Introduction

Elite crop plants are often more susceptible to insect pests than their wild progenitors and landraces, due to a reduction in plant defense caused by domestication, artificial selection and breeding (Chen et al. 2015; Dávila-Flores et al. 2013; Rosenthal and Dirzo 1997). For this reason, landrace accessions have been routinely mined for insect resistance traits in several crop species (Arnason et al. 1994; Laamari et al. 2008; Liu et al. 2009; Pelgrom et al. 2015; Valdez et al. 2012). Despite their value as genetic resource, mining plant defense traits in landraces is not an easy task, mainly because of the sheer number of accessions stored in germplasm repositories. A targeted geographical screening approach of landrace accessions was shown to be more successful, because geography underlies plant–insect interactions (El Bouhssini et al. 2011; Chen et al. 2015). Thus searching for new sources of insect resistance in geographic regions of the world where a plant species and its wild progenitors shared the longest evolutionary history with an insect pest can increase the likelihood of success.

Several species of grass sawflies in the family Cephidae infest wheat and other cultivated cereals throughout the world. Some of the most important species are in the genus Cephus. In North America, on the Canadian Prairies and the Northern Great Plains this is Cephus cinctus Norton, a native species also referred to as the wheat stem sawfly (WSS), which likely adapted to wheat from grass hosts already present in the prairie ecosystem shortly after tillage first began in the 1800s (Beres et al. 2011; Lesiuer et al. 2016). In east Asia, Cephus fumipennis Eversmann is the native and dominant species that infests cereal crops (Chen et al. 2004), while in Western Europe, North Africa and the Middle East, Cephus pygmaeus (L.), the European wheat stem sawfly, is the main pest of cereals (Shanower and Hoelmer 2004). In Eastern North America, the cosmopolitan C. pygmaeus is an introduced species that is infrequently encountered due to successful classical biological control using the parasitoid Collyria calcitrator (Gravenhorst) (Shanower and Hoelmer 2004). The life cycles of these closely related species of sawflies and the damage caused by their interaction with wheat are indistinguishable. Females lay eggs in the stem, where the larvae feed on plant tissue reducing grain quality and yield (Morrill et al. 1992, 1994). Prior to plant senescence, mature larvae cut a groove around the base of the stem just above their overwintering chamber. Cut stems lodge and are not harvested (Criddle 1922). Areas of the world infested by C. pygmaeus represent some of the earliest instances of wheat cultivation. Thus, landraces collected from these areas have been under pressure from wheat stem sawflies since the dawn of wheat domestication, and their wild progenitors before that.

Research on C. pygmaeus in the 1920s revealed that less damage was caused in plants that had stems filled with pith, referred to as solid stems (Schegolev 1926). In the following decade, screening efforts identified a Portuguese landrace accession, S-615, carrying the solid stem trait (Kemp 1934). This accession was used to develop ‘Rescue’ (CItr12435), the first solid-stemmed wheat variety (Platt et al. 1948). Stem solidness was shown to lessen plant damage by acting as a physical barrier reducing WSS egg deposition, larval movement and survival (Holmes and Peterson 1962; Wallace and McNeal 1966). As might be expected, the solid stem trait also provided resistance to C. fumipennis (Chen et al. 2004). In the 1950s, another screening effort involving 14,000 landrace accessions was undertaken by the USDA-ARS. Unfortunately, published information from these trials is sparse and lack of follow-up was probably related to the identification of solid stems as a simply inherited target for selection. Over the last six decades, the solid stem trait has been the base for WSS integrated pest management (Beres et al. 2011). The trait is conveniently measured in breeding programs at or near plant maturity.

