Fertilizer Effects on Endosperm Physicochemical Properties and Resistance to Larger Grain Borer, Prostephanus truncatus (Coleoptera: Bostrichidae), in Malawian Local Maize (Zea mays L.) Varieties: Potential for Utilization of Ca and Mg Nutrition

Abstract

Maize grain hardness influences storage pest resistance, a key characteristic valued by smallholder farmers. The structural changes in the endosperm determine grain hardness and are influenced by agronomic practices. The purpose of this study was to establish whether supply of calcium and magnesium based fertilizers can alter physicochemical properties of local and hybrid maize varieties and reduce the infestation by larger grain borer (Prostephanus truncatus (Coleoptera: Bostrichidae)) during storage. Two local and one hybrid maize varieties commonly grown by smallholder farmers in Malawi were cultivated under three fertilizer treatments (NPK (nitrogen, phosphorous, potassium), NPK plus gypsum, and NPK plus dolomite). After harvest, the grains were classified into flint and dent types, followed by P. truncatus infestation and determination of their physicochemical properties. The addition of gypsum and dolomite fertilizers led to higher levels of amylose, total zein and β-14 zein, traits associated with kernel hardness, compared to the application of NPK fertilizer. Moreover, local maize varieties showed higher resistance to P. truncatus infestation, hardness and biochemical properties associated with hardness (total zein, α-19 and β-14 zein, starch lysophosphatidylcholine, and non-starch free fatty acid) compared to hybrid variety. Our study suggests the potential for utilizing Ca and Mg nutrition in maize to improve kernel hardness, thus adoption of gypsum and dolomite by smallholder farmers may be beneficial against P. truncatus during storage.

1. Introduction

Maize (Zea mays L.) is an important staple food and the most widely grown crop by smallholder farmers in most African countries. Almost 77% of the maize grain harvested is used as food and 52% of the remaining proportion as livestock feed [1]. Smallholder farmers cultivate both local and hybrid varieties, with the former often retained because of the socio-economic and cultural interests, especially personal taste preference when preparing traditional foods [2]. Local maize varieties are also known for their high potential in carotenoids content, including provitamin A [3]. Our previous studies found that Malawian local maize varieties were more flinty and higher in nutritional content and possessed better cooking properties and storability traits compared to the hybrid varieties [4,5]. Smallholder farmers also believe that local maize varieties store better than hybrids and are resistant against the larger grain borer Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) and maize weevils (Sitophilus zeamais Motschulsky) during storage [6,7]. The P. truncatus and S. zeamais are the most important maize storage pests in Malawi [8,9]. Nevertheless, recent studies have shown that all local maize varieties grown by smallholder farmers in Malawi are susceptible to P. truncatus infestation [10].

The P. truncatus infestation causes both qualitative and quantitative losses by feeding on and tunneling into maize kernels for oviposition. Without insecticide application, postharvest losses of about 40% to 100% due to P. truncatus have been reported among smallholder maize farmers in Sub-Saharan Africa (SSA) regions including Malawi [11]. Currently, the majority of the smallholder farmers continue the use of insecticides as an effective control measure to contain the P. truncatus infestation in maize grain during storage [9]. However, the continued use of insecticides may not be sustainable under the smallholder farmers in the region due to the high possibility for insecticide resistance development in P. truncatus and the high costs associated with insecticide use. Furthermore, due to global human health and environmental concerns, insecticide use has become less appealing. Globally, human exposure to insecticides has been linked to an annual loss of approximately 300,000 lives [12]. Therefore, safer, low-cost, sustainable, and easily adaptable control measures under the smallholder farmers are urgently needed. The natural resistance of maize kernels to P. truncatus could provide the a most acceptable, economically viable, and sustainable control option for the smallholder famers in SSA region [13].

Maize grain physiochemical properties are known to affect the susceptibility of grains to P. truncatus damage. Grain hardness, amount of vitreous endosperm, and phenolic and amylose contents are some of the factors reported to affect the maize grain susceptibility to P. truncatus damage [14,15]. The maize grain composition, which is mostly 73% starch, 10% protein, 5% oil, and 2% of fiber, vitamins, and micronutrients combined, is mainly controlled by genetics of the endosperm sink, maternal plant, and the environment [16]. The maize kernel composition determines grain quality, an important factor to smallholder farmers who embrace low input farming system and rely on longer storability of maize for food security [17].

Crop agronomic practices such as fertilizer application manipulate the environment in which the crop is grown and thus have potential to significantly impact on maize production and kernel composition, though with associated costs [1]. Several studies have shown the effects of nitrogen, phosphorous, and potassium (NPK) compound-fertilizer-based agronomic practices on the maize grain hardness and susceptibility to storage pests [18,19,20,21,22,23]. Besides this, there is limited research on other equally important nutrients such as calcium (Ca) and magnesium (Mg). The use of agronomic N has been shown to influence grain quality especially on the protein and oil content though based on genotype [19]. An increase in N fertilizer application improves kernel hardness indicators (test weight, % vitreousness, and kernel density), which are associated with increased protein and zein profile contents [18,20]. In addition, phosphorus has been associated with the proportions of small and large starch granules in the kernels. Indeed, increased phosphorus application rates reduce the proportion of large starch granules and consequently reduces the kernel amylose content [23]. Previous studies on Ca and Mg based fertilizers have limited their investigations on the effect of these nutrients on grain yield rather than storage pest damage. These studies found significant effects of Ca and Mg on maize grain yield by utilizing gypsum and dolomite as sources of Ca and Mg, respectively [24,25,26]. Therefore, we postulated that Ca and Mg nutrients may also influence maize grain quality.

