Introduction
Maize (Zea mays L.) is the most important cereal crop among the resource-poor small-scale farmers in West Africa (WA) and ranks third after rice and wheat in the world (Olaniyan, Reference Olaniyan2015). It has rapidly gained popularity due to its high potential as source of calories in human diets and livestock feeds and raw materials for industrial products in the sub-region. Maize is the most widely grown cereal crop in Nigeria (FAOSTAT, 2016) due to its high productivity, wide adaptation, relative ease of cultivation, processing, storage, transportation, and income generation. Also, maize production is stimulated with the availability of high-yielding, pest-and disease-resistant cultivars. The availability of different maturity groups of maize varieties that can be consumed either as green maize or grain has helped to fill the hunger gap in the savannas of the sub-region in July when all other food reserves are depleted after the long dry period (Badu-Apraku et al., Reference Badu-Apraku, Oyekunle, Menkir, Obeng-Antwi, Yallou, Usman and Alidu2013).
Despite the potential of maize as a staple crop in WA, its production is hindered by several biotic and abiotic factors, including recurrent drought, poor soil fertility, and Striga hermonthica (Delile) Benth. parasitism, maize streak virus, and turcicum leaf blight, among others. Drought is the major abiotic factor contributing to maize yield loss in the lowland savanna belt of the sub region (Badu-Apraku et al., Reference Badu-Apraku, Fakorede, Menkir, Kamara, Akanvou and Chabi2004; NeSmith and Ritchie, Reference NeSmith and Ritchie1992). Drought causes major reduction in maize productivity, especially during the most drought-sensitive stages of maize growth and development. Yield loss as a result of drought at flowering and grain-filling periods ranged from 39 to 91% (Badu-Apraku et al., Reference Badu-Apraku, Lum, Akinwale and Oyekunle2011b; Badu-Apraku and Oyekunle, Reference Badu-Apraku and Oyekunle2012; NeSmith and Ritchie, Reference NeSmith and Ritchie1992; Oyekunle and Badu-Apraku, Reference Oyekunle and Badu-Apraku2014). The annual yield loss due to drought is about 24 million tons, which is equivalent to 17% of a normal year’s production in the developing world (Edmeades et al., Reference Edmeades, Bolanõs, Lafitte and Wilkinson1992).
Genetic gain studies comparing old and new cultivars have been conducted routinely in the temperate zones in an effort to understand how genetic selection has shaped important traits such as grain yield in maize (Campos et al., Reference Campos, Cooper, Edmeades, Löffler, Schussler and Ibañez2006; Wang et al., Reference Wang, Ma, Li, Bai, Liu, Liu, Tan, Shi, Song, Carlone, Bubeck, Bhardwaj, Jones, Wright and Smith2011). Maize breeders in developed countries have measured breeding progress by comparing the performance of cultivars developed and released over a long period of time in the same environments (Badu-Apraku et al., Reference Badu-Apraku, Oyekunle, Menkir, Obeng-Antwi, Yallou, Usman and Alidu2013; Tollenaar, Reference Tollenaar1989). Similar studies have also been carried out in other crops such as wheat (Triticum aestivum L.) (Lopes et al., Reference Lopes, Reynolds, Manes, Singh, Crossa and Braun2012; Xiao et al., Reference Xiao, Qian, Wu, Liu, Xia, Ji and He2012), oats (Avena sativa L.), and soybean [Glycine max (L.) Merr.] (Tefera et al., Reference Tefera, Kamara, Asafo-Adjei and Dashiell2009). Generally, the studies reported demonstrated that the varieties developed in later breeding eras (2001–2006 and 2007–2010) are superior in terms of grain yield and other agronomic traits. Similar studies have also been conducted in the tropics; Kamara et al. (Reference Kamara, Menkir, Fakorede, Ajala, Badu-Apraku and Kureh2004) reported a genetic gain of 0.4% per year for late-maturing maize cultivars released from 1970 to 1999 in the Nigerian savannas. In addition, Badu-Apraku et al. (Reference Badu-Apraku, Oyekunle, Menkir, Obeng-Antwi, Yallou, Usman and Alidu2013) reported genetic gain of 1.1% yr−1 under drought and 1.33% yr−1 under optimum growing conditions for early maize cultivars developed during three breeding eras from 1988 to 2007.
