1. Characteristics of the study sites in terms of agroecological zones (AEZs) and selected soil chemical characteristics
Tepid sub-moist Cool sub-moist Tepid moist Tepid sub-humid Tepid humid
mid-highlands mid-highlands mid-highlands mid-highlands mid-highlands
Variables (SM3) (SM4) (M3) (SH3) (H3)
Regions Tigray and Amhara Amhara, Oromia and Oromia
Oromia Oromia, and Benishangul
Southern Gumuz
Elevation (m.a.s.l) 1000–2000 1400–2200 1000–2000 2000–2800 2000–3000
Rain fall (mm) 900–1000 900–1000 900–1000 >1000 >1000
Temperature (0C) 16–21 11–15 16–21 16–21 16–21
Variable Cambisols Luvisols Nitisols Vertisols
pH H2O (1:2.5) 6.8 5.6 6.4 6.9
N (%) 0.18 0.26 0.29 0.16
P (mg kg−1) 15.97 4.69 13.25 11.61
K (mg kg−1) 5.81 4.52 5.66 4.43
SOC (%) 2.73 3.07 3.86 1.88
Materials and Methods
Site description
Field trials were conducted over three cropping seasons from 2013 to 2017 on 24 sites distributed
across 4 regions in Ethiopia. The experimental sites were Habru Seftu, Dambel, Gedeb Assasa, Ilu
Sambtu, Selka Jafara, Selka Mazoria, and Selka Odda in Oromia region; Adigolo, Ayba, Embahsti,
Freweyni, and Mesanu in Tigray region; Debre Guracha, Faji, and Meja in Amhara region, and
Alarigeta, Angecha, Abagada, Asheba, Bobicho, Boka, Hakmura, Mera, and Shomora in the
Southern Nations, Nationalities, and Peoples (SNNP) region. The sites are located within a
range of 1000–3000 m above sea level, with the annual mean temperature range of 11–27.5°
C and annual mean rainfall of <900 to >1000 mm (Table 1), with some areas characterized
by bimodal rainfall. The study sites fall under the following agroecological zones (AEZs): tepid
sub-moist mid-highlands (SM3), warm sub-moist lowlands (H3), cool sub-moist mid-highlands
(SM4), tepid moist mid-highlands (M3), and tepid sub-humid mid-highlands (SH3) (Table 1).
The dominant soil types in the study sites are Vertisols, Cambisols, Fluvisols, and Nitisols
according to IUSS Working Group World Reference Base (WRB) classification (IUSS
Working Group WRB, 2015) (Table 1).
Experimental design and treatment
The study was conducted on farmers’ fields. As indicated in Table 2, the treatments included six
rates of N (0, 23, 46, 69, 92, 115, and 138 kg N ha−1), six rates of P (0,10, 20, 30, 40, and 50 kg P ha−1),
eight rates of K (0, 15, 30, 45, 60, 75, 90, and 105 kg K ha−1), and six rates of S (0, 10, 20, 30, 40, and
50 kg S ha−1), all combined with the micronutrients boron (B) and zinc (Zn). The recommended NP
rates of 46 kg N ha−1 and 46 kg P2O5 ha−1 and the control with no applied nutrients were included as
the standard and negative control treatments, respectively. The treatments were laid out in random-
ized complete block design with three replications. On all sites, the plot size was 4 m by 3 m (12 m2),
and the spacing between rows, plots, and blocks were 20 cm, 1 m, and 1.5 m, respectively. Nitrogen
was applied in two splits, that is, half at planting and the other half at 35–45 days after planting, while
the full doses of P, K, or S were applied at planting close to the seed drilling row. Urea, triple super-
phosphate, murate of potash (KCl), and gypsum (CaSO4.2H2O) were used as source of N, P, K, and
S, respectively. Improved wheat varieties recommended for the area were sown at the seed rate of
125 kg ha−1 by usingmanual rowmaker. All other agronomic practices for wheat were applied as per
the recommendation for the area.
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
4 Sofonyas Dargie et al.
Table 2. Treatment setup for each nutrient and basal fertilizer application rates
Basal application (kg ha−1)
Nutrient Rates (kg ha−1) N P K S
N 0, 46, 92, 138, 176, 222 – 30 66 30
P 0, 10, 20, 30, 40, 50 92 – 66 30
K 0, 15, 30, 45, 60, 75, 90, 105 92 30 – 30
S 0, 10, 20, 30, 40, 50 92 30 57 –
Data collection and statistical analysis
To measure total above-ground biomass and grain yields, the central seven rows of each plot
(5 m × 1.4 m) were harvested at soil level when the crop reached physiological maturity. The
harvested plants were then weighed to determine the biomass yield and threshed and weighted
to determine the grain yield of each plot. Total biomass (dry matter basis) and grain yields
(adjusted to a moisture content of 12.5%) were recorded on plot basis and then converted to
kg ha−1 for statistical analysis.
