Sweet corn (Zea mays L. var saccharata), is a type of maize with kernels that are sweet as a result of high sugar content which when consumed in the fresh immature stage, has high levels of total sugars than field maize, rich in fibre, minerals, certain vitamins (Tracy, 1997; Lertrat and Pulam, 2007), and significant antioxidant properties (Dewanto et al., 2002). Sweet corn has its origin from a mutation that influences carbohydrate biosynthesis in the endosperm. In the genome of sweet corn, at least one of the eight mutant genes preventing the conversion of sugars to starch is present. These genes, which include shrunken-2 (sh2) on chromosome 3; brittle (bt), and amylose extender (ae) on chromosome 5; sugary enhancer (se), sugary (su), and brittle-2 (bt2) on chromosome 4; dull (du) on chromosome 10; and waxy (wx) on chromosome 9 (Tracy et al. 2006, Qi et al., 2009) are monogenic and recessive (Santos et al., 2014). Of all the mutant types, sh2 (Yousef and Juvik, 2002) and bt2 (Brewbaker, 1977) have the greatest commercial value.

The sh2 types of sweet corn contain about 29.9% sucrose, which is about ten times that of field maize. These types of sweet corn, generally referred to as super-sweet or extra-sweet corn, have extended shelf life due to the slow conversion of sugars to starch after harvest (Tracy, 1997), but with highly reduced levels of total carbohydrate. The sh2 gene reduces endosperm content of starch and water soluble polysaccharides which leads to a reduction in the energy content of maize kernels and the markedly collapsed physical appearance of the kernels when dry. Due to these features, sh2 varieties generally have a significantly reduced germination, seedling emergence, seedling vigour, plant development and growth and poor stand establishment.

In Nigeria, maize is grown throughout the country from the high rainfall forest of the southeast to the low rainfall Sudan savanna of the north; and with supplemental irrigation, maize can be grown throughout the year. Nigeria produces about 40% of the maize production in West and Central Africa (FAO, 2016). Great potential therefore exists for the production of super-sweet corn in Nigeria.  Sweet corn production is usually targeted at three distinct and largely independent markets, namely; fresh, canning and freezing, with the fresh market component accounting for more than 70% of the total (Lizaso et al., 2007; USDA, 2017). Imports of canned sweet corn to Nigeria have increased in recent years largely due to widening food preferences. Therefore, the development and production of super-sweet corn in large quantity and quality would impact positively on the social and economic life of Nigeria. However, sweet corn cultivars are virtually nonexistent in Nigeria, and coupled with poor adaptation to tropical environments, its cultivation and utility is limited.

To bridge the gap created by these challenges, a broad-base temperate super-sweet sh2 corn population was   introduced   into   the  country  and  adapted  to  the prevailing tropical environmental conditions by four cycles of mass selection (Adetimirin, 2008).  This was meant to serve both as an open pollinated variety, as well as basis for inbred line development and hybrid production. The performance of this sh2 super-sweet population in the growing conditions of Nigeria could be improved. One way of achieving this is by converting tropical field maize genotypes into sweet corn through backcrossing and selection for the sweet corn trait. This strategy will lead to improvement in the yield of the super-sweet corn varieties, broaden the narrow genetic base characteristic of sh2 (Tracy, 2001; Teixeira et al., 2013), and facilitate the development of sh2 inbred lines and hybrids. According to Entringer et al. (2017), the backcross method of breeding is efficient for obtaining super-sweet corn populations with good agronomic performance. Previous studies Cartea et al. (1996), Malvar et al. (1997, 2001), Tracy (2001), Butrón et al. (2008) and Entringer et al. (2017) reported the use of field corn to improve the agronomic performance of sweet corn. These studies have also shown that field maize genotypes could differ in their ability to improve the agronomic performance and quality of sweet corn. Therefore, tropical normal endosperm field maize genotypes have the potential to improve the adaptation and productivity of super-sweet corn. The objective of this study was to evaluate the agronomic performance of some sh2 super-sweet corn populations derived from crosses between a tropicalised sh2 population and tropical normal endosperm field corn genotypes.