Calcium is key in enhancing cell wall strength and maintaining cell membrane integrity [27,28]. This cell strength may influence maize grain quality and provide a mechanical barrier to insect pests, hence its susceptibility to storage pest damage. Magnesium is the central nutrient element for the chlorophyll molecule necessary for sugar synthesis. It is also one of the key nutrient elements required for the translocation of sugars from the source (leaves) to the storage organs such as the grain seed. These two functions of Mg could have an influence on the maize grain quality and susceptibility to storage pests. Both Ca and Mg are bridge components for the aggregation of ribosome subunits in the endoplasmic reticulum, and their shortage has been demonstrated to reduce protein synthesis [29]. Therefore, supply of adequate amounts of Ca and Mg nutrients is necessary to attain high maize quality and yield.

Presently, the application of nitrogen based fertilizers is the most widely used agronomic practice for maize production under smallholder farmers in Malawi. The government of Malawi recommends the use of NPK plus sulfur (S) compound fertilizer, with the majority of the farmers using the formulation 23% N, 21% P, 0% K, and 4 % S, along with urea (46% N) to attain a standard rate of 92 kg N ha⁻¹, 42 kg P₂O₅ ha⁻¹, 0 kg K₂O ha⁻¹, and 8 kg S ha⁻¹ [30]. This practice does not include other equally important nutrients such as Ca and Mg. Some studies have indicated widespread Ca and Mg deficiencies in maize grain grown under low-pH soils which are predominant in Malawi [31]. In order to achieve food and nutrition security among Malawi’s people, calcium, and magnesium based fertilizers would be necessary. Notably, gypsum (CaSO₂·2H₂O) and dolomite (CaMg(CO₃)₂) are naturally available and relatively cheap agro-minerals in Malawi [32] and could be used to supply Ca and Mg, respectively. The present need to improve maize grain quality and reduce postharvest losses from storage pests by utilizing Ca and Mg based fertilizers is valuable, where information is currently limited.

We therefore investigated the effect of gypsum and dolomite application in combination with NPK fertilizer on the development of hardness, biochemical properties, and P. truncatus resistance during storage in flint and dent kernel type maize varieties grown in Malawi. It was hypothesized that the application of gypsum and dolomite fertilizers in combination with inorganic NPK fertilizers would improve maize nutritional value and resistance to storage pests compared to the conventional practice. The present results demonstrate the importance of the inclusion of Ca and Mg based fertilizers in maize production systems for the improvement of the food and nutrition security with reduced postharvest losses.

2. Materials and Methods

2.1. Maize Varieties and Experimental Field Design

Two local maize varieties namely white local (ACC OU 549) and yellow local (ACC OU 546-2), and one hybrid maize variety, namely DKC-9089, hereafter referred to as local 1, local 2, and hybrid, respectively, were used in this study. The two local varieties were obtained from Malawi Plant Genetic Resources Center (Gene Bank) in Chitedze Agricultural Research Station (CARS), whereas the hybrid variety was obtained from DeKalb Genetics Corp., DeKalb, IL, USA.

A field experiment was conducted in CARS, Lilongwe, Malawi during the 2015–2016 cropping season. The soils in the study site were Oxisols with pH (H₂O) of 5.2 (acidic), low base saturation (40.3%), and low exchangeable Ca (4.0 cmol_c kg⁻¹). The available phosphate (Bray II) was 39 mg kg⁻¹, whereas exchangeable K and Mg of 0.08 and 0.49 cmol_c kg⁻¹, respectively. A Latin square split-plot design was used with the main plots allocated to the maize varieties and the subplots to fertilizer type. Each main plot was 10 m by 15 m, and the subplots were 3 rows by 3 columns. All agronomic practices were done according to the guide for maize production in Malawi [33]. Pesticides were not applied in all the treatments. The field was ploughed with machines before making the ridges using hand hoes. The maize seeds were planted on the ridges 75 cm apart using dibble stick method at a depth of 5–6 cm and at a distance of 25 cm between the planting stations (holes). Three fertilizer treatments, (1) NPK, (2) NPK plus gypsum, and (3) NPK plus dolomite, were evaluated and constituted the main plots. NPK fertilizers were applied on the ridges at a depth of 10 cm during planting period (basal dressing), and N was supplemented in-season at the same depth using UREA (top dressing) when the crop was at knee height. Both fertilizers were applied using the “spot” or “dollop’’ method. Dolomite and gypsum were applied on the ridges before seed sowing by broadcasting [33]. All the fertilizer treatments received the recommended rate of 92 kg N ha⁻¹, 42 kg P₂O₅ ha⁻¹, 0 kg K₂O ha⁻¹, and 8 kg S ha⁻¹ for maize production [30]. A compound NPK (23:21:0 + 4S) fertilizer and urea (46% N) locally purchased from Agricultural Trading Company Limited (Lilongwe, Malawi) were used to supply N, P, K, S, and N, respectively. The fertilizer treatments 2 and 3 received 36 kg of CaO ha⁻¹ from gypsum and dolomite, respectively (Agricultural Trading Company Limited, Lilongwe, Malawi). The fertilizer treatment 3, in addition to Ca, received 28 kg of MgO ha⁻¹ from dolomite. Both gypsum and dolomite were applied at the rate of 100 kg ha⁻¹ in the form of CaSO₄.2H₂O and CaMg (CO₃)₂, respectively. During the 2015/2016 cropping season, total precipitation was 619.3 mm, daily minimum and maximum temperatures averaged 19.3 to 29.9 °C, respectively, and daily relative humidity averaged 67.4% (Figure 1) (CARS weather station data).