A linear mixed model framework was used to determine the variation in yield with the different
levels of N, P, K, and S by soil type and AEZ over study locations and years. The linear mixed
model framework (PROC MIXED of the SAS system) was chosen for the different levels of anal-
yses, because it allows the analysis of hierarchical or clustered data arising from observational
studies through inclusion of both fixed and random effects. The mixed model approach was also
chosen to account for the imbalance in terms of sample size and confounding of responses by
uncontrolled variables. The fixed effects in the model were agroecology, soil type, nutrient rate,
and their interactions, while location was the random effect. In mixed models, the random com-
ponent specifies that the linear predictor contains a term that randomly varies with one or more
ecological correlates of crop yield, for example, location within an AEZ or soil type. This helps to
account for correlation, that is, observations in the same AEZ are likely to be more related than
observations in other zones, and that locations are nested within AEZs or soil types. The initial
model was of the following form:
Y 
 µ AEZ  Soiltype rate AEZ  rate Soiltype  rate Location ε (1)
where μ is the grand mean yield (kg ha−1), AEZ is agroecological zone, soil type is the soil type of
the location according to the WRB classification and correlation system (IUSS Working Group
WRB, 2015), rate is the nutrient application rate (kg ha−1) for the nutrient under study, location is
the random component, and ϵ is the error term. In many cases, however, sample sizes were not
adequate to accommodate this model, and parameters were not correctly estimated. In addition,
recent analyses show that the effect of soil type and AEZ cannot be separated because many soil
types are found in specific climatic zones (Sileshi et al., 2022). Consequently, analyses were done in
two steps, that is, for agroecology and soil type separately. As such, the models used were as
follows:
For AEZs:
Y 
 µ AEZ  rate AEZ  rate Location ε (2)
For soil type:
Y 
 µ Soiltype rate Soiltype  rate Location ε (3)
The variations in yield with fixed effects were considered significant when P≤ 0.05. Least
square estimates and their 95% confidence intervals (CIs) were used for statistical inference
(Supplementary Table S1 and S2). This is because the 95% CI functions as a very conservative
test of hypothesis, and it also attaches a measure of uncertainty to sample statistic (du Prel
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
Experimental Agriculture 5
et al., 2009). The means for two or more levels of a fixed effect were considered to be significantly
different from one another only if their 95% CI were non-overlapping.
In order to determine the optimum rate of the nutrient in question, nutrient dose–response
functions were compared and used as deemed appropriate. The first function chosen was the
asymptotic function given as yield (Y): Y 
 a  bcN , where a is yield at the plateau (i.e., expected
maximum), b is the amplitude (the gain in yield due to nutrient application), and c is a curvature
coefficient, and X is the nutrient rate applied. When the asymptotic function fails to converge,
other similar models, such as Mitscherlich, Gompertz, and logistic functions, which assume that
dose–responses follow Mitscherlich the law of diminishing return, were also applied (Sileshi,
2021). In addition, the quadratic function was compared with the other functions and the model
that fits the data well was chosen for determination of the optimum nutrient rate. The agronomic
optimum was defined as the nutrient rate at which the highest grain yield was obtained on the
dose–response curve, whereas the agronomic optimum is the rate at which the highest grain yield
was obtained, at the optimum pick on data point of response curve.
The agronomic efficiency of N (AEN) and P (AEP), defined as grain yield per unit of N or P
applied, was computed as follows:

   
 GY1 fGYuAEN kg kg (4)
Na

  GYfGYu
AEP kg kg1 
 (5)
Pa
where GY is the grain yield of the fertilized plot (kg ha−1f ), GYu is the grain yield of the unfertilized
plot (kg ha−1), and Na or Pa is the quantity of N or P applied as N or P fertilizer (kg ha−1).
Agronomic efficiency is the amount of additional yield obtained for each additional kg of nutrient
applied (Agegnehu et al., 2016; Fageria and Baligar, 2005).
Results
Response to N rates
The application of N fertilizer at the different rates significantly (p< 0.01) increased grain yield of
wheat across soil types compared to the recommended NP rate (Supplementary Table S3).