Location of experimental site

The study was conducted at the experimental field of the Department of Agronomy, Faculty of Agriculture, along Parry road, University of Ibadan (7°26' N, 3°54' E), Ibadan, Nigeria.

Generation of new homozygous sh2 super-sweet corn populations

A broad-based temperate super-sweet sh2 maize population, which was introduced and adapted to tropical environmental conditions of Ibadan, Nigeria after four cycles of mass selection (Adetimirin, 2008), was crossed as male (using bulk pollen) to six tropical field maize inbred lines (1368, 4001, 4008, 9613, KU1409 and KU1414) and five commercial/experimental hybrids [1368/9071 (Oba super 1), 4001/KU1414 (Oba super 2), KU1409/9613, KU1409/4008 and 4001/4008], all developed at the International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria

The resulting F1 populations from each of these initial crosses were sib-mated to generate F2 populations. The F2 seeds were thereafter selected for kernels homozygous for the sh2 phenotype. The selected homozygous sh2 derived from each of the F2 populations were backcrossed to the recurrent (field maize) parents. The BC1 plants from each population were sib-mated to generate BC1S1 which were thereafter selected for kernels homozygous for the sh2 phenotype. The selected homozygous sh2 derived  from  each  of  the  BC1S1 populations were then bulked to form new homozygous sh2 populations as shown in Table 1.

Evaluation of sh2 super-sweet corn populations

The eleven new homozygous derived sh2 populations alongside the original donor sh2 population were evaluated in replicated field trials under rain-fed conditions during the 2016 and 2017 cropping seasons. The soil at the experimental site is sandy-loam with 15.20 g kg^-1 organic carbon, 0.98 g kg^-1 total nitrogen, 14.79 mg kg^-1 available P (Bray-1), 0.26 cmol kg^-1K and a pH(H₂O) of 6.1. The twelve sh2 populations were arranged in a randomized complete block design with three replicates. Each plot consisted of four 5.0 m long rows, with plants spaced 0.50 m and rows 0.75 m apart. Planting was done on the flat. Plots were over-sown to ensure good plant stand (since germination failure is a common feature in super-sweet corn with the sh2 gene) and later thinned to two plants at two weeks after planting (WAP) to give a plant population of approximately 53,333 plants per hectare. NPK 15-15-15 fertilizer was applied at the rate of 300 kg ha^-1 at 2 WAP, which provided 45 kg N ha^-1, 20 kg P ha^-1 and 36 kg K ha^-1 and top-dressed using urea at the rate of 25 kg N ha^-1 at 4 WAP. Plots were kept weed free by hand weeding.

Data collection

Data were collected on days to anthesis (DA) as number of days from planting to when 50% of the plants in a plot shed pollen, and days to silking (DS) as number of days from planting to when 50% of the plants in a plot have emerged silks. Anthesis-silking interval (ASI) was calculated as the difference in days between DS and DA. Plant height (PH) and ear height (EH) were measured in meters 2 weeks after silking on all plants in the two middle rows of a plot, as the average distance from the soil level to the collar of the uppermost leaf and collar of the leaf bearing the uppermost ear, respectively. Husk cover (HC) was scored on plot basis on a scale of 1 to 9 (1 = husk tightly covers ear tip and extends beyond it; 9 = poor husk cover with ear tip clearly exposed). Harvesting for yield data was carried out between stages R4 and R5 when the ears were still green and fresh, 21 days after silking using plants in the two middle rows of a plot. Yield data included: (i) number of cobs (NC) recorded as total number of ears (fresh ears with husk removed) harvested per plot expressed per hectare; (ii) yield of cobs (YC) recorded as total weight of cobs (fresh ears with husk removed) harvested per plot expressed in tonnes per hectare; (iii) number of marketable cobs (NMC) recorded as number of cobs with approximately 250 filled edible kernels per plot, expressed per hectare; (iv) yield of marketable cobs (YMC) recorded as total weight of marketable cobs per plot expressed in tonnes per hectare; (v) number of kernel rows (NKR) recorded as the average number of kernel rows of 10 top cobs; (vi) cob length (CL) measured in cm as the average length of 10 top cobs; (vii) cob diameter (CD) was measured using an electronic 6 in digital calliper (Pittsburgh^®, Item #47257) as the average diameter of 10 top cobs taken at the middle portion of the cob.