2.2. Sample Collection, Preparation, and Kernel Classification

The maize grain samples were collected from each subplot after manual harvesting as done by small-scale farmers in Malawi. The samples were air dried and imported, under appropriate government permit, to Obihiro University of Agriculture and Veterinary Medicine in Japan for laboratory analyses. At the laboratory, the maize grains were disinfected by freezing at −30 °C for three weeks and thawed at 4 °C for one week. Moisture standardization was done in a hot air oven (WFO-700, Tokyo Rikakikai Co., Ltd., Tokyo, Japan) at 40 °C to achieve approximately 12.0 ± 1.0% moisture content, wet weight basis (ww). Dried samples were kept at 4 °C in an airtight container for one week. Classification of maize kernels into flint and dent types was done based on endosperm phenotypic characteristics [34]. To prepare samples for biochemical analyses, the maize grains were milled into flour using a blender (New Power Mill PM-2005, Osaka Chemical Co., Ltd., Osaka, Japan) and sieved through a 355 µm mesh sieve. Samples for endosperm starch and non-starch lipids analyses were prepared by removing the germ from the grains using hand dissection and then milled into flour using a blender and sieved through a 355 µm mesh sieve.

2.3. Prostephanus truncatus Resistance

Prostephanus truncatus resistance was evaluated as previously described with some modifications [35,36]. Two hundred grams of grain sample from each maize variety was placed in separate glass jars (500 mL) with steel screens at the lids (60 mesh), and then 30 adult P. truncatus plants of 0 to 7 days old were introduced into each jar and kept in an incubator (MIR-254, Sanyo Co., Ltd., Tokyo, Japan) at 28.0 ± 1.0 °C. After four and twelve weeks of incubation, the number of live adult and adult cadavers and number and weight of damaged grains were recorded. All P. truncatus were separated and removed from the maize using sieves of 1.00 and 4.00 mm mesh sizes (IIDA Testing Sieve, IIDA Manufacturing Co. Ltd., Osaka, Japan). The number of damaged grains was determined by randomly selecting 100 grains from the jar and counting all grains tunneled by P. truncatus and their weight determined in triplicate.

2.4. Physical Properties

Kernel hardness was determined using the previously described method [37]. Briefly, fifteen maize grains were punctured on the upper lateral surface using a texture analyzer (TA-XT2, Stable Micro System, Godalming, UK) with a flat probe P/2 (diameter 2 mm) and 25 kg load cell. Puncture test was performed at 1 mm/s of puncture speed and kernel punctured at the lateral center side, opposite to the embryo to a depth of 2 mm. Kernel hardness was expressed as the break force calculated in Newton (N).

2.5. Biochemical Properties

Moisture content and crude protein were determined using the standard AOAC methods [38]. Total starch in dry weight basis (dwb) and amylose contents were determined using Megazyme K-TSTA and K-AMYL Assay procedures, respectively (Megazyme International, Bray, Ireland). Total zein was extracted as previously described [38], and the supernatant representing the total zein fraction was dried at 80 °C in Kjeldahl tubes for 4 h and analyzed for nitrogen content by the Kjeldahl method [38].

Zein composition analysis was done by extracting the zein [39,40] with some modifications. Twenty-five milligrams of flour was transferred to a 2 mL micro-centrifuge tube, then mixed, vortexed with 2 mL of 70% ethanol with 2% 2-mercaptoethanol (v/v), and allowed to stand at room temperature overnight. The mixture was centrifuged at 13,000 rpm for 10 min; 100 µL of the supernatant was transferred into a new tube and mixed with 10 µL of 10% sodium dodecyl sulfate (SDS). The mixture was dried in a rotary vacuum (DNA Speed Vac DNA110, Savant Instruments, INC. Farmingdale, NY) and resuspended in 50 µL of water by vortexing. Further 50 µL of a sample-reducing buffer, composed of 4% SDS, 10% 2-mercapethanol, 10% sucrose, 0.05% Coomassie blue G-250 and 1.5M Tris-HCl pH 6.5 was added, and vortexed. Tubes were sealed and boiled for 5 min and frozen overnight. The extracted zein was separated by SDS-PAGE (49.5% T, 6% C mixture) [41], the gel was scanned at 312 nm (BioInstruments ATTO, AE-6933FXES) and quantification of the bands done using ImageJ 1.50i (National Institutes of Health, Bethesda, Maryland).

Mineral contents of magnesium (Mg), phosphorus (P), potassium (K), calcium (Ca), and zinc (Zn) were determined by wet-ashing 100 mg with concentrated sulfuric acid and hydrogen peroxide (2:1 v/v), the resulting solution was diluted 25-fold with deionized water and quantified by an inductively coupled plasma mass spectrometry (ICP; ICPS-8100, Shimadzu Co. Ltd., Japan) in triplicate.

Total starch and non-starch lipids fatty acid methyl esters (FAMEs) and composition in the endosperm flour were determined. Non-starch lipids were extracted [42], and the FAMEs contents quantified [43]. Starch lipids FAMEs were quantified from methyl esterification of the residual defatted flour by incubating in 2 mL of methanol-acetyl chloride (9:1 v/v) at 100 °C for 2 h in the presence of heptadecanoic acid as internal standard [44]. To determine starch and non-starch lipids composition, the lipids from endosperm flour were extracted [42]. Neutral lipids composed of free fatty acid (FFA), galactolipids composed of monogalactosyldiaglycerol (MGDG) and monogalactosylmonoacylglycerol (MGMG), and phospholipids composed of phosphatidylethanolamine (PE), phosphatidylcholine (PC), and lysoPC in non-starch lipid fractions were identified using thin-layer chromatography (TLC) after charring with 50% sulfate, anthrone and ittmer detection reagents, respectively. Images of the TLC plates were obtained and the bands quantified by Image J software. After extraction of starch lipids, their composition was determined by spotting in TLC, followed by charring with 50% sulfate and ittmer detection reagents.