The prevalence of solid stems in wheat from areas of the world with historic infestations of wheat stem sawflies, coupled with a limited understanding of its genetic basis, has impeded the identification of new sources of resistance. Recent research has elucidated the genetic architecture of the solid stem trait. Solid stems in hexaploid wheat cultivars is controlled by the solid stem quantitative trait locus (QTL), Qss.msub-3BL, which accounts for 70% of the genetic variation (Cook et al. 2004; Sherman et al. 2010; Kalous et al. 2011; Varella et al. 2015), and the allele for solid stems traces to landrace S-615. This locus has also been identified in durum wheat (T. turgidum var. durum) and is referred to as SSt1 (Houshmand et al. 2007). Varella et al. (2016) identified a source of resistance in the hexaploid wheat cultivar ‘Conan’ (PI607549) (WestBred, LLC). Resistance in Conan was due to both antibiosis, whereby the plant had a deleterious effect on pest development and survival, and antixenosis, whereby the plant had a deleterious effect on colonization by the pest. The resistance allele was also mapped to Qss.msub-3BL, and conferred high stem solidness early in plant development when WSS females are ovipositing and a low level of stem solidness at maturity. Cook et al. (2017) showed that the allele for early stem solidness had a different haplotype based on SNP markers than either the S-615-derived solid stem allele or the common alleles for hollow stems.

Besides the use of solid stem cultivars for WSS control, integrated pest management for this insect pest also involves enhancement of natural biological control. In wheat fields of Western North America, WSS is killed by two specialist congeneric larval parasitoids, Bracon cephi (Gahan) and B. lissogaster Muesebeck (Morrill et al. 1994; Weaver et al. 2004, 2005; Cárcamo et al. 2012), while in other regions of the world, C. pygmaeus and C. fumipennis are mainly killed by egg-larval parasitoids of the genus Collyria (Wahl et al. 2007, Shanower and Hoelmer 2004). Attempts at classical biological control of WSS using Collyria coxator (Villers) from C. pygmaeus (reviewed in Shanower and Hoelmer 2004) and Collyria catoptron (Wahl) from C. fumipennis failed due to host incompatibility (Rand et al. 2016a), even with favorable attributes for success (Rand et al. 2016a, b). An advantage of braconid species is that they kill damaging larvae, reducing yield loss (Buteler et al. 2008), while egg-larval parasitoids do not kill until the next crop season (after WSS larvae have cut the stem) and ongoing damage is not attenuated. Several other parasitoid species also kill WSS in feral grasses, but these are generalist parasitoids (Ainslie 1929). However, there are records of other species of Braconidae occurring in varying numbers in wheat fields infested by both C. pygmaeus (reviewed in Shanower and Hoelmer 2004) and C. fumipennis (Chen et al. 2004). Indirect defense due to braconid parasitoids is a desirable trait that can be readily detected during stem dissections, but only if the ambient populations of WSS and these parasitoids are both abundant.

The discovery of resistance unassociated with a high degree of stem solidness, coupled with the development of haplotype profiles for the known resistance gene, provides a basis for determining the potential novelty of newly identified sources of WSS resistance. In addition, research over the past decades has identified areas of the world most heavily impacted by wheat stem sawflies. The renewed interest in finding new types of resistance to WSS prompted a targeted screening of wheat landrace accessions from six distinct geographical regions of the world. A detailed characterization of WSS infestation, larval mortality and larval mortality due to parasitism allowed for identification of wheat landrace accessions showing antixenosis and/or antibiosis, as well as accessions displaying indirect plant defense characterized by increased recruitment of parasitoids. Resistant landrace accessions were genetically screened for the solid stem QTL, Qss.msub.3BL, to identify resistant accessions carrying the hollow stem haplotype for Qss.msub.3BL or other unique haplotypes, which may display resistance due to mechanisms other than stem solidness. This research lays the basis for a comprehensive discovery of novel direct and indirect defense genes for WSS resistance in wheat.

Materials and methods

Landrace accessions

To maximize the likelihood of identifying new genes for WSS resistance, selection of landrace accessions targeted geographical areas of the world where wheat stem sawflies are known to be endemic or where a high frequency of landrace accessions have been shown to possess the resistance characteristic of solid stems (Smith 1961; Wallace and McNeal 1966; Wallace et al. 1969; Damania et al. 1997; Chen et al. 2004; Frolov and Saulich 2004; Hoelmer and Shanower 2004; Shanower 2008; Korkmaz et al. 2010). In total, seeds from 1409 hexaploid wheat (Triticum aestivum L.) landrace accessions were obtained from the USDA-ARS National Small Grains Collection (NSGC) (Aberdeen, ID). Accessions originated from 30 countries located in South/Central/East Asia (n = 606), North Africa (n = 107), Europe (n = 35), Iberian Peninsula (n = 281), Levant (n = 215), and Persian Gulf Region (n = 165).