During the last two decades, the development of maize varieties with tolerance to drought, nutrient use efficiency, resistance to Striga hermonthica and major foliar diseases has been a major focus of the maize improvement program at International Institute of Tropical Agriculture (IITA) and Institute for Agricultural Research (IAR), Samaru. The maize breeders in IAR in collaboration with IITA conducted several researches to develop and release early, extra-early, and intermediate/late maturing maize varieties with high yield potential and resistance or tolerance to biotic and/or abiotic stresses for Nigerian farmers. These cultivars were developed, tested, and released under different environmental conditions over a period of times. However, no direct comparisons of grain yield potential and other agronomic traits of the released cultivars have been made under drought and well-watered conditions to justify the huge efforts and investments in maize breeding. It is therefore important to assess genetic gain in grain yield and associated changes in agronomic traits of maize cultivars released during the last two decades in order to assess progress made in breeding for improved maize varieties and to identify traits of potential value for accelerating genetic gains in future breeding as well as valuable information for the seed enterprise in Nigeria. The objectives of the study, therefore, were (i) to assess genetic gains in grain yield and other agronomic traits of released maize cultivars from 2001 to 2016 and (ii) to determine the relationship between grain yield and other agronomic traits of released maize cultivars.
Materials and Methods
Genetic materials and experimental procedures
Twenty-three maize cultivars registered and released in the IAR, Samaru from 2001 to 2016 were used for the present study. The 23 released cultivars along with one experimental variety (check) were evaluated under induced drought and well-watered conditions at Zaria (northern Guinea savanna, 11º11’N, 7º38’E, 640 m a.s.l., 1200 mm annual rainfall) and Kadawa (Sudan savanna, 12o00’N, 8o22’E, 580 m a.s.l., 800 mm annual rainfall) during 2015/2016 and 2016/2017 dry seasons, using 6 x 4 lattice design with three replications. The two experimental conditions were irrigated twice every week with furrow irrigation system. Sufficient irrigation water was channeled from the water source into the furrow of each planted ridge. In the drought experiment, the induced drought was achieved by withdrawing irrigation (furrow irrigation) water from 35 days after planting until maturity so that the maize plants relied on stored water in the soil for growth and development. In the well-watered experiment, the plants were irrigated throughout the growth period. The drought and well-watered experiments were planted in two adjacent blocks in the same field that received the different irrigation treatments during the dry seasons. Each experimental unit consisted of two-row plots 5 m long, with inter- and intra-row spacing of 0.75 m and 0.40 m, respectively. Three seeds were planted per hill, and the resulting maize plants were thinned to two per stand about 2 weeks after emergence to give a final plant population density of 66,000 plants ha−1. All trials received 60 kg NPK ha−1 in form of NPK (15–15–15) 2 weeks after planting (WAP). An additional 60 kg N ha−1 was supplied (top-dressed) at 5 WAP. Weeds were controlled with herbicides and/or manually when necessary.
Data collection and statistical analyses
Data were collected from each plot on the following traits: days to 50% anthesis and days to mid-silk were recorded as the number of days from planting to when 50% of plants shed pollen and had emerged silks, respectively. Anthesis-silking interval (ASI) was calculated as difference between number of days to mid-silk and days to 50% anthesis. Plant and ear heights were measured as the distance from the base of the plant to the first tassel branch and from the base to the node bearing the upper ear, respectively. Plant aspect was rated on scale of 1–5, where 1 is for plants with minimal reduction in height, ear size, low ear placement, resistance to foliar diseases, and lodging, and 5 is for plants with severally stunted growth, small ears, susceptible to foliar diseases, and lodging (Supplementary Material Table S1 available online at https://doi.org/10.1017/S0014479719000048). Ear aspect was scored from 1–5 scale, where 1 is clean, uniform, and large ears, and 5 is rotten, variable, and small ears (Supplementary Material Table S1). Husk cover was also rated on a scale of 1–5, where 1 is husks tightly arranged and extended beyond the ear tip, and 5 is very loosely arranged husk with ear tip exposed (Supplementary Material Table S1). Number of ears per plant (EPP) was calculated as number of ears harvested divided by the number of plants at harvest. Ears harvested from each plot were shelled to determine percent moisture and grain weight. Grain yield adjusted to 150 g kg−1 moisture was computed from grain weight.
