Identifying aphid resistance in the ancestral wheat Triticum monococcum under field conditions

Wheat is an economically, socially, and nutritionally important crop, however, aphid infestation can often reduce wheat yield through feeding and virus transmission. Through field phenotyping, we investigated aphid resistance in ancestral wheat Triticum monococcum (L.). Aphid (Rhopalosiphum padi (L.), Sitobion avenae (F.) and Metopolophium dirhodum (Wlk.)) populations and natural enemy presence (parasitised mummified aphids, ladybird adults and larvae and lacewing eggs and larvae) on two naturally susceptible wheat varieties, Triticum aestivum (L.) var. Solstice and T. monococcum MDR037, and three potentially resistant genotypes T. monococcum MDR657, MDR045 and MDR049 were monitored across three years of field trials. Triticum monococcum MDR045 and MDR049 had smaller aphid populations, whereas MDR657 showed no resistance. Overall, natural enemy presence was positively correlated with aphid populations; however, MDR049 had similar natural enemy presence to MDR037 which is susceptible to aphid infestation. It is hypothesised that alongside reducing aphid population growth, MDR049 also confers indirect resistance by attracting natural enemies. The observed resistance to aphids in MDR045 and MDR049 has strong potential for introgression into commercial wheat varieties, which could have an important role in Integrated Pest Management strategies to reduce aphid populations and virus transmission.

www.nature.com/scientificreports/ Plants have various defence mechanisms against aphid damage. Studies have shown that resistance traits are under selection from herbivores, and both the plant defence theory and plant-herbivore coevolutionary theory assume that resistance traits have evolved as adaptations to reduce herbivory [19][20][21][22][23] .
Resistance to aphids can be direct or indirect, where in the latter resistance is mediated via another organism. Direct resistance can be pre-alighting (antixenosis) where visual or olfactory cues from the plant affect insect behaviour, usually displayed as non-host behavioural responses [24][25][26] . Another form of direct resistance is postalighting (antibiosis), aphids feeding on these plants have reduced development and are less fecund. Antixenosis can be used in indirect resistance where visual and/or olfactory cues cause increased numbers of natural enemies, increasing predation and/or parasitism of pest insects 26 .
No commercial varieties have been bred specifically for R. padi, S. avenae or M. dirhodum resistance to date, but partial resistance has been identified in a number of varieties, including resistance to R. padi in Watkins land race collections 27,28 , north-eastern European Gediflux collection 27 and S. avenae resistance in the ancestral diploid Triticum urartu (T.) and Triticum boeoticum (L.) 29 . The diploid ancestral Triticum monococcum (L.) (genome A m A m ) 30,31 was cultivated in the Neolithic age 32 , genotypes within this species have shown to confer resistance to R. padi [33][34][35] , S. avenae 29,31,[35][36][37] and M. dirhodum 30,34 . Antibiosis to both R. padi and S. avenae has been shown in T. monococcum MDR045 and MDR049 33,38 (Simon et al., unpublished), and antibiosis to R. padi in T. monococcum MDR657 33 , under laboratory conditions. Aphid susceptible Triticum aestivum (L.) Solstice and T. monococcum MDR037 were used as controls in these studies. Both antixenosis and antibiosis has been observed towards M. dirhodum 30 and S. avenae 31 in T. monococcum under field conditions, however, the genotypes used in these studies were not disclosed 30,31 .
Indirect resistance is mediated via another organism such as a natural enemy. Aphid natural enemies include parasitoids, lacewings, and ladybirds which have been proven to be an effective biocontrol method of aphids in field settings on upland cotton (Gossypium hirsutum (L.)) 39 and barley (Hordeum vulgare (L.)) 40 . In a field study on T. aestivum, predator and parasitoid presence resulted in 2.9 to 11.2 times lower Russian wheat aphid (Diuraphis noxia (K.)) populations 41 . Ladybirds were one of the most abundant predators observed in this study 41 . It is unknown whether any of the T. monococcum genotypes confer indirect resistance to reduce aphid populations through increased presence of natural enemies.
The aim of this study was to determine whether T. monococcum MDR657, MDR045 and MDR049 confer direct and/or indirect resistance to economically important aphid species R. padi, S. avenae and M. dirhodum under field conditions. Wheat with natural resistance traits would be desirable candidates for breeding wheat cultivars with resistance to aphids, therefore identifying T. monococcum genotypes with such resistance is of great importance.

