Quantitative Trait Loci Analysis of Maize Husk Characteristics Associated with Gibberella Ear Rot Resistance

Abstract

Maize (Zea mays L.) is a vital crop susceptible to Gibberella ear rot (GER), a disease caused by Fusarium graminearum, resulting in significant yield losses and mycotoxin production. This study aimed to investigate the correlation between ear characteristics and GER resistance in 74 maize inbred lines (42 with non-stiff stalks and 32 stiff stalks) adapted to the northern Italian environment. Mycotoxin analysis was performed to assess the presence of deoxynivalenol (DON) and zearalenone (ZEA). The results showed a positive correlation between the husk traits, like the husk number and husk cover, and GER resistance in both heterotic groups. A positive correlation was also found between the DON and ZEA concentrations. In addition, we conducted a genome-wide association study (GWAS) which identified novel quantitative trait loci (QTLs) associated with the husk number, husk cover, ear attitude, and infection score. These QTLs can be utilized in marker-assisted selection for breeding new GER-resistant maize varieties. Our study provides valuable insights into the genetic basis of ear traits and their relationship with GER resistance, which can contribute to an improvement in the environmental and economical sustainability of the corn growing system.

1. Introduction

Maize (Zea mays L.) is a major cereal crop which can grow in different areas with different climates. It is an important nutritive source used as food and feed. In recent years, maize use for industry and energy production has gained importance [1,2]. In research, it is used as a model species due to its genotypic and phenotypic diversity. This shows how versatile this crop is and its importance in a wide range of markets.

Like other crops, a large number of pests and diseases can affect maize production. Even though different management methods exist to reduce their effects, diseases are still a concern, causing yield losses and reducing grain quality [3,4]. Fungi are the primary pests responsible for maize diseases [5], and the genus Fusarium is one of the most serious types. In this genus, different species are known to infect maize, and two of the most important ones are Fusarium verticilloides and Fusarium graminearum.

Fusarium verticilloides (Sacc.) Nirenberg (syn. F. moniliforme Sheldon, teleomorph G. fujikuroi (Sawada) Wr.) can infect different parts of the plant, like the stalk, ear, and roots [6], and is the causative agent of Fusarium ear rot (FER), an impactful disease which can reduce yields by 10–50% and is also associated with the production of fumonisin, a harmful mycotoxin which can cause disease in both animals and humans [7,8,9]. The interaction between F. verticilloides and F. graminearum is complex, and the presence of one can reduce the effect of the other, sometimes leading to difficulties in distinguishing between the two species [10].

Fusarium graminearum (Schw.) (Ascomycota) is a fungus of the genus Fusarium diffused around the world which can infect different plant species like maize, barley, wheat, oats, rice, and rye [11,12]. This fungus is sometimes reported with its teleomorph name of Gibberella zeae (Schwabe Petch). Due to the complex biology of this pathogen, it is more common to refer to it as the Fusarium graminearum species complex (FGSC), the cause of different diseases in cereals [11]. It can grow in a temperature range of 15–29 °C. If the temperature is higher than 30 °C, then its development is rather limited, while on the contrary, it can grow at temperatures below 15 °C [13,14,15]. In maize, it is known as the causative agent of Gibberella ear rot (GER), a key disease in temperate areas. In particular, infection can occur through silk or kernel wounds caused by insect feeding, weather damage, or mechanical injury [16,17,18]. Ostrinia nubilalis (or European corn borer (ECB)) is one of the most important insects associated with GER infection in maize. Larval feeding activities create wounds and tunnels which can favor F. graminearum infection [19]. Yield loss associated with this disease is estimated to range from 12% to 48%, with alternating years of higher impact and years without infection [20,21,22].

Mycotoxin production is associated with yield loss in maize infected by F. graminearum [23,24]. Infected ears can develop various types of mycotoxins, the most important of which are deoxynivalenol (DON) and zearalenone (ZEA) [24,25]. Mycotoxins are secondary metabolites of fungi which are harmful to animals and humans [15,26]. Deoxynivalenol causes gastrointestinal irritation (feed refusal and vomiting), altered immune function, and decreased milk and meat production, while zearalenone has estrogenic-like effects which can cause fertility and reproductive problems [27,28]. Due to this risk associated with mycotoxin intake for both animals and humans, the European Union set maximum amounts and guidance levels for a wide range of mycotoxins in grain and its derived products [29].

To reduce the impact of this disease, different methods can be used. Agronomic practices like crop residue management and crop rotation are the most important techniques for reducing sources of inoculum. This fungus can overwinter in cereal debris, and a tillage system like ploughing, which permits faster residue degradation, is effective in the control of this pathogen [18,30,31]. Crop rotation with susceptible species has shown an increase in infection and symptoms of this disease, demonstrating the importance of growing a succession of non-host species as an effective control method [31,32]. Another method used to mitigate GER effects is the use of insecticide against ECB. In countries where fungicides against F. graminearum are registered, control with these products is widely used and is increasing. GMO Bt technology is also effective against ECB, but due to the fact that this type of genetic control Bt technology is forbidden in Italy and other European Union countries [33,34], genetic selection of resistant hybrids of maize with a classic breeding strategy plays the crucial role in the struggle against GER [35,36,37,38,39].

