Agron. Sustain. Dev.

(2017) 37:25
DOI 10.1007/s13593-017-0431-0

REVIEW ARTICLE

Application of secondary nutrients and micronutrients increases

crop yields in sub-Saharan Africa
Job Kihara 1 & Gudeta Weldesemayat Sileshi 2 & Generose Nziguheba 3 &
Michael Kinyua 4 & Shamie Zingore 5 & Rolf Sommer 1

Accepted: 1 June 2017

# The Author(s) 2017. This article is an open access publication

Abstract Secondary and micronutrients are important in en- Africa Soil Information Service were all for S and micronutri-
hancing crop productivity; yet, they are hardly studied in sub- ent combinations. Of the two sources, most yield data are for
Sahara Africa. In this region, the main focus has been on maize (73.6%), followed by sorghum (6.7%) and wheat
macronutrients but there is emerging though scattered evi- (6.1%) while rice, cowpea, faba bean, tef, and soybean each
dence of crop productivity limitations by the secondary and accounted for less than 5%. The major points are the follow-
micronutrients. Elsewhere, widespread deficiencies of these ing: (1) application of S and micronutrients increased maize
nutrients are associated with stagnation of yields. Here, we yield by 0.84 t ha−1 (i.e., 25%) over macronutrient only treat-
undertake a meta-analysis using 40 articles reporting crop re- ment and achieved agronomic efficiencies (kilograms of grain
sponse to secondary and micronutrients to (1) determine the increase per kilogram of micronutrient added) between 38 and
productivity increase of crops and nutrient use efficiency as- 432 and (2) response ratios were >1 for S and all
sociated with these nutrients, and (2) provide synthesis of micronutrients, i.e., the probability of response ratio exceed-
responses to secondary nutrients and micronutrients in sub- ing 1 was 0.77 for S and 0.83 for Zn, 0.95 for Cu, and 0.92 for
Sahara Africa. This study used 757 yield data rows (530 from Fe, and indicates positive crop response for a majority of
publications and 227 from Africa Soil Information Service) farmers. We conclude that S and micronutrients are holding
from field trials carried out in SSA between 1969 and 2013 in back crop productivity especially on soils where response to
14 countries. Data from publications constituted response to S macronutrients is low and that more research is needed to
(49.4%), Zn (23.0%), S and micronutrient combinations unravel conditions under which application of S and
(11.5%), and <10% each for Cu, Mo, Fe, and B. Data from micronutrients may pose financial risks.

* Job Kihara Keywords Agronomic efficiency . Boundary analysis .

j.kihara@cgiar.org Meta-analysis . Sulfur . Effect size . Response ratio

1
International Center for Tropical Agriculture (CIAT) c/o ICIPE Contents
Duduville Complex, Off Kasarani Road, PO Box 823-00621,
Nairobi, Kenya
1. Introduction
2
2. Materials and methods
Plot 1244 Ibex Hill, Lusaka, Zambia
2.1 Choice of crops
3
International Institute of Tropical Agriculture (IITA) c/o ICIPE 2.2 Data retrieval
Duduville Complex, Off Kasarani Road, PO Box 823-00621,
Nairobi, Kenya
2.3 Data analysis
4
3. Results and discussion
School of Environmental Studies, Kenyatta University,
P.O. Box 43844-00100, Nairobi, Kenya
3.1 Overall crop yield response
5
3.2 Agronomic efficiency
International Plant Nutrition Institute, Sub-Saharan Africa Program,
IFDC—East and Southern Africa Division, ICIPE Complex,
4. Conclusions
Duduville-Kasarani, Thika Road, P.O. Box 30772-00100, 5. References
Nairobi, Kenya 6. Acknowledgements
25 Page 2 of 14 Agron. Sustain. Dev. (2017) 37:25

1 Introduction

For the last three decades, crop nutrition research in sub-

Saharan Africa (SSA) has largely focused on macronutri-
ents, i.e., nitrogen, phosphorus, and potassium (Chilimba
and Chirwa 2000; Stoorvogel et al. 1993; Voortman 2012;
Vanlauwe et al. 2015). However, there are strong indica-
tions from some studies of the effects of deficiencies in
secondary nutrients and micronutrients limiting crop pro-
ductivity in SSA, especially under continuous cropping
without nutrient replenishment (Vanlauwe et al. 2015;
Van Asten et al. 2004; Fig. 1). For example, low produc-
tive spots related to zinc (Zn) deficiency have been ob-
served in rice (Van Asten et al. 2004) following continu-
ous cultivation without micronutrient application, but Fig. 1 Experimental field showing a control treatment (foreground) and
could also relate to interactions with other factors such a fertilizer treatment including sulfur and micronutrients in Ethiopia in
as soil alkalinity and high plant available soil P. In 2014
Ethiopia, Rwanda, and Burundi, sulfur (S), Zn, and boron
(B) are deficient in large proportions of arable land
(Vanlauwe et al. 2015). In other words, nutrients other cereals (maize and wheat) in Ethiopia, Ghana, Malawi,
than the macronutrients are becoming important limita- Nigeria, Sierra Leone, Tanzania, and Zambia, and
tions in crop production in SSA. Deficiencies of second- highlighted the need for micronutrient supply especially
ary nutrients and micronutrients have also been observed copper (Cu), B, Zn, and Mo to address observed
in other parts of the world such as Campo Cerrado soils of deficiencies and to realize full potential in agriculture
Brazil, IIanos of Colombia, and calcareous soils of productivity in developing countries.
Pakistan, India, and the Philippines (Lopes 1980; Ryan Recently, Kihara and Njoroge (2013) noted that max-
et al. 2013). Stagnation of yields some 10 years ago in imum maize grain yield obtained in fields under re-
India was primarily because of widespread deficiencies of searcher management in western Kenya, a region with
these important secondary and micronutrients, mainly S, potential yield of at least 10 t ha −1 , stagnated at
B, and Zn (Sahrawat et al. 2010). In comparison to other 7 t ha−1. This is likely to be the result of deficiencies
regions of the world, relatively little research has been in secondary and micronutrients during crop growth, or
done on secondary and micronutrient deficiencies in soils the effect of low soil pH and associated toxicities (e.g.,
and the corresponding crop responses in SSA. aluminum). Indeed, in the last decade, there has been
The lack of sufficient focus on micronutrients and the growing evidence from the region of soils that barely
related crop response in SSA have been echoed by ex- responds to application of the commonly emphasized
perts in different generations (Lopes 1980; Kang and macronutrient-based fertilizers under smallholder farming
Osiname 1985; Weil and Mughogho 1999; Gungula and (Tittonnell and Giller 2013), and researchers note that
Garjila 2006). As early as the 1970s, molybdenum (Mo), this is likely due to deficiencies of the secondary and
Zn, and B were identified as the most important micronutrients (Vanlauwe et al. 2015). Although there
micronutrients but only for selected crops such as cotton are increasing calls to include secondary and
and groundnuts mainly in areas under intensification micronutrients in fertilizers in SSA (Chianu et al.
(Drosdoff 1972). Apart from such areas of intensifica- 2012), the limited and scattered research on potential
tion, long fallow periods under the then shifting cultiva- responses to these nutrients has not been reviewed to
tion replenished sufficient macro-, secondary, and understand their contribution to crop productivity. The
micronutrients, making it unlikely to experience deficien- aim of this study was to determine the productivity in-
cies within that traditional agricultural system. In the crease in cereal and grain legume crops and nutrient use
review by Lopes (1980), only a few micronutrient (Mn, efficiency associated with application of secondary and
Zn, B, Fe, and Mo) studies in sub-Sahara Africa are micronutrients through a meta-analysis of available pub-
mentioned and, similarly to Drosdoff (1972), none of lished data, and to provide synthesis of responses to sec-
the studies included cereals (maize, wheat, rice, sorghum, ondary nutrients and micronutrients in SSA. We aimed to
or millet) under field conditions. A decade later, answer the specific research question: “What is the mag-
Sillanpaa (1982), in a global study involving 30 coun- nitude of change in yield due to application of S and
tries, identified micronutrient deficiencies in for two micronutrients applied alone or in combination?”
Agron. Sustain. Dev. (2017) 37:25 Page 3 of 14 25