Screening of landraces in WSS-infested sites

Landraces were planted in late April, in the midst of stubble from a previous WSS-infested wheat crop, at sites with a history of WSS infestation located near Amsterdam, MT, USA, in 2012, 2014, and 2015, and at a second site near Loma, MT, USA, in 2013. Each landrace accession was screened in a single year. Plots consisted of ten seeds per entry planted in individual hills with spacing of 0.8 m between adjacent hills. Trials were conducted using an augmented design in which entries were unreplicated and checks were replicated in each block. Check entries included the WSS susceptible varieties, ‘Reeder’ (PI613586) and ‘McNeal’ (PI574642), and the solid-stemmed WSS resistant variety, ‘Choteau’ (PI633974). Accessions were scored for stem solidness at maturity by rating a cross-section of the third or fourth internode of a stem using a 1 (completely hollow) to 5 (completely solid) scale, as described by Varella et al. (2016). At maturity (Zadoks growth stage 99, ripening), plots were visually evaluated for percent cut stems. Accessions that had percent cut stems less than the susceptible check entries were collected and dissected to determine WSS infestation, larval mortality, and parasitism for further characterization as per Talbert et al. (2014). Accessions not harvested were characterized as susceptible. Harvested accessions showing WSS infestation below 25% were characterized as displaying antixenosis, accessions showing WSS mortality (due to plant resistance) over 25% were characterized as displaying antibiosis, and accessions that fit both criteria were characterized as displaying both antixenosis and antibiosis. Accessions showing parasitism levels over 45% were characterized as displaying indirect plant defense. Landraces that headed after the susceptible check varieties (Julian heading date of 191 days) were considered late heading and were not included in the resistant group as they may have missed the WSS flight period and escaped infestation.

Replicated trial with selected landrace accessions

Based on field data from the first year of the experiments, twelve resistant wheat landrace accessions from Iran (n = 1), Japan (n = 1), Tajikistan (n = 1), Turkey (n = 8), and Turkmenistan (n = 1) were selected to be used in replicated trials. Selection of accessions aimed to capture multiple resistance types (e.g. antibiosis, antixenosis, indirect defense due to parasitoids) present at distinct geographic regions of the world. A susceptible accession from Nepal was included in the trial as a susceptible check along with Reeder and McNeal. The stem dissection procedure from the first year of experiments revealed variation in WSS reaction between different plants within a hill plot, suggesting the presence of genetic heterogeneity. Such variation was observed only for some accessions, from which seeds from different heads were selected for further screening to increase the chances of finding different alleles conferring resistance to WSS. Therefore, a total of 23 unique resistant lines were evaluated in replicated trials. Accessions were planted in hill plots as described above in Amsterdam, MT, USA, in 2014 and 2015, and at a second site near Loma, MT, USA, in 2013 and 2014. Trials were conducted in a randomized complete block design with four replications in 2013, and two replications in 2014 and 2015. At maturity (Zadoks growth stage 99, ripening), accessions were collected and dissected to determine WSS infestation level, larval mortality, and parasitism.

DNA preparation and Kompetitive Allele Specific PCR (KASP) assay

Genomic DNA was isolated from young leaf tissue using a 96-well plate protocol (Chao and Somers, http://maswheat.ucdavis.edu/PDF/DNA0003.pdf, accessed February 2017) modified from Pallotta et al. (2003). Five KASP markers, BS00022441, BS00079029, IWB72657, BS00073411, and BS00074345 were used to characterize the haplotype diversity for the solid stem QTL, Qss.msub.3BL, as identified by Cook et al. (2017). KASP assays were performed following the protocol and conditions given in the KASP manual (http://www.lgcgroup.com/) as described by Cook et al. (2017). Check entries included the hollow-stemmed variety, McNeal, semi-solid stemmed variety, Conan, and solid-stemmed variety, Choteau.

