Zinc and Provitamin A Biofortified Maize Genotypes Exhibited Potent to Reduce Hidden—Hunger in Nepal

Zinc deficiency affects one third of the population worldwide, and vitamin A deficiency is a prevalent public health issue in Sub-Saharan Africa and South-Asia, including Nepal. Crop biofortification is the sustainable solution to these health—related problems, thus we conducted two different field trials in an alpha lattice design to identify zinc and provitamin A biofortified maize genotypes consistent and competitive in performance over the contrasting seasons (Season 1: 18 February to 6 July 2020 and Season 2: 31 August to 1 February, 2020/21). In our study, the performance of introduced maize genotypes (zinc—15 and provitamin A biofortified—24) were compared with that of the local check, focusing on the overall agro-morphology, yield attributes, yield, and kernel zinc and total carotenoid content. Zinc and total carotenoid in the tested genotypes were found in the range between 14.2 and 24.8 mg kg−1 and between 1.8 and 3.6 mg 100 g−1. Genotypes A1831-8 from zinc and EEPVAH-46 from provitamin A biofortified maize trial recorded kernel zinc and total carotenoid as high as 52.3, and 79.5%, respectively, compared to the local check (DMH849). The provitamin A genotypes EEPVAH-46 and EEPVAH-51 (total carotenoid: 3.6 and 3.3 mg 100 g−1), and zinc biofortified genotypes A1847-10 and A1803-42 (20.4 and 22.4 mg kg−1 zinc) were identified as superior genotypes based on their yield consistency over the environments and higher provitamin A and zinc content compared to the check. In addition, farmers can explore August sowing to harvest green cobs during December-January to boost up the emerging green cob business.


Introduction
Micronutrient deficiency attributes to the global burden of diseases by elevating the instances of illness and mortality from disease infection and mental disabilities [1]. The extent of micronutrient deficiency in South-Asia, including Nepal, is alarming. Nepal reports 32 to 35% and 11.3 to 23.3% stunting and wasting in different age groups due to a low intake of essential micronutrients in the daily diet [2,3]. Zinc deficiency is widespread among children (6 to 59 months-21%) and non-pregnant woman (15 to 49 years-24%), and vitamin A deficiency is 4.2% and 3% respectively for the children and non-pregnant woman [3]. Earlier reports by the World Health Organization indicated that more than 32% of pre-school children had vitamin A deficiency disorders [4]. However, nutrient supplementation programs have helped to reduce vitamin A deficiency in recent years as the program is reported to cover more than 90% of children in Nepal [3,4].
Zinc deficiency affects one third of the population worldwide, and vitamin A deficiency is a prevalent public health issue in Sub-Saharan Africa and South-Asia, including The Field trials were conducted at Directorate of Agricultural Research, Karnali Province, Dasharathpur, Surkhet, Nepal. The geographical coordinates and elevation of the study site are 28 • 30 N, 81 • 47 E, and 490 masl, respectively. Analysis of the composite soil sample taken from the experimental site (in different years) indicated that the soil was slightly acidic to neutral (pH-6.11 to 6.45), medium organic matter (1.87 to 2.02%), nitrogen (0.09 to 0.10%), phosphorous (30.57 to 96.12 mg kg −1 ), and low to medium potassium (104.50 to 125.80 mg kg −1 ) with sandy loam texture [19,23,24]. The details of agro-meteorological parameters recorded during two growing seasons are displayed in Figure 1.