2 Materials and methods multiple combinations was undertaken. Also, search was
made for titles of publications in reference lists of journal
2.1 Choice of crops articles referring to micronutrient use. A study was included
if it was conducted in SSA and met three conditions as fol-
This meta-analysis focused on maize, rice, wheat, sorghum, lows: (i) included maize, sorghum, rice, tef, cowpea, soybean,
cowpea, and soybean because of data availability and the im- faba bean, or wheat as test crops; (ii) included both a treatment
portance of these crops in SSA. Maize is the key focus of most with macronutrients only (fertilized control) and a similarly
studies on crop response to S and micronutrients in SSA with a managed treatment but with S and/or micronutrients in addi-
limited number of studies on wheat, rice, cowpea, sorghum, tion to the macronutrients in the fertilized control treatment
and soybean responses. Maize, constituting 45% of the cereal (i.e., same rate of macronutrients in both control and test);
production in SSA in 2014, is the staple food crop (accounting and,(iii) reported yield data for treatments under condition ii.
for over 40% of the calories consumed in some countries such Crop response to Ca and Mg is not commonly investigated
as Malawi and Zambia (http://www.fao.org/faostat/en/#data)) and therefore has not been considered here. Data for absolute
and the main component of food aid interventions in SSA control treatments (i.e., unfertilized) were also obtained, al-
(Leonardo et al. 2015). It is also grown under widely varying though these were not available in all studies. Pot and green-
climatic, soil, and altitudinal conditions ranging from sea level house experiments, and studies where yields were reported in
(the coastal zones) to elevations above 2400 m (Sileshi et al. units per plant without providing data on plant population, as
2010). Rice, the most rapidly growing food source in Africa, well as publications that reported yields as percentages were
is a strategic crop and a staple food in many countries excluded as these lacked true productivity information. In
(Wopereis et al. 2013). Sorghum accounted for 16% of all some cases, the source of P (e.g., SSP) and sources of
cereal production in Africa in 2014. Wheat is also an impor- micronutrients (e.g., ZnS) contained S. As such, we assumed
tant staple crop whose consumption steadily increased during that if the S contained in the macronutrient control treatment
the past 20 years as a result of growing population, changing was at least 40 kg S ha−1, further S in the micronutrient treat-
food preferences, and socioeconomic transformations associ- ment can be ignored since no further response to S is expected
ated with urbanization. It constituted 14% of all cereal pro- (Kang and Osiname 1976; Ojeniyi and Kayode 1993). When
duction in Africa in 2014 (FAOSTAT). Soybean is an impor- no S was included in the macronutrient control treatment, but
tant pulse, currently widely promoted for its high protein con- the micronutrient source also contained S (e.g., in the form of
tent and superior biological nitrogen fixation (Adesoji et al. zinc sulfate, in the case of Abbas et al. (2007)), we considered
2009). Like soybean, cowpea is an important source of pro- this as “combined” rather than single micronutrient response.
teins especially for poor rural families in dry sub-humid and In total 530 rows of yield data were extracted from the 40
semi-arid zones of SSA (Gungula and Garjila 2006). papers of which 49.4% were on S response, 23.0% on Zn,
7.4% on Cu, 3.0% on Mo, 4.5% on Fe, 1.1% on B, and
2.2 Data retrieval 11.5% involved two or more, i.e., S and micronutrient combi-
nations. These data are from 14 countries namely Nigeria
This study used data from field trials carried out in SSA be- (36.6%), Malawi (12.8%), Ethiopia (21.8%), Kenya (3.4%),
tween 1969 and 2013. The data were derived from 40 peer- Côte d’Ivoire (3.0%), Ghana (4.3%), Zambia (3.0%), Sudan
reviewed publications accessed online (Table 1). Basically, on (2.3%), Zimbabwe (2.3), Togo (6.0%), Mozambique (1.5%),
24 November 2016, advanced search was made in the Web of Tanzania (1.3%), Benin (1.1%), and Burkina Faso (0.4%).
Knowledge using the equation: TS=(micronutrient* AND The data were derived from both on- farm and on-station trials
crop yield AND Africa) OR TS=(sulfur AND crop yield under researcher management. All the data obtained are on
AND Africa) OR TS=(sulphur AND crop yield AND improved crop varieties except for wheat where 18 of the 46
Africa) OR TS=(boron AND crop yield AND Africa) OR data rows used local wheat variety in Ethiopia. From all the
TS=(zinc AND crop yield AND Africa) OR TS=(manganese publications, data on soil types were obtained; if presented
AND crop yield AND Africa) OR TS=(molyb*AND crop otherwise, the information was obtained from ISRIC maps
yield AND Africa) OR TS=(copper AND crop yield AND based on locational information presented in the publications.
Africa) OR TS=(calcium AND crop yield AND Africa) OR The studies cover most soil types prevalent in SSA namely
TS=(magnesium AND crop yield AND Africa) OR TS=(iron Acrisols, Cambisols, Ferralsols, Fluvisols, Gleysols,
AND crop yield AND Africa), where OR and AND are bool- Leptosols, Lixisols, Luvisols, Nitisols, Plinthosols, and
ean operators and TS=topic. Although time span was set to all Vertisols. The sites also represent different climates with mean
years, it only returned articles published between 2005 and annual rainfall from 255 to 2860 mm. The intensity of re-
2016. As such, we also used the Google Scholar literature search on micronutrient responses varied with time, and
search engine where search for the keywords micronutrients, 19.2% of the studies were conducted in 1969–1980, 15.8%
sulfur, zinc, boron, crop response, and sub-Sahara Africa in in 1981–1990, 7.9% in 1991–2000, 48.1% in 2000–2010, and
25 Page 4 of 14 Agron. Sustain. Dev. (2017) 37:25

Table 1 List of publications, the study locations within sub-Saharan Africa, and secondary and micronutrients used in this study of crop responses

