Genetic Gains of Grain Yield among the Maize Cultivars Released over a Century from the National Breeding Program of Zimbabwe

Abstract

Monitoring genetic gain is required in crop improvement programs in order to assess the effectiveness of breeding initiatives. The periodic measurement of genetic gain will quantify the efficiency of new technologies incorporated into the program. Here, a total of 24 cultivars (20 released from Zimbabwe’s National Breeding program (ZNBP) plus 4 released by the largest and oldest private seed company in Zimbabwe, Seed Co) over the period of 1900–2016, were evaluated across 10 locations in Zimbabwe. The testing locations represented agro-ecologies where maize is optimally grown and where maize production is under threat from climate change-induced abiotic stresses, particularly drought and heat stress, in Zimbabwe. The 24 cultivars were laid out across all the locations using the alpha (0.1) lattice design replicated two times with six incomplete blocks nested in each of the replicates. The genetic gains were estimated at 61 kg ha⁻¹ yr⁻¹, 25 kg ha⁻¹ yr⁻¹, 6 kg ha⁻¹ yr⁻¹, and 2 kg ha⁻¹ yr⁻¹ under optimal, random drought stress, heat stress, and managed drought stress conditions, respectively. The results were comparable with those from other studies where newly released cultivars yielded more than old cultivars. Overall, the results demonstrated that over a century, the ZNBP has been making significant progress in yield increments in its maize genotypes.

1. Introduction

Maize (Zea mays L.) is a food source for the most vulnerable people in sub-Saharan Africa (SSA), occupying 50% of land that is devoted to the production of cereals in over 50% of countries in this region [1]. According to Masuka et al. [2], maize provides food and earnings to millions of people in the developing world. Therefore, in order to meet these socio-economic demands, there is an urgent need to improve global maize productivity to sustain and improve economies. In agricultural research, breaking the cycle of poor maize yields and food insecurity is crucial. This cannot be achieved with a single intervention. An important strategy that can be used to increase maize yield is through the growing of enhanced, high-yielding cultivars designed for SSA’s rain-fed regions. The incorporation of stress tolerance in such varieties will help to mitigate the potential effects of climate change and variability on maize yield [3]. The new hybrids currently developed are selected simultaneously for their performance under optimal, random stress, drought, and heat stress conditions [2].

The Zimbabwe National Breeding Programme (ZNBP) has been in existence for more than 100 years. In the CIMMYT maize breeding program, which has a history of less than 60 years in Zimbabwe, significant genetic gains have since been reported [2]. Genetic gain is defined as the amount of increase in performance achieved through artificial selection per unit of time [4]. Crop improvement projects must monitor genetic gain in order to assess the effectiveness of selection procedures that would have been employed at various points in time. The efficiency of new technologies incorporated into breeding programs can thus be quantified by the periodic measurement of genetic gain. The evaluation of genetic gain is conducted in period (also, popularly known as era) studies [5,6]. An era study refers to the evaluation of crop genotypes released across different time points (or years) in a single trial. In these trials, the crop management will be uniform such that differences in performance can only be attributed to genetic differences [7]. A popular method for evaluating how genetic selection contributed to improvements in economically significant features, like grain production in maize, is to compare earlier and newer hybrids using period studies [8]. The main aim of plant breeders is to supply farmers with superior varieties which are high yielding and with high potential under stress conditions and are able to withstand adverse environmental conditions [1]. Every genotype might have a specific environment where it can perform best, but the successful new varieties must show high performance for yield and other essential agronomic traits, and their superiority must be reliable over a wide range of environments [9]. Several researchers reported different genetic gains in grain yield and secondary traits in maize [5]. The objective of this study was to determine genetic gains realised in the grain yield performance of maize cultivars released from the ZNBP between the years 1900–2016 under stress (i.e., managed heat stress and managed drought as well as random heat and drought stress) and optimal conditions in Zimbabwe. We hypothesize that maize cultivars from the later periods perform better in terms of GY than cultivars from earlier periods under both stress and non-stress production environments, and that gains in GY performance will be realised under both stress and non-stress conditions.