Experimental Detail and Genetic Materials
Two field experiments were conducted simultaneously in two contrasting seasons; season 1 (18 February-6 July) and season 2 (31 August-1 February) of 2020/21 in the same block. The first season resembles spring, and the second season had the Summer month; September, and rest of Winter; October to February. In the experiment, one trial set consisted of regional extra early multi-stress tolerant provitamin A biofortified maize genotypes (hereafter EEPVAH trial), whereas the other composed of three-way cross zinc enriched maize genotypes (hereafter TWC trial). The hybrids in both the trials were developed by the International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria and introduced in Nepal by the International Maize and Wheat Improvement Centre (CIMMYT), Kathmandu, Nepal through the Nepal Seed and Fertilizer Project.
The EEPVAH and TWC maize trial consisted of 25 and 16 genotypes, respectively, ( Table 1). DMH849 was used as a local check, which is an Indian hybrid widely grown by the farmers around the experimental location. Check-RE (TZEE-Y Pop STR C5 x TZEEI 58) in EEPVAH trial was an extra early provitamin A genotype (11.4 µg g −1 ) released in Ghana, Nigeria and Mali and has 5 t ha −1 yield potential. Alpha lattice design was employed (5 × 5 lattice in EEPVAH, and 4 × 4 lattice in TWC) and genotypes were replicated twice. The experimental plots were composed of two rows of 5 m length, and inter and intra-row spacing of 75 cm and 40 cm (50 cm in TWC trial), respectively. Intra-row spacing (plant to plant) in our experiment was nearly double than the recommended spacing for maize (20 to 25 cm) due to the limited seed available to conduct the trial. Two seeds were placed in each hill during sowing, and later thinning was done to maintain a single stand at the 3-4 leaf stage. Fertilizer doses of 120:60:40 NPK kg ha −1 and 8 tons ha −1 farm yard manure (FYM) were applied. A half dose of nitrogen, full doses of phosphorus, potassium, and FYM were applied as a basal dose during the final land preparation. The remaining dose of nitrogen was applied in two equal splits at V6 and V10 stages of the crop growth. The government's protocol for integrated pest management of the fall armyworm was employed to control the fall armyworm during the growing seasons [25]. and Mali, potential yield and provitamin A content are 5 t ha −1 and 11.4 µg g −1 , respectively, TWC: three-way cross zinc biofortified hybrid maize lines, DMH849 in both the trial was used as local check which indicates widely grown hybrid maize around the experimental area.

Data Recording Procedures
We recorded data related to phenology, growth and yield attributing traits, cob characteristics, and grain yield. Days to tasseling and silking were recorded when 50% of plants shed pollen grains, and with the emergence of 2-3 cm long silk in 50% of plants in the plot. Plant and ear heights were recorded in centimeters from five central plants. The distance from plant base to the first tassel branch was recorded as plant height, whereas ear height was measured as a distance between plant bases to the node bearing the uppermost ear. The number of plants with a stalk broken below the ear were counted in each plot and converted into a percentage to record the stalk lodging. At the time of harvest, we recorded the total number of plants and ears harvested in each plot. Cob characteristics; cob length and diameter (cm), number of kernel rows per cob, and kernel number in a row were recorded from five randomly selected cobs in each plot as a post-harvest recording. Grain yield, shelling percentage, harvest index, and hundred kernel weights were recorded after proper drying and shelling of the cobs. The ratios of grain recovered to the total dry cob weight, and total grain weight to the biological yield (grain and stover yield) in the plot were recorded as shelling percentage and harvest index, respectively. The grain yield is reported in tons per hectare by adjusting at 14% moisture content (Wile 55 moisture meter used). One hundred dried kernels in each plot were counted and weighed to determine one hundred kernel weights in grams. The standard data recording procedure was followed for recording all the studied parameters [26].

Kernel Nutrient Analysis (Total Carotenoid and Zinc)
After harvesting, composite grain samples were taken from each genotype of EEP-VAH and TWC trials to analyze the total carotenoid and zinc content, respectively. The well-dried grain samples (65 • C for 72 h) were grounded, packed, and then sealed in a polythene bag with respective leveling. Each sample was at least 200 g in weight when submitting to the laboratory (National Food Research Centre, Khumaltar, Nepal). For carotenoid analysis (EEPVAH trial), the pigment present in the grounded sample was extracted in di-acetone alcohol, and later was transferred to petroleum ether. Methanolic KOH was used for saponification of chlorophyll and then washed with water to remove it from the mixture. Finally, the carotenoid amount was estimated through the spectrophotometric method at 450 nm β-carotene standard [27].
The USEPA method 3050B (second revised version of method 3050) was employed to determine zinc content of the grain samples taken from TWC trials [28]. The quantitative amount of the crushed, sieved, and dried plant matter sample was digested with tri-acid solution (nitric acid, hydrochloric acid, and per chloric acid) in a beaker followed by digestion over a controlled hot plate and quantitative filtration into a known volumetric flask. The aliquot was aspirated into the AAS at an air-acetylene flame after blank, and three consecutive working metal standard solutions. The final concentration of the respective element was calculated from the values obtained through the software-displayed values on the basis of the calibration curve.