Authors Source Country of study Yield data Nutrient of focus

rows

Abbas et al. 2007 Gezira J. of Agri Sci 5 (2) Sudan 12 Zn

Abunyewa and Mercer-Quarshie Asian J. of Plant Sci 3 (1): 1–5 Ghana 12 Zn
2004
Adesoji et al. 2009 American-Eurasian J. of Sus Agri, 3(4): 845–849 Nigeria 6 Mo
Admas et al. 2015 18
Afolabi et al. 2014 Int. J. of Agri & Rural Dev. 17(3): 2006–2011 Nigeria 4 B, Mo, Zn,
combination
AfSIS 227
Allan 1970 FAO Soils Bulletin 14 Kenya 18 S
Chaguala et al. 2011 Afr Crop Sci Proc. 10: 611–615 Mozambique 8 S
Chiezey 2014 J. of Agri Sci. 6(3);2014 Nigeria 9 Zn
Chilimba and Chirwa 2000 Sulfur Nutr Def Amendment for Maize Pro in Malawi 48 S
Malawi
Chude et al. 2003 Badu-Apraku B, Fakorede MAB, Ouedraogo M, Nigeria 4 Zn
Carsky
RJ, Menkir A. (eds) (2003) p 201–207
Desta et al. 2015 J Soil Sci Environ Manag (2015) 6:9–15 Ethiopia 12 Zn
Desta 2015 Int J Recent Sci Res (2015) 6: 5689–5698 Ethiopia 8 S
Eteng et al. 2014 Open J. of Soil Sci, 4: 235–245 Nigeria 12 Cu, Zn
Florent et al. 2014 J. of Advances in Agri 3(1): 129–141 Côte d’Ivoire 9 Zn, combination
Friesen 1991 Mokwunye AU (Ed.) (1991) Kluwer Academic Niger, Burkina Faso, 20 S
Publishers p 59–68 Togo
Gungula and Garjila 2006 American-Eurasian J. of Agri & Envi Sci, 1(2): Nigeria 3 Mo
96–101
Habtegebrial and Singh 2006 Nutr Cycl Agroecosyst (2006) 75:213–222 Ethiopia 24 S
Habtegebrial and Singh 2009 J. of Plant Nutr, 32(10): 1768–1787 Ethiopia 36 S
Habtegebrial et al. 2013 J. of Soil Sci and Mgt 4(3): 62–70 Ethiopia 10 S
Habtemichial et al. 2007 J. Plant Nutr Soil Sci (2007) 170, 412–418 Ethiopia 6 S
Haileselasssie et al. 2011 Soil Sci & Plant Nutr, 57(4): 587–596 Ethiopia 1 Combined
Kang and Osiname 1976 Agro J., 68: 333–336 Nigeria 24 S
Kayode 1984 Expl Agric (1984) 20: 335–337 Nigeria 16 Fe
Kayode and Agboola 1985 Fert Res 8:129–135 Nigeria 36 Cu, Fe, Zn,
combination
Kone et al. 2011 Archives of Agro Soil Sci (2011) 57(7): 763–774 Benin 6 Zn
Kone et al. 2014 Archives of Agron Soil Sci (2014) 60: 735–746 Côte d’Ivoire 3 Zn
Kurwakumire et al. 2015 Agron J (2015) 107: 1068–1076 Zimbabwe 6 Combination
Lisuma et al. 2006 J. of Soil & Crop Manag, 98:402–406 Tanzania 7 B, Cu, Zn,
combination
Manzeke et al. 2014 Field Crops Res 166 (2014) 128–136 Zimbabwe 6 Zn
Nesgea et al. 2012 Int.J. Agr & Res, 2(9): 14–32 Ethiopia 1 S
Nyalemegbe et al. 2011 Int. Res J. of Agri Sci & Soil Sci, 2(1): 008–016 Ghana 5 S
Nziguheba et al. 2009 Plant & Soil DOI 10.1007/s11104-008-9714-1 Togo 12 B, S, Zn
Ojeniyi and Kayode 1993 J. of Agr Sci, Cambridge 120: 295–299 Nigeria 40 Cu, S
Olivier et al. 2014 Ind. J. Sci. Res.& Tech. 2(2): 18–24 Côte d’Ivoire 3 Zn
Oseni 2009 World J. of Agr Sci. 5(6): 670–674 Nigeria 16 Zn
Osiname et al. 1973 Agron J (1973) 65: 875–877 Nigeria 24 Zn
Rechiatu et al. 2015 J Plant Sci (2015) 3: 64–70 Ghana 6 Mo
Van Asten et al. 2004 Land Degrad. Dev. 15: 383–396 Burkina Faso 2 Zn
Weil and Mughogho 1999 Agro J. 92: 649–656 Malawi 20 S
Yerokun and Chirwa 2014 Afr J. of Agr Res 9(11): 963–970 Zambia 16 Zn
Agron. Sustain. Dev. (2017) 37:25 Page 5 of 14 25

8.9% in 2010–2016. The proportions of yield data per indi- In the subsequent analysis, the effects of the S and
vidual crop species reported from the various studies varied micronutrients on maize yield (crop with sufficient data
greatly with maize having the highest (70%), followed by points) were estimated using LMM of the form. Yield is a
wheat (8.7%) and rice (6.6%). Cowpea, tef, faba bean, sor- function of fertilizer treatment with experimental site as the
ghum, and soybean each accounted for less than 3% of the random variable. Here, treatment refers to either individual or
yield data. None of the studies reported using organic re- combined application of S and micronutrients. Control refers
sources as a nutrient management strategy. to the crop that received N, P, and/or K fertilizer. In the mixed
In addition to the review, we include a dataset from multi- model, the site was used as random effect and parameters were
locational trials implemented within the framework of Africa estimated via restricted maximum likelihood (REML). The
Soil Information System (AfSIS) and that provides a unique obtained means were used to estimate percentage yield in-
opportunity to relate response to soil micronutrients. The crease following application of S and micronutrients.
dataset consist of 227 data rows with response to combined As the index of effect size, we used the response ratio (RR)
secondary and micronutrients namely Ca, Mg, S, Zn, and B calculated as the ratio of the yields from the treatment (i.e., S
(implemented as one treatment). This is the only large and and/or micronutrient) to yields from the fertilized control (i.e.,
consistent dataset that provides observed yield with actual soil plots receiving only macronutrients). RR is a measure of the
analysis. The data are from diagnostic trials implemented in proportionate change resulting from a treatment and expressed
Kenya, Malawi, Mali, Nigeria, and Tanzania and conducted as a natural log for meta-analysis (Hedges et al. 1999).
for 1–2 seasons; the test crop was maize except in Mali where Absence of publication bias for this measure was confirmed
sorghum was used. The macronutrient control treatment had using funnel plot (Fig. 2). Since our objective was to obtain
100 kg N ha−1, 30 kg P ha−1, and 60 kg K ha−1 for maize and the mean responses and their confidence limits, our analysis of
60 kg N ha−1, 20 kg P ha−1, and 30 kg K ha−1 for sorghum. response ratio is averaged over different sites and application
The sources of N, P, and K were urea, triple super phosphate, rates. Effect sizes in meta-analyses are normally weighted by
and muriate of potash, respectively. Secondary and study precisions (Philibert et al. 2012), most commonly by
micronutrients in the response treatment were applied at inverse of study variance, or the sample size. Weighting is
10 kg Ca ha−1, 5 kg Mg ha−1, 5 kg S ha−1, 3 kg Zn ha−1, important, because it increases the precision of the mean effect
and trace amounts of B. These were from a blended basal estimate and the power of the tests and improves sampling
fertilizer, Mavuno. Detailed description of the implementation distribution of the test statistics (Gurevitch and Hedges
methods for these trials is reported by Huising et al. (2013) 1999). However, nearly all of the studies we found did not
and is also accessible online (http://afsis-dt.ciat.cgiar.org) and report any measure of variance, but the numbers of replica-
from Kihara et al. (2016). tions were available. Before analysis, RR was log transformed
(lnRR) to ensure normality (Hedges et al. 1999). Then lnRR
2.3 Data analysis values were weighted by the number of replicates, so that
studies based on larger sample sizes are given more weight
Data analysis, as also data retrieval, considered the six quality
criteria of meta-analysis recommendations of Philibert et al.
(2012). Most of the analysis involved a linear mixed model-
ling (LMM) approach implemented in the R statistical soft-
ware (www.r-project.org). LMM was chosen, because the data
gathered across studies were unbalanced with respect to
sample sizes and treatments.
Initially, yield data following S and micronutrient applica-
tion were plotted against the fertilized control yield in a scatter
to demonstrate the distribution of crop response to applied
nutrients. Also, means and the confidence limit of maize,
wheat, and rice grain yield for the absolute control, the fertil-
ized control (i.e., the macronutrient treatment), and the S and
micronutrient treatment were obtained from the raw data. For
this, bootstrap confidence limit was obtained using boot R
package with 100 replications. Yield data for the absolute
control were not available in some cases: of the 530 cases of
yield data, the absolute control was available in 232 cases
which represented 80% of the cases for both rice and wheat Fig. 2 Funnel plot of relationships between response ratio (RR) and
and 39% for maize but these were available for all AfSIS data. standard deviations for the studies used
25 Page 6 of 14 Agron. Sustain. Dev. (2017) 37:25