2.6. Statistical Analysis

Means and standard deviations were used for descriptive statistics. Standard deviations presented were calculated and then pooled between replicates within variety and fertilizer, as these were independent observations. Matched differences between flint and dent were calculated for data from each subplot for all P. truncatus resistance, physical and biochemical properties parameters. Equality of means between flint and dent kernel types was tested by a, modified for within replicate correlation, paired t-test across the 45 differences. The adjusted p-values took into account that the observations within an individual replicate (row within variety) may be correlated. The following non-parametric methods were used: measures were averaged across flint and dent for each subplot, using the corresponding proportions of flint and kernel as weights. The resulting weighted subplot means were used to test differences between varieties and fertilizer types. Based on the geography of the test site, rows within each variety were treated as blocks for fertilizer (subplot) comparisons and as replicates for variety (main plot) comparisons. Non-parametric methods were chosen as more robust against unequal in variances. Differences between varieties were tested by the exact Kruskal–Wallis test, and difference between fertilizers were tested by Friedman’s test, appropriate for blocked data. These tests were based on ranking observations and testing whether higher or lower ranks are more likely for certain varieties or fertilizers, respectively.

3. Results

3.1. Effect of Fertilizer Type on Maize Resistance to P. truncatus Infestation

The percentage and weight of damaged grains after 12 weeks of infestation was significantly differed among the kernel type and maize varieties (

Table 1. Differences between flint and dent kernel types in the number of live adult P. truncatus and adult P. truncatus cadavers, percentage of damaged grains, and weight of damaged grains after an P. truncatus infestation period of 12 weeks.

Variable	N	Mean Difference	SD Across Reps	Adjusted p
Live Adult P. truncatus	27	−7.6	216.9	0.877
Adult P. truncatus Cadavers	27	−2.2	25.2	0.657
% Damaged Grains	27	−7.0	8.2	0.0039
Damaged Grains (g)	27	−1.5	1.4	0.0016
Flour (g)	27	−3.9	12.2	0.243

). Flint type kernels had lower values compared to the dent type kernels. The local varieties, especially the Local 2 variety, experienced the lowest P. truncatus infestation by showing the lowest percentage of damaged grains, weight of damaged grains (p = 0.01), and weight of flour (

Table 2. Differences among maize varieties (A) and among fertilizer treatments (B) in the number of live adult P. truncatus, and adult P. truncatus cadavers, percentage of damaged grains, and weight of damaged grains after an P. truncatus infestation period of 12 weeks.

Variable	A				B
	Hybrid	Local 1	Local 2	p-Value	NPK	NPK + Dolomite	NPK + Gypsum	p-Value
Live Adult P. truncatus	693 (43)	652 (55)	633 (49)	0.07	848 (197)	774 (192)	656 (180)	0.72
Adult P. truncatus Cadavers	74 (2.2)	67 (10)	61 (9.4)	0.13	67 (23)	79 (20)	59 (19)	0.12
% Damaged Grains	87 (2.2)	85 (10)	82 (2.1)	0.17	86 (3.1)	84 (5.3)	85 (2.3)	0.64
Damaged Grains (g)	16 (1.5)	14 (0.3)	13 (0.6)	0.01	14 (1.2)	15 (0.8)	15 (2.0)	0.72
Flour * (g)	62 (4.1)	59 (5.4)	58 (3.4)	0.63	22 (2.8)	21 (2.9)	19 (3.4)	0.032

Means (SD) and p-values. In varieties, p-value is for overall comparison by exact Kruskal–Wallis test. In fertilizer types, the treatments were compared within rows of each variety, pooled by Friedman’s test. * Grain dust produced by larger grain borer when tunneling into the grain.

A). The application of gypsum fertilizer in combination with NPK fertilizer showed with significance the lowest weight of flour (p = 0.032) after the infestation period and relatively the lowest number of adult P. truncates and P. truncatus cadavers compared to the other fertilizer treatments (Table 2B). In contrast, the application of NPK fertilizer alone tended to be the highest in key variables; percentage of damaged grains, live adults P. truncatus, and weight of flour.

3.2. Effect of Fertilizer Type on Physicochemical Properties of Maize Grains

Flint type of kernels showed significantly higher hardness levels compared to dent type (

Table 3. Differences between flint and dent kernel types in physical and chemical properties.