Statistical Analysis
Recorded raw data were entered and processed in Microsoft Excel 2007, and ANOVA was generated with ADEL-R software version 2.0 [29]. Statistical analysis was performed individually for the two separate seasons in both the trials. We chose the randomized incomplete block design (RIBD) while operating data analysis in ADEL-R. The Fisher least significant difference test (p < 0.05) was performed in statistically significant response variables to separate treatment means. In order to identify consistent performers (in terms of grain yield) in both the seasons, AMMI analysis was performed using GEA-R software version 4.1 [30].

Phenological Traits
The evaluated genotypes were highly significant for the days to tasseling and silking in both the trials and in both the seasons. However, anthesis-silking interval was found non-significant (Tables 2 and 3). The genotypes took a short time to tassel and silk during season-2 in both TWC and EEPVAH trials compared to season-1 (Tables 2 and 3). The check (RE) recorded the earliest tasseling (70 days) and silking (73 days) during season-1, and was among the early group during season-2 too (tasseling-51 days, and silking-54 days). It was observed that mean anthesis-silking interval during season-2 (4 days) was longer than that in season-1 (2 days). Among the genotypes of the TWC trial, A1830-14 took the longest time for tasseling (86 and 62 days in season-1 and season-2) and silking (88 and 67 days in season-1 and season-2) in both the seasons (Table 3).  Remarks: StdMSE-standard mean sum of error, Std-standard deviation, LSD-least significant difference, CV-coefficient of variation, p-value < 0.01 indicates highly significance, <0.05 significant effect, >0.05-nonsignificant effect, similar statistical letters within a parameter (column) indicates non-significant difference among the genotypes.

Growth and Yield Attributing Traits
Growth and yield attributing traits viz. plant (PHT), ear height (EHT), number of plant (PHPP), ear harvested (EHPP), shelling percentage (SHELP), harvest index (HI), and hundred kernel weight (HKW) were observed in both the trials. A highly significant (p < 0.01) effect was observed in all the growth and yield attributing traits in the EEP-VAH trial, while the majority of traits were significantly different (p < 0.05) in TWC trial (Tables 4 and 5). The number of plants harvested (PHPP-during season-1 in EEPVAH trial), and hundred kernel weight (HKW-during season-2 in TWC trial) were found nonsignificant among the tested genotypes (Tables 4 and 5). Performance of the EEPVAH and TWC genotypes for growth and yield attributing traits were reduced more during season-2 than in season-1 (Tables 4 and 5). The fall armyworm infestation during season-2 affected the overall vegetative growth and development of the maize genotypes ( Figure 2). Remarks: StdMSE-standard mean sum of error, Std-standard deviation, LSD-least significant difference, CV-coefficient of variation, p-value < 0.01 indicates highly significance, <0.05 significant effect, >0.05-non-significant effect, similar statistical letters within a parameter (column) indicates non-significant difference among the genotypes. Remarks: StdMSE-standard mean sum of error, Std-standard deviation, LSD-least significant difference, CV-coefficient of variation, p-value < 0.01 indicates highly significance, <0.05 significant effect, >0.05-non-significant effect, similar statistical letters within a parameter (column) indicates non-significant difference among the genotypes.  Three-way cross zinc biofortified genotype A1803-37 reported the highest dry matter deposition in the kernel (recorded as HKW) during season-1 (36.5 g) as well as in season-2 (30.7 g). A1830-6 and A1831-3 produced statistically similar dry matter content as in A1803-37 ( Table 5). The harvest index was recorded highest in the local check (0.6) during season-1 while A1803-37 had the highest (0.6) dry matter deposition in the kernel during season-2. Shelling percentage in the TWC genotypes ranged between 78.7 and 86.1% (mean-82.3%), and between 73 and 80.5% (mean-77.5%) during season-1 and season-2, respectively ( Table 5). The mean number of plant and ears harvested were highest during season-1 than in season-2, and similar observations were found for plant and ear height. Plant stature was tallest in 1803-42 (261 cm), and was recorded shortest in A1830-6 (200 cm), and the local check had 202.5 cm during season-1. The plant (172.4 cm) and ear height (69.9 cm) were observed shortest in A1830-9 during season-2 (Table 5).
Among the EEPVAH genotypes, EEPVAH-44 reported maximum dry matter deposition in the kernel in both seasons. The local check also produced higher dry matter (in the form of HKW) in the kernel (35.8 g and 32.4 g), statistically similar to EEPVAH-44 (Table 4). EEPVAH-8 consistently recorded the highest harvest index in both seasons (0.6 in season-1 and 0.7 in season-2). The shelling percentage in EEPVAH genotypes was higher as compared to TWC genotypes. The mean shelling percentage during season-1 and season-2 were 84.7 and 82.5%, respectively (Table 4). Similarly, mean plant and ear height, the number of plant and ears harvested during season-2 were comparatively lower than in season-1. During season-1, the plant stature observed in EEPVAH genotype was as high as 242 cm, and the mean height was 212.2 cm in the overall genotypes. EEPVAH-50 had the shortest plant height (195.4 cm), and the local check recorded 208.5 and 186.6 cm plant and ear height, respectively during season-1 and season-2. Genotype EPVAH-67 reported the tallest plant and ear height during season-2 (Table 4).