than those based on small size. These values were subjected to cowpea, and soybean, responses to S and micronutrients were
LMM, and the means and 95% confidence intervals for each low. For maize and wheat, responses to S and micronutrients
study were computed via the REML method. These were then decreased with increasing macronutrient control yields
presented as forest plots (see also Hossard et al. (2016)). (Fig. 4). When all studies were combined (cross crops and
Traditionally, inferences from meta-analyses are based on sites), the overall effect size (RR = 1.20; 95% CL 1.18–
the mean and its 95% confidence interval. However, inference 1.23) was significantly greater than unity (Fig. 5), indicating
based on the mean alone can be misleading if the probability an overall positive crop response to S and micronutrients. In
distribution of the response is not known (Sileshi et al. 2010). 19 out of the 41 studies, RR values were significantly larger
Therefore, we estimated the probability of obtaining a given than 1, while in 18 studies, RR was not significantly different
response ratio. For this purpose, we generated the frequency from 1. Only one study had an effect size significantly lower
distribution of response ratios and then calculated the cumu- than 1, indicating significant reduction in yields due to appli-
lative probabilities. Then, we calculated the probability of cation of S or micronutrients (Fig. 5a). The largest increase in
exceeding a given level of response, for example, RR >1, 2, response was noted in Burkina Faso (Fig. 5b).
etc., under application of a given nutrient. The overall positive response to micronutrients indicates
We calculated the agronomic efficiency (AE = kilograms of that these nutrients are holding back crop productivity, partic-
yield increase per kilogram of applied nutrient) of S and ularly in areas with low response to macronutrients, and that
micronutrients by maize as the difference between yield with their application can have huge effect in some locations. Low
fertilizer together with S or micronutrients and yield of the crop productivity under macronutrient application in SSA has
fertilized control and dividing by the quantity of applied S or often been reported (Kihara and Njoroge 2013), and it has
micronutrient. AE is an integrated index of nutrient recovery been suggested that deficiencies of secondary and
efficiency and physiological or internal nutrient use efficiency micronutrients could be one of the causes to limited crop re-
(Ladha et al. 2005). Yield data where nutrients were applied in sponse to macronutrients (Vanlauwe et al. 2015). In the case of
a combination were not included, because it was not possible maize, application of S and micronutrients resulted in average
to isolate agronomic efficiency associated with the individual 0.7 t ha−1 (i.e., 20%) more yield compared to the macronutri-
nutrients. To establish the maximum AE that can be expected ent control treatment (Fig. 6). Similarly, wheat and rice yield
based on available data, boundary analysis (see also Kihara were increased over macronutrient only treatment by 27 and
and Njoroge (2013)) was used on the three nutrients with a 12%, respectively, following additional application of S and
majority of yield data, namely S, Cu, and Zn. Here, a bound- micronutrients. Thus, although application of macronutrients
ary line was fit on the mean AE of the three highest points at resulted in 1.87 t ha−1 (i.e., 122%) more maize, 0.85 t ha−1
every 10 kg ha−1 for S and 1 kg ha−1 intervals for Cu and Zn (i.e., 41%) more wheat, and 1.3 t ha−1 (i.e., 81%) more rice
applied. These boundary lines are four-parameter log logistic grain yield increase over the unfertilized (no-input) treatment,
curves fit using the dose response curve (drc) package in the R application of S and micronutrients still led to an additional
statistical software (www.r-project.org) that uses the formula: yield increase. Addressing such secondary and micronutrient
deficiencies is critical to resolving the recurrent food insecu-
d−c
y ¼ f ð xÞ ¼ c þ rity challenge facing SSA that is heightened by burgeoning
1 þ expðbðlogðxÞ−logðeÞÞÞ population and climate change. Understanding the conditions
where, under which different responses occur and designing fertilizer
solutions to address the limitations are critical in maintaining
y Agronomic efficiency (AE) and or increasing crop productivity. The responses to S and
x Amount of nutrient applied micronutrients vary between crops, and it can be expected that
c Lower horizontal asymptote variations in responses exist between crop varieties within a
d Upper horizontal asymptote specific crop, though these may translate more in nutrient
e Slope steepness contents than in crop yields (De Valenca et al. 2017). The
b ED50 value studies included in this paper are from researcher-managed
trials, and improved crop varieties were used. Improved vari-
eties have gained focus in initiatives to combat food insecuri-
ty, owing to their high yielding capacity. Given that crop re-
3 Results and discussion sponse to fertilizers (macronutrients) has been shown for such
improved varieties than for local varieties (Vanlauwe et al.
3.1 Overall crop yield response 2011), it is important to understand the responses among va-
rieties also in relation to S and micronutrients.
Maize, wheat, and rice showed positive yield responses to Overall, mean yield increment (over the mean of fertilized
applied S or micronutrients (Fig. 3). In the case of sorghum, control and based on mixed models) realized by the
Agron. Sustain. Dev. (2017) 37:25 Page 7 of 14 25

Fig. 3 Effect of S and/or Grain yield Grain yield

micronutrient application on yield (t ha-1) of
of (a) maize, (b) wheat, (c) rice, (t ha-1) of
(d) sorghum, (e) soybean and (f) micronutrient micronutrient
cowpea across a range of control 16 6
yields in SSA. The fertilized
control was the crop that received 14 5
the recommended rate of N, P,
and K fertilizer. N represents the 12
(a) Maize (N=508) 4 (b) Wheat (N=45)
sample size
10

8 3

6 2

4
1
2

0 0
0 2 4 6 8 10 12 14 16 0 1 2 3 4 5 6

8 3.0
(d) Sorghum (N=51)
7 (c) Rice
2.5
6
2.0
5

4 1.5

3
1.0
2
0.5
1

0 0.0
0 1 2 3 4 5 6 7 8 0.0 0.5 1.0 1.5 2.0 2.5 3.0

2.4 1.4

(e) Soybean (N=10) 1.2 (f) Cowpea (N=11)

2.0

1.0
1.6
0.8
1.2
0.6
0.8
0.4

0.4 0.2

0.0 0.0
0.0 0.4 0.8 1.2 1.6 2.0 2.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Grain yield (t ha-1) in the fertilized control Grain yield (t ha-1) in the fertilized control

application of specific nutrients are 26% for S, 6% for B, 15% These deficiencies can be widespread as noted by Abe et al.
for Zn, 13% for Cu, 9% for Fe, and 20% for combinations of S (2010) for Zn and S deficiencies in the lowland soils of West
and micronutrients (see also Fig. 7). Several studies (Sillanpaa Africa, by Chilimba and Chirwa (2000) for S in Malawi, and
1982; Van Asten et al. 2004; Oyinlola and Chude 2010) have by Vanlauwe et al. (2015) for S, Zn, and B in Ethiopia. Buri
reported deficiencies of one or more S and micronutrients in et al. (2000) reported that over 66% of West Africa lowland
specific regions in Ivory Coast, Nigeria, Togo, Democratic soils had available Zn below the critical level of 0.8 mg kg−1.
Republic of the Congo, Kenya, Sudan, Ethiopia, Ghana, Although relative yield increase due to combined S and
Malawi, Sierra Leone, Tanzania, Zambia, and Burkina Faso. micronutrients is not different from single nutrients (e.g., S
25 Page 8 of 14 Agron. Sustain. Dev. (2017) 37:25

Response
ratio
4.0 4.0
4.0
3.5 3.5
3.5

3.0 3.0 (b) Wheat

(a) Maize 3.0 (c) Rice
2.5 2.5 2.5

2.0 2.0 2.0

1.5 1.5 1.5

1.0 1.0 1.0

0.5 0.5 0.5

0.0 0.0 0.0

0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 0 2 4 6 8
Grain yield (t ha-1) of control Grain yield (t ha-1) of control Grain yield (t ha-1) of control
Response
ratio
1.4
2.0 1.4 (f) Cowpea
(d) Sorghum (e) Soybean
1.2
1.2
1.6
1.0 1.0