2. Materials and Methods

2.1. Plant Materials

A total of 24 maize cultivars were used in the experiment. The cultivars consisted of 20 genotypes released by the ZNBP and 4 check varieties from the largest and oldest private seed company in Zimbabwe, Seed Co (

Table 1. Cultivars and period of release.

Genotype	Source	Year of Release	Period
Hickory King (open-pollinated variety)	Crop Breeding Institute, ZNBP	1900	Before 1960
Salisbury White (open-pollinated variety)	Crop Breeding Institute, ZNBP	1900	1st period
Southern Cross (open-pollinated variety)	Crop Breeding Institute, ZNBP	1900
R200 (hybrid)	Crop Breeding Institute, ZNBP	1971	1960–1980
R201 (hybrid)	Crop Breeding Institute, ZNBP	1971	2nd period
R215 (hybrid)	Crop Breeding Institute, ZNBP	1971
SR52 (hybrid)	Crop Breeding Institute, ZNBP	1960
ZS107 (hybrid)	Crop Breeding Institute, ZNBP	1985	1981–2000
ZS108 (hybrid)	Crop Breeding Institute, ZNBP	1998	3rd period
ZS255 (hybrid)	Crop Breeding Institute, ZNBP	1998
ZS206 (hybrid)	Crop Breeding Institute, ZNBP	1988
ZS232 (hybrid)	Crop Breeding Institute, ZNBP	1985
ZS240 (hybrid)	Crop Breeding Institute, ZNBP	1992
SC513 * (hybrid)	Seed Co	1999
ZS263 (hybrid)	Crop Breeding Institute, ZNBP	2011	2001–2016
ZS265 (hybrid)	Crop Breeding Institute, ZNBP	2011	4th period
ZS269 (hybrid)	Crop Breeding Institute, ZNBP	2014
ZS271 (hybrid)	Crop Breeding Institute, ZNBP	2014
ZS273 (hybrid)	Crop Breeding Institute, ZNBP	2014
ZS275 (hybrid)	Crop Breeding Institute, ZNBP	2014
ZS242 (hybrid)	Crop Breeding Institute, ZNBP	2015
SC529 * (hybrid)	Seed Co	2012
SC533 * (hybrid)	Seed Co	2007
SC543 * (hybrid)	Seed Co	2009

* Check varieties.

). Seed Co hybrids were selected as checks because the seed house is one of the oldest after the national breeding program and has released superior cultivars that can represent the genetic gain that has been realised in the industry.

2.2. Testing Locations

Ten locations in Zimbabwe were designated as trial sites for the 2019–2020 summer and winter seasons. The testing locations represented agro-ecologies where maize is optimally grown and where maize production is under threat from climate-induced abiotic stresses, particularly drought and heat stress, in Zimbabwe (

Table 2. Climatic and geo-physical characteristics of testing sites that were used.

Site Name	Latitude	Longitude	Soil Type	Rainfall (mm)	Altitude (masl)	Natural Region	Management	Temperature (min~max, °C)
Chiredzi Research Station	21°58′ S	31°17′ E	Vertisols	<450 m	425 m	IV	Managed heat stress	12 min~35 max
Chiredzi Research Station	21°58′ S	31°17′ E	Vertisols	<450 m	425 m	IV	Managed drought stress	12 min~35 max
Chiredzi Research Station	21°58′ S	31°17′ E	Vertisols	<450 m	425 m	IV	Random drought stress	12 min~35 max
Kadoma Research Station	18°19′ S	29°53′ S	Ferralsols	550–800	1149 m	IIB	Optimal	20 min~28 max
Kadoma Research Station	18°19′ S	29°53′ S	Ferralsols	550–800	1149 m	IIB	Random drought	20 min~28 max
Art Farm	17°26′ S	31°05′ E	Ferralsols	750–1000	1480 m	IIA	Optimal	18 min~25 max
Ratray Arnold Research Station	17°14′ S	31°14′ E	Ferralsols	918	1300 m	IIB	Optimal	12.8 min~28.6 max
CIMMYT Harare	17°49′ S	31°01′ E	Ferralsols	750–1000	1480 m	IIA	Optimal	10 min~25 max
Bindura	17°30′ S	31°32′ S	Ferralsols	400–600	1070 m	IIB	Random drought	17 min~28 max
Muzarabani Estate	16°33′ S	31°17′ E	Ferralsols	<450	400 m	V	Managed drought stress	15 min~40 max

).

