lncluding predator presence in a refined model for assessing resistance of alfalfa cultivar to aphids

TU Xiong-bing, FAN Yao-li, Mark McNeill, ZHANG Ze-hua

1 State Key Laboratory for Biology of Plant Diseases and Insect Pests, Institute of Plant Protection, Chinese Academy of Agricultural Sciences, Beijing 100193, P.R.China

2 AgResearch, Canterbury 8140, New Zealand

1. lntroduction

Aphids are widespread throughout the world and cause severe damage to a variety of economically important crops(Gutierrez and Ponti 2013; Yang 2013). For this reason,control of aphids using plant resistance and biological control as part of integrated pest management (IPM) programmes has been the focus of extensive and ongoing research in many economically important crops (Smith and Clement 2012). Statistical models of pest population density are widely used to evaluate variety resistance in IPM strategies(Luginbill 1969). Numerous researchers have worked to improve the accuracy of such models over the years.For example, Rana (1999) suggested that aphid quantity on plants could be used to estimate cultivar resistance toSitobion avenae(Fabricius). Havlí?ková (1997) analyzed tolerance differences of five winter wheat cultivars to cereal aphids using aphid population density. Thackrayet al.(1990) evaluated wheat cultivar tolerance toRhopalosiphum padiL. using the intrinsic growth rate of populations on the plants, while Souzaet al.(1991) proposed an aphid damage index to evaluate wheat cultivar tolerance toDiuraphis noxiaMordvilko. More complex evaluations, such as that undertaken by He and Zhang (2006), have evaluated alfalfa(Medicago sativaL.) resistance to aphids using an aphid damage index and plant infestation index. Hesleret al.(1999) proposed a method to study resistance differences of wheat cultivars toR.padiusing aphid quantity, aphid development duration, and intrinsic growth rate. All of these authors made important contributions to the methods of evaluating crop variety resistance to aphids. However,the accuracy of these models is limited mainly due to plant variety and different levels of resistance (Chenet al.1997).

In an effort to provide a more robust assessment ofM.sativaresistance to aphids, the aphid quantity ratio (AQR)model was developed in China. AQR has been used in theoretical and applied research on resistance to plant pests(Sun 2006; Quet al.2012), including studies of aphids onM.sativa(Huang 2007) and oat bird-cherry aphids on wheat cultivars (Li and Ye 2002). Chen (2005) evaluated 12 goat weed resistance toS. avenaeusing a combination of the AQR model, weight loss of plants, and changes in chlorophyll content. The AQR model has played an important role in assessing plant resistance to pests, but it does not indicate which factor produced resistance of the cultivars (Farmer 2001; Schneeet al.2006; Shiojiriet al.2006; Zeng 2008).Natural enemies affect pest population dynamics, but they are only a part of the intrinsic components of crop resistance to pests, and the relationship between natural enemies and pests is not entirely clear (Huanget al.2008; Fanget al.2010). For example, host plant resistance has been clearly defined as inherited qualities that result in less damage by pests (Souzaet al.1991). Therefore, evaluation methods should explore parameters, including natural enemies that reflect inherent traits of the plant.

Unfortunately, the current AQR model leads to incorrect conclusions about plant susceptibility to aphid pests,with discrepancies between the predictedvs. actual field observation of plant species resistance (Temueret al.2005; Chenget al.2009; Liuet al.2012). In this paper,we investigated the value of a modified AQR model for evaluating resistance ofM.sativavarieties to aphids by monitoring populations of aphids and their predators over growing seasons in two consecutive years.

2. Materials and methods

2.1. Region description and cultivar source

The field experiment was conducted at Cangzhou City,Hebei Province, China (39°37′N, 98°30′E, 40 m above sea level) from May 2013 to August 2014. The site was on cultivated farmland, with a saline-alkali soil (Bohai alkaline moisture soil) and it was unirrigated. The plots were sown in May 2012 using a precision seeder, with N, P, K fertilizer applied once prior to sowing and once after sowing. The experiment used a randomized complete block design and evaluated 28M.sativacultivars (Table 1) with three replications per cultivar. Each plot measured 4 m by 5 m,with the seed planted at a rate of 15 kg ha–1, a seeding depth of 2 cm and a between-row spacing of 20 cm. There was a 1-m spacing between plots which was kept clear of weedsby regular surface cultivation or herbicide application. No insecticides were used over the period of the experiment.