Statistical analysis

Results from the replicated trial with selected landrace accessions were analyzed using PROC GLIMMIX in SAS 9.3 (SAS Institute Inc. 2012). Variables were modelled using the LINK = LOGIT function and denominator degrees of freedom were determined using the KENWARDROGER option. Plant ID was considered a fixed effect and replication within environment was a random effect.

Results

Phenotypic variation for WSS reaction in wheat landrace accessions

Natural WSS infestation was variable across the four experimental years. The highest levels of infestation were recorded in 2015, when the mean value was 22% and variation among lines ranged from 0 to 72%. The lowest infestation levels were recorded in 2014, when infestation ranged from 0 to 48% with a mean of 14% (data not shown). From the 1409 landrace accessions screened between 2012 and 2015, 204 (14.47%) showed resistance to WSS. The stem dissection procedure allowed for characterization of accessions as presenting antibiosis, antixenosis, or indirect defense due to parasitoids (Table 1). Half of the resistant lines showed a combination of both antibiosis and antixenosis, 19.6% showed only antixenosis, and 26.5% showed only antibiosis. Resistance type was not determined in 5.4% of accessions due to an insufficient number of stems (<15). Approximately 43% of the resistant accessions came from the Levant (25%) and Persian Gulf (18.13%) regions, though resistant accessions were found in all targeted regions of the world (Table 1). Among all 204 resistant accessions, WSS infestation ranged from 0 to 72%, WSS parasitism and mortality ranged from 0 to 100%, and stem cutting ranged from 0 to 44% (Table 2). As stated in the methods section, the range of variation for WSS infestation among accessions displaying antixenosis was 0 to 25%, whereas variation in WSS mortality for accessions displaying antibiosis was 25–100%, meaning that some accessions, such as PI262628 and PI537968, which showed high levels of WSS infestation (68 and 48%, respectively), were classified resistant because they caused high levels of WSS mortality (antibiosis) (64.7 and 67%, respectively). Susceptible accessions accounted for 37.5% of the screened material, while late heading and missing (no germination) accounted for 39.2 and 9.9% of accessions, respectively (Table 1).

Table 1 Geographical origin of spring wheat landrace accessions and their reaction to the wheat stem sawfly
Full size table
Table 2 Range, mean, standard deviation, and coefficient of variation for traits associated with wheat stem sawfly resistance in landrace accessions resistant to this insect pest
Full size table

Evaluation of haplotypes corresponding to the Qss.msub.3BL region in resistant accessions

The solid stem trait was commonly observed among the resistant landrace accessions, with 41% of the resistant accessions showing either solid or semi-solid stems. Since stem solidness is known to cause antibiosis and antixenosis (Varella et al. 2017), all resistant accessions were screened for the major solid stem QTL, Qss.msub.3BL, haplotype. There were 17 Qss.msub.3BL haplotypes among the resistant accessions. The solid stem haplotype identified in Northern Great Plains WSS resistant varieties and originally derived from S-615 (haplotype A from Cook et al. 2017) was found in 29 accessions from the Iberian Peninsula (Spain and Portugal), North Africa (Morocco), Levant (Turkey and Israel), Persian Gulf Region (Iran), Europe (Greece), and South/Central/East Asia (Turkmenistan and Georgia), whereas the haplotype conferring early stem solidness (haplotype B) was present in two accessions from Iran (Table 3). The haplotype for hollow stems present in the susceptible wheat variety McNeal (haplotype C) was present in 82 landrace accessions. Among these 82 accessions, 31 displayed antibiosis, 14 displayed antixenosis, and 34 had both types of resistance. Among the resistant landrace accessions selected for the replicated trials, there were 7 haplotypes (Table 4). In general, seeds from different heads of an accession had the same Qss.msub.3BL haplotype. Exceptions were PI639349-1 and PI639349-4, which had haplotypes AB and T, respectively, and PI565386-1, PI565386-2, PI565386-3, which had haplotypes AD, C, and AB, respectively, as indicated by Cook et al. (2017) (Table 4). After multiple genotyping attempts, 12 landrace accessions showed no amplification for at least one KASP assay and, therefore, were excluded from the haplotype results. A complete list of resistant accessions, their haplotypes and resistance type is available in Supplementary Table 1.