Results
Aphid populations. The aphids investigated were Rhopalosiphum padi, Sitobion avenae and Metopolophium dirhodum, and made up 15%, 62.3% and 22.7% of field trial aphid populations respectively. To identify resistance to these three cereal aphid species, their population densities were investigated.
Alate aphid populations. Alate (winged) aphids are often the initial colonisers of cereal crops and in our studies made up 6.5% of the field trial aphid populations. To determine the presence of pre-alighting resistance (antixenosis) the alate aphid populations were investigated. Alate aphid density varied depending on the wheat genotype (F 4, 599 = 8.55, P < 0.001), the highest density of alate aphids was observed on Solstice (Fig. 4). Alate aphid population density varied by week (F 11, 599 = 4.86, P < 0.001), the highest number of alate aphids was observed in week 23 (Fig. 4). There was no difference in alate aphid density between the three years of field trials (F 2, 599 = 1.65, P = 0.193). There was no interaction between the wheat genotype and week (F 44

Natural enemies.
To determine whether any of the wheat genotypes conferred indirect resistance by attracting aphid natural enemies, natural enemy presence were investigated. The natural enemies investigated were parasitised mummified aphids, ladybirds adults and larvae and lacewing eggs and larvae which made up 64.7%, 23.5% and 11.8% of field natural enemy presence respectively. Due to low number the presence of these three insects were combined in analysis. The density of natural enemies differed between wheat genotypes (F 4, 599 = 28.34, P < 0.001), the highest natural enemy presence per 25 tillers was observed on Solstice (Fig. 5). www.nature.com/scientificreports/ Natural enemy density varied by week (F 11, 599 = 6.53, P < 0.001), the highest numbers were observed in weeks 24-28 ( Fig. 5). There was no difference in natural enemy presence across the three field trial years (F 2, 599 = 3.44, P = 0.052). There was, however, an interaction between the wheat genotype and week (F 44, 599 = 1.88, P < 0.001) and week and year (F 16, 599 = 2.12, P = 0.006). There was no interaction between the wheat genotype and year (F 8, 599 = 1.88, P = 0.059) nor the wheat genotype, week and year (F 64, 599 = 0.98, P = 0.519). The average number of natural enemies per 25 tillers across all three field trials is presented (Fig. 5).
To determine whether there was a relationship between the aphid populations and natural enemy presence observed throughout all three field trials, a simple linear regression was carried out.
There was a positive relationship between the densities of aphids and natural enemies in the crop (R 2 = 0.28, P < 0.001) (Fig. 6).   www.nature.com/scientificreports/ There was a positive correlation between aphid density and rainfall duration (R 2 = 0.109, P = 0.042) (Fig. 7a), and between natural enemy density and minimum temperature (R 2 = 0.101, P = 0.048) (Fig. 7b). Correlations between aphids and natural enemies and all meteorological factors are presented in Supplementary Material.  www.nature.com/scientificreports/

Discussion
Populations of aphids and natural enemies were recorded on five wheat genotypes over three years to test whether plant resistance to aphids observed in laboratory experiments would also hold under field conditions. The field trials confirmed natural resistance against the aphid species Rhopalosiphum padi and Sitobion avenae for Triticum monococum (L.) MDR045 and MDR049, identified resistance to Metopolophium dirhodum in T. monococum MDR045, and showed that natural enemy numbers were positively correlated with aphid densities.  www.nature.com/scientificreports/ A consideration to factor in is the small size of the field plots (0.6-1 m 22 ) and ideally this kind of work will be repeated at larger field scale in the future. Under laboratory conditions, antibiosis in T. monococcum has been shown to vary by genotype 33,37,38 . Triticum monococcum genotypes MDR657 and MDR049 have been shown to confer partial resistance to R. padi 33 , MDR045 and MDR049 have shown partial resistance to S. avenae 38