Combined analyses of variance for all the traits measured under drought and well-watered conditions were performed separately. A random model of the PROC GLM in SAS was used (SAS Institute, 2002), in which cultivars, location–year combination (environments), block, and replications were considered as random factors. The linear model for the combined ANOVA is as follows:
$${Y_{bklmi}} = {\mu _i} + {\rm{ }}{E_{ki}} + {\rm{ }}B{\left( {RE} \right)_{b\left( {kl} \right)i}} + R{\left( E \right)_{l\left( k \right)i}} + {G_{mi}} + {\rm{ }}G{E_{kmi}} + {\rm{ }}{\varepsilon _{bklmi}}$$
(1)
where Ybklmi is the observed measurement of trait i of m genotype within l replicate, in k environment, b block within l replicate and k environment, μi is mean effect, Eki is the effect of environment k on trait i, B(RE)b(kl)i is the effect of block b within replicate l and environment k on trait i, R(E)l(k)i is the effect of replication l within environment k on trait i, Gmi is the effect of genotype m on trait i, GEkmi is the effect of the interaction between genotype m and environment k on trait i, and εbklmi is the experimental error effect associated with genotype m and block b within replication l and environment k on trait i.
Pearson correlation coefficients were computed between grain yield and other measured traits of maize cultivars under drought and well-watered conditions. In addition, the mean values of yield under drought were regressed on yield under well-watered conditions and vice-versa. The relationship between cultivar grain yield and year of released (expressed as number of years since 2001) was determined using regression analysis. The mean grain yield of the maize cultivars under drought and well-watered conditions was used as the dependent variable and regressed on the year of released as independent variables to obtain regression coefficient (b-value). The b-value was then divided by the intercept (a) and multiplied by 100 to obtain the relative genetic gain per year (Badu-Apraku et al., Reference Badu-Apraku, Oyekunle, Menkir, Obeng-Antwi, Yallou, Usman and Alidu2013). Both correlation and regression analyses were carried out using SAS version 9.2 (SAS Institute, 2002). Grain yield reduction due to drought was calculated as follows:
$${\rm Yield \ reduction} = {{\left( {{Y_{\rm{W}}}{\rm{-}}{Y_{\rm{D}}}} \right){\rm{ }} \times {\rm{ 100}}} \over {{Y_{\rm{W}}}}}{\rm{ }}$$
(2)
where Y W = yield under well-watered conditions and Y D = yield under drought.
Results
Analysis of variance and mean performance of maize cultivars
The combined analysis of variance for grain yield and other agronomic traits of the 24 maize cultivars revealed that environment mean squares were significant (P < 0.01) for all measured traits, except husk cover and EPP under drought and well-watered conditions (Table 1). Similarly, cultivar mean squares were significant for all measured traits, except ASI, husk cover, and EPP under drought, and ASI, husk cover, and ear aspect under well-watered conditions (Table 1). In contrast, cultivar × environment interaction mean squares were significant only for grain yield and plant height under drought and days to anthesis and mid-silk, and ASI under well-watered conditions (Table 1).
Analysis of variance of yield and other agronomic traits of released maize cultivars evaluated under drought and well-watered conditions at Zaria and Kadawa during 2015/2016 and 2016/2017 dry seasons
*, ** Significant difference at P < 0.05 and P < 0.01 levels, respectively. ASI = anthesis-silking interval; EPP = number of ears per plant.
The mean grain yield of the cultivars ranged from 2251 kg ha−1 for SAMMAZ 31 to 4938 kg ha−1 for SAMMAZ 19 with an average of 3131 kg ha−1 under drought and from 3082 kg ha−1 for SAMMAZ 37 to 5689 kg ha−1 for SAMMAZ 51 with an average of 4373 kg ha−1 under well-watered conditions (Table 2). The highest yielding cultivar SAMMAZ 19 out-yielded the check by 41% under drought and SAMMAZ 51 out-yielded the check by 28% under well-watered conditions. The grain yield reduction under drought ranged from 8.8% for SAMMAZ 37 to 48.1% for SAMMAZ 15 when compared with grain yield under well-watered conditions. On average, grain yield of cultivars under drought was 71.6% of the yield obtained under well-watered conditions.
Grain yield and other agronomic traits of released maize cultivars evaluated under drought (DS) and well-watered (WW) conditions at Zaria and Kadawa during 2015/2016 and 2016/2017 dry seasons
ASI = anthesis-silking interval; EPP = number of ears per plant.