In the last few decades, the different genetics behind GER resistance have been explored, and different quantitative trait loci (QTLs) were found across the 2.5 billion base pairs in all 10 maize chromosomes [40,41,42,43,44,45,46]. All of these studied QTLs offer an opportunity to breeders to apply genomics-assisted selection techniques. These techniques can help with the improvement in Gibberella ear rot resistance of new maize varieties, but due to the complex nature of this disease and the influence of the environment the application of these loci in breeding programs is still difficult. This quantitative nature of this trait and the complex interaction between plant and pathogen create obstacles to the full implementation of genomics-assisted selection techniques in breeding programs for this character [47,48].

In this context, the aim of this work was to understand the correlation between inbred corn ear characteristics and Gibberella ear rot resistance. In addition, through a genome-wide association study (GWAS), this study highlights the genetic basis of these characteristics with the identification of chromosomal regions associated with resistance to this disease, providing results useful for the selection of new resistant varieties, starting with inbred lines already adapted to the Italian environment.

2. Materials and Methods

2.1. Genetic Materials and Experimental Design

In the present study, 74 inbred lines (42 non-stiff stalk and 32 stiff stalk) were selected according to Gibberella ear rot resistance and their adaptation to the northern Italian environment (Table S1). The experiment was carried out during the 2022 and 2023 seasons in 3 different locations in northern Italy: Cremona (Location 1), Torino (Location 2), and Cuneo (Location 3). Sowing was performed with a plot planter (0.75 m × 0.14 m spacing) in a randomized complete block with two replications for each location. Each location was designed for a single-row plot experiment, and each plot was about 3 sq m (0.75 m × 4 m). The sowing, flowering, and harvesting dates of the three sites for both years are shown in

Table 1. Planting, flowering of 50% of plots, and harvesting dates of the three locations for the 2022 and 2023 seasons.

Season	Location	Planting Date	Harvest Date	Flowering of 50% of Plots
2022	Location 1	11 May 2022	28 August 2022	18 July 2022
	Location 2	16 May 2022	19 October 2022	24 July 2022
	Location 3	26 May 2022	24 October 2022	2 August 2022
2023	Location 1	24 May 2023	2 October 2023	28 July 2023
	Location 2	21 June 2023	14 November 2023	22 August 2023
	Location 3	27 June 2023	22 November 2023	30 August 2023

.

The experimental fields were grown with conventional farming methods in a maize–maize succession with standard soil fertilization (about 220 kg/ha of nitrogen). Pre-emergence herbicide was applied, and no insecticides were used. Flood irrigation was also applied every 15 days in all locations to avoid drought.

The following traits were evaluated in each plot: (1) the silk flowering date, or the date when 50% of the ears in the plot were producing silks; (2) the husk number, which was counted from manually harvested ears (average); (3) the husk coverage score, a semi-quantitative parameter measured using a scale from 0 to 9; (4) the ear attitude score, a visual evaluation of 5 ears on a scale from 1 to 9, where 1 = >90% of the ears being upright, 5 = 50% of the ears being upright, and 9 = >90% of the ears being pendant; and (5) the Fusarium graminearum infection score, a visual evaluation of 5 ears scored on a scale from 1 to 9 (disease severity rating (DSR)) where 1 = >75% of the ear pile being infected and 9 means no infection.

All traits were collected in both replications, and data were collected from the five central ears (marked before data collection) of the plot.

Data analysis was performed using SPSS software (IBM SPSS Statistic 20) and Paleontological Statistics (PAST, version 4.12).

2.2. Mycotoxin Analysis

At harvest, five ears per plot were collected with husks, and after data collection, they were dried to reach a relative humidity of 13%. After manual shelling and the creation of a bulk, the kernels were analyzed for mycotoxin contamination to assess the concentration of deoxynivalenol (DON) and zearalenone (ZEN).

The bulk seeds were milled in a Pulverisette 19 electric mill (Fritsch GmbH, Idar-Oberstein, Germany) to reach a final particle size of approximately 4 mm. Five grams of flour for each sample were prepared for the mycotoxin analysis using liquid chromatography-tandem mass spectrometry. Both mycotoxins were extracted from the same sample using 30 mL of a solution of acetonitrile/water/acetic acid (73:25:2 v/v). Extraction was performed by mechanical shaking for 30 min at 2300 rpm in a multitube vortex mixer (Benchmark Scientific, Sayreville NJ, USA). The extracts were filtered through a 0.2 µm polytetrafluoroethylene (PTFE) filter (Phenomenex, Torrance, CA, USA), and an aliquot of 1.8 mL was transferred to an HPLC vial and analyzed.

Analysis was carried out with a Vanquish Core HPLC (Thermo Fisher Scientific Inc., Waltham, MA, USA) in a Luna Omega Polar C18 100 × 2.1 mm column (Phenomenex, USA) coupled to an Orbitrap Explore 120 detector (Thermo Fisher Scientific Inc., USA). The total chromatographic run was 12.5 min at a flow rate of 400 µL/min. The column temperature was maintained at 45 °C, and the injection volume was 2.5 µL. The autosampler was operated at 10 °C. Data acquisition was performed using Xcalibur software (version 4.3). The system was equipped with an electrospray ionization (ESI) interface, and nitrogen was used as the drying and collision gas. The ion source parameters were positive ion spray voltage of 3700 V, negative ion spray voltage of 2300 V, sheath gas flow rate of 55 Arb, auxiliary gas flow rate of 15 Arb, an ion transfer tube temperature of 325 °C, and a vaporizer temperature of 350 °C.

2.3. Genotyping Data

Genomic DNA extraction was performed on one kernel per inbred line. All 74 materials were analyzed using an Infinium XT BeadChip platform (Illumina, San Diego, CA, USA). Genotyping was based on 24,000 authenticated SNPs distributed across all 10 maize chromosomes derived from the B73 reference sequence. This high-density genotyping was performed at the Corteva Agriscience LLC Johnston Laboratory using the protocol developed by the Illumina Company.