2.3. Experimental Design

Using an alpha (0.1) lattice design replicated twice with six incomplete blocks nested in each replicate, the 24 maize cultivars were established throughout all 10 sites. A cultivar was planted in a four-meter-long, one-row plot with an intra-row spacing of 0.25 m and an inter-row spacing of 0.75 m in each replicate.

2.4. Trial Management

Land was ploughed using a tractor to a depth of 30 cm. A disc harrow was used to break the clods to a fine tilth to obtain good seed–soil contact. Holing out of planting stations was performed manually using hand hoes. Planting was carried out manually by placing two seeds per hole, followed by thinning to one plant per planting station two weeks after emergence to give a final plant population of approximately 53,000 plants per hectare.

The trials were established under different environmental regimes. Plants were exposed to managed heat stress, managed drought stress, random drought stress, as well as optimal conditions in Zimbabwe (

Table 3. Traits and their methods of measurement.

Traits	Method of Measurement
Grain yield (GY)	Shelled grain weight per plot adjusted to 12.5% grain moisture and converted to tons per hectare.
Anthesis date (AD)	Number of days from planting to the time when 50% of the tassels of plants shedding pollen.
Silking date (SD)	Number of days from planting to the time when 50% of plants have emerged silks.
Anthesis silking interval (ASI) h	Calculated as the difference between days to 50% silking and 50% anthesis.
Cobs per plant (EPP)	Obtained by dividing the total number of cobs per plot by the number of plants harvested.
Number of plants (NP)	The total number of plants harvested per plot.
Root lodging (RL)	Percentage of plants that show stems that are inclining by more than 45°.
Plant height (PH)	Average height of five randomly selected plants in centimetres from the ground level to the node bearing of the flag leaf using graduated measuring stick.
Cob height (EH)	Average height of the upper most cobs of the same plants used for plant height measurement from ground level to the node bearing the cob, measured in cm using graduated measuring stick.
Leaf senescence (Sen)	The proportion of dead leaves to that of the whole plant expressed as a percentage using a scale of 1–10. A score of 10 is when 100% or all the leaves have senesced while 1 is when 10% of leaves senesce.
Moisture (MOI)	Water content percentage of grain as measured at harvest.
Tassel blasting (TB)	Tassel blast is the drying of the complete tassel (or most of it) without pollen extrusion. It will be recorded at 1–2 weeks after tassel emergence. Tassel blast is recorded as the percentage of plants in a plot with tassel blast symptoms.
Leaf firing (Lf)	Recorded as the percentage of plants in a plot with leaf symptoms, i.e., leaves burned starting from the top of the plant and progressing downward due to heat stress. This will be recorded 1–2 weeks after anthesis.

). Briefly, optimum trials were rain-fed and irrigation water was applied to supplement the rains. Irrigation was conducted at planting and emergence. However, supplementary irrigation was applied as needed to avoid drought stress. For managed drought trials, stress was imposed by stopping irrigation at around three weeks before anthesis until four weeks after anthesis [10]. Rescue irrigation was then applied only when necessary to avoid total crop loss. For managed heat trials, the planting was carried out from mid-July to mid-August to coincide flowering with the highest temperatures (above 35 °C) that are normally received between 10 and 20 October in Chiredzi. In the heat experiment, irrigation was supplied normally to physiological maturity. For random drought stress evaluations, the establishment of the trials was performed in drought-prone locations known in history to experience mid-season or terminal droughts while maize is growing. Random drought sites were rain-fed. For managed drought and heat stress trials, rainfall was least expected because the sites used have long-term meteorological records of being dry during winter. Compound D fertilizer (7%N: 14%PO₅: 7% K₂O) was applied as a basal dressing at a rate of 400 kg/ha for optimum trials. Top dressing was performed using ammonium nitrate (34.5% N) where half (200 kg/ha) was applied at four weeks after crop emergence (WACE) and at eight WACE for the optimum trials. For drought and heat trials, 400 kg/ha of compound D fertilizer was applied, and 120 kg/ha of nitrogen was split applied at four weeks and eight weeks after planting. All other standard agronomic management practices were performed at all sites.