However, yield increments were not consistent with the increase in N rate across soil types, rather
yield declined at the highest N rate (222 kg N ha−1) except on Vertisols (Table 3). The lowest yield
was recorded in the control (without N fertilizer) across soil types (Table 3). Higher yield incre-
ments were recorded on Cambisols and Vertisols, while yield increment was very low on Luvisols
(Supplementary Table S2). Compared to the control, N application increased grain yield by 43%
on Cambisols and 22% on Luvisols at 138 kg N ha−1, 26% on Nitisols at 176 kg N ha−1, and 51% on
Vertisols at 92 kg N ha−1. The asymptotic dose–response function was found to be appropriate for
describing yield response to N on Cambisols, Nitisols, and Vertisols, while a polynomial function
was more suitable on Luvisols (Figure 1). The optimum wheat grain yields were obtained with
138 kg N ha−1 on Cambisols and Luvisols, 176 kg N ha−1 on Nitisols, and 92 kg N ha−1 on
Vertisols (Figure 1).
The N dose–response curve showed a similar trend on Cambisols and Nitisols where distinct
increases were observed up to 46 kg N ha−1 and then leveled off until it reached the maximum
yield. In contrast, on Luvisols a gradual increase was observed in yield response to N rate until the
optimum grain yield was attained at ∼92 kg N ha−1 followed by a decrease consistent with a qua-
dratic function. On Vertisols, the yield response to N rate sharply increased up to ∼90 kg N ha−1
and continued with a slight increase up to the maximum N application rate of 222 kg ha−1
(Figure 1).
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
6 Sofonyas Dargie et al.
Table 3. Response of wheat grain yield (kg ha−1) to N, P, K, and S fertilizer (see Supplementary Table S1 and 2 for details)
Soil type AEZ
Nutrient Rate Cambisols Luvisols Nitisols Vertisols H3 M3 SH3 SM3 SM4
N 0 2212 2930 2634 2702 2865 1425 2977 1789 2574
46 2888 3319 3008 3189 3220 2703 3400 2141 3338
92 2997 3483 3226 4098 3458 3145 3987 2549 3757
138 3170 3582 3044 3538 3252 3014 3638 2475 3733
176 2981 3557 3316 3820 3562 2923 3496 2712 4046
222 2806 2603 3179 3948 3420 2978 2890 2998 3978
P 0 2860 3176 2822 2774 3285 2817 2867 2123 3261
10 3357 3346 2991 2777 3477 2822 3131 2561 3406
20 3636 3306 3251 2933 3795 3373 3210 2724 3415
30 3325 3267 2990 2931 3466 2432 3099 2659 3467
40 3601 3296 3212 2970 3731 3320 3165 2738 3526
50 3720 2738 3369 3043 3869 3626 2906 2862 3639
K 0 3828 3516 2951 3111 3260 1929 3524 2984 4518
15 4065 3338 3266 3616 3571 2906 3561 3263 4881
30 3866 3616 3200 3377 3541 2350 3544 3165 4616
45 3775 3527 3057 3268 3340 1891 3618 2864 4906
60 3890 3345 2974 3709 3196 2252 3544 3343 5282
75 3804 3519 3206 3521 3483 Na 3620 3169 Na
90 4107 3290 3171 3639 3421 Na 3459 3340 Na
105 3557 2579 3256 3493 3509 Na 2878 3079 Na
S 0 3327 3462 2988 3330 3371 2554 3457 2810 3371
10 3419 3376 2901 3344 3232 3074 3426 2838 3429
20 3336 3540 3095 3294 3442 2898 3476 2757 3404
30 3369 3499 3067 3225 3420 2806 3401 2703 3379
40 3284 3547 3141 3172 3418 2564 3569 2650 3311
50 3421 2825 2967 3252 3335 2853 3190 2701 3450
Wheat grain yield significantly (p≤ 0.01) varied among AEZs with N rates. Yield increment
due to applied N relative to the control was higher in M3 than all other AEZs (Supplementary
Table S3). The highest yield of 4046 kg N ha−1 was attained in SM4 at 176 kg N ha−1, while lowest
grain yield was recorded in SM3 (Table 3).
The results revealed a significant (p< 0.01) variation in total biomass yield with N rates across
soil types. Total biomass yield increased with N application rates, and the highest values were
recorded at 176 kg N ha−1 on Cambisols, 138 kg ha−1 on Luvisols, and at 222 kg ha−1 on both
Nitisol and Vertisols. Generally, the higher biomass yield was obtained on Cambisols (Table 4).
Total biomass yield response to N also significantly varied (p< 0.01) with AEZs. As in grain yield,
highest total biomass yield was recorded in SH3 compared to SM3.
Response to P rates
The mean wheat grain yield varied significantly (p< 0.05) with P application rates under balanced
fertilization across soil types and AEZs (Table 3). However, total biomass yield did not signifi-
cantly vary with P rates across soil types and AEZs. Across soil types, increasing P rates increased
wheat grain yield by 5–27% compared to the control. Higher yield increments were recorded on
Cambisols followed by Nitisols, while yield increment was very low on Luvisols and Vertisols
(Supplementary Table S2). The highest yield increment (30.1%) was recorded with 50 kg P ha−1
on Cambisols, followed by 50 kg P ha−1 on Nitisol. The highest grain yield was recorded at the
highest P rate on Nitisols. On all soil types, significant yield increments were not observed with
P rates beyond 30 kg P ha−1 (Table 3). The leveling of response for yield with increased rates of
P fertilizer application resulted in a significant quadratic component to the model. However,
the patterns of variation were also similar for Cambisols, Nitisols, and Vertisols (Table 3. b).