Variable	N	Mean Difference	SD across Reps	Adjusted p
Physical Property
Hardness	27	46.4	22.1	<0.0001
Biochemical Properties
Total Starch, % dwb	27	−1.0	2.7	0.154
Amylose, % ww	27	1.5	3.9	0.143
Crude Protein, % dwb	27	−0.2	0.2	0.127
Total zein, % db	27	0.2	0.4	0.340
γ-27kDa, % Total Zein dwb	27	0.1	0.1	0.150
α-19kDa, % Total Zein dwb	27	1.0	0.6	<0.0001
β-14kDa, % Total Zein dwb	27	0.1	0.1	0.011
Total α-Zein, % Total Zein dwb	27	1.1	1.0	0.0004
Ash (%)	27	0.03	0.04	0.88
Mg (mg/100 g, dwb)	27	2.8	4.0	0.33
P (mg/100 g, dwb)	27	6.9	6.4	0.29
K (mg/100 g, dwb)	27	13.2	6.0	0.036
Ca (mg/100 g, dwb)	27	2.0	1.4	0.0005
Zn (mg/100 g, dwb)	27	0.3	0.25	0.052
Starch Palmitic Acid (mol%)	27	−0.2	1.9	0.4687
Starch Stearic Acid (mol%)	27	−0.2	1.3	0.4026
Starch Oleic Acid (mol%)	27	0.03	1.1	0.9933
Starch Linoleic Acid (mol%)	27	0.2	2.0	0.7465
Starch Linolenic Acid (mol%)	27	0.2	0.5	0.1790
Starch FFA (% of Total Starch)	27	−3.3	4.6	0.0093
Starch lysoPC (% of Total Starch)	27	3.1	4.2	0.0109
Non-Starch Palmitic Acid (mol%)	27	−0.5	1.8	0.2195
Non-Starch Stearic Acid (mol%)	27	0.2	1.5	0.2616
Non-Starch Oleic Acid (mol%)	27	0.3	2.1	0.3326
Non-Starch Linoleic Acid (mol%)	27	−0.1	3.0	0.8933
Non-Starch Linolenic Acid (mol%)	27	0.02	0.2	0.6030
Non-Starch FFA (%)	27	2.6	7.2	0.1011
Non-Starch MGDG (%)	27	1.7	1.7	0.4661
Non-Starch MGMG (%)	27	−3.6	7.4	0.2004
Non-Starch PE (%)	27	−3.2	5.5	0.0054
Non-Starch PC (%)	27	3.6	5.2	0.0032
Non-Starch lysoPC (%)	27	−0.2	1.4	0.4174

dwb, dry weight basis; ww, wet weight.

). Moreover, flint type showed significantly higher α-19 kDa zein, β-14 kDa zein, total α-zein, Ca, K, Zn, starch lysoPC, and non-starch PE biochemical parameters compared to dent type kernels, while starch FFA and non-starch PE were significantly higher in dent kernels compared to flint kernels. The application of gypsum and dolomite fertilizer in combination with NPK fertilizer significantly increased the levels of amylose, crude protein, total zein, γ-27 kDa zein, β-14 kDa zein, ash content, and non-starch linolenic acid compared to the application of NPK fertilizer alone (

Table 4. Differences among maize varieties (A) and among fertilizer treatments (B) in physical and chemical properties.