Data analyses

All data were subjected to analysis of variance for a randomized complete block using the PROC. GLM procedure in SAS (SAS Institute, 2003) assuming a mixed model with genotype and the interaction between genotype and environment were considered random. The years were considered as separate environments. Significant means were separated using DMRT (p=0.05).

Evaluation of sh2 super-sweet corn populations

The eleven new homozygous derived sh2 populations alongside the original donor sh2 population were evaluated in replicated field trials under rain-fed conditions during the 2016 and 2017 cropping seasons. The soil at the experimental site is sandy-loam with 15.20 g kg^-1 organic carbon, 0.98 g kg^-1 total nitrogen, 14.79 mg kg^-1 available P (Bray-1), 0.26 cmol kg^-1K and a pH(H₂O) of 6.1. The twelve sh2 populations were arranged in a randomized complete block design with three replicates. Each plot consisted of four 5.0 m long rows, with plants spaced 0.50 m and rows 0.75 m apart. Planting was done on the flat. Plots were over-sown to ensure good plant stand (since germination failure is a common feature in super-sweet corn with the sh2 gene) and later thinned to two plants at two weeks after planting (WAP) to give a plant population of approximately 53,333 plants per hectare. NPK 15-15-15 fertilizer was applied at the rate of 300 kg ha^-1 at 2 WAP, which provided 45 kg N ha^-1, 20 kg P ha^-1 and 36 kg K ha^-1 and top-dressed using urea at the rate of 25 kg N ha^-1 at 4 WAP. Plots were kept weed free by hand weeding.

Data collection

Data were collected on days to anthesis (DA) as number of days from planting to when 50% of the plants in a plot shed pollen, and days to silking (DS) as number of days from planting to when 50% of the plants in a plot have emerged silks. Anthesis-silking interval (ASI) was calculated as the difference in days between DS and DA. Plant height (PH) and ear height (EH) were measured in meters 2 weeks after silking on all plants in the two middle rows of a plot, as the average distance from the soil level to the collar of the uppermost leaf and collar of the leaf bearing the uppermost ear, respectively. Husk cover (HC) was scored on plot basis on a scale of 1 to 9 (1 = husk tightly covers ear tip and extends beyond it; 9 = poor husk cover with ear tip clearly exposed). Harvesting for yield data was carried out between stages R4 and R5 when the ears were still green and fresh, 21 days after silking using plants in the two middle rows of a plot. Yield data included: (i) number of cobs (NC) recorded as total number of ears (fresh ears with husk removed) harvested per plot expressed per hectare; (ii) yield of cobs (YC) recorded as total weight of cobs (fresh ears with husk removed) harvested per plot expressed in tonnes per hectare; (iii) number of marketable cobs (NMC) recorded as number of cobs with approximately 250 filled edible kernels per plot, expressed per hectare; (iv) yield of marketable cobs (YMC) recorded as total weight of marketable cobs per plot expressed in tonnes per hectare; (v) number of kernel rows (NKR) recorded as the average number of kernel rows of 10 top cobs; (vi) cob length (CL) measured in cm as the average length of 10 top cobs; (vii) cob diameter (CD) was measured using an electronic 6 in digital calliper (Pittsburgh^®, Item #47257) as the average diameter of 10 top cobs taken at the middle portion of the cob.

Data analyses

All data were subjected to analysis of variance for a randomized complete block using the PROC. GLM procedure in SAS (SAS Institute, 2003) assuming a mixed model with genotype and the interaction between genotype and environment were considered random. The years were considered as separate environments. Significant means were separated using DMRT (p=0.05).