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
Experimental Agriculture 7
(a) (b)
5000 Cambisols 5000 Luvisols
4500 4500
4000 4000
3500 3500
3000 3000
2500 2500
2000 2000
1500 1500
0 46 92 138 176 222 0 46 92 138 176 222
(c) (d)
5000 Nitisols 5000 Vertisols
4500 4500
4000 4000
3500 3500
3000 3000
2500 2500
2000 2000
1500 1500
0 46 92 138 176 222 0 46 92 138 176 222
N rate (kg ha–1) N rate (kg ha–1)
Figure 1. Dose–response of wheat N rates in different soil type in Ethiopia. Circles represent measured yield, while black
solid lines and gray dotted lines represent the predicted yields and their 95% confidence limits, respectively.
On Cambisols and Luvisols, the highest grain yield was obtained from the application of 20 kg
P ha−1. However, yields consistently and linearly increased as the P rate increased on Nitisols
and Vertisols; thereby, the diminishing rate of return was not reached (Figure 2).
Wheat grain yield response to P also significantly varied (p≤ 0.01) with AEZs. Yield increment
due to applied P relative to the control was higher in SM3 than the other AEZs (Supplementary
Table S3).
Response to K rates
The mixed-effect model analysis results did not reveal significant (p= 0.37) variation in yield with
K. The 95% confidence limits of grain yields of wheat with the different K rates overlapped with
yields of the control yields and the recommended NP rate on all soil types and AEZs (Table 3). The
changes in yield over the control were very small (<10%) on all soil types except on Vertisols,
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
Grain yield (kg ha –1) Grain yield (kg ha–1)
8 Sofonyas Dargie et al.
Table 4. Yield advantages recorded due to the application of agronomic optimum N rate over the control, without N only
and with recommended NP fertilizer
Yield increase over (%)
Agronomic optimum
Yield advantages Soil type N rate (kg ha−1) Without fertilizer Without N Recommended NP
Grain yield advantage Cambisols 138 49.3 43.3 13.0
Luvisols 138 34.7 22.3 19.2
Nitisols 176 67.1 25.9 42.0
Vertisols 92 75.7 51.7 32.6
Biomass yield advantage Cambisols 138 53.9 45.0 17.9
Luvisols 176 46.3 23.3 21.3
Nitisols 222 55.7 18.1 25.2
Vertisols 222 112.1 93.3 33.8
(a) (b)
4000 Cambisols 4000 Luvisols
3500 3500
3000 3000
2500 2500
2000 2000
1500 1500
0 10 20 30 40 50 0 10 20 30 40 50
(c) (d)
4000 Nitisols 4000 Vertisols
3500 3500
3000 3000
2500 2500
2000 2000
1500 1500
0 10 20 30 40 50 0 10 20 30 40 50
P rate (kg ha–1) P rate (kg ha–1)
Figure 2. Dose–response of wheat P rates in different soil type in Ethiopia. Circles represent measured yield, while black
solid lines and gray dotted lines represent the predicted yields and their 95% confidence limits, respectively.
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
Grain yield (kg ha) Grain yield (kg ha)
Experimental Agriculture 9
where K application increased yields by up to 19% (Supplementary Table S3). Similarly, the
changes in yield over the control were very small in all AEZs except in M3, where K application
increased yields by up to 50%. The K dose–response curve also did not reveal clear trends either
with soil type or AEZ (Supplementary Figure S1). Consequently, the optimum K rate could not be
determined on all soil types.
Response to S rates
The application of S fertilizer at different rates had no significant effect on grain yield of wheat
across the different soil types and AEZs (Table 3). The yields recorded at different S rates also did
not show clear trends on the different soil types and AEZs (Supplementary Figure S1). Overall, the
changes in yield over the control due S application were negligible (<5%) on the different soil
types (Supplementary Table S3). Similarly, the changes in yield over the control were negligible
in the different AEZs except in M3, where S application increased yields by up to 20%.
The dose–response curves also did not reveal any clear pattern for S on all soil types and AEZs
(Supplementary Figure S1). Consequently, the optimum S rate could not be determined on all
soil types.
Agronomic efficiency of N and P
The AEN and AEP significantly (p≤ 0.05) varied with N and P application rates across soil types
and AEZs. The mean AEN varied from 1.3 to 15.2 kg grain kg−1 N applied across soil types, with
the highest value being on Vertisols (15.2 kg grain kg−1 N) followed by Cambisols (14.7 kg kg−1).