1.2 0.8 0.8

0.6 0.6
0.8

0.4 0.4
0.4
0.2 0.2

0.0 0.0 0.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 0.2 0.4 0.6 0.8 1.0 1.2

Grain yield (t ha-1) of control Grain yield (t ha-1) of control Grain yield (t ha-1) of control

Fig. 4 Variations in response of (a) maize, (b) wheat, (c) rice, (d) sorghum, (e) soybean and (f) cowpea with the level of control yield. The red line
represents RR = 1 while the black line represents the trend condition. The control is the plot that received macronutrients

and Zn), arguments favoring combined rather than the single only limit profitability and economic returns for farmers but
secondary and micronutrients applications have been made also lead to toxic levels of some micronutrients if not well
(Vanlauwe et al. 2015). On the other hand, some studies report monitored. This highlights the need for identification of mi-
sufficiency of S and micronutrients such as Cu, Fe, and Zn in cronutrient limitations and recommending management that is
Western Usambara Mountains of Lushoto in Tanzania focused on the specific nutritional problems in different re-
(Ndakidemi and Semoka 2006); Zn and B in Gongola River gions/sites. Also, as crop production increases from the cur-
Basin (Adeboye 2011); and Fe and Mn at Bauchi in Guinea rent low-input levels following green revolution initiatives,
Savanna, Samaru in northern Guinea, and Kadawa in Sudan the deficiencies could become more significant and spread
Savanna (Oyinlola and Chude 2010). Also, Snapp (1998), in a out (Current practices are low yielding and may not reveal
study involving 1130 soil samples in Malawi, observed limit- crop growth limitations.).
ed deficiencies in Zn with at least 90% of the cases having Overall, estimates and 95% confidence limits of RR were
more Zn than the critical limit of 0.8 mg Zn kg−1. Thus, while 1.35 (1.28–1.41) for S, 1.54 (1.35–1.74) for Cu, 1.34 (1.23–
combined S and micronutrients applications may be the best 1.44) for Fe, and 1.40 (1.29–1.51) for Zn. The probability of
option in some cases, there is no need for applications of those RR exceeding 1 (i.e., yields increasing over the macronutrient
nutrients available in soil in sufficient quantities as this can not control) was 0.77 for S and 0.83 for Zn (Fig. 8). The
Agron. Sustain. Dev. (2017) 37:25 Page 9 of 14 25

Fig. 5 Forest plot of the variation Studies

in effect sizes with study (a) and All studies combined
Eteng et al. 2014
(a)
from country to country (b). Abunyewa and…
Horizontal bars represent 95% Van Asten et al. 2004
Friesen, 1991
confidence limits. The vertical Chilimba and Chirwa 2011
bar in (b) represents RR = 1 (i.e., Abbas et al. 2007
Habtegebrial and singh 2006
no effect of micronutrients) Habtegebrial and singh 2009
Rechiatu et al. 2015
Habtemichial et al. 2007
Manzeke et al. 2014
Ojeniyi and Kayode 1993
Kayode 1984
Kang and Osiname 1976
Allan 1971
Habtegebrial et al.2013
Lisuma et al. 2006
Desta, 2015
Weil and Mughogho 2000
Gungula and Garjila 2006
Chude et al. 1991
Desta et al.2015
Olivier et al. 2014
Chiezy 2014
Osiname et al. 1973
Kone et al. 2011
Nyalemegbe et al. 2012
Chaguala et al. 2011
Admas et al. 2015
Yerokun and Chirwa 2014
Florent et al. 2014
Nesgea et al. 2012
Kurwakumire et al. 2015
Adesoji et al. 2009
Kone et al. 2014
Afolabi et al. 2014
Nziguheba et al. 2009
Oseni 2009
AfSIS
Haileselassie et al. 2011
Kayode and Agboola 1985
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Effect size (RR)

Countries
Burkina Faso (b)
Ghana
Sudan
Togo
Malawi
Ethiopia
Kenya
Nigeria
Tanzania
Zimbabwe
Benin
Mozambique
Zambia
Cote d'Ivoire
Mali
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Effect size (RR)

probabilities of RR exceeding 1 for Cu and Fe were 0.95 and been made: for example, Weil and Mughogho (1999), in their
0.92, respectively. In other words, RR <1 that constitute risks work on S in four regions of Malawi, concluded that the 80–
to the farmer were very low for micronutrients. This highlights 160-kg maize grain gained per kilogram S applied would be
the fact that the benefits of applying macronutrients in com- economically attractive. Based on improved agronomic efficien-
bination with S and micronutrients far outweigh the risk to cy of fertilizer macronutrients as a result of micronutrient appli-
farmers. Ironically, only one of the 41 studies whose data are cation in deficient soils, Vanlauwe et al. (2015) conclude on
used in the current analysis, i.e., Van Asten et al. (2004), did blending micronutrients with fertilizers as a cost-effective way
an economic analysis of micronutrient used (Zn) with results to overcome micronutrient deficiency. Since micronutrients are
being highly profitable (value cost ratio >2). The lack of eco- often viewed for their role in human health, and not for their
nomic analysis in most of these studies translates a general gap contribution in enhancing crop yields, there are opinions that
of knowledge on the profitability of secondary and micronu- overcoming soil micronutrients deficiency is a main path to
trient fertilization in SSA. Indirect assumptions have often overcoming deficient of those nutrients in human health
25 Page 10 of 14 Agron. Sustain. Dev. (2017) 37:25

Fig. 6 Grain yield under different fertilizer regimes as observed in the average 91 (±67) kg N ha − 1 , 25 (±9) kg P ha − 1 , and 36
reported studies. Error bars are confidence intervals of the means. For (±8.6) kg K ha−1 applied while +S or Micronutrient also had 39 (±16)
maize, macronutrient treatment had on average 97 (±36) kg N ha−1, 34 S, 30 (±5.7) Zn. For rice, macronutrient treatment had on average 66
(±20) kg P ha−1 , and 45 (±23) kg K ha −1 applied while +S or (±29) kg N ha−1, 52 (±33) kg P ha−1, and 59 (±34) kg K ha−1 applied
Micronutrient also had 32 (±21) S, 5.6 (±3.2) Zn, 6 (±3.5) Fe, 4.1 while +S or Micronutrient also had 36 (±25) S, 10 (±0) Zn. Data used are
(±2.7) Cu, and 1 (±0) B. For wheat, macronutrient treatment had on both from publications and AfSIS

(Barret and Bevis 2015; de Valenca et al. 2017). As such, the the price of fertilizers is not clear. However, in the context of
profitability of micronutrient fertilization has been assessed for current efforts of increasing access to fertilizers by farmers, in-
their role in reducing the health burden associated with micro- cluding micronutrient-enriched fertilizers in various subsidy pro-
nutrient deficiency in human. Joy et al. (2015) reported that grams has potential to increase widespread application of
enriching granular fertilizers with Zn would be cost-effective micronutrients without direct additional costs for farmers cov-
in reducing healthy life lost due to micronutrient deficiency in ered by such programs.
human, particularly if embedded into ongoing subsidy pro- As expected, maize yield response to macronutrients and to
grams. The extent to which enrichment of fertilizers could affect S and micronutrients significantly varied with soil type
(Fig. 9). Average response ratios to S and micronutrients were
positive in all soil types, varying from 0.9 in Luvisols to 2.5 in
Gleysols. The lowest yield gains due to S and micronutrient
applications on Luvisols are related to their high inherent fer-
tility. Although yield responses to macronutrients were very
low on Gleysols (soils affected/influenced by shallow ground-
water) and Vertisols (poor drainage and difficult workability),
yield gains due to the addition of micronutrients were high on
these soils (Fig. 9b). Trial locations for both soils were char-
acterized by very low soil micronutrients (<0.26 Cu and 0.23
Zn mg kg−1, data not shown). These results indicate the im-
portance of inherent soil properties on crop yields and yield
responses to macronutrients, S, and micronutrients, which, in
combination with other explaining factors, may provide a
good basis for targeting nutrient recommendations.

