1. Introduction
The Hessian fly (HF), or
Mayetiola destructor Say [
1], is one of the oldest recorded invasive species in North America. It can cause substantial economic damage wherever wheat is grown [
2]. The US experienced major HF epidemics in the past, so the government started programs to control this catastrophic pest in large wheat-growing regions [
1]. Sixteen million acres of HF-resistant wheat were planted nationwide in 1974 [
1]. HF infestation usually lowers grain yield more than quality. If over five percent of tillers are infested during the early tillering stage, yield loss can be considered significant [
3]. HF has caused millions of dollars in damage to US wheat. South Carolina lost approximately
$4 million annually from 1984 to 1989, and Georgia lost roughly
$20 million from 1988 to 1989 [
4]. HF can cause annual GA field losses of up to 10% [
5]. In Plains and Tifton, GA, single-stem samples of wheat averaged 1.97 larvae for each infested stem [
5]. HF lowered the average grain weight of infested stems by 41.3% [
5]. In Oklahoma, one immature HF per tiller can lead to approximately 31.27 kilograms (kg) per hectare (ha) in losses [
4]. This pest can infest quack grass [
6], barley, rye, and wheat [
7].
HF adult females may lay 100–400 eggs on leaf adaxial surfaces for around three hours [
8]. After three to four days, larvae emerge from the eggs at 20 °C [
8] to crawl down to the closest node to feed for two to three weeks in the first and second instar stages [
4]. Larvae are more likely to survive on younger leaves because there are more responsive cells that larvae can convert to galls for better nutrition [
9]. Third-instar larvae pupate, and this stage lasts for 7–35 days [
4]. However, pupae can remain dormant for three to four months in wheat stubble [
10]. If temperatures stay at least 21 °C and humidity remains high for 10–14 days, adults usually emerge from pupae to mate and lay eggs. This temperature is ideal for HF to develop [
4]. The larval stage is when damage is inflicted on wheat. As larvae feed, they turn the base of wheat plants into nutritive tissue, stunting their tiller growth [
3]. They can also cause lodging, smaller wheat kernels and spikes, and fewer kernels per spike [
1]. Other infestation symptoms include unusually short leaf blades, sheaths, and internodes, as well as darker green leaf color [
3].
Resistant cultivars are the most cost-effective control option [
3], especially in the Southeast (SE) of the US where fly-free dates are less effective [
11]. The SE has a more optimal climate for HF to produce more generations than in other regions [
11]. While 37 HF R genes were identified [
12], only a few, such as
H13, work well in the SE. Genes that used to confer higher resistance to HF in soft red winter wheat (SRWW) in the SE, including
H3,
H5,
H6,
H7, and
H8, are no longer as effective [
13].
H9 is losing its efficacy in parts of the SE, and
H18 is temperature-sensitive [
14]. Microsatellite data [
15] and virulence assays [
13] indicated that there is just one major population of HF in the SE US with population structure as well as microscale diversity [
13,
15]. HF populations from Holmes County, Mississippi, and Florence County, South Carolina, had a different population identity when compared to populations from other counties from SE states [
16]. The
H12,
H13,
H18,
H24–
H26,
H31–
H33, and
Hdic genes are still effective in multiple SE counties, but
H24–
H26 can be associated with undesirable agronomic traits [
14]. Recently discovered quantitative trait locus (QTL),
QHft.nc-7D, has been linked with partial field resistance in North Carolina (NC) [
17]. More sources of resistant germplasm are needed to combat HF and avoid overcoming the few available effective R genes in the SE.