Correlation between grain yield and other agronomic traits
Results of Pearson’s correlation analysis revealed significant positive correlations between grain yield and EPP (r = 0.74**) but negative correlations with days to anthesis (r = −0.20**) and mid-silk (r = −0.37**), plant aspect (r = −0.53**), and ear aspect (r = −0.63**) under drought (Table 3). On the other hand, grain yield had significant positive correlations with plant height (r = 0.32**), ear height (r = 0.35**), and EPP (r = 0.56**) but negative correlations with days to anthesis (r = −0.37**) and mid-silk (r = −0.40**), plant aspect (r = −0.67**), and ear aspect (r = −0.70**) under well-watered conditions. Plant height had significant positive correlations with ear height under drought (r = 0.65**) and well-watered conditions (r = 0.72**). Similarly, days to anthesis had significant positive correlations with days to mid-silk under drought (r = 0.82**) and well-watered conditions (r = 0.93**) (Table 3). Plant aspect had significant positive correlations with ear aspect under drought (r = 0.65**) and well-watered conditions (r = 0.61**). The results of regression analysis of grain yield under drought and well-watered conditions showed positive predictive relationship between one level and the other, although performance under drought predicted performance under well-watered conditions better than vice-versa (Figure 1).
Correlation coefficients of grain yield and other agronomic traits of released maize cultivars evaluated under drought and well-watered conditions at Zaria and Kadawa during 2015/2016 and 2016/2017 dry seasons
*, **Significant difference at P < 0.05 and P < 0.01 levels, respectively. ASI = anthesis-silking interval; EPP = number of ears per plant.
Regression of (a) grain yield of well-watered conditions on grain yield of drought environments and (b) grain yield of drought environments on grain yield of well-watered conditions.
Genetic gain in grain yield of released maize cultivars
Substantial increase in the grain yield of released maize cultivars was observed during the breeding period (Table 4, Figure 2). The average rate of increase in grain yield was 50 kg ha−1 yr−1, corresponding to a genetic gain of 1.93% yr−1 under drought (Table 4). Similarly, the average rate of increase in grain yield was 70 kg ha−1 yr−1, corresponding to a genetic gain of 1.93% yr−1 under well-watered conditions. Plant aspect had a genetic gain of −0.94% yr−1, −0.36% yr−1 for ear aspect, 0.55% yr−1 for plant height, and 0.75% yr−1 for EPP under drought. On the other hand, ear aspect had a genetic gain of −1.76% yr−1, −0.38% yr−1 for ear height, −0.29% yr−1 for days to anthesis, −0.21% yr−1 for days to mid-silk, and 0.69% yr−1 for EPP under well-watered conditions.
Relative genetic gain, coefficient of determination (R 2), slope (a), and regression coefficients (b) of grain yield and other agronomic traits of released maize cultivars evaluated under drought and well-watered conditions at Zaria and Kadawa during 2015/2016 and 2016/2017 dry seasons
ASI = anthesis-silking interval; EPP = number of ears per plant.
Regression of grain yield of released maize varieties under drought and well-watered conditions on years of release.
Discussion
The present study provided an opportunity to assess the performance of the maize cultivars released in Nigeria within the last two decades under drought and well-watered conditions. The presence of significant difference among environments for most of the measured traits indicated that the test environments were unique under both research conditions in identifying high-yielding cultivars. The presence of significant difference among the cultivars for grain yield and most other traits under both drought and well-watered conditions indicated genetic variability among the cultivars released during the two decades of maize breeding. The existence of variability among the cultivars would allow significant progress from selection for improvements in most of the measured traits and identification of source of genetic materials for development of inbred lines and populations. In addition, the differences observed among the cultivars would allow identification of superior genotypes for varietal replacement and commercialization in the country. The information generated in the present study is invaluable in guiding the small- and medium-scale seed companies and farmers in Nigeria for adoption and commercialization of improved maize cultivars. The differential response of genotypes to varying environmental conditions constitutes a major challenge in the identification of superior maize cultivars for narrow or wide adaptation. The significant cultivar × environment interaction observed for grain yield and plant height under drought (Table 1) indicated that the expression of these traits would not be consistent in varying test environments. This result suggests the need for extensive testing of the cultivars in multiple environments for identification of genotypes with consistent performance under varying resource availability, such as water. In contrast, the lack of significant cultivar × environment interaction for grain yield under well-watered conditions (Table 1) indicated that the trait was not affected by cultivar × environment interaction and hence the expression of these traits would be consistent in varying test locations. The grain yield reduction of 8.8–48.1% observed among the cultivars fell within the ranged reported by Oyekunle and Badu-Apraku (Reference Oyekunle and Badu-Apraku2014), who reported yield reduction of 4–84% among the early-maturing inbreds evaluated under drought and well-watered conditions.