3.2 Agronomic efficiency

As expected, and for all S and micronutrients, the highest agro-

Fig. 7 Boxplots of the effect of S and micronutrient application on yield
nomic efficiency (kilograms of maize grain per kilogram of
of maize as observed in SSA. Plus and minus signs indicate if nutrient nutrient applied) was observed at low application rates, and
was added or omitted. S&MN combined S and micronutrients decreased with increasing nutrient application. For example,
Agron. Sustain. Dev. (2017) 37:25 Page 11 of 14 25

Probability Probability
of exceeding of exceedig
1.0 1.0
0.9
(a) (b)
0.9
0.8 0.8 Copper
0.7 Sulfur 0.7 Iron

0.6 Zinc
0.6
0.5 0.5
0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0.0 0.0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Response ratio Response ratio
Fig. 8 The probability of exceeding a given response ratios in maize yield following application of (a) sulfur and (b) micronutrients (Cu, Fe, and Zn)

maximum agronomic efficiency plotted against the amount of S macronutrients (Vanlauwe et al. 2015), and higher macronutrient
and Zn applied and defined by the four-parameter dose response apparent recoveries have been observed (Girish et al. 2015).
function is characterized by decline within the first 30 kg S ha−1 Response to nutrients including micronutrients depends on,
and throughout Zn application rates (Fig. 10). In other words, among other factors (e.g., soil acidity and interactions between
application of S beyond this rate for maize runs the risk of being nutrients), the level of crop available nutrients in soil.
wasteful. On average, and based on our data, agronomic effi- Excluding AfSIS, only 57% of the included studies reported
ciency of 432 kg grain kg−1 Cu, 254 kg grain kg−1 Fe, micronutrient concentrations in soils. Since methods of soil
203 kg grain kg − 1 B, 151 kg grain kg − 1 Zn, and analysis varied from one study to another, relating soils test
38 kg grain kg−1 S were observed. Further, notable residual values with crop responses was not undertaken in our study.
effects on crop yield can still be expected in succeeding seasons Studies on crop response to secondary nutrients and
after application to the soil as observed for Zn (a 25% yield micronutrients are scanty in SSA, and research is needed not
increment after 38–75 kg Zn ha−1 applied in the previous sea- only to further quantify potential responses under different
son) by Soleimani (2012). The highest level of agronomic effi- soils conditions (e.g., pH and micronutrient levels), crop
ciency followed the application of copper at 0.8 kg Cu ha−1 in types, and varieties but also to understand residual effects
Nigeria, and is in line with reports of Cu deficiency in eastern during subsequent seasons, the effectiveness and use efficien-
Nigerian states (Sillanpaa 1982). Application of micronutrients cy of different secondary and micronutrient sources, their in-
is known to increase also the use efficiency of the teractions, and appropriate application methods. This is

Fig. 9 Variation in maize yield Soil type

Soil type
(a) and response ratios (b) with
Ferralsols Gleysols
soil type. Horizontal bars (a) (b)
represent 95% confidence limits. Acrisols Vertisols
The vertical bar in (b) represents Luvisols Cambisols
RR = 1 (i.e., no effect of
Lixisols Fluvisols
micronutrients)
Nitisols Arenosols
Vertisols Lixisols
Arenosols Ferralsols
Plinthosol Nitisols
Fluvisols Leptosols
Leptosols Plinthosol
Cambisols Acrisols
Gleysols Luvisols
0 1 2 3 4 5 6 7 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Yield (t ha-1) in fertilized control Response ratio
25 Page 12 of 14 Agron. Sustain. Dev. (2017) 37:25

Fig. 10 Boundary lines of

agronomic efficiency of sulfur
and zinc for maize at different
application rates in SSA. Curves
are boundary lines of maximum
agronomic efficiency

needed in order to ascertain where their applications are re- uptake of S and micronutrient technologies by farmers are
quired, guide fertilizer blending targeted to the specific pro- needed.
duction systems, and to recommend best practices to small-
holder farmers in SSA. Such research should also go further to Acknowledgements This work was conducted through the support of
relate soil micronutrients or applied S and micronutrients to two projects namely the USAID Feed the Future’s Africa RISING
Program under the project “transforming key production systems: maize
nutritional quality of harvest products from different crops and
mixed east and southern Africa” and the “sustainable intensification of
genotypes for human and animal nutrition. Undoubtedly, SSA maize-legume cropping systems for food security in eastern and southern
is in need of micronutrients and more so with the expected Africa—Phase II (SIMLESA-2).” We acknowledge insights provided by
increase in crop production level. It is noteworthy that micro- Steve Ichami during formative stages of this publication.
nutrient fertilizers are not available in all countries or regions
Open Access This article is distributed under the terms of the Creative
where these are needed (see Van Asten et al. (2004)) and Commons Attribution 4.0 International License (http://
efforts are needed by governments to reverse the situation. creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided you give appro-
priate credit to the original author(s) and the source, provide a link to the
Creative Commons license, and indicate if changes were made.

4 Conclusions
References
The application of S and micronutrients resulted in 0.84 t ha−1
more maize yield compared to the recommended N, P, and K
Abbas AE, Medani W, Hamad ME, Medani W, Babiker HM, Medani W,
fertilizers. This represents a 25% yield increment over what was Nour AE (2007) Effects of added phosphorus and zinc on yield and
achieved using the recommended fertilizer, which only contains its components of corn. Gezira J Agric Sci 5(2)
macronutrients. Although there is high variability in crop re- Abe SS, Buri MM, Issaka RN, Kiepe P, Wakatsuki T (2010) Soil fertility
sponse to S and micronutrients, the high proportions of cases potential for rice production in West African lowlands. Jpn Agric
Res q 44:343–355
with increased yield underscore the need for these nutrients. S
Abunyewa AA, Mercer-Quarshie H (2004) Response of maize to mag-
and micronutrients are holding back crop productivity especial- nesium and zinc application in the semi-arid zone of West Africa.
ly on soils where response to macronutrients is low. Coupling Asian J Plant Sci 3(1):1–5
the ongoing efforts towards the African green revolution, with Adeboye MKA (2011) Status of total and available boron and zinc in the
the supply of S and micronutrients in most fertilizers, is vital for soils of Gongola river basin of Nigeria. Savannah J Agric 6:47–57
Adesoji AG, Abubakar IU, Ishaya DB (2009) Performance of soybean
enhanced agricultural productivity. Considering also the ob-
(Glycine max(L.) Merrill) as influenced by method and rate of mo-
served incidences of low responses and variations among crops, lybdenum application in Samaru, northern Guinea Savanna of
more research is needed to unravel conditions under which Nigeria. Am-Eurasian J Sustain Agric 3(4):845–849
application of S and micronutrients is beneficial to farmers. Admas H, Gebrekidan H, Bedadi B, Adgo E (2015) Effects of Organic
The results of this study highlight the need for policy support and Inorganic Fertilizers on Yield and Yield Components of Maize
at Wujiraba Watershed, Northwestern Highlands of Ethiopia.
for the research on secondary nutrients, particularly S, and American J. Plant Nutr and Fert Tech 5:1–5
micronutrients and for blending them in currently used fertil- Afolabi SG, Adekanmbi AA, Adeboye MK, Bala A (2014) Effect of
izers. Further, decision support tools and training that can aid various input combinations on the growth, nodulation and yield of
Agron. Sustain. Dev. (2017) 37:25 Page 13 of 14 25