Cob Characteristics
During season-1, EEPVAH and TWC genotypes were found highly significant (p < 0.01) for most of the cob characteristics (cob length-CL and diameter-CD, and number of kernel-rows per cob-NOKRC) except for the number of kernels per row-NOKPR (Tables 6 and 7). The EEPVAH genotypes were significantly different in cob characteristics during season-2, however a non-significant difference was observed for CD and NOKRC in TWC genotypes (Tables 6 and 7). It was observed that the performance of EEPVAH and TWC genotypes were inferior in terms of cob characteristics during season-2 than in season-1. Local check-DMH849 produced the longest (18.1 cm in season-1 and 17.1 cm in season-2) and bigger cob (5.1 cm in season-1 and 4.7 cm in season-2) during both seasons in the EEPVAH trial (Table 6). In the same trial, EEPVAH-67 recorded highest number of kernels per row (41 in season-1, and 39 in season-2) in both seasons. The genotype EEPVAH-41 had the highest number of kernel-rows per cob during season-1 while in season-2, EEPVAH-53 recorded the highest NOKRC ( Table 6). The cob length (18.5 cm in season-1 and 17.1 cm in season-2) was recorded highest in A1803-37 in both seasons in the TWC trial. The local check produced highest NOKRC (17), and NOKPR (40), and cob length and diameter were statistically similar to top genotypes during season-1, however, the cob characteristics were inferior during season-2 (Table 7).

Stalk Lodging
Stalk lodging was statistically non-significant in both the trials during season-1, and the genotypes recorded up to 6% and 9% lodging in TWC and EEPVAH trials, respectively (Figure 3). Only a few genotypes (four) recorded lodging in the TWC trial in season-1 whereas lodging recorded was within 2 to 9% in 13 genotypes of the EEPVAH trial ( Figure 3). The local check (DMH849) recorded comparatively low lodging; 2.8% in TWC season-1, and 6.6% in EEPVAH season-2 ( Figure 3). In contrast, lodging was significantly higher during season-2 and statistically significant among the maize genotypes in both the trials. Three-way cross hybrid maize genotype A1803-37 produced 16% lodging (2 to 16% in overall), while the highest lodging recorded in EEPVAH genotypes was 20% (EEPVAH-55) (Figure 3).