and R. padi (Simon et al. unpublished).
Here we show that smaller populations of cereal aphid species, R. padi and S. avenae were also observed on MDR045 and MDR049 in field conditions, and for the first time demonstrating MDR045 confers resistance to M. dirhodum. However, as all aphid populations investigated on MDR657 were similar to those observed on the susceptible MDR037, it is concluded that this genotype did not confer resistance to these aphid species in field conditions. Lower aphid populations observed on MDR045 and MDR049 could be due to direct and/or indirect resistance. Direct resistance can be via antixenosis or antibiosis, or both. Antixenosis relies on semiochemical and/ or visual cues that prevent alate aphids alighting on the plant 24,26,42 . It has been shown that semiochemical production can vary between crop genotypes [43][44][45] and that this can affect cereal aphid behaviour towards different crop genotypes 46 . The highest number of alate aphids was observed on Solstice, suggesting that this genotype was attractive to alate aphids and/or a consequence of antixenosis from T. monococcum genotypes. The role of semiochemicals in this interaction would be an important area for further research.
Antibiosis in plants causes reduced development and fecundity in aphids. As noted above, this type of resistance against aphids has been observed under laboratory conditions for MDR657, MDR045 and MDR049 33,38 . Smaller populations of all three aphid species were observed on MDR045 and smaller R. padi and S. avenae populations were observed on MDR049 in the field trials. It is suggested therefore that the resistance shown by MDR045 and MDR049 is antibiosis, making both these genotypes strong candidates for introgression into commercial wheat varieties to develop resistant plants as part of Integrated Pest Management strategies. During the introgression process it would be imperative to determine whether the reduced aphid populations would result in less virus transmission as well and consider any changes in yield and grain quality. This work confirms that the host genotype can affect aphid populations on these wheat lines under field conditions.
To determine whether any of the genotypes studied had indirect resistance (resistance mediated via another organism), natural enemy presence was observed. Natural enemies recorded included parasitised mummified aphids, ladybird adults and larvae and lacewing eggs and larvae. Overall, natural enemy densities were positively correlated with aphid densities in the field trials. This positive correlation has also been observed between ladybird and aphid densities 47 , and parasitised aphids and aphid densities 48 .
Surprisingly, natural enemy densities on MDR049 were comparable to their presence on MDR037 and MDR657. This indicates that MDR049 recruited more of these natural enemies in relation to its aphid population. It has previously been shown that natural enemies perform better on resistant plants; in a glasshouse study, lacewing Chrysoperla plorabunda (F.) larvae reduced aphid populations more on a resistant transgenic wheat line 49 . It has also been shown that lacewing eggs on resistant plants had higher larval emergence and the proportion of female larvae increased on transgenic resistant T. aestivum than on the isogenic susceptible plants 50 . Therefore, it is hypothesised that the smaller aphid populations observed on MDR049 are due to a combination of slower aphid population growth, and more natural enemy presence and/or more effective parasitism. To further www.nature.com/scientificreports/ investigate natural enemy population effects on MDR049, cage studies where natural enemies can be controlled should be carried out in laboratory and/or field conditions. Aphid populations varied between years, the lowest number of aphids was observed in the second field trial year, 2018, which had the highest average maximum and minimum temperature, and lowest rainfall duration and average rainfall in June and July. The UK Meteorological Office declared a heatwave during the second field trial year. A heatwave is "A marked unusual hot weather (Max, Min and daily average) over a region persisting at least two consecutive days during the hot period of the year based on local climatological conditions, with thermal conditions recorded above given thresholds. " (Guidelines on the Defintion and Monitoring of Extreme Weather and Climate Events, 2016). Whilst aphid populations were not correlated with temperature, there was a positive correlation between aphid population densities and rainfall duration.
Within the T. monococcum genotypes, MDR037 had the highest number of aphids in the first and third year of field trials. MDR037 has also been shown to be susceptible to R. padi and S. avenae infestation under controlled conditions 33,38 . However, in the second year, MDR037 had the second lowest aphid population and a lower aphid presence than MDR049. The reduced rainfall was correlated with reduced aphid population densities, although it seems that aphids on MDR037 were more adversely affected by the heatwave conditions than those on the other genotypes. Drought stress has previously been shown to affect aphid feeding behaviour, for example in an experiment where R. padi was feeding on drought stressed Hordeum vulgare, the aphids had a lower proportion of probes where phloem feeding occurred and salivated for longer than those on well-watered plants 51 . In another study on drought stressed H. vulgare, the greenbug (Schizaphis Graminum (Rondani)) had slower development rates than those on well-watered plants 52 . It has also been shown that R. padi and S. avenae, on drought-stressed T. aestivum produced fewer nymphs and had a lower intrinsic rate of increase 53 . Therefore, it is hypothesised that drought stress reduced MDR037 susceptibility as an aphid host and that drought differentially affects the T. monococcum genotypes.
Natural enemy presence was positively correlated with minimum temperature; however, it is unknown how much this is impacted by aphids and other factors. A number of studies have reported positive correlations between temperature and a natural enemy; temperature has a positive effect on the development rates of the lacewing Chrysopodes lineafrons (A.P.) 54 , and the parasitoid Aphidus ervi (Haliday) 55 . Temperature can also influence predator success, as in the example of the aphid endosymbiont Hamiltonella defensa (M.) that has been shown to provide protection against parasitism against A. ervi at 20 °C was reduced at 25 °C and aphids were susceptible to parasitism at 30 °C 56,57 . Therefore, increased natural enemy presence observed at higher temperatures may be due to a combination of reduced aphid defences against parasitism as well as increased parasitoid and lacewing development. Climate change scenarios predict more frequent and intense summer droughts 58 , leading to increased aphid and parasitoid abundance 17 . Here we show that aphid densities were positively correlated with rainfall duration, and natural enemy presence can increase with higher temperatures. These meteorological interactions should be further investigated to better understand the impacts of climate change on wheat-aphid-natural enemy interactions.
We have shown that aphid population densities were dependent on wheat genotype, confirming that the partial aphid resistance to R. padi and S. avenae observed in T. monococcum MDR045 and MDR049 under laboratory conditions, are also present under field conditions. We also show for the first time that T. monococcum MDR045 confers resistance to M. dirhodum. In addition to reduced aphid population growth, MDR049 also recruited natural enemies although it is unknown how much the predation and/or parasitism impacted the observed aphid populations.