Similarly, the AEP ranged from 1.2 to 49 kg grain kg−1 P, with the highest value being on
Cambisols (49 kg grain increase kg−1 P) followed by Nitisols (21 kg grain kg−1 P) (Figure 3).
The highest AEN and AEP were obtained with the application of 92 kg N ha−1. The agronomic
efficiency increments over the highest N rate (222 kg N ha−1) were 177%.
Agronomic efficiency decreased with N rates on all soil types (Figure 3). Except on Vertisols,
the highest agronomic efficiency of wheat was recorded at 46 kg N ha−1, which then decreased
with increase in N rate. On Luvisols, application of at 222 kg N ha−1 resulted in the lowest AEN.
Similarly, AEP decreased with increasing P rates, with the lowest AEP being on Luvisols, Vertisols,
and Nitisols (Figure 3). Low amount of available soil P on Luvisols (Table 1) possibly contributed
to the low AEP. On Cambisols, application of 10 kg P ha−1 resulted in the highest AEP. On
Cambisols, the highest AEP was recorded with the application rate of 10 kg P ha−1.
Discussion
Deficiencies of N and P, S, B, and Zn are widespread in Ethiopian soils, while some soils are also
deficient in K, Cu, Mn, and Fe (Habte and Boke, 2017). Nitrogen (N) is often the most limiting
nutrient for crop yield in Ethiopia. The present study has demonstrated significant improvement
in wheat grain and total biomass yield with increasing rates of N application on different soil types.
The results are consistent with previous research documenting significant responses in wheat
grain yield to N application on different soil types (Guarda et al., 2004; Mansoor et al., 2000).
The remarkable yield increments registered by the new wheat cultivars have been enhanced by
the progressively higher N-inputs (Guarda et al., 2004). The results of this study show that N rates
as high as 222 kg N ha−1 give substantial yield increments on some soils (e.g., Vertisols). Mansour
et al. (2017) similarly reported that increasing N level up to 280 kg N ha−1 significantly increased
grain yield.
This study also revealed significant increases in grain yield with increasing P rates on all soils
except Luvisols. Agegnehu et al. (2015) similarly found that application of P fertilizer at different rates
increased wheat grain yield, up to 30% over the control. However, our results indicate that P levels
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
10 Sofonyas Dargie et al.
(a)
16 Agronomic efficiency of N
14 Cambisols
Luvisols
12 Nitisols
10 Vertisols
8
6
4
2
0
0 46 92 138 176 222
Nitrogen rate (kg ha–1)
(b)
60 Agronomic efficiency of P
Cambisols
50 Luvisols
40 Nitisols
Vertisols
30
20
10
0
0 10 20 30 40 50
Phosphorus rate (kg ha–1)
Figure 3. Nutrient use efficiency of bread wheat under balanced fertilization in four soil types in Ethiopia.
between 10 and 30 kg ha−1 may be adequate, depending on soil type. P rates above 30 kg ha−1 appear
to depress yields on Luvisols,
The application of different rates of K fertilizer did not significantly improve grain yields of
wheat on most soils except on Vertisols, which had a slightly lower K than the other soils
(Table 1). This is probably because the test soil already had sufficient K for plant growth and might
be some of K released from non-exchangeable sources. According to the EthioSIS soil map, most
Ethiopian soils are not deficient in K. This result is in agreement with the finding of Amare et al.
(2010) who reported that application of K fertilizer on maize grain yield had no significant effect.
Another study also reported that application of K did not significantly affect wheat yield (Tariq
and Shah 2002), as the experimental soil already had sufficient K for plant growth. However, we
recommend application of the minimum rate of 15 kg K ha−1 for maintenance of soil K reserves.
The results showed negligible improvement in grain yields of wheat with the different S appli-
cation rates on all soils and AEZ, except in M3. This is probably because the sites have adequate
soil S contents. According to Itanna (2005), all soil types other than Nitisols, Andosols, and
Vertisols contain soluble sulfate in adequate amount for crop production in Ethiopia.
However, Nitisols had the lowest soluble sulfate, which is below the critical level for crop produc-
tion (Itanna, 2005). Previous studies have indicated that land degradation, removal of crop res-
idues, crop uptake, and use of non-S fertilizers are major causes of sulfur deficiency (Dibabe et al.,
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
AEP (kg grain per kg P) AEN (kg grain per kg N)
Experimental Agriculture 11
2007; Itanna, 2005). According to Weil and Mughogho (2000), failure to supply S in the form of
urea or diammonium phosphate, which contain little S, contributes to S deficiency in Africa.