Stalk Lodging
Stalk lodging was statistically non-significant in both the trials during season-1, an the genotypes recorded up to 6% and 9% lodging in TWC and EEPVAH trials, respective (Figure 3). Only a few genotypes (four) recorded lodging in the TWC trial in season whereas lodging recorded was within 2 to 9% in 13 genotypes of the EEPVAH trial (Figu 3). The local check (DMH849) recorded comparatively low lodging; 2.8% in TWC seaso 1, and 6.6% in EEPVAH season-2 ( Figure 3). In contrast, lodging was significantly high during season-2 and statistically significant among the maize genotypes in both the tria Three-way cross hybrid maize genotype A1803-37 produced 16% lodging (2 to 16% overall), while the highest lodging recorded in EEPVAH genotypes was 20% (EEPVAH 55) (Figure 3).  est kernel zinc content (24.86 ppm) while DMH849 (local check), A1803-37, A1831-9 A1830-14 had kernel-zinc content ≤ 16.5 ppm. It was observed that EEPVAH and biofortified maize hybrids had 9.5 to 79.5%, and 2.5 to 52.3% higher total carotenoid kernel-zinc content, respectively, compared to the local check ( Figure 4).

Grain Yield
The tested EEPVAH maize genotypes were significantly different in terms of yield, producing 3.3 to 8.1 t ha −1 in season-1 and 1.4 to 4.7 t ha −1 during season-2. O other hand, the yield ranges produced by TWC genotypes were 3.7 to 7.1 and 1.8 to ha −1 , respectively, for season-1 and season-2 ( Figure 5). The local check (DMH849ha −1 in TWC, and 8.1 t ha −1 in EEPVAH trial) out yielded all the introduced maize g types during season-1 (except TWC genotype A1831-3), however, the performance du season-2 was greatly reduced; TWC genotype by 52.7, and EEPVAH by 44.3% (Figu The majority of introduced maize genotypes recorded a higher yield advantage ove local check in season-2; yield advantage over the local check was 1.6 to 54.5% in genotypes while EEPVAH genotypes recorded even higher-4.8 to 97.9% ( Figure 6) cording to the AMMI analysis, TWC genotypes A1803-42, A1847-10, and A1803-13 EEPVAH genotypes EEPVAH-8, EEPVAH-46, and EEPVAH-67, were found consiste performance (in terms of grain yield) with higher average grain yield over the sea and were placed nearest to the origin in the graph (Figure 7).

Grain Yield
The tested EEPVAH maize genotypes were significantly different in terms of grain yield, producing 3.3 to 8.1 t ha −1 in season-1 and 1.4 to 4.7 t ha −1 during season-2. On the other hand, the yield ranges produced by TWC genotypes were 3.7 to 7.1 and 1.8 to 3.7 t ha −1 , respectively, for season-1 and season-2 ( Figure 5). The local check (DMH849-6.8 t ha −1 in TWC, and 8.1 t ha −1 in EEPVAH trial) out yielded all the introduced maize genotypes during season-1 (except TWC genotype A1831-3), however, the performance during season-2 was greatly reduced; TWC genotype by 52.7, and EEPVAH by 44.3% ( Figure 5). The majority of introduced maize genotypes recorded a higher yield advantage over the local check in season-2; yield advantage over the local check was 1.6 to 54.5% in TWC genotypes while EEPVAH genotypes recorded even higher-4.8 to 97.9% ( Figure 6). According to the AMMI analysis, TWC genotypes A1803-42, A1847-10, and A1803-13, and EEPVAH genotypes EEPVAH-8, EEPVAH-46, and EEPVAH-67, were found consistent in performance (in terms of grain yield) with higher average grain yield over the seasons and were placed nearest to the origin in the graph (Figure 7). Plants 2022, 11, x FOR PEER REVIEW 14 of 22