Variable	A				B
	Hybrid	Local 1	Local 2	p-Value	NPK	NPK + Dolomite	NPK + Gypsum	p-Value
Hardness (N)	197.9 (7.7)	207.2 (2.9)	233.4 (7.2)	0.0036	210.0 (14.6)	215.2 (10.1)	213.2 (15.5)	1.00
Total Starch, % dwb	71.0 (0.7)	70.4 (1.2)	69.0 (1.5)	0.439	69.6 (1.7)	70.4 (1.5)	69.4 (1.8)	0.1690
Amylose, % ww	19.8 (0.9)	21.0 (1.1)	19.1 (1.2)	0.254	18.04 (1.2)	20.8 (1.5)	21.1 (1.6)	0.0048
Crude Protein, % dwb	11.9 (0.1)	12.3 (0.04)	11.8 (0.1)	0.011	11.7 (0.1)	12.1 (0.1)	12.2 (0.1)	0.0011
Total Zein, % db	10.0 (0.1)	10.1 (0.04)	10.4 (0.1)	0.050	10.1 (0.2)	10.3 (0.1)	10.2 (0.1)	0.0319
γ-27kDa, % Total Zein dwb	0.3 (0.03)	0.2 (0.01)	0.2 (0.01)	0.025	0.2 (0.03)	0.3 (0.1)	0.2 (0.04)	0.0018
α-19 kDa, % Total Zein dwb	3.9 (0.1)	3.5 (0.2)	3.7 (0.1)	0.011	3.6 (0.5)	3.7 (0.1)	3.8 (0.4)	0.6412
β-14 kDa, % Total Zein dwb	0.5 (0.04)	0.8 (0.01)	0.8 (0.1)	0.050	0.6 (0.1)	0.8 (0.1)	0.7 (0.04)	0.0183
Total α-zein, % Total Zein dwb	8.1 (0.01)	8.0 (0.4)	8.0 (0.4)	0.1964	7.7 (0.8)	8.0 (0.4)	7.8 (0.8)	0.3679
Ash (%)	1.1 (0.01)	1.2 (0.01)	1.2 (0.01)	0.0038	1.1 (0.06)	1.2 (0.04)	1.2 (0.02)	0.0446
Mg (mg/100 g, dwb)	100.0 (0.8)	107.9 (0.9)	108.1 (0.6)	0.0036	107.9 (1.4)	109.3 (1.5)	108.9 (1.7)	0.0970
P (mg/100 g, dwb)	243.7 (0.7)	289.2 (1.1)	292.2 (0.9)	0.0029	280.5 (1.7)	284.2 (3.5)	283.7 (1.5)	0.8948
K (mg/100 g, dwb)	145.6 (0.5)	146.9 (1.1)	148.1 (0.8)	0.0033	148.9 (1.5)	150.3 (2.4)	149.7 (1.6)	0.3679
Ca (mg/100 g, dwb)	5.5 (0.1)	5.8 (0.4)	5.8 (0.4)	0.2964	5.9 (0.4)	5.6 (0.5)	5.7 (0.6)	0.2359
Zn (mg/100 g, dwb)	1.5 (0.1)	1.9 (0.1)	1.9 (0.1)	0.0714	1.8 (0.2)	1.8 (0.2)	1.8 (0.1)	1.0000
Starch Palmitic Acid (mol%)	42.1 (0.4	39.9 (0.6)	42.7 (0.7)	0.2050	41.2 (0.7)	41.8 (0.6)	41.6 (1.1)	0.0970
Starch Stearic Acid (mol%)	2.7 (0.6)	3.6 (0.7)	3.0 (0.2)	0.1964	3.0 (0.6)	3.2 (0.5)	3.1 (0.3)	0.8948
Starch Oleic Acid (mol%)	7.8 (0.4)	8.1 (0.3)	8.5 (0.1)	0.0857	8.3 (0.4)	8.0 (0.4)	8.1 (0.4)	0.2636
Starch Linoleic Acid (mol%)	45.4 (1.5)	49.9 (1.0)	43.3 (0.8)	0.1000	45.3 (1.1)	44.5 (0.9)	44.9 (1.9)	0.7165
Starch Linolenic Acid (mol%)	2.1 (0.5)	2.4 (0.02)	2.5 (0.03)	0.0036	2.3 (0.5)	2.5 (0.04)	2.3 (0.4)	0.5945
Starch FFA (% of Total Starch)	48.2 (12.3)	53.9 (2.8)	46.6 (12.4)	0.7214	48.4 (11.2)	51.1 (11.4)	49.9 (8.9)	0.4594
Starch lysoPC (% of Total Starch)	15.4 (8.4)	12.6 (2.0)	17.2 (9.7)	0.9286	5.4 (3.3)	14.7 (8.7)	4.8 (2.8)	0.4594
Non-Starch Palmitic Acid (mol%)	16.3 (0.5)	17.2 (0.6)	18.5 (1.3)	0.0857	17.1 (1.6)	17.2 (1.0)	17.6 (0.6)	0.0970
Non-Starch Stearic Acid (mol%)	3.1 (0.4)	3.2 (0.2)	3.2 (0.2)	0.8286	3.1 (0.5)	3.3 (0.5)	3.1 (0.3)	0.6412
Non-Starch Oleic Acid (mol%)	25.7 (0.5)	25.6 (0.4)	24.5 (0.9)	0.1321	26.2 (1.9)	25.1 (0.6)	24.4 (0.5)	0.0084
Non-Starch Linoleic Acid (mol%)	52.9 (0.2)	51.9 (0.5)	51.4 (0.6)	0.0520	21.5 (0.8)	52.2 (0.8)	52.6 (0.6)	0.1211
Non-Starch Linolenic Acid (mol%)	2.0 (0.03)	2.2 (0.04)	2.4 (0.02)	0.0036	2.1 (0.1)	2.2 (0.1)	2.3 (0.1)	0.0446
Non-Starch FFA (%)	28.6 (12.4)	34.7 (4.5)	40.1 (17.0)	0.6286	36.8 (12.3)	34.0 (14.6)	32.6 (12.3)	0.2359
Non-Starch MGDG (%)	27.8 (8.1)	30.2 (2.2)	23.0 (2.5)	0.5333	28.5 (7.1)	22.4 (1.5)	30.1 (14.3)	0.8465
Non-Starch MGMG (%)	22.3 (2.4)	13.0 (4.2)	13.6 (3.0)	0.050	19.4 (7.5)	15.5 (6.0)	13.8 (5.6)	0.2636
Non-Starch PE (%)	30.4 (0.8)	24.8 (0.3)	26.8 (1.7)	0.0107	25.6 (3.9)	27.7 (1.4)	28.7 (4.1)	0.0622
Non-Starch PC (%)	67.8 (0.6)	73.7 (0.1)	70.9 (1.4)	0.0036	72.6 (4.1)	70.6 (0.7)	69.2 (3.8)	0.1211
Non-Starch lysoPC (%)	1.8 (0.3)	1.6 (0.3)	2.4 (0.3)	0.0714	1.9 (0.5)	1.7 (0.8)	2.1 (0.5)	0.2636

Means (SD) and p-values. In varieties, p-value is for overall comparison by exact Kruskal–Wallis test. In fertilizer types, the treatments were compared within rows of each variety, pooled by Friedman’s test. dwb, dry weight basis; ww, wet weight.

B). In addition, gypsum and dolomite fertilizer application increased the hardness of the kernels.

The local varieties had significantly higher levels of hardness, crude protein, total zein, β-14 kDa zein, minerals contents, starch linoleic acid (18:2 n-6), non-starch linolenic acid (18:3 n-3), and non-starch PC and significantly lower levels of γ-27 kDa zein, non-starch linoleic acid, non-starch MGMG, and non-starch PE compared to the hybrid variety (Table 4A). On the other hand, gypsum and dolomite fertilizer treatments showed significantly higher levels of amylose, crude protein, total zein, γ-27 kDa zein, β-14 kDa zein, ash content, and non-starch linolenic acid compared with NPK fertilizer treatment (Table 4B).

3.3. Pearson’s Correlation Coefficients of Kernel Hardness with P. truncatus Resistance and Biochemical Parameters

After twelve weeks of infestation, maize kernel hardness had a negative and significant correlation with percentage and weight of damage grains at r = −0.801 (p < 0.01) and −0.477 (p < 0.05), respectively (

Table 5. Pearson’s correlation coefficients of kernel hardness with P. truncatus and biochemical properties of the maize varieties with three fertilizer treatments.