Therefore, we recommend application of S at the minimum rate of 10 kg ha−1 for maintenance
of soil S reserves and sustain wheat production.
The AEN recorded on Cambisols and Vertisols with application of 46 and 92 kg N ha−1 was in
the range of commonly reported values (Agegnehu et al., 2016; Tarekegne and Tanner, 2001).
Dargie et al. (2018) also reported that the AEN decreased with increasing rates of N on
Vertisols and Cambisols of Tigray. The highest AEP obtained at 20 kg P ha−1 may be due to
the high amount of active iron and aluminum which often results in P-fixation (Agegnehu
et al., 2019; Batjes, 2011).
The agronomic efficiency of wheat recorded in this study up to 92 kg N ha−1 falls within the
range of value reported for cereals (Dobermann (2005) and wheat elsewhere. For wheat, Guarda
et al. (2004) found mean agronomic efficiency of 2–18 kg kg−1 N, while Tarekegne and Tanner
(2001) reported agronomic efficiency of 12.6–29 and 15–26 kg yield increase kg−1 applied N on a
Vertisols and Nitisol, respectively. Similar research was also reported by Haileselassie et al. (2014),
where AE of wheat decreases with P rates in sandy soils. The highest agronomic efficiency was
obtained at a rate of 20 kg P ha−1.
Conclusions
From the results of this study, it can be concluded that the balanced application of N, P, K, S
combined with B and Zn significantly increases wheat yield. Across all four soil types and
AEZs, application of 46–92 kg N ha−1 and 10–30 kg P ha−1 with balanced application of K, S,
B, and Zn could be sufficient. The application of 20 kg P ha−1 was optimum for wheat production
on Cambisols and Luvisols. Application of K and S at different rates did not show clear trends on
the different soil types. However, application of K and S at their minimum rates of 15 kg K ha−1
and 10 kg S ha−1 should be promoted for the maintenance of soil K and S reserves and sustain
productivity.
We recommend further field trials involving different N and P levels under balanced fertiliza-
tion, climatic conditions, and soil types to enhance our understanding of limiting factors and facil-
itate formulation of site-specific fertilizer recommendations.
Supplementary material. To view supplementary material for this article, please visit https://doi.org/10.1017/
S0014479722000151
Acknowledgement. The authors would like to acknowledge the Ethiopian Institute of Agricultural Research (EIAR), Oromia,
Amhara, Southern and Tigray Agricultural Research Institutes (RARIs) for funding this research and generating data. The
International Center for Tropical Agriculture (CIAT) and Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) are
highly appreciated for organizing the data, and facilitating and financing the preparation of this manuscript.
Funding Support. Funding for this work came from the Ethiopian Institute of Agricultural Research (EIAR), Oromia,
Amhara, Southern and Tigray Agricultural Research Institutes (RARIs), the International Center for Tropical Agriculture
(CIAT), and Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ). This work was supported, in whole or in part,
by the Bill & Melinda Gates Foundation [INV-005460]. Under the grant conditions of the Foundation, a Creative Commons
Attribution 4.0 Generic License has already been assigned to the Author Accepted Manuscript version that might arise from
this submission.
Competing Interest. No competing interest among the authors and the organization.
References
Agegnehu G., Nelson P.N. and Bird M.I. (2016). The effects of biochar, compost and their mixture and nitrogen fertilizer on
yield and nitrogen use efficiency of barley grown on a Nitisol in the highlands of Ethiopia. Science of the Total Environment
569–570, 869–879. doi: 10.1016/j.scitotenv.2016.05.033
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
12 Sofonyas Dargie et al.
Agegnehu G., Nelson P.N., Bird M.I. and vanBeek C. (2015). Phosphorus response and fertilizer recommendations for
wheat grown on Nitisols in the central Ethiopian highlands. Communications in Soil Science and Plant Analysis 46,
2411–2424. doi: 10.1080/00103624.2015.1081922
Agegnehu G., Yirga C. and Erkossa T. (2019). Soil Acidity Management. Addis Ababa, Ethiopia: Ethiopian Institute of
Agricultural Research (EIAR), pp. 1–56. ISBN:9789994466597.
Amare T., Feyisa T. and Gebreselasie Y. (2010). Response of Maize, Malt Barley and tomato to potassium. In Yitaferu B.,
Tesema T. and Ayalew Z. (eds), The Third Annual Regional Conference on Completed Research Activities on Soil and Water
Management, Forestry, and Agricultural Mechanization. Bahir Dar, Ethiopia: Amhara Region Agricultural Research
Institute (ARARI), pp. 67–75.
Batjes N. (2011). Global Distribution of Soil Phosphorus Retention Potential. ISRIC Report 2011/06, ISRIC-World Soil
Information (with dataset), ISRIC Report 2011/06: Wageningen.