2.4. QTL Detection

Identification of the QTL signals was performed with a developed Corteva Agriscience model based on previously published works [49,50,51].

Signal detection was performed with a two-step analysis model. The first step was a genome-wide association study (GWAS). Markers were tested by fitting them one at a time while evaluating the increase in likelihood provided by each marker separately. Signals with a −Log10 value (p value) equal to three were marked as significant.

The second step was selection of the most important signals in relation to all of the others. If a signal was selected, then no other markers were selected within 10 cM left and right of the already-selected signal.

Since the inbred lines were analyzed by dividing them only by heterotic group, no clear family structure was present in the germplasm studied. The population structure term was included in the model as a random effect with marker-based relationships to complete the relationship information of these materials.

3. Results

3.1. Phenotypic Analysis

Seventy-four inbred lines were planted for 2 years (2022 and 2023) in three locations with two replications each to characterize the ear parameters associated with resistance to F. graminearum.

Table 2. Summary of agronomic parameters, shown as average ± SD, collected in three locations over 2 years for 74 inbred lines. DON = deoxynivalenol concentration (ppm); EAS = ear attitude score; FGIS = Fusarium graminearum infection score; HCS = husk cover score; HN = husk number; ZEA = zearalenone concentration (ppb). For each parameter, different letters indicate statistically significant differences according to Dunn’s test (p < 0.05).