Several diversity panels as well as biparental populations were developed to conduct GWAS and QTL analysis and identify genomic regions involved with HF resistance. A diversity panel of hard red spring (HRS), soft white spring (SWS), and soft white club spring (SWC) wheat evaluated for seedling HF resistance in Moscow, Idaho (ID), revealed
IWA6803, a significant SNP closely linked to
H34 on chromosome 6B, and a novel QTL,
QHf.pnw.2B, on chromosome 2B [
18]. Winter wheat diversity panel AM203 in Manhattan, Kansas, was used to validate KASP-3B3797431 and KASP-3B4525164, which could be near diagnostic markers to detect
QHf.hwwg-3B, a QTL on chromosome 3B mapped to 6.79 Mb, explaining up to 46.7% phenotypic variation (PV) for HF resistance [
19]. Analysis from a diversity panel, biparental population, and elite ICARDA lines all of durum wheat revealed
QHara.icd-6B, a locus explaining 83% PV with a 54.5 logarithm of odds (LOD) value that did not demonstrate yield drag when evaluated across locations [
20].
Recombinant inbred line (RIL) population Seneca, developed from a cross between HF-resistant spring wheat variety Bobwhite and winter wheat variety Seneca (CI 12529), was used in conjunction with genotyping-by-sequencing (GBS) SNPs for mapping to reassign
H7, a major gene explaining up to 78.3% PV, from chromosome 5D to 6A [
21].
H35 from chromosome 3BS and
H36 from chromosome 7AS were two HF resistance genes detected using a 154 RIL population generated from resistant HRWW line SD06165 and susceptible line OK05312.
H35 was a major QTL explaining up to 36% PV, and
H36 was a minor QTL explaining up to 13.1% PV [
12]. The major QTL
QHf.wak-1A was mapped in a registered spring wheat RIL population [
22] produced from a cross between resistant line Louise and susceptible line Penawawa, which explained up to 90% PV for HF resistance [
23].
Breeding efforts in the US have led to the release of several HF-resistant cultivars. In the SE region, resistance to HF is necessary, and breeding programs releasing cultivars for this region must incorporate genes for HF resistance in the newly developed cultivars. The UGA breeding program has released numerous cultivars adapted to the SE region with
H13 and
H9 genes that confer resistance to HF in GA and the SE. This includes recently released cultivars in 2020 (AGS 2021, PGX 20-15, and AP 1983) and in 2022 (AGS 3026, AGS 4023, and USG 3725) (Mergoum, Personal communication).
H24 is one of the R genes still highly effective against SE HF field collections [
14]. HF resistance on wheat line KS89WGRC06 was deemed to be governed by
H24 on the long arm of chromosome 3D via monosomic analysis [
24,
25] and RFLP marker validation in the early 1990s. The
H24 linked RFLP markers were
XcnlBCD451,
XcnlCDO482, and
XksuG48 [
26].
Chromosome 3D also has R genes
H26 and
H32, derived from wheat lines KSWRCG26 and W-7984, respectively [
14]. Seedlings with single R genes
H24,
H26, or
H32 exposed to HF populations from AL, GA, and NC demonstrated 75–100%, 87.8–100%, and 83.2–99.5% resistance, respectively [
14].
H32 has been mapped in between flanking simple-sequence repeat (Xgwm3 and Xcfd223) [
27], sequence-tagged site (
Xrwgs10 and
Xrwgs12) [
28], and SNP markers (
IWB65911 and
IWB37580) [
29].
Xrwgs10 and
Xrwgs12 are also tightly linked to
H26 [
30], so
H26 and
H32 can be easily introgressed together [
10]. Since
H24 and
H26 can be associated with unideal agronomic traits due to linkage drag,
H32 may be an alternative that does not lower yield as much [
14].
Kompetitive Allele-Specific PCR (KASP) markers were created for some R genes to quickly, cheaply, and accurately screen cultivars and accelerate marker-assisted selection (MAS) for plant breeding. KASP primer sets were developed for SNP
IWB65911 that cosegregates with
H32, and it has differentiated HF-susceptible cultivars from resistant ones with high sensitivity as well as specificity [
29]. R genes
h4,
H7,
H35, and
H36 also have KASP markers, but validation is needed before using these markers for MAS [
10]. KASP-6B7901215 and KASP-6B112698 were validated to deploy
QHf.hwwg-6BS, a major QTL explaining up to 84% PV that was derived from the cultivar Chokwang [
31]. Using KASP markers and other techniques, such as crossing durum wheat to bread wheat and doubling F1 chromosomes via colchicine, can expedite the introgression of new HF R genes [
10,
32].