2.5. Data Collection

Comprehensive data was collected following standard procedures used by CIMMYT [11], as shown in Table 3.

2.6. Statistical Analysis

Analysis of variance (ANOVA) for grain yield and all other secondary traits for single-site and across-site analysis was conducted using GenStat software version 18 [12]. To estimate the annual genetic gain achieved in GY, the mean values of GY were regressed against the year of cultivar release using ggplot2 v3.3.3 R- software [13]. Principal component and biplot analyses were performed using GenStat software version 18.2.0, VSN International [12]. Means were separated using the Fisher’s Protected LSD at 5% significance level.

3. Results

3.1. Grain Yield Performance under Optimal and Stress Conditions

There were highly significant differences (p < 0.001) among hybrids for across-site analysis. Genotypic variation was higher than G x E variation for the across-site analysis, random drought stress site, and optimal management site. Heritability was high across sites (89.9%), under optimal management (81.9%), and under a random drought stress environment (74.8%). However, heritability was low for managed heat stress (46.6%) as well as managed drought stress environments (13.2%) (

Table 4. Analysis of variance for grain yield data collected in trials established under diverse stress and non-stress conditions in Zimbabwe during the 2019–2020 summer and winter seasons.

	Across-Site Analysis		Managed Drought Stress		Random Drought Stress		Optimal Management		Managed Heat Stress
Source of variation	DF	MS	DF	MS	DF	MS	DF	MS	DF	MS
Site	9	222.8 ***	1	30.2 ***	2	57.4 ***	2	63.7 ***	1	6.3
Replication (Site)	10	3.828	2	0.2375	3	5.513	3	1.23	2	8.788 ***
Block (Replication × Site)	100	3.368 ***	20	1.9615	30	3.026	30	5.205	20	2.5315
Cultivar	23	17.483 ***	23	2.396	23	4.98 ***	23	16.6 ***	23	2.5349
Cultivar × Site	207	2.422 ***	23	1.610	46	1.313	46	3.479	23	1.090
Heritability		0.8999		0.1326		0.7487		0.8197		0.4669
Genotypic variance		1.0548		0.0795		0.797		2.8056		0.2447
Error variance		1.1759		0.6114		1.0434		2.1279		0.7824
G x E variance		0.7807		0.8368		0.281		0.7875		0.2982
CV		34.637		45.279		56.087		23.662		40.453
LSD		2.1254		1.5325		2.002		2.8591		1.7336
Grand mean		3.1308		1.7268		1.8212		6.1647		2.1865
Minimum		31		0.3		−0.54		0.69		0.4
Maximum		163		2.57		7.22		10.58		6.45

DF—degrees of freedom; MS—Mean square values; ***—shows values which are significant at p < 0.001; ns—not significant.

).