Heterotic Group	Inbred	EAS	FGIS	HCS	HN	DON (ppm)	ZEA (ppb)
NSS	PHAP9	5.00 ± 3.62 abcde	5.92 ± 2.50 a	5 ± 1.91 acdfgh	6.50 ± 0.90 bcfghk	0.38 ± 1.07 a	30.14 ± 92.05 a
	PHG29	6.00 ± 3.46 abcde	5.67 ± 1.30 a	4.17 ± 1.59 cfh	6.92 ± 1.51 abcdefghijk	0.31 ± 0.63 a	12.36 ± 34.04 a
	PHJ90	4.33 ± 3.34 abcde	5.42 ± 2.07 a	4.17 ± 1.80 cdfh	6.33 ± 0.98 bcfgk	0.75 ± 1.24 a	76.48 ± 185.57 a
	PHK42	7.17 ± 3.13 bde	6.25 ± 2.09 a	4.17 ± 1.59 cfh	6.08 ± 1.62 bcfgk	0.04 ± 0.06 a	1.28 ± 2.25 a
	PHKW3	4.00 ± 3.36 abcde	6.00 ± 2.52 a	8.00 ± 1.35 abeg	7.25 ± 0.97 abcdefghijkl	2.32 ± 5.87 a	93.5 ± 224.41 a
	PHR31	8.56 ± 0.88 e	6.13 ± 1.55 a	4.11 ± 1.05 cdfh	6.63 ± 0.74 abcdefghijk	0.09 ± 0.23 a	2.24 ± 5.45 a
	PHT77	6.09 ± 3.02 bcde	7.91 ± 0.70 a	5.18 ± 1.08 abcdfgh	6.58 ± 1.44 abcdfghijk	0.02 ± 0.02 a	1.44 ± 4.27 a
	PH5 HK	2.67 ± 2.53 abcde	7.58 ± 1.38 a	7.17 ± 1.59 abcdefgh	6.83 ± 1.19 abcdefghijk	0.17 ± 0.36 a	6.72 ± 16.53 a
	PHK56	5.67 ± 3.55 abcde	6.17 ± 2.21 a	6.33 ± 1.56 abcdefgh	7.00 ± 0.85 abcdefghijk	0.04 ± 0.06 a	0.94 ± 2.03 a
	PHN11	7.33 ± 2.67 de	6.33 ± 2.06 a	4.00 ± 2.17 cdfh	5.50 ± 1.57 f	0.85 ± 1.49 a	41.00 ± 80.13 a
	PHN46	4.67 ± 2.53 abcde	6.67 ± 1.61 a	5.67 ± 1.3 abcdefgh	6.09 ± 0.94 bcfk	0.38 ± 1.28 a	10.44 ± 32.06 a
	PHN73	5.00 ± 2.83 abcde	5.63 ± 2.92 a	6.20 ± 1.69 abcdefgh	6.60 ± 0.84 abcfghijk	0.26 ± 0.54 a	6.24 ± 10.80 a
	PHP76	5.67 ± 3.23 bcde	6.17 ± 1.75 a	4.00 ± 1.35 ch	7.33 ± 1.07 abcdefghijkl	0.24 ± 0.51 a	5.09 ± 10.15 a
	PHPP8	4.33 ± 2.99 abcde	7.08 ± 1.51 a	6.83 ± 1.80 abcdefgh	6.83 ± 1.03 abcdefghijk	0.94 ± 2.94 a	101.87 ± 235.27 a
	PHR62	5.17 ± 3.24 abcde	7.50 ± 1.00 a	6.50 ± 1.93 abcdefgh	5.92 ± 0.90 bf	0.16 ± 0.51 a	4.60 ± 13.91 a
	PHW30	6.33 ± 2.99 bde	7.42 ± 0.90 a	8.17 ± 1.80 abe	7.25 ± 1.22 abcdefghijkl	0.28 ± 0.73 a	2.49 ± 6.94 a
	PHW53	2.00 ± 1.81 abcd	6.25 ± 1.91 a	6.33 ± 1.56 abcdefgh	8.25 ± 1.54 abcdefghijkl	0.05 ± 0.08 a	0.98 ± 2.40 a
	PHZ51	1.67 ± 1.30 abc	6.00 ± 2.41 a	6.00 ± 1.35 abcdefgh	8.25 ± 1.29 abcdefghijkl	0.08 ± 0.19 a	5.78 ± 11.62 a
	PHBE2	7.18 ± 3.28 bde	7.18 ± 0.98 a	4.82 ± 1.08 cdfgh	7.50 ± 1.31 abcdefghijkl	0.21 ± 0.54 a	3.96 ± 8.18 a
	PHJ33	2.33 ± 1.78 abcde	7.91 ± 1.04 a	5.50 ± 1.51 abcdefgh	7.17 ± 1.11 abcdefghijkl	0.10 ± 0.21 a	2.64 ± 7.22 a
	PHPM0	1.60 ± 1.35 abc	7.20 ± 1.62 a	7.40 ± 1.58 abcdefgh	8.40 ± 1.58 abcdefghijkl	0.06 ± 0.08 a	13.27 ± 28.00 a
	PHR03	4.17 ± 3.13 abcde	7.33 ± 1.67 a	5.67 ± 1.78 abcdefgh	7.50 ± 1.17 abcdefghijkl	4.85 ± 11.24 a	370.64 ± 1021.91 a
	PHR55	4.33 ± 3.34 abcde	5.09 ± 2.07 a	4.17 ± 1.03 ch	8.64 ± 1.21 adeghijkl	0.08 ± 0.17 a	1.11 ± 1.97 a
	PHVB2	5.17 ± 4.04 abcde	5.83 ± 1.95 a	5.67 ± 1.78 abcdefgh	6.58 ± 1.08 abcfghik	0.04 ± 0.07 a	19.47 ± 62.09 a
	PHBG4	1.00 ± 0.00 a	6.17 ± 2.33 a	4.17 ± 1.03 ch	7.42 ± 1.51 abcdefghijkl	0.04 ± 0.06 a	2.25 ± 5.34 a
	PHBR2	1.83 ± 1.03 abcd	7.50 ± 1.08 a	6.83 ± 1.03 abcdefgh	10.00 ± 2.52 adeijl	1.16 ± 2.62 a	7.58 ± 14.73 a
	PHBV8	4.17 ± 1.8 abcde	7.55 ± 2.34 a	6.08 ± 2.15 abcdefgh	7.82 ± 1.33 abcdefghijkl	0.03 ± 0.05 a	3.24 ± 10.72 a
	PHGV6	1.73 ± 1.85 abc	8.25 ± 1.04 a	7.73 ± 1.01 abdefg	9.45 ± 1.69 adehijl	2.11 ± 4.59 a	1.76 ± 1.05 a
	PHHH9	3.17 ± 2.33 abcde	6.89 ± 1.96 a	8.67 ± 0.78 e	9.83 ± 1.59 del	0.20 ± 0.48 a	10.46 ± 34.01 a
	PHJ31	1.33 ± 0.78 ac	6.92 ± 1.51 a	6.83 ± 2.62 abcdefgh	8.42 ± 1.44 abcdefghijkl	0.01 ± 0.01 a	0.55 ± 0.95 a
	PHJ65	1.20 ± 0.63 ac	7.17 ± 2.32 a	7.80 ± 1.40 abdefg	9.78 ± 2.28 adeghijl	7.80 ± 18.57 a	195.58 ± 476.51 a
	PHK46	2.17 ± 1.80 abcde	6.08 ± 1.98 a	8.00 ± 1.35 abeg	8.67 ± 1.50 acdeghijkl	0.99 ± 3.29 a	36.48 ± 122.12 a
	PHM57	2.00 ± 1.60 abcd	7.56 ± 0.88 a	6.83 ± 1.80 abcdefgh	7.33 ± 1.07 abcdefghijkl	0.69 ± 1.75 a	44.63 ± 134.08 a
	PHNB7	2.33 ± 1.56 abcde	6.91 ± 1.64 a	5.50 ± 1.51 abcdefgh	7.67 ± 1.15 abcdefghijkl	0.03 ± 0.07 a	0.69 ± 1.06 a
	PHP60	1.00 ± 0.00 ac	7.00 ± 2.00 a	7.20 ± 1.48 abcdefgh	9.60 ± 0.70 el	0.10 ± 0.17 a	3.04 ± 5.08 a
	PHR32	1.67 ± 1.03 abcde	7.00 ± 1.73 a	8.67 ± 0.82 abeg	7.67 ± 3.83 abcdefghijk	0.01 ± 0.01 a	0.01 ± 0.00 a
	PHR58	2.33 ± 1.30 abcde	7.27 ± 1.62 a	7.00 ± 1.48 abcdefgh	6.92 ± 1.16 abcdefghijk	0.01 ± 0.00 a	0.57 ± 0.80 a