UGA 111729 and AGS 2038 are elite SRWW breeding lines developed by the University of Georgia (UGA) Small Grains Breeding Program (UGA-SGBP). AGS 2038 was developed from a cross between Pioneer 26R61 and GA 961581, and it was released to AGSouth Genetics in 2011 [
33]. Pioneer 26R61 has HF resistance QTL
QHf.uga-2AS,
QHf.uga-3DL, and
QHf.uga-6AL on chromosomes 2A, 3DL, and 6A, respectively.
QHf.uga-6AL is a major QTL flanked by SSR marker
Xgwm427 and DArT marker wPt-731936 that explained up to 63% average PV.
QHf.uga-2AS is a minor QTL flanked by SSR markers
Xgwm359 and
Xbarc124.
QHf.uga-3DL was considered different from
H24,
H26, and
H32 and significant during the late seedling growth stage since PV and LOD values increased as seedlings aged [
34]. UGA 111729 was developed by backcrossing KS89WGRC06 to AGS 2038. Thus, it is presumed to carry
H24. These two cultivars were crossed to develop a biparental RIL population that was first studied for leaf rust resistance [
33].
Therefore, in this study, the objectives were to identify genomic regions involved with HF resistance using the SRWW UGA 111729 × AGS 2038 biparental RIL population for QTL analysis and to determine the most significant marker intervals influencing HF resistance that could be used for MAS.
4. Discussion
HF is a highly damaging insect species to wheat in the US SE. Only six R genes are highly effective for that region, and three of them may lower agronomic traits [
14]. HF can easily overcome these genes under high host selection pressure for HF virulence [
32]. HF biotypes are constantly evolving and overcoming introduced R genes. Biotype L HF is currently the dominant biotype in the US SE [
17]. While HF R gene
H13 is still effective against it [
53], biotype
vH13 is overcoming this R gene, necessitating a search for novel resistance [
17]. HF-resistant cultivars were demonstrated during an infestation to save
$100/ha–
$240/ha in damages compared to HF-susceptible cultivars [
54], so finding novel HF resistance benefits farmers.
In this study, an SRWW biparental population was used to identify genomic regions involved with HF resistance. Yearly Res X
2 values ratio were lower for a 1:1 segregation than a 1:1:1:1 ratio, meaning observed Res values were more likely to fit 1:1 ratio X
2. This result means that one major QTL expressing HF resistance in UGA 111729 is more likely than multiple QTL. This is similar to X
2 results from Zhang et al. [
31], which found a major QTL that explained resistance to HF in the cultivar Chokwang.
Our QTL results and KASP and phenotypic validations revealed one major gene for HF resistance in 3DL associated with the
IWB65911 marker. The HF resistance in UGA 111729 was detected both in the growth chamber and field trials, suggesting it is expressed from the seedling to the adult stage.
IWB65911 was used for previous KASP marker validation for HF studies, and they co-segregated with
H32 [
29]. This finding is interesting, considering that UGA 111729 is supposed to carry
H24, inherited from its progenitor KS89WGRC06.
H24 and
H32 are at least 20 cM away from each other [
27]. Since Tan et al. [
29] reported that
IWB65911 demonstrated a specificity of 1 and sensitivity of 0.93–0.94, our findings validate the efficacy of
IWB65911 due to our 96.3% result. Since
H24 does not currently have a publicly available SNP marker, RFLP marker validation with flanking markers,
Xcdo428 and
Xbcd451, would need to be conducted to confirm that UGA 111729 has
H24 [
26]. Since
Xrwgs10 and
Xrwgs12 are linked to
H26 as well as
H32, STS marker validation can be used to determine if UGA 111729 also has
H26 [
28].