3.2. Estimated Genetic Gains under Optimal and Stress Conditions

Under optimal conditions, GY increased significantly (p < 0.001) from the 1st period to the 4th period. The trend showed a positive and linear increase in genetic yield gains. The highest-yielding cultivars were SC543 (check), ZS265, ZS269, and ZS108 and all these are from the 4th period except for ZS108, which is from the 3rd period. The genetic gain was estimated at 61 kg ha⁻¹ yr⁻¹ (Figure 1). Under managed drought stress conditions, the newest cultivars (i.e., 4th-period cultivars) exhibited the greatest positive linear and significant (p = 0.015) increase in GY performance across the periods. The highest-yielding cultivars were SC543 (check), ZS265, ZS269, and ZS263 and all these are from the 4th period. The genetic gain was estimated at 2 kg ha⁻¹ yr⁻¹ (Figure 2). Under managed heat stress conditions, again, the newest hybrids exhibited the greatest positive linear and significant increase (p = 0.077) in GY performance, having an average yield ranging from 2.5 t ha⁻¹ to 2.8 t ha⁻¹. The highest-yielding cultivars were ZS263, ZS269, ZS273, and SC543 (check) and all these are from the 4th period. The genetic gain was estimated at 6 kg ha⁻¹yr⁻¹ (Figure 3). Under random drought stress conditions, a positive and significant (p < 0.001) linear increase in GY performance was observed across the periods. The highest-yielding cultivars were ZS265, ZS271, ZS263, and SC543 (check) and all these cultivars again are from the 4th period. A positive genetic gain was estimated at 25 kg ha⁻¹yr⁻¹. The new cultivars yielded better than the old cultivars on average (Figure 4). Lastly, across the optimal and stress sites, GY increased significantly (p < 0.001) from the old to the new breeding periods. The newest hybrids from the 4th period exhibited the greatest linear increase. The highest-yielding cultivars under all the environments were ZS265, ZS269, ZS263, and ZS108 and all these are from the 4th period except for ZS108, which is from the 3rd period. The most recently released varieties yielded more on average from 3.6 t ha⁻¹ to 4.9 t ha⁻¹ with a mean yield of 37 kg ha⁻¹ yr⁻¹. The genetic gain was estimated at 37 kg ha⁻¹ yr⁻¹ (Figure 5).

4. Discussion

Genetic gain analysis helps to assess the progress and success of a breeding program. Genetic gains realised over time in a breeding program are helpful for evaluating the effectiveness of breeding methodologies. One could argue that having access to this kind of information aids breeders in redefining and implementing effective breeding techniques. These investigations demand a lot of resources and they come at a hefty cost [14]. Since 1905, a variety of strategies have been employed, including hybridization. Public studies on maize hybridization began in 1910, and in the 1930s, the production of hybrid maize grew quickly [15]. In the second decade of the twenty-first century, the science and art of plant breeding underwent a remarkable 120-year evolution that culminated in a revolution [16]. A number of new strategies, concepts, tools, and techniques have been developed from time to time. Traditional plant breeding has been crucial in the development of several variations in nearly all crops during this time, especially with the rediscovery of laws of inheritance [16]. This includes rice that was essential to the so-called “green revolution”, high-yielding wheat cultivars, and maize hybrids. During this time, traditional breeding techniques included polyploidy, hybridisation, mutation, and selection [16]. This study was to determine genetic gains realised in the grain yield performance of maize cultivars released from the ZNBP between the years 1900–2016, under stress (i.e., managed heat stress and managed drought as well as random heat and drought stress) and optimal conditions in Zimbabwe. The study aimed to test the hypotheses that gains in GY performance were realised under both stress and non-stress conditions over more than a century of maize breeding research in Zimbabwe.

The results show that all sources of variation except for replication/site were highly significant (p < 0.001). For the genotypes, it means there was great genetic variation that enabled selection to be carried out. Environments were significantly different, hence the genotypes were exposed to different environments and their reactions to different conditions can thus be effectively assessed. This necessitates the exploitation of G x E, where specific environmental recommendations are issued rather than broad recommendations. When the G x E interaction is highly significant it should not be ignored, instead, different methods should be considered and applied for further understanding of the G x E interaction and subsequent exploitation [17]. Error can be reduced by using strategies such as good weed management, proper data collection, and data entry, gap filling, uniform fertilizer application, and pest control.