	PHR63	3.83 ± 1.99 abcde	6.64 ± 1.63 a	8.67 ± 0.78 e	7.67 ± 1.37 abcdefghijkl	2.04 ± 6.86 a	433.41 ± 1465.63 a
	PHV53	1.00 ± 0.00 a	6.45 ± 2.07 a	6.00 ± 1.04 abcdefgh	6.50 ± 1.09 bcfghik	0.03 ± 0.05 a	33.75 ± 104.14 a
	PHW65	2.17 ± 2.33 abcd	4.44 ± 2.96 a	7.00 ± 0.00 abcdefgh	7.00 ± 1.18 abcdefghijkl	2.19 ± 6.49 a	165.75 ± 492.88 a
	PHW79	2.00 ± 1.60 abcd	6.58 ± 1.78 a	6.00 ± 2.17 abcdefgh	6.67 ± 0.78 abcfghik	0.06 ± 0.11 a	0.22 ± 0.50 a
	PHWG5	1.50 ± 0.90 abc	8.00 ± 1.41 a	8.50 ± 0.90 be	10.50 ± 1.73 l	0.20 ± 0.60 a	10.41 ± 34.16 a
SSS	PH42B	5.67 ± 2.99 bcde	6.08 ± 2.07 a	5.00 ± 1.71 acdfgh	8.92 ± 1.98 adeghijkl	0.13 ± 0.26 a	4.90 ± 10.20 a
	PHR47	2.17 ± 1.34 abcde	7.00 ± 1.56 a	8.00 ± 1.35 abeg	7.92 ± 0.79 abcdefghijkl	0.27 ± 0.57 a	15.22 ± 33.57 a
	PHR61	2.00 ± 1.04 abcde	7.67 ± 1.61 a	7.00 ± 1.21 abcdefgh	9.83 ± 1.80 dejl	6.31 ± 20.88 a	26.50 ± 86.10 a
	PHVA9	1.17 ± 0.58 ac	7.33 ± 1.83 a	7.67 ± 1.56 abdefg	7.50 ± 0.52 abcdefghijkl	3.42 ± 11.66 a	45.91 ± 151.01 a
	PHAG6	1.50 ± 0.90 abc	6.17 ± 1.99 a	7.50 ± 1.51 abcdefg	6.67 ± 0.78 abcfghik	0.20 ± 0.42 a	2.21 ± 6.38 a
	PHBW8	4.33 ± 2.73 abcde	5.17 ± 0.98 a	4.67 ± 0.82 abcdfgh	7.33 ± 1.21 abcdefghijkl	0.01 ± 0.01 a	0.01 ± 0.00 a
	PHEW7	2.67 ± 1.87 abcde	7.80 ± 0.79 a	6.00 ± 1.81 abcdefgh	8.42 ± 1.56 abcdefghijkl	0.04 ± 0.08 a	2.35 ± 5.73 a
	PHHB4	2.83 ± 1.59 abcde	5.50 ± 2.50 a	5.17 ± 1.8 abcdfgh	6.00 ± 0.45 bf	0.09 ± 0.16 a	4.19 ± 8.31 a
	PHJR5	3.67 ± 2.31 abcde	6.92 ± 1.73 a	5.83 ± 1.8 abcdefgh	8.42 ± 1 acdeghijkl	0.23 ± 0.69 a	3.91 ± 12.86 a
	PHK29	2.00 ± 1.04 abcde	6.91 ± 1.76 a	7.33 ± 1.15 abcdefg	8.00 ± 0.85 abcdefghijkl	0.07 ± 0.19 a	0.55 ± 1.00 a
	PHK35	2.00 ± 1.35 abcd	8.10 ± 0.99 a	7.00 ± 1.48 abcdefgh	6.83 ± 0.94 abcdfghijk	0.08 ± 0.13 a	1.79 ± 4.66 a
	PHN29	5.00 ± 2.56 abcde	7.55 ± 1.13 a	6.5 ± 1.24 abcdefgh	7.18 ± 0.75 abcdefghijkl	0.09 ± 0.15 a	3.65 ± 10.29 a
	PHNJ2	1.17 ± 0.58 ac	7.78 ± 0.67 a	7.83 ± 1.34 abeg	8.08 ± 1.83 abcdefghijkl	2.59 ± 5.39 a	520.11 ± 1456.31 a
	PHP38	3.83 ± 1.80 abcde	7.64 ± 1.63 a	6.83 ± 1.99 abcdefgh	6.83 ± 2.44 abcdefghijk	0.08 ± 0.17 a	1.58 ± 3.15 a
	PHT10	4.00 ± 3.02 abcde	6.50 ± 2.33 a	6.00 ± 2.34 abcdefgh	6.00 ± 0.95 bcf	2.59 ± 7.69 a	93.99 ± 243.48 a
	PHT11	1.91 ± 1.04 abcde	7.40 ± 1.07 a	6.82 ± 1.89 abcdefgh	9.55 ± 1.51 adejl	0.02 ± 0.02 a	5.42 ± 8.59 a
	PHT47	2.50 ± 2.43 abcde	6.00 ± 1.89 a	3.50 ± 2.11 h	6.83 ± 1.03 abcdefghijk	0.03 ± 0.04 a	0.56 ± 1.17 a
	PHTP9	1.00 ± 0.00 a	6.25 ± 1.42 a	5.83 ± 1.99 abcdefgh	8.08 ± 1.08 abcdefghijkl	0.40 ± 1.06 a	24.14 ± 60.63 a
	PHTV7	2.5 ± 2.11 abcde	6.73 ± 2.05 a	4.83 ± 2.17 acdfgh	9.08 ± 1.62 adehijl	0.11 ± 0.26 a	1.96 ± 4.11 a
	PHV07	1.00 ± 0.00 a	6.58 ± 1.88 a	5.50 ± 1.51 abcdefgh	7.17 ± 1.11 abcdefghijkl	0.92 ± 2.99 a	88.03 ± 298.01 a
	PHBB3	1.67 ± 1.56 abc	6.18 ± 1.94 a	6.50 ± 0.90 abcdefgh	7.92 ± 1.73 abcdefghijkl	0.15 ± 0.28 a	2.21 ± 3.59 a
	PHG86	5.00 ± 2.95 abcde	5.00 ± 2.68 a	7.33 ± 1.67 abcdefgh	8.09 ± 0.94 abcdefghijkl	0.70 ± 1.50 a	184.56 ± 422.86 a
	PHHB9	2.17 ± 1.59 abcde	6.18 ± 1.72 a	5.00 ± 1.71 acdfgh	7.33 ± 0.78 abcdefghijkl	0.07 ± 0.18 a	0.46 ± 1.56 a
	PHP85	1.73 ± 1.35 abcd	6.40 ± 2.01 a	7.73 ± 1.85 abdeg	6.91 ± 0.94 abcdefghijk	0.01 ± 0.00 a	0.68 ± 0.90 a
	PHPR5	4.67 ± 3.06 abcde	7.33 ± 1.30 a	6.17 ± 1.59 abcdefgh	7.17 ± 1.53 abcdefghijkl	0.14 ± 0.36 a	1.88 ± 3.25 a
	PHW51	2.50 ± 1.51 abcde	6.50 ± 1.73 a	5.00 ± 1.21 acdfgh	8.75 ± 0.75 adehijl	0.12 ± 0.26 a	2.54 ± 5.35 a
	PHW52	2.50 ± 1.51 abcde	5.83 ± 1.85 a	6.33 ± 1.56 abcdefgh	7.08 ± 1.00 abcdefghijkl	1.14 ± 3.61 a	51.89 ± 162.52 a
	PHEG9	1.17 ± 0.58 ac	4.91 ± 1.87 a	4.50 ± 0.90 cdfh	7.73 ± 0.9 abcdefghijkl	4.97 ± 15.60 a	446.73 ± 1468.60 a
	PHGF5	2.82 ± 3.16 abcde	6.38 ± 1.85 a	5.36 ± 1.75 abcdefgh	8.45 ± 1.04 abcdeghijkl	0.03 ± 0.04 a	2.29 ± 5.25 a
	PHJ70	3.00 ± 2.31 abcde	7.13 ± 0.99 a	6.00 ± 1.05 abcdefgh	9.63 ± 1.77 adehijl	0.11 ± 0.27 a	488.22 ± 1379.85 a
	PHMK0	1.00 ± 0.00 a	7.40 ± 1.34 a	8.50 ± 0.90 be	8.64 ± 1.57 abcdefghijkl	0.06 ± 0.13 a	2.62 ± 3.48 a
	PHT55	2.50 ± 2.28 abcde	7.90 ± 1.45 a	7.67 ± 1.30 abdefg	7.83 ± 1.19 abcdefghijkl	0.06 ± 0.10 a	3.10 ± 6.81 a