Variable	Kernel Hardness	Variable			Kernel Hardness
	p Value				p Value
P. truncatus Resistance		Starch FAME (mol%)	Palmitic Acid (16:0)		0.047
Live Adult P. truncatus	−0.091		Stearic Acid (18:0)		−0.165
Adult P. truncatus Cadavers	−0.309		Oleic Acid (18:1 N-9)		0.035
% Damaged Grains	−0.801 **		Linoleic Acid (18:2 N-6)		−0.109
Damaged Grains (g)	−0.477 *		Linolenic Acid (18:3 N-3)		0.585 *
Flour (g)	−0.349	Starch Lipids (%)	FFA		−0.588 *
Biochemical Properties			LysoPC		0.623 **
Total Starch, % dwb	−0.608 **	Non-Starch FAME (mol%)	Palmitic Acid (16:0)		0.183
Amylose, % ww	0.198		Stearic Acid (18:0)		0.207
Crude Protein, % dwb	−0.292		Oleic Acid (18:1 N-9)		−0.156
Total Zein, % dwb	0.495 *		Linoleic Acid (18:2 N-6)		−0.130
γ-27 kDa	0.423		Linolenic Acid (18:3 N-3)		0.465
α-19 kDa	0.686 **	Non-Starch Lipids(%)	Neutral Lipids	FFA	0.668 **
β-14 kDa	0.593 **		Galactolipids	MGDG	−0.279
Total α-Zein	0.670 **			MGMG	−0.319
Ash, %	0.196		Phospholipids	PE	−0.481 *
Mg (mg/100 g)	0.257			PC	0.433
P (mg/100 g)	0.330			LysoPC	0.180
K (mg/100 g)	0.492 *
Ca (mg/100 g)	0.749 **
Zn (mg/100 g)	0.592 **

Correlation coefficients with * and ** are significant at p < 0.05 and 0.01, respectively. dwb, dry weight basis; ww, wet basis.

). For biochemical parameters, kernel hardness was positively and significantly correlated with the contents of total zein, α-19 kDa zein, β-14 kDa zein, total α-zein, K, Ca, Zn, starch linolenic acid, starch lysoPC, and non-starch FFA. Total starch, starch FFA, and non-starch PE had a negative correlation with hardness.

4. Discussion

4.1. Fertilizer Type and Maize Varietal Character Influence Maize Grain Resistance to P. truncatus Infestation

The influence of fertilizer type on resistance of commonly grown maize varieties in Malawi to infestation with P. truncates, for 12 weeks, are highlighted in this study. Gypsum and dolomite in combination with NPK inorganic fertilizer application resulted in lower weight of flour (p = 0.032) in comparison with the application of NPK fertilizer alone. The combination of gypsum and NPK fertilizer showed the lowest weight of flour and live adults, generally suggesting that the application of this combination has the potential to confer P. truncatus resistance traits to maize kernels and hence reducing damage during the storage period. This could be attributed to the grain hardness impacted by the Ca supply as shown by their significant and strong correlations (Table 5). Calcium is important for strengthening the cell walls, [27,28] which could provide a mechanical resistance to P. truncatus feeding and tunneling through the grain. These traits may be explained by the grain-filling mechanisms that are based on physicochemical properties of the grains. Flint type kernels showed superiority over dent type on the insusceptibility to damage by P. truncatus during storage (Table 1). Moreover, the local varieties showed higher resistance to damage by P. truncatus (Table 2). This may be attributed to the low number of live adult P. truncatus at the end of the infestation period, which was in agreement with previous work [13] and generally relatively higher grain P, K, Ca, and Mg concentration (Table 4A). These results show the complexity associated with nutrient demands between locals and hybrid, of which the overall effects may be dependent on individual nutrient effect and thus susceptibility to larger grain borer. Therefore, our results suggest that different fertilizer application rates are required in order to improve grain quality as well as reduce the storage losses, and thus a focus on Ca and Mg is hereby emphasized. Furthermore, this supports the effect of the genetic properties of a variety to resistance when exposed to similar environmental conditions.

4.2. Fertilizer Type Influence the Physicochemical Properties in Maize Grain

Maize grain hardness has been established as a major factor contributing to the resistance of storage pests in flint corn [34]. Moreover, flint kernels are known to have higher ratio of vitreous to floury fraction of the endosperm than dent type, contributing to their hardness [45]. In this study, the local varieties that were flinty had higher hardness compared to the hybrid variety, and a slight increase in hardness was observed where gypsum and dolomite were applied compared to where NPK fertilizer was applied. The grain Ca concentration was also slightly higher in the local varieties compared to the hybrid (Table 4A). Since significant positive correlation was observed between the Ca concentration in the grain and the grain hardness, our results suggest that Ca supply affected the physiochemical properties of the grain. The Ca nutrition might have an effect on the metabolism pathways of starch, protein, and lipids in the local varieties, which influenced the compactness of the endosperm, resulting in the increased hardness.

Fertilizer type affected the amylose content with the application of gypsum and dolomite, showing significantly higher levels compared to application of NPK fertilizer alone, and varietal effect was not significant. Amylose in maize is synthesized by granular-bound starch synthase and its activity is regulated by Ca and Mg availability during kernel development [46]. Grain crude protein is influenced by variety and differs depending on grain maturity and agricultural practices [20]. In our study, crude protein and total zein, the main storage protein in maize were significantly affected by both variety and fertilizer type (Table 4). Local 1 variety had the highest crude protein content, while Local 2 had the lowest. This inconsistency may be related to genetic factors such as the variation in prolamins, albumins, globulins, and glutelin, which make up crude protein in maize endosperm [40]. Gypsum and dolomite application in combination with NPK fertilizer led to higher crude protein and total zein contents compared to application of NPK fertilizer alone. This may be attributed to the role of Ca and Mg, which act as bridging elements for the aggregation of ribosome subunits in the endoplasmic reticulum, and their deficiency has been shown to decrease protein synthesis [29]. On the other hand, zein composition (α-, γ-, β-zeins) was mostly affected by variety compared to fertilizer treatment, depicting the dependency of mineral supply during zein class accumulation on the genotype and other factors as the kernel develops.