Materials and methods
Plant material. Triticum aestivum seeds were provided by Rothamsted Research, Hertfordshire, UK.
Triticum monococcum MDR037, MDR657, MDR045 and MDR049 seeds were provided by the Wheat Genetic Improvement Network (WGIN). All plants used in these field trials complied with international, national and institutional guidelines.
Field trials experimental design. Observations of aphid and natural enemy presence on Triticum monococcum genotypes MDR037, MDR657, MDR045, MDR049 and Triticum aestivum cv. Solstice were carried out at Rothamsted Research, Hertfordshire, UK. Seeds were sown in Autumn at 250 seeds per plot for the first field trial in 2017, and 300 seeds per plot for the second and third field trials in 2018 and 2019 respectively. The first field trial was sown in a field named Long Hoos 4, in 0.6 × 1 m plots sown at a drill speed of 0.7miles per hour (mph), plots were separated by 0.5 m unsown ground. The second and third field trials were at Long Hoos 5 and Claycroft respectively, with 1 m 2 plots sown at a drill speed of 1.2mph, plots were separated by 0.75 m unsown ground. The reason for the smaller plot size in the first year was due to a limited availability of seed. Subsequent years used saved seeds hand harvested from previous field season. The plots were sown in a randomised block design with four replicate plots for each genotype. At each side of the field trial was a strip of T. aestivum var. Hereward.
Measuring aphid populations and natural enemy presence. The number of cereal aphids (Sitobion avenae, Rhopalosiphum padi and Metopolophium dirhodum), and natural enemies (ladybird adults and larvae, lacewing eggs and larvae and parasitised mummified aphids) per tiller was counted on 25 tillers in each plot on a weekly or biweekly basis, depending on weather. Each tiller was chosen at random using Genstat (2016, 18th Edition, VSN International Ltd, Hemel Hempstead, UK) statistical software. The sampling started either the first or second week of May when the aphid populations started to increase and finished in the last week of July when aphid populations had diminished. www.nature.com/scientificreports/ Meteorological data on daily maximum temperature, minimum temperature, and rainfall were provided by Rothamsted Research Long-term Experiments National Capabilities (LTE-NCG) Electronic Rothamsted Archive (e-RA).
Statistical analysis. Data were subjected to transformation to comply with assumptions of normality and equal variance. Normailty was determined through the Shapiro-Wilk test and equal variances through the homogeneity of variance test. The number of R. padi, S. avenae, M. dirhodum, alate aphids and natural enemy presence per 25 tillers were subject to square root transformation. Genstat software (19th Edition, VSN International Ltd, Hemel Hempstead, UK) was used for the following analysis. The number of R. padi, S. avenae, M. dirhodum, alate aphids and natural enemy presence per 25 tillers was analysed using linear mixed model (LMM) form of the restricted maximum likelihood model (REML) as there were some missing values. The wheat genotype, field trial year and the week of recording were the fixed model. The random model included blocks of the field trial design and the time of assessments.
To determine relationships between aphid populations and natural enemy presence, a simple linear regression was carried out. Relationships between weather conditions and aphid or natural insect presence were determined through simple linear regression analysis.
Ethical approval. This article does not contain any studies on human participants or cells performed by any of the authors.

Data availability
All original data within this manuscript is available in FigShare, https:// figsh are. com/.