As for our candidate genes from significant QTL,
IWA6387, associated with
QHf.ga.srww.3A, was a flanking SNP with SSR marker
Xbarc12 as part of an additive QTL,
QShi.hwwgr-3AS, that explained up to 5.6% PV for wheat grain quality in a F
10–12 RIL winter wheat population [
55]. For
QHF.ga.srww.6B, no candidate genes were found for SNPs
IWB62788 and
IWB59262.
QHf.ga.srww.6B was a major growth chamber QTL in this study; however, it was only detected in one replicate for 2019 results and not for 2021.
QHf.ga.srww.6B should be further investigated.
Data from Plains had higher heritability than Williamson, indicating Williamson had higher environmental variance. Res and PIT
H2 were higher than all other traits for growth chamber data, meaning these two traits are more replicable than the other traits. In Plains, genetic causes from multiple genes can explain NOPPT and PIT better than NOPIT since NOPPT and PIT had higher
H2 and
h2. Winn et al. [
17] had a similar observation and suggested that PIT and NOPPT continue to be used to assess pest instance and pest severity, respectively. This study is the first to assess the correlations between PIT, NOPPT, NOPIT, and Res for HF resistance. All Res replicates and averages across years were negatively correlated with all other trait replicates and averages across years. This is expected since higher values for Res mean higher resistance vs. lower resistance for those higher values for PIT, NOPPT, and NOPIT. Res was the most strongly correlated with PIT across the years. When looking at correlations for averages within each individual year, PIT vs. NOPPT consistently had stronger correlations than PIT vs. NOPIT. PIT is likely to help determine how high or low NOPPT values will be. PIT should be the priority trait when phenotyping because it is easier to assess than NOPPT and NOPIT and it is highly correlated with NOPPT and Res.
The parents had smaller differences between the measured traits in Williamson vs. Plains due to lower insect pressure. There was evident G × E interaction between Plains and Williamson field results. One explanation for this G × E interaction and differences in insect pressure could be the number of acres planted in Pike County and Sumter County, where Williamson and Plains are located, respectively. The most recent publicly available data on acres harvested for individual Georgia counties dates to 2017. For 2012, in which there is data for both counties, Sumter County harvested 11,133 acres of wheat, and Pike County harvested 492 acres of wheat. In 2017, Sumter County harvested 2,523 acres, while there is no available information for Pike County (
https://www.nass.usda.gov (accessed on 9 August 2023)). HF are more likely to reproduce in areas with more wheat planted and warmer climates (Mergoum Lab, Personal communication) [
3,
4]. There could have also been a difference in biotype composition per county, considering that lines with
H32 were shown to be more effective against Sumter County biotypes than Tift County biotypes [
13]. However, Cambron et al. [
13] did not have any results for the effect of Pike County biotypes on lines with
H32.
There is not much literature to directly compare our LD decay results among RIL populations for HF response QTL. Bassi et al. [
20] studied durum wheat, and Ando et al. [
18] and Joukhadar et al. [
56] used bread wheat diversity panels, but they did not compare subgenome or individual linkage group LD decay. Also, these studies did not assess PIT, NOPPT, or NOPIT. Pariyar et al. [
57] used LD decay analysis with GWAS, considered 0.1 as their
r2 critical value, and had LD decay values of 2 cM and 6 cM for chromosomes 3A and 3D, respectively. Although their 3A LD decay value was smaller than ours at 23 cM, our 3D LD decay value was equal to theirs [
57].
H32 is still effective against HF biotypes in the US SE [
14], which was also confirmed in our study. Despite KS89WGRC06 (known to carry
H24) being a progenitor of UGA 111729, this paper validated the presence of
H32 in UGA 111729. This novel finding is valuable, considering that a KASP marker was developed for
H32 detection. Since the US SE is losing effective HF R genes,
IWB65911 should be used for MAS to introgress
H32 into new varieties for HF resistance, and the effect of
H32 on yield should be evaluated.
H32 should also be pyramided with other HF R genes for better resistance management against quickly evolving biotypes. This study demonstrates the efficacy of
QHf.ga.srww.3DL and that breeders can use
IWB65911 for MAS.