Phenological Traits
The synchrony of tasseling and silking determines the outcomes of hybridization in maize. There is a higher possibility of seed setting when anthesis-silking days are closer as it increases effective pollination and eventually the grain yield [31]. Thus, narrowing the tasseling silking gap is of prime importance in hybrid development programs. As observed in the present study, Dhakal et al. [19] and Kandel and Shrestha [32] reported a non-significant effect on the anthesis-silking interval in the evaluated hybrid genotypes. It is observed that silking and tasseling takes double or more time in the winter season compared to summer or spring sowing [19,33], however, the tasseling and silking took a comparatively shorter duration in our experiment during season-2 (Tables 2 and 3). In general, October sowing is practiced for winter cropping, and we tested August sowing in our experiment (August 31). The daily average temperature remained within 24 to 30 • C for up to mid-October after sowing (Figure 1), by the time tasseling was initiated in several genotypes in our experiments, thus shortened the tasseling and silking compared to February sowing, and other winter sowings (October/November).
Right from germination to different growth stages and eventually to the physiological maturity, temperature plays an important role in maize crop. Maize requires around a 20 • C to 30 • C temperature for the rapid emergence of seedlings, and 28 • C is considered optimum for both tasseling and silking in maize [34][35][36]. The emergence and overall growth of the plant is faster when an optimum temperature is provided, and it slows down when the threshold is not met [37]. In addition, the short duration nature of the variety obtains earlier fulfillment of growing degree days required for tasseling and silking [37]. Regarding the tasseling and silking in hybrid maize, the mean value ranged between 66 and 108 days, and between 70 and 119 days, respectively, under Terai and Inner-Terai regions during the winter season in Nepal [32,33,38,39]. Corroborating our results, anthesis and silking days in provitamin A tropical maize genotypes were reported at 55-65 days, and 56-69 days, respectively, from Africa [40,41].

Growth, Yield Attributing, and Cob Characteristics
The overall growth, yield attributing traits, and cob characteristics of maize genotypes were reduced during season-2 comparative to season-1. The physiological stress attributed to fall armyworm damage during the vegetative stage, and drought following tasseling had great effect in our experiment in season-2. Fall armyworm larvae, right from the first instar, is most damaging, and feed on foliage of maize plants. Larger larvae feed extensively and cause heavy defoliation, and at a higher damage level, only the ribs and stalk remain in the plant [42]. Thus, its infestation reduces the photosynthetic area, elevates physiological stress, and affects normal growth and development of the plants. From a study conducted to assess the effect of the fall armyworm at an early vegetative stage, it was reported that the late whorl stage was most sensitive, and mean larvae 0.2 to 0.8 per plant at this stage could result in a 5 to 20% yield reduction in maize [43].
In maize, the number of kernel-rows in the cob is determined at the V5 to V8 growth stage, which is controlled mostly by genetic factors, and to a lesser degree, by the growing environment [44,45]. Thus, the number of kernel-rows per cob (NOKRC) was found significantly different in EEPVAH genotypes in both the seasons. The non-significant effect during season-2, in TWC genotypes (Table 7), might be due to the influence of drought and physiological stress caused by fall armyworm infestation. On the other hand, the number of kernels per row (NOKPR), heavily influenced by the growing environment and crop management practices, is determined between the V12-V15 growth stages [46]. For this reason, we observed a reduction in the mean number of kernels per row (Tables 6 and 7) during season-2 due to drought ( Figure 1) and fall armyworm damage in our experiment ( Figure 2). The observed reduction in cob size (cob length and diameter) during season-2 (Tables 6 and 7), in both EEPVAH and TWC trials, was due to the corresponding reductions in NOKRC and NOKPR as the cob size of maize is influenced by these traits [45]. Prolonged drought (Figure 1) immediate after the silking stage during season-2 might have resulted in comparatively smaller and lighter grains in the cob (reported as hundred kernel weight in Tables 4 and 5) due to less dry matter deposition in the kernels [44]. It was reported that long duration of drought stress following silking affects normal cob development and reduces kernel size and weight in maize [47,48]. The continued drought stress across several growth stages in maize attributes to a complete crop failure [48]. Thus, we observed a reduction in performance of the genotypes in terms of yield attributing traits (Tables 4 and 5), cob characteristics (Tables 6 and 7), and grain yield ( Figure 5) during season-2 due to prolonged drought following silking to the crop maturity stages (Figure 1).