CSA. (2018). Report on Area and Production of Major Crops. Addis Ababa, Ethiopia: Agricultural Sample Survey 2017/18,
Central Statistical Agency (CSA), Volume 1.
Dargie, S., Wogi L. and Kidanu S. (2018). Response of Bread Wheat (Triticum aestivum L.) to Application of Slow Releasing
Nitrogen Fertilizer in Tigray. Ethiopian Journal of Agricultural Sciences 28, 111–126.
Dawson J.C., Huggins D.R. and Jones S.S. (2008). Characterizing nitrogen use efficiency in natural and agricultural eco-
systems to improve the performance of cereal crops in low input and organic agricultural systems. Field Crops Research 107,
89–101. doi: 10.1016/j.fcr.2008.01.001
Dibabe A., Bekele T. and Assen Y. (2007). The status of micronutrients in Nitisols, Vertisols, Cambisols and Fluvisols in
major Maize, Wheat, Teff and Citrus growing areas of Ethiopia. In Proceedings of the Workshop on Status of Micronutrients,
pp. 77–96. Addis Ababa, Ethiopia.
Dobermann A.R. (2005). Nitrogen use efficiency-state of the art. In IFA International Workshop on Enhanced Efficiency
Fertilizers, 28–30 June 2005. Frankfurt, Germany: Agronomy and Horticulture – Faculty Publications.. https://
digitalcommons.unl.edu/agronomyfacpub/316 (accessed 21 December 2021).
du Prel J.B., Hommel G., Röhring B. and Blettner M. (2009). Confidence interval or-pvalue?: part 4 of a series on evaluation
o scientific publications. Deutsches Ärztteblatt International 106, 100–105. doi: 10.3238/arztebl.2009.0100
Fageria N. and Baligar V. (2005). Enhancing nitrogen use efficiency in crop plants. Advances in agronomy 88, 97–185. doi: 10.
1016/S0065-2113(05)88004-6
FAO. (2016). FAO Stat 2016. Rome, Italy: Food and Agriculture Organization of the United Nations (FAO).
Guarda G., Padovan S. and Delogu G. (2004). Grain yield, nitrogen-use efficiency and baking quality of old and modern
Italian bread-wheat cultivars grown at different nitrogen levels. European Journal of Agronomy 21, 181–192. doi: 10.1016/j.
eja.2003.08.001
HabteM. and Boke S. (2017). Influence of balanced nutrients on growth performance and yield of Teff (Eragrostis tef (Zucc.))
in the midland of Bensa, Southern Ethiopia. Journal of Scientific and Innovative Research 6, 101–103. www.jsirjournal.co
Haileselassie B., Habte D., Haileselassie M. and Gebremeskel G. (2014). Effects of mineral nitrogen and phosphorus fer-
tilizers on yield and nutrient utilization of bread wheat (Tritcum aestivum) on the sandy soils of Hawzen District, Northern
Ethiopia. Agriculture, Forestry and Fisheries 3, 189–198.
Hirel B., Le Gouis J., Ney B. and Gallais A. (2007). The challenge of improving nitrogen use efficiency in crop plants: toward
a more central role for genetic variability and quantitative genetics within integrated approaches. Journal of Experimental
Botany 58, 2369–2387. doi: 10.1093/jxb/erm097
Itanna F. (2005). Sulfur distribution in five Ethiopian Rift Valley soils under humid and semi-arid climate. Journal of Arid
Environments 62, 597–612. doi: 10.1016/j.jaridenv.2005.01.010
IUSS Working Group WRB. (2015). World Reference Base for Soil Resources 2014, Update 2015: International Soil
Classification System for Naming Soils and Creating Legends for Soil Maps. World Soil Resources Reports 106.
Rome: FAO.
Jones A., Breuning-Madsen H., Brossard M., Dampha A., Deckers J., Dewitte O., Gallali T., Hallett S., Jones R. and
Kilasara M. (2013). Soil atlas of Africa. Luxembourg: European Commission, Publications Office of the European
Union, 176.
Lopez-Bellido R.J. and Lopez-Bellido L. (2001). Efficiency of nitrogen in wheat under Mediterranean conditions: Effect of
tillage, crop rotation and N fertilization. Field Crop Research 71, 31–46. doi: 10.1016/S0378-4290(01)00146-0
Macauley H. and Ramadjita T. (2015). Cereal crops: rice, maize, millet, sorghum, wheat. Feeding Africa: Background Paper
for an Action Plan for African Agricultural Transformation, 21–23 Oct. 2015. Abdou Diouf International Conference
Center: Dakar, Senegal. https://www.afdb.org/fileadmin/uploads/afdb/Documents/Events/DakAgri2015/Cereal_Crops-_
Rice__Maize__Millet__Sorghum__Wheat.pdf (accessed 06 September 2021).