Table 3 Summary of haplotype diversity for the solid stem QTL, Qss.msub.3BL, on wheat landrace accessions resistant to the wheat stem sawfly
Full size table
Table 4 LS mean ± SE of wheat stem sawfly infestation, stem cutting, parasitism, and larval mortality in selected wheat landrace accessions
Full size table

Replicated trials with selected landrace accessions

Mean values of WSS infestation across replicated trials ranged from 13.1 to 46.9%, with 6 of 12 accessions showing significantly (p < 0.05) less infestation than the susceptible check from Nepal (Table 4). Stem cutting was significantly (p < 0.05) reduced in all resistant accessions and mean values ranged from 0.8 to 11.3%. One of the Turkish accessions had significantly (p < 0.05) more parasitism than the susceptible check from Nepal and two accessions from Turkey and Turkmenistan showed significantly (p < 0.05) higher WSS larval mortality (Table 4). When compared with susceptible wheat varieties, Reeder and McNeal, landrace accessions PI623445-2 (t = 2.03, p = 0.04), PI382150-2 (t = 2.20, p = 0.03), PI576680-2 (t = 2.24, p = 0.03) and PI576680-3 (t = 2.05, p = 0.04), PI576688-4 (t = 2.12, p = 0.04), PI166471-1 (t = 2.10, p = 0.04), PI166477-2 (t = 2.68, p = 0.008), PI178055-1 (t = 2.75, p = 0.007), PI565386-2 (t = 2.11, p = 0.04) and PI565386-3 (t = 2.49, p = 0.01) had significantly more indirect defense due to parasitism by braconids. Mean parasitism of WSS in Reeder and McNeal was 24.60%. In general, seeds from different heads of an accession showed similar values for WSS infestation, stem cutting, parasitism, and larval mortality. Exceptions were values for stem cutting for PI565386-1, which was significantly higher (p < 0.05) than PI565386-2 and PI565386-3, and value for parasitism in PI178055-1, which was significantly higher (p < 0.05) than PI178055-2 (data not shown).

Discussion

Phenotypic variation was observed for all traits evaluated, highlighting the potential of using landrace accessions for improving WSS resistance in North American wheat breeding programs. Previous studies aimed at identifying landrace accessions showing resistance to WSS reported resistance in approximately 10% of the screened material (Wallace et al. 1969). In the present study, resistance was identified in about 14% of the screened material, probably because of the targeted geographical screening approach used to select accessions (Damania et al. 1997; El Bouhssini et al. 2011; Chen et al. 2015). Resistance was predominant in areas where C. pygmaeus is endemic, which is consistent with the expectation that sources of resistance would likely be found in geographical regions where plant and insect species shared the longest evolutionary history. Also consistent with this generalization is the observation that the Levant, which is a center of origin for wheat, harbors the greatest diversity of Qss.msub-3BL haplotypes (n = 10) among WSS resistant accessions. The Levant was previously reported to harbor the greatest diversity of Qss.msub-3BL haplotypes in tetraploid wheat landrace accessions (Cook et al. 2017). Indeed, the rare haplotype B associated with early stem solidness expression and pith retraction (Varella et al. 2016) was only observed in the Persian Gulf region, also a center of origin for wheat.