Stalk Lodging
The breaking of the stalk below the ear is recorded as stalk lodging in maize [26]. Stalk lodging, in addition to yield loss up to 5 to 25%, increases harvest loss, harvest time and decreases the quality of grains [49,50]. Genotypes with a short stature suffered less lodging as in the local check and other maize genotypes during season-1 (Tables 4 and 5), however, season-2 lodging was exacerbated by the combined stress caused by fall armyworm infestation and drought. The continuous infestation of the fall armyworm from the early stage resulted in weaker (small stalk diameter) plants (Figure 2), and drought stress following tasseling resulted in higher lodging in season-2. Continuous infestation of fall armyworm feeding on vascular tissues of plant might have increased the physiological stress which favored the stalk lodging in plants [49]. Additionally, it is reported that drought stress during the grain filling period increases the potential for stalk lodging in maize [44,51]. In our experiment, the effect of fall armyworm infestation and drought stress in season-2 cropping is observed clearly in growth and yield attributing traits (Tables 4 and 5), cob characteristics (Tables 6 and 7), stalk lodging (Figure 3), and grain yield ( Figure 5).

Kernel Zinc and Total Carotenoid
Discriminating the zinc content of tested genotypes in our experiment, it is evident that the local check had low zinc content in its kernel (16.3 mg kg −1 ), while some introduced genotypes were highly enriched (more than 50%). It can be inferred that about 500 g grain (not including cooking or other processing losses) of high zinc maize line (A1831-8, 24.86 mg kg −1 ) could supply the daily recommended dietary intake for adults (up to 12 mg day −1 ). On the other hand, provitamin A enriched maize lines (total carotenoid-3.59 mg 100 g −1 ) seem promising to supply the daily dietary requirement of vitamin A; the adult human body requires 900 and 700 microgram (µg) retinol activity equivalents (RAE) per day, respectively, for men and women [8]. Thus, genotypes with a significantly higher amount of total carotenoid and zinc content were found in EEPVAH and TWC trials of maize compared to the local check which can be used as important genetic resources in future breeding programs and recommend for cultivation after further evaluation.
The yellow or orange endosperm of provitamin A contains carotenoids, which is the precursor of vitamin A [52]. The pollen gene from a non provitamin A parent affects development of fruit or seeds, and hence, kernel micronutrient content, which is known as the xenia effect [53]. The total carotenoid content, overall, in our experiment was lower ( Figure 4) in comparison to the breeding target; 15 µg g −1 of β-carotene standard [54]. This might be attributed to the xenia effect of adjacent white kernel hybrids from the three-way cross hybrid maize trial which contained mixture of white and yellow kernel hybrids. Similarly, it has been demonstrated that drought also influences provitamin A and β-carotene content in maize. Ortiz-Covarrubias et al. [55] reported lower provitamin A (12.9 µg g −1 ) and β-carotene content (16.5% lower) under drought than that in optimum growing contitions (14.1 µg g −1 ). In another study, the β-carotene was reported within the range between 1.67 and 3.39 µg g −1 under drought conditions [52].
The maximum zinc content of the genotype in our experiment was 24.86 mg kg −1 (Figure 4), which is lower (32.81%) than the breeding target for zinc in maize-37 mg kg −1 on the dry matter basis [13]. As soil zinc availability greatly influences kernel-zinc content in maize, the lower zinc content in our experiment might be associated with low soil zinc levels. Researchers have reported that breeding efforts for increasing kernel zinc concentration is highly dependent on agronomic interventions for most of the developing countries report deficient zinc levels in soil [13]. As a proof, several papers have reported widespread zinc deficiency in different agro-ecology of Nepalese soils, including similar to our study area [56][57][58][59][60].