Information on the relationships among traits is vital for designing effective breeding programs for maize improvement. The significant correlations observed between grain yield and EPP, days to anthesis and mid-silk, plant and ear aspects under both drought and well-watered conditions (Table 3) were desirable for improvement of grain yield under both research conditions. The presence of significant correlations between grain yield and days to anthesis and mid-silk indicated that later maturing cultivars tend to give higher yields than those earlier maturing. On the other hand, significant correlations between grain yield and EPP, and plant and ear aspects under both drought and well-watered conditions (Table 3) indicate the possibility of using secondary traits (plant and ear aspects, and EPP) in improving grain yield and thus, justifying the inclusion of the traits in the selection index for the identification and improvement of drought-tolerant cultivars. These results are in agreement with the findings of Badu-Apraku et al. (Reference Badu-Apraku, Akinwale, Ajala, Menkir, Fakorede and Oyekunle2011a, Reference Badu-Apraku, Akinwale, Franco and Oyekunle2012). However, the lack of significant correlation between grain yield and plant and ear heights under drought was in disagreement with previous research (Badu-Apraku et al., Reference Badu-Apraku, Akinwale, Ajala, Menkir, Fakorede and Oyekunle2011a, Reference Badu-Apraku, Akinwale, Franco and Oyekunle2012, Reference Badu-Apraku, Oyekunle, Menkir, Obeng-Antwi, Yallou, Usman and Alidu2013). The presence of significant correlations between pair of traits such as plant and ear heights, days to anthesis and mid-silk, and plant and ear aspects under both water conditions (Table 3) indicated that improving one of the traits would lead to improvement in the other. This is advantageous in breeding and would reduce cost of measuring two different traits that provide similar information. The positive predictive relationship observed between grain yield under drought and well-watered conditions (Figure 1) indicated that performance under drought could be utilized in predicting the performance under well-watered conditions better than vice-versa.
An important objective of the present study was to assess the progress made in maize breeding during the last two decades. It is important to determine the magnitude of the increase in grain yield and genetic gain so as to effectively assess progress made within a period of time of breeding. In fact, substantial increase in the grain yield of released maize cultivars was observed under drought and well-watered conditions during the breeding period. The genetic gain in grain yield was 1.93% yr−1, regardless of water condition (Table 4). Such gains are higher than 1.1 and 1.33% yr−1 reported by Badu-Apraku et al. (Reference Badu-Apraku, Oyekunle, Menkir, Obeng-Antwi, Yallou, Usman and Alidu2013) for 50 early-maturing open-pollinated maize cultivars developed between 1988 and 2007. Similarly, the genetic gains obtained here are substantially higher than the 0.41% yr−1 reported by Kamara et al. (Reference Kamara, Menkir, Fakorede, Ajala, Badu-Apraku and Kureh2004) for late-maturing maize cultivars developed from 1970 to 1999 in the West African savannas. The genetic gain in grain yield under drought was associated with improvement in plant and ear aspects, plant height, EPP, and earliness, whereas gain in grain yield under well-watered conditions was associated with improvement in ear aspect, good standability, ear height, EPP, and earliness (Table 4). Our results suggest that the breeding strategies including recurrent selection, backcrossing, hybridization, and selection indices utilized in developing improved maize cultivars over a breeding period of 16 years were effective. The results clearly indicated that new cultivars possess favorable genes that make them better performing than the old cultivars. In addition, accumulation of new favorable alleles through rapid breeding cycles is one of the possible scenarios that substantially boost the rate of gain, and modern crop breeding and advances in management practices have contributed substantially to the annual gain in crop productivity.
In conclusion, we found that the average rate of increase in grain yield varied between 50 (drought) and 70 kg ha−1 yr−1 (well-watered conditions), corresponding to a genetic gain of 1.93% yr−1 in both water regimes. Grain yield had significant correlations with all measured traits under both water conditions, except husk cover, and plant and ear heights under drought. The substantial genetic gain in grain yield was associated with improvement in agronomic traits and SAMMAZ 19 and SAMMAZ 51 were the highest yielding cultivars under drought and well-watered conditions, respectively. Those cultivars should serve as genetic source for development of inbred lines, synthetic varieties, and populations.
Supplementary materials
For supplementary material for this article, please visit https://doi.org/10.1017/S0014479719000048.
Acknowledgements
The authors are grateful to the staff of IAR’s Maize Improvement Program for technical support. The financial support of STMA Project and IAR for this research is also gratefully acknowledged.
Financial support
STMA-IITA project and IAR, Samaru.
Disclosure statement
No potential conflict of interest.