shows the averages of all the parameters measured with the relative standard deviation.

High variability in measurements was found among these materials; the number of husks ranged from an average of 5.50 ± 1.57 (PHBR2) to 10.50 ± 1.73 (PHWG5), and the husk coverage score ranged from 3.5 ± 2.11 (PHT47) to 8.67 ± 0.82 (PHR32), while the ear attitude score ranged from 1 ± 0 (PHBG4, PHMK0, PHP60, PHTP9, PHV07, PHV53) to 8.56 ± 0.88 (PHR31).

As reported in the Section 2, the harvested ears were visually scored to evaluate F. graminearum infection. This score, on average, ranged between 4.44 ± 2.96 (PHW65) and 8.25 ± 1.04 (PHGV6), with the higher score associated with less fungal infection on the ear.

The average of the infection score parameter by location was lower in the 2022 season compared with the 2023 season (

Table 3. F. graminearum infection score average (x⁻) and standard deviation (σ) for the three locations in two different years. Different letters indicate statistically significant differences according to Dunn’s test (p < 0.01).

Location	Year	x−	σ
Location 1	2022	5.16 a	1.90
	2023	6.94 b	1.97
Location 2	2022	6.51 c	1.71
	2023	7.25 d	1.89
Location 3	2022	6.49 e	1.40
	2023	7.36 f	1.76

), indicating higher infection rates in the first testing season.

Mycotoxin analysis showed an almost complete absence of DON and ZEA for the first testing season, except for rather few samples, namely 16 for ZEA and 3 for DON (Figure 1). In 2023, the mycotoxin concentration was higher, even if only 19 samples for the ZEA analysis and 35 for the DON analysis (1.75 ppm) were above the threshold set by the European Union for human consumption (350 ppb) (Table S2).

3.2. Multi-Year Multi-Site Data Correlation

The collected data were not normally distributed according to a Shapiro–Wilk test (Table S3). Therefore, a Spearman’s rank correlation test was used to find possible relations between the ear parameters (

Table 4. Spearman’s rank correlation between ear parameters and mycotoxin concentration. DON = deoxynivalenol concentration (ppm); EAS = ear attitude score; FGIS = Fusarium graminearum infection score; HCS = husk cover score; HN = husk number; ZEA = zearalenone concentration (ppb). * Correlation significant at p < 0.05.