Grain Yield
Grain yield is a key parameter for identifying and recommending genotypes for cultivation at the farmer level. The performance of the local check, in season-1, was superior to all the introduced maize genotypes, while in season-2, a drastic reduction was observed ( Figure 5). The reduction in overall performance of the genotypes could be associated with physiological stress (caused by fall armyworm damage) during the vegetative stage, and drought following the tasseling stage. No supplementary irrigation was provided and there was no rainfall recorded ( Figure 1) following tasseling in season-2. After the entry (9 May 2019) in Nepal [25], the fall armyworm has been widespread and its infestation was observed continuous in season-2 (August sowing), higher before the tasseling stage, in our experiment ( Figure 2). The intensity was higher as there was no maize cultivation around the study location and the chemical control was ineffective. Thus, overall growth and yield attributing (Tables 4 and 5), cob characteristics (Tables 6 and 7), and grain yield of maize ( Figure 5) were significantly reduced in both the trials during season-2. Fall armyworm larvae at a mean density of 0.2 to 0.8 per plant could reduce grain yield by 5 to 20%, if it occurred at the late whorl stage [43]. The reduction in yield was higher in the TWC trial than in EEPVAH, about a 10% reduction ( Figure 5). This might be due to the genotypes in the EEPVAH trial being stress tolerant as stated by IITA, Nigeria.
In a recent study conducted by the National Maize Research Program, Rampur, Nepal, the highest mean grain yield obtained from provitamin A biofortified maize, combining the spring and winter season, was 8.2 t ha −1 (crop geometry-0.75 m × 0.20 m and fertilizer dose of 180:60:40 NP 2 O 5 K 2 O ha −1 ). It was also reported that the provitamin A hybrids yielded 3.5 to 8.5 t ha −1 in spring, while the yield range was higher in winter-3 to 10.6 t ha −1 [33]. In the same location, the mean grain yields of the top four three-way cross white and yellow kernel hybrid (non-biofortified) maize were 9.5 and 8.8 t ha −1 in the winter season (crop geometry-0.60 m × 0.25 m and fertilizer dose of 180:60:40 NP 2 O 5 K 2 O ha −1 ) [61]. Similarly, the mean yield of the top performing three-way cross (non-biofortified) hybrid reported from Surkhet (current study location) was 8.4 t ha −1 in winter and 9.8 t ha −1 in summer [19]. The provitamin A tropical maize inbreed lines were reported to yield up to 6.2 t ha −1 on average (range-3.9 to 7.9 t ha −1 ) in Africa [40]. From the recent study, it was also disseminated that three-way cross provitamin A hybrids yielded between 1.1 and 5.1 t ha −1 in Nigeria [41]. The grain yield reported in season-1, in our study, corroborates with the findings of the previous studies and even indicates the scope of increasing it by maintaining plant population and fertilization as per recommendation.
Drought and fall armyworm infestation significantly reduced the grain yield in season-2 in our experiment ( Figure 5). It was reported that the fall armyworm can reduce maize grain yield by 5-40% and even cause heavy losses at the higher infestation level [62][63][64][65]. In Africa, annual maize production loss due to fall armyworm damage was estimated at 25 to 53%, which was around 2.48 to 6.19 billion USD [66]. Similarly, researchers reported that prolonged drought stress following silking affects normal cob development, and reduces kernel size and weight [47,48], and moderate to extreme drought could reduce grain yield in maize by 64 to 74% [67].

Conclusions
The provitamin A genotypes EEPVAH-46 and EEPVAH-51 (total carotenoid 3.6 and 3.3 mg 100 g −1 ), and zinc biofortified genotypes A1847-10 and A1803-42 (20.4 and 22.4 mg kg −1 zinc) were identified as superior genotypes based on their yield consistency over the seasons and higher provitamin A and zinc content compared to the local check. The plant population in the present experiment was almost half (40-50 cm intra-row), in comparison to the government recommendation (25 cm for hybrids), and we applied 120:60:40 NP 2 O 5 K 2 O ha −1 instead of 180:60:40 NP 2 O 5 K 2 O ha −1 (for hybrids), thus there is scope to increase the grain yield by maintaining plant population and fertilizer doses. The new sowing time (August) was tested for winter harvesting of maize, which enables farmers to harvest green cobs during December-January. Additionally, micronutrient rich maize is highly desired in the poultry industry to prepare feeds rich in vitamins and minerals. Thus, the identified provitamin A and high zinc maize genotypes could equally be effective in reducing hidden-hunger, enhancing feed nutrient value for poultry and livestock sectors, and also has commercial value in a green cob business.