inoculated soybean in Minna. Nigeria Int J Agric Rural Dev 17(3): Habtegebrial K, Singh BR (2009) Response of wheat cultivars to nitrogen
2006–2011 and sulfur for crop yield, nitrogen use efficiency and protein quality
Allan AY (1971) Fertilizer use on maize in Kenya. In FAO. Improving Soil in the semiarid region. J. Plant Nutr 32:1768–1787
Fertility in Africa. FAO Soils Bulletin 14. Rome, Italy. pg 10-25. Habtegebrial HK, Ram SB, Aune JB (2007) Wheat response to N2 fixed
Barret CB, Bevis LEM (2015) The micronutrient deficiencies challenge by faba bean (Vicia faba L.) as affected by sulfur fertilization and
in African food system. In: Sahn DE The fight against hunger and rhizobial inoculation in semi-arid Northern Ethiopia. J Plant Nutr
malnutrition: the role of food, agriculture and targeted policies. Soil Sci 170(3):412–418
Oxford University Press, 61–84 Habtegebrial K, Mersha S, Habtu S (2013) Nitrogen and sulphur fertil-
Buri MM, Masunaga T, Wakatsuki T (2000) Sulfur and zinc levels as izers effects on yield, nitrogen uptake and nitrogen use efficiency of
limiting factors to rice production in West Africa lowlands. upland rice variety on irrigated Fluvisols of the Afar region,
Geoderma 94:23–42. doi:10.1016/S0016-7061(99)00076-2 Ethiopia. J Soil Sci Env Mgt 4:62–70
Chaguala PA, Nobela L, Aline S, Maria R, Buque I (2011) Maize re- Haileselassie B, Stomph TJ, Hoffland E (2011) Teff (Eragrostis tef) pro-
sponse to macro-and micro-nutrients in two different agro-ecologi- duction constraints on Vertisols in Ethiopia: farmers' perceptions and
cal zones in Mozambique. In 10th African Crop Science Conference evaluation of low soil zinc as yield-limiting factor. Soil Sci Plant
Proceedings, Maputo, Mozambique (10-13 October 2011). Afr Crop Nutr 57:587–596
Sci Soc. Hedges LV, Gurevitch J, Curtis PS (1999) The meta-analysis of response
ratios in experimental ecology. Ecology 80:1150–1156
Chianu JN, Chianu JN, Mairura F (2012) Mineral fertilizer in the farming
Hossard L, Archer DW, Bertrand M, Colnenne-David C, Debaeke P,
systems of sub-Saharan Africa. A review. Agron Sustain Dev 32:
Ernfors M, Jeuffroy M, Munier-Jolain N, Nilsson C, Sanford GR,
545–566. doi:10.1007/s13593-011-0050-0
Snapp SS, Jensen ES, Makowski D (2016) A meta-analysis of maize
Chiezey UF (2014) Field Performance of Quality Protein Maize with and wheat yields in low-input vs. conventional and organic systems.
Zinc and Magnesium Fertilizers in the Sub-Humid Savanna of Agron J 108:1155–1167. doi:10.2134/agronj2015.0512
Nigeria. J Agric Sci 6(3):84 Huising J, Zingore S, Kihara J, Nziguheba G (2013) Diagnostic trials: a
Chilimba ADC, Chirwa IMD (2000) Sulphur nutrient deficiency amend- practical guide and instruction manual. African Soil Information
ment for maize production in Malawi. Report for Chitedze Research Service, Nairobi, 53 p
Station. Ministry of Agriculture, Lilongwe Joy EJM, Stein AJ, Young SD, Ander EL, Watts MJ, Broadley MR
Chude VO, Iwuafor ENO, Amapu IY, Pam SG, Yusuf AA (2003) (2015) Zinc-enriched fertilisers as a potential public health interven-
Response of maize to zinc fertilization in relation to Mehlich III tion in Africa. Plant Soil 389:1–24
extractable zinc. In Badu-Apraku B, Fakorede MAB, Ouedraogo Kang BT, Osiname OA (1976) Sulfur response of maize in Western
M, Carsky RJ, Menkir A (eds) Maize revolution in West and Nigeria. Agron J 68:333–336
Central Africa, p 201-207. Kang BT, Osiname OA (1985) Micronutrient problems in tropical Africa.
Desta HA (2015) Response of maize (Zea mays L.) to different levels of Fertil res 7(1–3):131–150
nitrogen and sulfur fertilizers in Chilga district, Amhara national Kayode GO (1984) Effect of iron on maize yields in forest and Savanna
regional state, Ethiopia. Int J Recent Sci Res 6:5689–5698 zones of Nigeria. Exp Agric 20:335–337
Desta Y, Habtegebrial K, Weldu Y (2015) Inoculation, phosphorous and Kayode GO, Agboola AA (1985) Macro and micro-nutrient effects on the
zinc fertilization effects on nodulation, yield and nutrient uptake of yield and nutrient concentration of maize (Zea mays L.) in South
Faba bean (Vicia faba L.) grown on calcaric cambisol of semiarid Western Nigeria. Nutr Cycl Agroecosyst 8(2):129–135
Ethiopia. J Soil Sci Env Mgt 6(1):9–15 Kihara J, Njoroge S (2013) Phosphorus agronomic efficiency in maize-
De Valenca AW, Bake A, Brouwer ID, Giller KE (2017) Agronomic based cropping systems: a focus on western Kenya. Field crops Res
biofertilization of crops to fight hidden hunger in sub-Saharan 150:1–8. doi:10.1016/j.fcr.2013.05.025
Africa. Gobal Food Sec 12:8–14 Kihara J, Nziguheba G, Zingore S, Coulibaly A, Esilaba A, Kabambe V,
Drosdoff M (1972) Soil micronutrients. In: Soils of the humid tropics. Njoroge S, Palm C, Huising J (2016) Understanding variability in
National Academy of Sciences, Washington, D.C., pp 150–162 crop response to fertilizer and amendments in sub-Saharan Africa.
Eteng EU, Asawalam DO, Ano AO (2014) Effect of Cu and Zn on Maize Agric Ecosyst Environ 229:1–12. doi:10.1016/j.agee.2016.05.012
(Zea mays L.) Yield and Nutrient Uptake in Coastal Plain Sand Koné B, Amadji GL, Aliou S, Diatta S, Akakpo C (2011) Nutrient con-
Derived Soils of Southeastern Nigeria. Open J Soil Sci 4:235–245 straint and yield potential of rice on upland soil in the south of the
Florent Z, Kone B, Jeremie GT, Fulgence AE, Firmin KK, Joachim TM, Dahoumey gap of West Africa. Arch Agron Soil Sci 57(7):763–774
N’ganzoua KR, Albert YK (2014) Lowland Rice Yield as Affected Kone B, Fofana M, Sorho F, Diatta S, Ogunbayo A, Sie M (2014)
by Straw Incorporation and Inorganic Fertilizer Over Cropping Nutrient constraint of rainfed rice production in foot slope soil of
Seasons in Fluvisol. J Adv Agri 3(1):129–141 Guinea Forest in Côte d’Ivoire. Arch Agron Soil Sci 60(6):735–746
Kurwakumire N, Chikowo R, Zingore S, Mtambanengwe F, Mapfumo P,
Friesen DK (1991) Fate and efficiency of sulfur fertilizer applied to food
Snapp S, Johnston A (2015) Nutrient Management Strategies on
crops in West Africa. Fert Res 29(1):35–44
Heterogeneously Fertile Granitic-Derived Soils in Subhumid
Girish C, Suhas PW, Kanwar LS, Rajesh C (2015) Enhanced nutrient and
Zimbabwe. Agron J 107(3):1068–1076
rainwater use efficiency in maize and soybean with secondary and
Ladha JK, Pathak H, Krupnik TJ, Six J, van Kessel C (2005) Efficiency of
micro nutrient amendments in the rainfed semi-arid tropics. Arch
fertilizer nitrogen in cereal production: retrospects and prospects.
Agron Soil Sci 61(3):285–298. doi:10.1080/03650340.2014.
Adv Agron 87:85–156. doi:10.1016/S0065-2113(05)87003-8
928928
Leonardo WJ, van de Ven GWJ, Udo H, Kanellopoulos A, Sitoe A, Giller
Gungula DT, Garjila Y (2006) The effects of molybdenum application on KE (2015) Labour not land constrains agricultural production and food
growth and yield of cowpea in Yola. J Agric Environ Sci 1(2):96– self-sufficiency in maize-based smallholder farming systems in
101 Mozambique. Food Sec 7:857–874. doi:10.1007/s12571-015-0480-7
Gurevitch J, Hedges LV (1999) Statistical issues in ecological meta-anal- Lisuma JB, Semoka JM, Semu E (2006) Maize yield response and nutri-
yses. Ecology 80(4):1142–1149 ent uptake after micronutrient application on a volcanic soil. Agron J
Habtegebrial K, Singh BR (2006) Effects of timing of nitrogen and sul- 98(2):402–406
phur fertilizers on yield, nitrogen, and sulphur contents of Tef Lopes AS (1980) Micronutrients in soils of the tropics as constraints to
(Eragrostis tef (Zucc.) Trotter). Nutr Cycl Agroecosyst 75:213–222 food production. In: Drosdoff M, Zandstra H, Rockwood WG (eds)
25 Page 14 of 14 Agron. Sustain. Dev. (2017) 37:25