Table 1 Origin of alfalfa cultivars assessed in the experiment to evaluate aphid resistance

2.2. Data collection and organisms

Data on populations of aphids and their natural enemies were monitored to identify the best time within a growing season to carry out field evaluations that would be used in the modified AQR model. To assess aphids and their predators, we selected five points randomly within each plot, and 20 stems were collected at each point. Stems were carefully cut at ground level and placed into a separate plastic bag. The natural enemy complex was sampled by taking 10 sweeps in each plot using a 50-cm sweep net,following the protocols of Cuperus and Radcliffe (1982) and Girousseet al.(2003). All samples were returned to the laboratory, frozen to kill them, then counted and identified under a stereo microscope. Aphids and predators were identified using arthropod keys (Feng 1990; Qiao 2009).Parasitoids were not collected in this study. The mean number of aphids per plot and numbers of predators per plot and per cultivar were determined.

2.3. Evaluation of period selection

Generally,M.sativais harvested four times per year in Cangzhou City, with cuttings taken in spring (May), early summer (June), mid-summer (August) and early autumn(September). In this study, insect sampling commenced 10 May and continued through to 30 August for both 2013 and 2014. This covered the second (10 May to 10 June), third(1–30 July) and fourth (1–30 August) harvests. Samples were taken every 7 days, which permitted a census of insect species across the growing season covering the vegetative growth stage to the full bloom stage.

2.4. Predator, aphid, and cultivar correlation analysis

It is known that different numbers of predators can be found in different types of crops, and that the number of predators can also be described as a function of host availability (McLaren and Craven 2008). For this study,aphid and predator numbers and their diversity in each of the 28 cultivars were analyzed to assess the main influential factor on aphid population dynamics over theM.sativagrowing season. To determine the role of each factor, we re-estimated the correlation between aphids, their predators andM.sativacultivar.

2.5. Statistical analysis

The AQR model for 28 cultivars was determined using the approaches of Tonget al.(1991) and Liet al.(1998). The improved AQR model differed from the AQR because it allowed for assessment of the abundance of the predators associated with each plot and therefore provided a correction factor to estimate resistance of the 28 cultivars to aphids.AQR, α, and improved AQR were calculated as follows:

We defined the predator ratio, ‘α’, as the predator quantity on each cultivar divided by the predator quantity on all cultivars.

The improved AQR (αAQR) is the two original components multiplied (AQR×α). To determine an optimal period to estimate aphid resistance, the AQR values for the second,third, and fourthM.sativaharvests were calculated and the standard errors (SE) of each period were compared.To further determine the appropriate period to carry out an evaluation, we also compared the SE of each period using the αAQR model. Pearson correlation analysis and the median clustering method were used to compare/correlate the aphid and predator numbers among cultivars, then the median clustering method was also used to divide the cultivars into different groups (Tuet al.2016). Grey correlation analysis of standardization and distinguishing the coefficient at the 0.1 level were used to compare the contribution of each predator to the calculation of aphid cultivar resistance (Wanget al.1993). Analysis was performed using SAS (ver. 8.0)statistical software package.

3. Results

3.1. Data from both growing seasons

Overall, in 2013, aphid numbers across all plots peaked in late July, while the number of natural enemies peaked in early August. In 2014, aphid and predator numbers both peaked in August. Of the four species of aphids collected across both years, 80% wereTherioaphis trifoliiand the remaining 20% were a mix ofAcyrthosiphon craccivoraKoch,Acyrthosiphon pisum(Harris), andAcyrthosiphon kondoiShinji et Kondo (Table 2). The aphid predator complex consisted of several taxa, including four species of Coccinellidae (Coccinella septempunctataL.,Adalia bipunctataL.,Propylaea japonicaThunberg, andHarmonia axyridisPallas), two species of Chrysopidae (Chrysoperla sinicaTjeder,Chrysopa pallensRambur) and one species of Anthocoridae (Orius minutusL.). Spiders were the most abundant taxa collected (Table 2) and were represented by five species:Pardosa astrigeraL. Koch,Misumenopos tricuspidataFahricius,Xysticus croceusFox,Singa hamataClerck, andErigonidium graminicolumSundevall.