Material and Year	Variable	DON (ppm)	EAS	FGIS	HCS	HN	ZEA (ppb)
NSS 2022	DON (ppm)	1
	EAS	0.05	1
	FGIS	−0.07	0.02	1
	HCS	−0.02	−0.22 *	0.19 *	1
	HN	−0.04	−0.31 *	0.20 *	0.34 *	1
	ZEA (ppb)	0.32 *	0.07	0.03	−0.04	−0.01	1
NSS 2023	DON (ppm)	1
	EAS	−0.18 *	1
	FGIS	−0.08	0.00	1
	HCS	0.02	−0.19 *	−0.02	1
	HN	0.09	−0.17 *	0.07	0.20 *	1
	ZEA (ppb)	0.70 *	−0.18 *	−0.07	0.01	0.02	1
SSS 2022	DON (ppm)	1
	EAS	0.09	1
	FGIS	−0.12	−0.06	1
	HCS	0.08	−0.07	0.14	1
	HN	−0.04	−0.03	0.21 *	0.09	1
	ZEA (ppb)	0.21 *	0.00	0.01	0.00	0.04	1
SSS 2023	DON (ppm)	1
	EAS	−0.15	1
	FGIS	−0.06	0.07	1
	HCS	0.08	−0.08	0.19 *	1
	HN	−0.08	−0.06	0.02	0.01	1
	ZEA (ppb)	0.68 *	−0.19 *	−0.10	0.06	0.15	1

).

Correlation analysis was performed while considering the materials’ genetic backgrounds, which in this case were their heterotic group (NSS and SSS) and also taking into account the testing year.

Positive and significant correlation between the DON and ZEA concentrations was found in both heterotic groups and for both testing years (NSS 2022: p < 0.05, r_s = 0.32; NSS 2023: p < 0.05, r_s = 0.70; SSS 2022: p < 0.05, r_s = 0.21; SSS 2023: p < 0.05, r_s = 0.67). A significant negative correlation was found between the ear attitude and ZEA concentration in both heterotic groups for the 2023 season (NSS 2023: p < 0.05, r_s = −0.18; SSS 2023: p < 0.05, r_s = −0.19). The ear attitude was also found to significantly correlate negatively with the husk cover score (NSS 2022: p < 0.05, r_s = −0.22; NSS 2023: p < 0.05, r_s = −0.19) and husk number (NSS 2022: p < 0.05, r_s = −0.31; NSS 2023: p < 0.05, r_s = −0.17) for the NSS materials in both testing years. In the 2022 season, for both heterotic groups, a significant positive correlation was found between the F. graminearum infection score and husk number (NSS 2022: p < 0.05, r_s = 0.20; SSS 2022: p < 0.05, r_s = 0.21). The F. graminearum infection score was also found to have a significant positive correlation with the husk cover score in the NSS heterotic group in 2023 and SSS in 2022 (NSS 2023: p < 0.05, r_s = 0.19; SSS 2022: p < 0.05, r_s = 0.19).

3.3. QTL Analysis

QTL analysis was performed using phenotypic data recorded over 2 years (2022 and 2023) in three locations, with two replications each. Ear characteristic QTL identification was performed separately for the SSS and NSS materials (

Table 5. Quantitative trait locus (QTL) mapping of ear traits. EAS = ear attitude score; FGIS = Fusarium graminearum infection score; HCS = husk cover score; HN = husk number. Chromosome positions and genome positions are expressed in centimorgans (cM). The effect and standard error of the effect (SE effect) are expressed in each trait’s unit of measure. Markers listed are the ones which were above the −log (p value) threshold of three.

Heterotic Group	Trait	Model	Marker	Chromosome	Chromosome Position (cM)	Genome Position (cM)	Effect	SE Effect	−log (p Value)
NSS	EAS	GWAS	C10550D-001	7	185.98	1675.98	1.33	0.4	3.07
	HCS	GWAS	C104E04-001	1	264.81	264.81	1.01	0.3	3.16
			C1052NP-001	6	86.75	1392.75	−0.86	0.25	3.27
	HN	GWAS	C104W1Y-001	3	153.15	732.15	−1.02	0.31	3.08
			C104XB3-001	4	110.41	940.41	−1.07	0.31	3.19
			C1054NE-001	7	163.06	1653.06	−0.96	0.26	3.58
			C104PTU-001	10	71.27	2173.27	−0.88	0.26	3.13
SSS	EAS	GWAS	C104NB1-001	1	79.39	79.39	−1.48	0.42	3.33
			MZA15082-13	4	13.06	843.06	−1.57	0.47	3.04
			C104XFF-001	4	117.82	947.82	−1.25	0.32	4.02
			C104XTF-001	4	148.22	978.22	−1.31	0.37	3.4
			C105186-001	5	181.04	1259.04	−1.31	0.39	3.07
			MZA15414-29	6	52.79	1358.79	−1.63	0.43	3.76
			C104EJV-001	7	34.26	1524.26	−1.29	0.39	3.01
SSS	FGIS	GWAS	C104UT6-001	2	109.39	431.39	0.19	0.06	3.17
			C105025-001	4	97.58	927.58	0.17	0.05	3.27
			C1053JV-001	6	24.03	1330.03	0.16	0.04	3.83
			C1053E9-001	6	180.5	1486.5	0.19	0.06	3.14

, Figure 2).

The best linear unbiased prediction (BLUP) values for the ear traits were calculated to conduct a GWAS with 24,000 SNPs. The threshold for QTL detection was determined as a −log (p value) equal to three. No peak was selected if it was closer than 10 cM left or right to an already-selected peak.