Mansoor H.A., Ndondi R.V., Tanner D.O., Ndakidemi P. and Ngatokewa R.T. (2000). On-farm evaluation of the response
of four bread wheat varieties to nitrogen fertilizer in karatu district in northern Tanzania. In CIMMYT (eds.), The Eleventh
Regional Wheat Workshop for Eastern, Central and Southern Africa. Addis Ababa, Ethiopia: CIMMYT, pp. 352–359.
ISBN:92-9146-087-7.
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press
Experimental Agriculture 13
Mansour E., Merwad A., Yasin M., Abdul-Hamid M., El-Sobky E. and Oraby H. (2017). Nitrogen use efficiency in spring
wheat: genotypic variation and grain yield response under sandy soil conditions. The Journal of Agricultural Science 155,
1407–1423. doi: 10.1017/S0021859617000600
Mason N.M., Jayne T.S. and Shiferaw B.A. (2012). Wheat Consumption in Sub-Saharan Africa: Trends, Drivers, and Policy
Implications. MSU International Development Working Paper 127, December 2012, Michigan State University, East
Lansing, Michigan 48824. https://ageconsearch.umn.edu/record/146936 (accessed 06 September 2021).
Minot N. (2014). Food price volatility in sub-Saharan Africa: Has it really increased? Food Policy 45, 45–56. doi: 10.1016/j.
foodpol.2013.12.008
Oerke E.C. (2006). Crop loses to pests. The Journal of Agricultural Science 144, 31–43. doi: 10.1017/S0021859605005708
Regassa H. and Agegnehu G. (2011). Potentials and limitations of acid soils in the highlands of Ethiopia: a review. In
Mulatu B. and Grando S. (eds), Barley Research and Development in Ethiopia. Aleppo, Syria: ICARDA, pp. 103–112.
ISBN:92-9127-249-3
Sharma I., Tyagi B., Singh G., Venkatesh K. and Gupta O. (2015). Enhancing wheat production-A global perspective. Indian
Journal of Agricultural Sciences 85, 3–13.
Sileshi G.W. (2021) Dose-response models to guide site-specific nutrient management and lessons for fertilizer trial design in
sub-Saharan Africa. Experimental Agriculture, 1–18. doi: 10.1017/S0014479721000193
Sileshi G.W. and Gebeyehu S. (2021). Emerging infectious diseases threatening food security and economies in Africa. Global
Food Security 28, 100479. doi: 10.1016/j.gfs.2020.100479
Sileshi G.W., Kihara J., Tamene L., Vanlauwe B., Phiri E. and Jama B. (2022) Unravelling causes of poor crop response to
applied N and P fertilizers on African soils. Experimental Agriculture. doi: 10.1017/s0014479721000193
Tarekegne A. and Tanner D.R. (2001). Effects of Fertilizer application on N and P uptake, recovery and use efficiency of
bread wheat grown on two soil types in central Ethiopia. Ethiopian Journal of Natural Resources 3, 219–244.
Tariq M. and Shah M. (2002). Response of wheat to applied soil potassium. Asian Journal of Plant Sciences 1, 470–471.
Weil R.R. and Mughogho S.K. (2000). Sulfur nutrition of maize in four regions of Malawi. Agronomy Journal 92, 649–656.
doi: 10.2134/agronj2000.924649x
Yirga C., Tesfaye A., Agegnehu G., Keneni G., Kassa B. and Asefa G. (2002). Crop-livestock farming systems of the high-
lands of Welmera Wereda: the case of Welmera Goro bench-mark site. In Keneni G., Gojjam Y., Bedane K., Yirga C. and
Dibabe A. (eds.), Towards Farmers’ Participatory research: Attempts and Achievements in the Central Highlands of Ethiopia.
Proceedings of Client-Oriented Research EvaluationWorkshop, 16-18 October 2001, Holetta Agricultural Research Center,
Holetta, Ethiopia, pp. 147–174.
Zeleke G., Agegnehu G., Abera D. and Rashid S. (2010). Fertilizer and Soil Fertility Potential in Ethiopia: Constraints and
Opportunities for Enhancing the System. Working Paper, Washington, DC, USA: International Food Policy Research
Institute (IFPRI), pp. 1–42. www.ifpri.org
Cite this article: Dargie S, Girma T, Chibsa T, Kassa S, Boke S, Abera A, Haileselassie B, Addisie S, Amsalu S, Haileselassie M,
Soboka S, Abera W, and Weldesemayat SG. Balanced fertilization increases wheat yield response on different soils and
agroecological zones in Ethiopia. Experimental Agriculture. https://doi.org/10.1017/S0014479722000151
https://doi.org/10.1017/S0014479722000151 Published online by Cambridge University Press