It is a known fact that the contribution of genetics to the overall yield increase has been analysed by numerous authors [18]. For the period from 1901 to 2016, genetic gain in grain yield by the ZNBP was estimated at 61 kg ha⁻¹ yr⁻¹ under optimal management, which was higher than under random drought stress (25 kg ha⁻¹ yr⁻¹), managed heat stress (6 kg ha⁻¹ yr⁻¹), and managed drought stress environments (2 kg ha⁻¹ yr⁻¹). Newer hybrids produced noticeably more than previous hybrids produced on average. For the national program in Zimbabwe, this is positive progress. These results show that gains in GY performance have been realised under both stress and non-stress conditions as hypothesised. The results in this study show almost the same trend as observed by Masuka et al. [2], where the genetic gains were estimated at 109. 4 kg ha⁻¹ yr⁻¹, 32.5 kg ha⁻¹ yr⁻¹, and 22.7 kg ha⁻¹ yr⁻¹ under optimal management, a managed drought stress environment, and a random drought stress environment, respectively. The results of the study show that the genetic gain in yield was higher than the one reported in West and Central Africa by Badu-Apraku et al. [6], which was 61 kg ha⁻¹yr⁻¹ compared with 40 kg ha⁻¹yr⁻¹ under optimal management and 25 kg ha⁻¹yr⁻¹ under random drought stress compared to 13.5 kg ha⁻¹yr⁻¹, which helps to reflect the higher yield levels presented in this study. Under optimum conditions, the yield gain is similar to other US experiments where Duvick [8] reported gains of 77 kg ha⁻¹ yr⁻¹, and in China, where 60 kg ha⁻¹ yr⁻¹ was reported. Duvick [8] reported linear genetic gain ranges from approximately 65 kg ha⁻¹ yr⁻¹ to 75 kg ha⁻¹ yr⁻¹ in the US for the past 70 years of breeding, and this is comparable to the results in this study. The gains across non-stress and stress conditions showed a positive and linear trend. Similar results, with a mean of 66 kg ha⁻¹ yr⁻¹, were reported by Duvick [8], showing positive and linear gains in genetic yield. The values ranged from 33 to 92 kg ha⁻¹ yr⁻¹.

Despite positive genetic gains across all the environments, relatively lower genetic gains were realised under stressful environments than in optimum environments. This may be because stressful surroundings include certain variables that limit crop growth and cause crop yield gains to be less than ideal. On the other hand, optimal conditions offer the best conditions for crop growth. The other reason could be that little attention has been given by breeders to breeding for tolerance to stressful environments. This calls for more effort in that direction in order to sustain maize productivity under harsh climatic conditions. This implies that there is more need to breed for abiotic stress tolerance in the wake of climate change in order to realise comparable genetic gains. Duvick [8], from their findings, also stated that hybrids representing the period from 1930 to 1991 showed a linear gain for grain yield of 74 kg ha⁻¹ yr⁻¹. The genetic gain in the maize crop is due to improvement in stress tolerance [8]. The same season length was used for the newer and the old hybrids, which means the hybrids can still grow in the same length of season, yet the newer hybrids yielded better. According to Duvick [8], the selection was started by farmers but they targeted yield, and yet the selection is now targeted for both yield and stress tolerance. This could have contributed to the higher yields in the newer hybrids as compared to the older hybrids. In this current study, new hybrids showed better stress tolerance than old hybrids as had been hypothesised, which shows the success of the breeding program in its strategies for stress tolerance breeding. Overall, the results in this study show the success of breeding in the ZNBP in Southern Africa.

The results obtained showed that genetic gains in grain yield performance were realised under both stress and non-stress production conditions. However, more genetic gains were realised under non-stress conditions as compared to stressed conditions. This could be due to the reasons that optimum conditions provide the most suitable conditions for crop growth whilst stressed environments have some restricting factors affecting crop growth, resulting in lower yield gains being realised. The other reason could be that little attention has been given by breeders to breeding for tolerance to stressful environments. This calls for more efforts in that direction in order to sustain maize productivity under harsh climatic conditions.

5. Conclusions

The study showed a significant estimated gain in yield of 61 kg ha⁻¹ yr⁻¹ under optimum conditions for the past 100 years. The gains realised are higher than gains recorded in other studies. The increase is a result of improved heat and drought stress tolerance in the ZNBP varieties released in the latest periods. This shows the success of the breeding program in its breeding strategies. The new hybrids, overall, yielded better than the old hybrids under the same conditions and within the same season length as the old hybrids and the commercial checks. The ZNBP should put in place more strategies to allow the dissemination of the new hybrids which are drought- and heat-tolerant to reach farmers to improve production, thereby improving food security.