The majority of the resistant accessions showed both antixenosis and antibiosis. Some insect resistance traits, such as stem solidness, simultaneously affect insect biology and behavior (Varella et al. 2017), while others may affect one or the other. Mechanisms of resistance associated with antibiosis could have reduced plant palatability (Michell et al. 2016) and increased production of anti-nutritive substances (Tamiru et al. 2015), increased egg mortality through hypersensitive response (Hilker and Meiners 2002), or increased larval mortality due to presence of toxic compounds (Fürstenberg-Hägg et al. 2013) and physical barriers (Michell et al. 2016). Likewise, antixenosis could have been caused by many mechanisms of resistance involving volatile compounds associated with host plant selection (Weaver et al. 2009) and non-volatile deterrent compounds that discourage probing and oviposition (Tamiru et al. 2015). Further studies will be necessary to determine the mechanisms of resistance displayed by these landrace accessions. The fact that antibiosis and antixenosis were often expressed in the same accession suggests that multiple resistance genes might exist in a single accession. Alternatively, as indicated by Varella et al. (2017) for the allele conferring early stem solidness, both types of resistance might be conferred by the same gene. Either scenario would provide parents for introgressing resistance into breeding materials.

The genetic characterization of the solid stem QTL, Qss.msub-3BL, revealed 82 resistant accessions with the hollow stem haplotype C. Most of these had stem solidness scores below 2 (hollow), although there were a few exceptions, such as accessions PI499971 and PI623599 that showed scores of 5 (solid) and 3.5 (semi-solid), respectively. These exceptions may harbor other minor genes for stem solidness (Larson and MacDonald 1962; Lanning et al. 2006; Varella et al. 2015). On the other hand, hollow-stemmed accessions carrying haplotype C have great potential for harboring new mechanisms of resistance to WSS. The effects of haplotypes A, B, and C on stem morphology and WSS biology and behavior have already been investigated (Varella et al. 2016, 2017). Such investigation has provided evidence that haplotype A, derived from S-615 and commonly found in North American solid-stemmed varieties, might not provide the best WSS control (Varella et al. 2016, 2017).

Besides antibiosis and antixenosis, resistance to the WSS can be provided by indirect plant defense mechanisms, which result in increased recruitment of parasitoids. Two species of native parasitoids, B. cephi and B. lissogaster, were identified inside wheat stem of landrace accessions. Therefore, despite differences in reproductive strategies shown by the dominant parasitoids species of WSS and the two congeners, results presented here indicate that the North American parasitoid species can successfully locate the WSS larvae in landrace accessions. By comparing WSS parasitism rates in landraces and wheat varieties, it was possible to identify accessions that were more efficient in recruiting parasitoids. This may be of value in enhancing the contributions of braconid species in the overall kill of C. pygmaeus and C. fumipennis by parasitoids. Crop domestication and artificial selection can disrupt tritrophic interactions and unintentionally reduce the effectiveness of natural enemies by negatively altering important cues used by predators and parasitoids for host location (Chen et al. 2015; Chen and Welter 2005; Wang et al. 2009). Thus, it is reasonable to speculate that some landraces might harbor genes or favorable alleles for indirect plant defense that were lost during domestication and artificial selection. These accessions may have the potential benefit of improving WSS control and enhancing our understanding of tritrophic interactions in wheat.

Within the North American wheat producing area there are three distinct genetic clusters of WSS, the Northern (Canada and central and eastern Montana), Southern (Wyoming, Nebraska, and Colorado), and Mountain (western Montana and Idaho) clusters (Lesiuer et al. 2016). These distinct genetic groups could potentially show different responses to resistance genes and/or alleles. Also, high genetic diversity among clusters may facilitate the development of WSS populations that are adapted to resistant varieties (Lesiuer et al. 2016). Thus, the development of WSS resistant varieties harboring multiple resistance genes might provide a more durable control strategy that could be implemented in a larger wheat producing area. Landrace accessions identified here should be a great source of multiple genes for resistance.

In conclusion, wheat landraces harbor phenotypic variation for WSS resistance that may be used in breeding programs to improve pest control. Exploitation of accessions displaying distinct types of plant defense may expand the genetic diversity for WSS resistance currently present in elite breeding lines. Furthermore, genetic resistance harbored by landraces was shown to go beyond the Qss.msub.3BL solid stem QTL, as multiple resistant accessions carried the hollow stem Qss.msub.3BL haplotype associated with susceptibility in North America. Accessions showed different types of host plant defense that were not genetically linked to the commonly used allele for solid stems. These accessions provide a basis for development of bi-parental populations to identify new QTL for WSS resistance for use in wheat breeding programs.