Priorities for alleviating soil-related constraints to food production in in rainfed agroecosystems: case study from Indian semi-arid tropics.
the tropics. International Rice Research Institute, Los Banos, pp Commun Soil Sci Plant Anal 41:346–360. doi:10.1080/
277–298 00103620903462340
Manzeke GM, Mtambanengwe F, Nezomba H, Mapfumo P (2014) Zinc Sileshi G, Akinnifesi FK, Debusho LK, Beedy T, Ajayi OC, Mong’omba
fertilization influence on maize productivity and grain nutritional S (2010) Variation in maize yield gaps with plant nutrient inputs, soil
quality under integrated soil fertility management in Zimbabwe. type and climate across sub-Saharan Africa. Field Crops Res 116:1–
Field Crop Res 166:128–136 13. doi:10.1016/j.fcr.2009.11.014
Ndakidemi PA, Semoka JMR (2006) Soil fertility survey in western Sillanpaa M (1982) Micronutrients and the nutrient status of soils: a
Usambara Mountains, northern Tanzania. Pedosphere 16(2):237– global study, Issue 48. FAO, Rome
244. doi:10.1016/S1002-0160(06)60049-0 Snapp SS (1998) Soil nutrient status of smallholder farms in Malawi.
Nesgea S, Gebrekidan H, Sharma JJ, Berhe T (2012) Effects of nitrogen Commun Soil Sci Plant Anal 29:2571–2588. doi:10.1080/
sources and application time on yield attributes, yield and grain 00103629809370135
protein of rain-fed NERICA-3 rice in Gambella, Ethiopia. Int J Soleimani R (2012) Cumulative and residual effects of zinc sulfate on
Agro Agri Res 2(9):14–32 grain yield, zinc, iron, and copper concentration in corn and wheat. J
Nyalemegbe KK, Hotsonyame GK, Ofori F, Osakpa TY (2012) Plant Nutr 35:85–92. doi:10.1080/01904167.2012.631669
Comparative study of the efficacy of actyva compound fertilizer
Stoorvogel JJ, Smaling EMA, Hanssen BH (1993) Calculating soil nutri-
(N23 P10 K5 3S 2MgO 0.3 Zn) on maize cultivation in the coastal
ent balances in Africa at different scales. Fert Res 35:227–285. doi:
savanna and the humid forest ecologies of Ghana. Int Res J Agric
10.1007/BF00750641
Sci Soil Sci 2(1):8–16
Nziguheba G, Tossah BK, Diels J, Franke AC, Aihou K, Iwuafor EN, Tittonnell P, Giller KE (2013) When yield gaps are poverty traps: the
Nwoke C, Merckx R (. (2009) Assessment of nutrient deficiencies in paradigm of ecological intensification in African smallholder agri-
maize in nutrient omission trials and long-term field experiments in culture. Field Crops Res 143:76–90. doi:10.1016/j.fcr.2012.10.007
the West African Savanna. Plant Soil 314(1-2):143 Van Asten PJA, Barro SE, Wopereis MCS, Defoer T (2004) Using farmer
Ojeniyi SO, Kayode GO (1993) Response of maize to copper and sulfur knowledge to combat low productive spots in rice fields of a
in tropical regions. J Agric Sci 120:295–299. doi:10.1017/ Sahelian irrigation scheme. Land degrad Develop 15:383–396.
S0021859600076450 doi:10.1002/ldr.619
Olivier K, Cherif M, Kone B, Emmanuel D, Firmin KK (2014) Growth, Vanlauwe B, Kihara J, Chivenge P, Pypers P, Coe R, Six J (2011)
yields and ratooning performance of lowland rice Nerica l14 as Agronomic use efficiency of N fertilizer in maize-based systems in
affected by different fertilizers. Ind J Sci Res Tech 2(2):18–24 sub-Saharan Africa within the context of integrated soil fertility
Oseni TO (2009) Growth and zinc uptake of sorghum and cowpea in management. Plant Soil 339:35–50. doi:10.1007/s11104-010-
response to phosphorus and zinc fertilization. World J Agric Sci 0462-7
5(6):670–674 Vanlauwe B, Descheemaeker K, Giller KE, Huising J, Merckx R,
Osiname OA, Kang BT, Schulte EE, Corey RB (1973) Zinc response of Nziguheba G, Wendt J, Zingore S (2015) Integrated soil fertility
maize (Zea mays L.) grown on sandy Inceptisols in western Nigeria. management in sub-Saharan Africa: unravelling local adaptation.
Agron J 65(6):875–877 Soil 1:491–508. doi:10.5194/soil-1-491-2015
Oyinlola EY, Chude VO (2010) Status of available micronutrients of the Voortman RL (2012) Micronutrients in agriculture and the world food
basement Complex rock-derived Alfisols in northern Nigeria system—future scarcity and implications. In: Scarcity of
Savanna. Trop Subtrop Agroecosyst 12:229–237 micronutrients in soil, food and mineral resources—background re-
Philibert A, Loyce C, Makowski D (2012) Assessment of the quality of port of the platform Agriculture, innovation and society (LIS)
meta-analysis in agronomy. Agric Ecosyst Environ 148:72–82. doi: Weil RR, Mughogho SK (1999) Sulfur nutrition of maize in four regions
10.1016/j.agee.2011.12.003 of Malawi. Agron J 92(4):649–656. doi:10.2134/agronj2000.
Rechiatu A, Ewusi-Mensah N, Abaidoo RC (2015) Response of Soybean 924649x
(Glycine max L.) to Rhizobia Inoculation and Molybdenum Wopereis CSM, Diagne A, Johnson DE, Seck PA (2013) Realizing
Application in the Northern Savannah Zones of Ghana. J Plant Sci Africa’s rice promise: priorities for action. In: Wopereis CSM,
3(2):64–70. doi:10.11648/j.jps.20150302.14 Johnson DE, Nourollah A, Tollens E, Jalloh A (eds) Realizing
Ryan J, Rashid A, Torrent J, Kwong YS, Ibrikci H, Sommer R, Bulent EE Africa’s rice promise. CAB International. CPI Group (UK) Ltd,
(2013) Micronutrient constraints to crop production in the Middle Croydon, pp 424–436
East–west Asia region: significance, research, and management. Yerokun OA, Chirwa M (2014) Soil and foliar application of Zinc to
Adv Agron 122:1–84. doi:10.1016/b978-0-12-417187-9.00001-2 maize and wheat grown on a Zambian Alfisol. Afr J Agric Res
Sahrawat KL, Wani SP, Pardhasaradhi G, Murthy KVS (2010) Diagnosis 9(11):963–970
of secondary and micronutrient deficiencies and their management