The YC ranged from 5.82 to 8.16 t/ha. The range in   NC   and   NMC   were  43,290  to  53,330 and 36,710 to 45,300 cobs/ha, respectively. On the average, 83.1%  of  the  total  NC  harvested  was marketable. The NKR, CL and CD averaged 14.2, 15.7 cm and 45.2 mm, respectively.

One important quality attribute in fresh market sweet corn production is HC. In the present study, HC scores ranged from 2.8 to 6.8 with a mean of 4.2. All the derived sh2 super-sweet corn populations had significantly lower HC scores than the donor sh2 population, eight of which had scores lower than the mean. The PH and EH of the populations ranged from 1.9 to 2.1 m (mean = 2.0 m) and 0.9 to 1.2 m (mean = 1.0 m), respectively. On the average, ears were placed mid-way the height of the plants.

All the population including the sh2 donor population exhibited intermediate maturity with DA and DS ranging from 53.8 to 56.7 days and 56.3 to 59.0 days, respectively. However, nine of the derived populations shed pollen significantly later than the donor sh2 population. On the average, the derived sh2 super-sweet corn populations flowered significantly later than the donor sh2 population. The ASI, which is an indication of the extent of synchrony in flowering ranged from 1.5 to 3.5 days with a mean of 2.6 days. Eight of the derived sh2 populations had significantly lower ASI that the donor sh2 composite population.

In this study, orthogonal comparison revealed significant differences between the derived sh2 populations and the donor sh2 composite population for all the traits, with the derived sh2 populations being superior to the donor sh2 composite (Table 3).

Sweet corn cultivars are virtually non-existent in Nigeria, have narrow genetic base and poor adaptation to tropical environments. One strategy that could be adopted to broaden its genetic base, improve its adaptation and enhance its agronomic performance under tropical conditions is to cross with tropical field maize. In this study, some sh2 super-sweet corn populations derived from crosses between sh2 super-sweet corn population and tropical field maize genotypes were evaluated for their agronomic performance.

The significant effects of genotype and the higher contribution of the main effects of genotype to the total sum of squares observed in this study was a manifestation of the fact that the populations varied greatly in their performance. These genetic differences could be exploited in further breeding programmes. These significant genetic differences also indicated that the conversion of the field corn genotypes into super-sweet corn was effective in the development of new sweet corn populations different from the sh2 donor parent. This was furthermore evidenced in the superiority of most of the derived populations over the sh2 donor parent population for most of the traits studied. In this study, yield of marketable cobs averaged 6.84 t/ha. This was higher than the average yield of marketable fresh cobs of 5.03 t/ha reported by Kim et al.  (2007) and  4.72  t/ha reported by Abe and Akinrinola (2015) for open pollinated varieties of normal endosperm tropical field maize. This observed yield potential reflects the great prospects that abound for sweet corn production in Nigeria. The findings of this study confirmed previous reports (Cartea et al., 1996; Malvar et al., 1997, 2001; Tracy, 2001; Butrón et al., 2008; Santos et al., 2014; Entringer et al., 2017) on the potential of tropical field maize genotypes in the improvement of sweet corn.

Significant genotypic differences were observed among the populations studied. Most of the derived sh2 populations exhibited significantly superior performance relative to the sh2 donor population for most of the traits. The results of the present study indicated that new super-sweet corn populations different from the donor parent could be developed by converting field corn genotypes into super-sweet corn through backcrossing and selection for the sweet corn trait. The observed genetic differences could be exploited in further breeding programmes. The superior populations could be further improved and used as base populations for inbred line development.

The authors have not declared any conflict of interests.

The authors wish to thank Professor V.O. Adetimirin of the Department of Agronomy, University of Ibadan for generously providing the sh2 donor population used in this study. The normal endosperm tropical field maize genotypes were sourced from the Maize Improvement Programme of IITA, Ibadan. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.