Seven QTLs for the ear traits were found for the NSS heterotic group. Four QTLs for the husk numbers were located on chromosomes 3, 4, 7, and 10, with an effect which ranged from −0.43 to −0.25 in the number of husks. For the husk cover score, two QTLs were located on chromosomes 1 and 6, with effects on the score of 0.62 and −0.56, respectively. A QTL for the ear attitude score was located on chromosome 7, with an effect of 1.33 on the ear attitude score scale. In the SSS heterotic group, different QTLs were found for the ear attitude score and infection score. Seven QTLs were found for the ear attitude score on chromosomes 1, 4, 5, 6, and 7. The effects of these markers ranged between −1.63 and −1.25 on the scale of the score. Four QTLs were found for the infection score, with an effect on the score which ranged between 0.16 and 0.19.

4. Discussion

This work aimed to evaluate the ear’s characteristics and their impact on the mitigation of GER infection in 74 maize inbreds selected for their adaptation to the northern Italian environment. A quantitative trait loci analysis was performed to understand the genetic background of these ear traits.

In this project, different parameters were collected to evaluate the ear traits, and the data of two seasons were analyzed to understand possible correlations between the characteristics. The F. graminearum infection score severity was higher in the 2022 season compared with 2023 in all locations, but extremely small differences were found in the mycotoxin concentrations in both seasons, and no correlation was found between the infection score and mycotoxin concentration for either DON and ZEA. A positive and statistically significant correlation was therefore found between the DON and ZEA concentrations. This correlation was driven mostly by the few samples which presented high mycotoxin concentrations, confirming the already-known fact that F. graminearum can produce both of these mycotoxins [52,53].

To prevent fungal infection, different methods can be implemented, and among them, the most common are cropping techniques, crop residue management, and chemical or biological control [17,30,54,55]. However, with the optics of input reduction and environmental preservation, the best method for controlling F. graminearum is the selection of resistant genotypes, even if no completely resistant materials are available [56].

From the standpoint of the selection of resistant materials, it was observed that the ear morphology plays an important role in the susceptibility of corn to fungal diseases [38,39,57]. In this study, different ear characteristics were taken into account. Positive statistically significant correlations, even if not strong (r_s = 0.20 in NSS materials and r_s = 0.21 in SSS materials), were found between the F. graminearum infection score collected in terms of a reverse scale, where a higher number means less infection, and the husk number. A higher husk number can be associated with less insect ear damage due to the barrier effect which it represents [58]. Another husk trait found to be positively correlated (r_s = 0.19 in NSS in 2023; r_s = 0.19 in SSS in 2022) with the F. graminearum infection score was the husk cover score. Husks which can cover the ear completely play an important role in protection against both insect and fungi, as reported by several studies [59,60,61].

Other significant correlations found were the one between the ear attitude and husk cover score and between the ear attitude and husk number, but these were only in the SSS heterotic group. Even if no correlation between the F. graminearum infection score and ear attitude was found, it has been seen that these traits are correlated with less ear rot infection due to the pendant position, which accelerates kernel dry down [62].

The genetic basis of the ear characteristics is still not well understood due to the quantitative nature of these traits. To investigate this, we performed a genome-wide association study (GWAS) to identify possible QTLs associated with ear traits which can be useful for selecting resistant against materials GER. Other studies have already explored the presence of possible QTLs associated with husk morphology, and a wide range of markers was found [63].

In this work, four QTLs were found to be associated with the husk number, and two were associated with the husk cover. Other authors have already explored these traits, but none of the markers found in our study were found before. These other studies also explored a wide range of maize agronomic characteristics like the days to silk, kernel dry down, silk resistance, kernel resistance, and specific husk traits like the husk length and width. None of these studies correlated their findings with resistance to GER, but they were concentrated on identification of the molecular pathways involved in husk development. The total phenotypic variation explained by the QTLs found in these studies which was associated with the husk number and husk cover traits ranged between 8.9% and 44.5% for the husk number and between 3.49% and 10.19% for the husk cover score [57,64,65,66]. Candidate genes related to metabolism, gene expression regulation, signal transduction, and flowering time regulation were also found in these studies [64,66].

For the first time, eight QTLs were found to be associated with the ear attitude score in both heterotic groups, namely one in the NSS group and seven in the SSS group. Four QTLs were also found to be associated with the F. graminearum infection score trait in the SSS heterotic group associated with resistance to GER. Like previous traits, our QTLs were found in different genome positions compared with other works which explored the scores associated with infection resistance. The total phenotypic variation explained ranged between 0.46% and 21.80% in these different studies [42,43,44]. In these other works, interesting QTLs associated with GER resistance were found. Specifically, Galiano-Carneiro reported a stable QTL across environments and populations useful for increasing GER resistance. Other studies reported finding different markers which could be used in genomic prediction-assisted breeding [42,43,44]. Candidate genes were also found, but in this case, they were associated with responses to broad spectrum resistance to fungi, bacteria, and oomycetes [42,67].

All of these newly found QTLs can be explained by the difference in materials, environments, and growing seasons in our experiment, confirming the strong influence of genotype–environment interaction which lies behind the phenotypic expression of all these characters [47]. However additional studies are needed to better understand the importance of these markers and their effect on resistance against GER. The use of larger populations of inbred lines or studying the resistance in hybrids can greatly improve our understanding of these highly complex traits.

In conclusion, the current study reported the relationships between the ear characteristics and their correlation with F. graminearum infection and identified QTLs unique to each heterotic group associated with the ear morphology and disease resistance. Correlation was observed between the husk number, husk cover, and infection score. Another significant correlation was observed between ear attitude and both the husk cover and husk number. Our results can help with the selection of more resistant materials, and from a marker-assisted selection standpoint, these QTLs can be useful for breeders in their activity. These findings can be crucial for better control of fungal diseases without the use of additional inputs, with positive effects on the environmental and economical sustainability of corn growing systems.