Ash content in maize varieties varies with genotype and is known to influence the mineral composition [47,48]. In the current study, the local varieties had significantly higher levels of ash, Mg, P, and K compared to the hybrid one. Concurrently, fertilizer type significantly affected the ash content, with the addition of gypsum and dolomite having higher levels than the application of NPK fertilizer alone. These results indicate that the flinty local varieties have higher potential to assimilate minerals during kernel development. Mineral composition in maize may influence the survival of storage pests and alongside, mineral malnutrition is a widespread problem in developing countries where maize is the staple food [49,50]. Therefore, utilization of the local varieties combined with application of Ca and Mg based fertilizers may provide suitable remedial agricultural practice for the smallholder farmers in addressing mineral undernourishment and P. truncatus damage during storage.

Starch and non-starch lipid FAMEs were characterized by having palmitic acid (16:0), stearic acid (18:0), oleic acid (18:1 n-9), linoleic acid (18:2 n-6), and linolenic acid (18:3 n-3), which was in agreement with previous works [51,52]. Starch lipid composition was mainly characterized by FFA and lysoPC while non-starch lipid composition mainly consisted of neutral lipids and membrane lipids, which include galactolipids and phospholipids as minor compounds of the endosperm [42]. Variety had an effect on non-starch linoleic and linolenic acids contents, which indicates that genotype influences the metabolic pathways of unsaturated fatty acids. Concomitantly, fertilizer type showed an effect on non-starch oleic and linolenic acids contents, which suggest that Ca and Mg assimilation affects the synthesis of membrane lipids of the amyloplast. The significantly higher levels of non-starch MGMG in the hybrid variety compared to local varieties may be related to the higher content of starch levels in the hybrid variety, since galactolipids are known to be specific lipid compounds of the amyloplast membrane [45]. The local varieties had higher levels of non-starch PC synthesis compared to the hybrid variety (Table 5). A decrease in the synthesis of PC has been shown to decrease protein synthesis by activating unfolding protein response (UPR) [53]. Thus, the lower content of PC in the hybrid variety may have activated UPR leading to the observed lower content of total zein.

4.3. Relationship between Physicochemical Properties and Maize Grain Resistance

P. truncatus tunnel maize grains as part of their feeding process and for oviposition [35]. Natural hardness of maize grains acts as a physical barrier from damage by P. truncatus during this pre-ingestion stage. Therefore, harder grains could limit food availability and multiplication sites for P. truncatus, hence the reduction in the percentage and weight of damaged grains observed in our present study as the kernel hardness increased (Table 5). Local maize varieties are known to be less susceptible to storage pests, particularly the maize weevil, hence their preference by smallholder farmers in Malawi [54]. In our study, we have shown that the local varieties had higher hardness than the hybrid variety, which resulted in the lesser damage observed in P. truncatus during the storage period.

Zein content in maize endosperm is associated with hardness of kernels. If there is proper zein accumulation during the kernel filling process, the protein body formation fully fills the starch intergranular spaces, resulting in a vitreous or hard endosperm phenotype [40,55]. This explains the positive correlation of zeins and hardness observed in our present study. Minerals play an important role in most of the enzymatic processes; thus from this study, we postulate K, Ca, and Zn may have define the biosynthesis of amino acids and lipids, hence contributing to hardness. However, the mechanism of minerals directly contributing to compactness of the endosperm and eventual kernel hardness remains unknown. LysoPC forms complexes with amylose, increasing rigidity of starch granules, hence the impact on the starch–protein matrix, while non-starch FFA, which are localized on the surface of starch granules may enhance the adhesion of protein bodies on starch granules contributing to a compactness of the endosperm [42], hence the hardness of the kernels. The observed negative and significant correlation of starch FFA and non-starch PE is illustrated by their higher contents in dent type kernels (Table 3).

5. Conclusions

Contents of starch, zein profile, and endosperm lipids directly influenced kernel hardness. Total zein, α zein, and individual zein sub-classes and α-19 kDa zein and β-14 kDa zein correlated positively with kernel hardness. As a result, the accumulation of these proteins during kernel filling process until maturity altered the endosperm compactness, resulting in increased hardness in the local varieties. Despite of the low contents of lipids in the endosperm, starch lysoPC, non-starch FFA, and PC lipids may have a role in the starch architecture and protein body adhesion on starch granules, respectively, leading to the formation of a compact endosperm and therefore hardening the kernels. Application of gypsum and dolomite fertilizer had an effect on protein and lipid contents in the endosperm as well as mineral composition (Mg, P, and K contents). This shows that the amount of macro- and micronutrients in maize kernels is influenced by the amount of Ca and Mg in the kernel, presumably through altering their biosynthesis pathways. Therefore, supplementation of Ca and Mg based fertilizers may be a recommended agronomic practice for maize cultivation in Malawi. Maize variety selection is important since it showed greater effect on macro- and micronutrients in the kernels; hence it may be a crucial factor in endosperm compactness. The hardness of kernels, which was related to their biochemical composition, was attributed to the ability of flinty local varieties to show resistance to P. truncatus during storage. As a result, using recommended NPK fertilizers in combination with gypsum and/ or dolomite, especially in local varieties, may result in superior hardness. This may be utilized by the economically constrained smallholder farmers as a natural barrier to P. truncatus during storage to minimize losses.