Table 2 Mean (±SE) number of individuals collected per sampling occasion, and total number of individuals collected across all the 28 alfalfa plots in 2013 and 20141)

3.2. Selection of the appropriate period to estimate aphid resistance of alfalfa cultivars

When the AQR and αAQR for the pre-period, mid-period,and late-period for both years were calculated, and the SE of each period was compared (Table 3, Appendices A and B), results indicated that the SE for the AQR and αAQR was the lowest at the start of the season (F=3.17,P=0.0475 andF=6.48,P=0.0024, respectively). Hence, the early stage(May to June) of the year was considered the most suitable period for analyzing cultivar resistance to aphids.

3.3. Resistance of alfalfa cultivars to aphids using the AQR model

Results showed that the 28 alfalfa cultivars could be clustered into three broad resistance categories, including resistant, tolerant and susceptible cultivars using the AQR model (Fig. 1). Seven resistant cultivars were subdivided into four subgroups: (1) Zhongmu 1; (2) WL354HQ and Sanditi; (3) Gongnong 2 and WL440HQ; (4) SARDI 10 and SARDI 5. Ten cultivars assessed as tolerant were part of one group: SARDI 7, Cangzhou, WL319HQ, Queen, Sitel,WL343HQ, Apex, Zhongmu 3, 53HR and WL363HQ. Eleven susceptible cultivars were divided into four sub groups: (1)Gongnong 1, Farmers Treasure, and WL323; (2) WL168HQ and Derby; (3) Zhongmu 2, Algonquin, and FD 4. and (4)KRIMA, SARDI 10, and Alfaking (Fig. 1).

3.4. Contributions of predators to cultivar resistance to aphids

Grey correlation analysis was used to investigate the contributions of each predator to control aphids, and found that spiders were the most significant predator across the plots (correlation coefficient value of 0.4200), followed in decreasing order by Chrysopidae (0.3923), Coccinellidae(0.3738) andO.minutus(0.3482) (Table 4, Appendix C).This information also indicated predators of aphids are present in high numbers across the growing season, and are assumed to have an important role in controlling aphid populations.

3.5. lmprovement of the AQR model to assess alfalfa cultivar resistance to aphids

The correlation analysis between the predator complex and aphid populations for each cultivar showed that the predator population is an important control factor in decreasing aphid population density. The αAQR model that included predators showed that the 28 cultivars could be clustered into three broad resistance categories (resistant,tolerant and susceptible) and eight sub groups. The resistant class comprised three groups: (1) Zhongmu 1; (2)WL354HQ, Zhongmu 3, and Gongnong 2; (3) Gongnong 1,WL343HQ, WL440HQ, SARDI 5, and Queen. The tolerant class consisted of only one group, including Cangzhou,WL319HQ, Sanditi, Apex, SARDI 10, 53 HR, Sitel, SARDI 7,and WL363HQ. The susceptible class again included four groups: (1) Farmers Treasure, Zhongmu 2, Algonquin, and WL168HQ; (2) KRIMA; (3) FD 4, WL323 and Derby; (4)Alfaking and SOCA (Fig. 2).

3.6. Standard evaluation method for alfalfa cultivar resistance to aphids based on the αAQR model

Comparing the AQR model and the αAQR model, we found that values of the αAQR were near the numerical value‘1’. Most importantly, it was found that the tolerant class divided by the αAQR consisted of the middle branch (Fig. 2).Hence, we defined the numerical value from (1–0.1)≤(αAQR)≤(1+0.1) as the tolerant class, then took the numerical value‘0.1’ as an index to divide the resistant and susceptible classes into four groups. Based on this new method, 17

cultivars were reassigned from one resistant classification to another compared with the AQR value (Table 5, Appendix D). Zhongmu 1 was re-classified as highly resistant (HR);WL354HQ as resistant (R); Zhongmu 3, Cangzhou, and Gongnong 2 as moderately resistant (MR); and Gongnong 1,WL343HQ, WL440HQ, SARDI 5, Sanditi, and Queen as the low resistant group (LR); these four groups belong to the resistant class (Table 5, Appendix D). WL319HQ,Apex, SARDI 10, 53HR, Sitel, SARDI 7, and WL363HQ were classified as tolerant (M), which was the same as the classification by the improved AQR model and only included one group (Table 5, Appendix D). The susceptible class including four groups, WL168HQ, WL323, Zhongmu 2,and Farmers Treasure, was classified as low susceptible(LS); Algonquin and Derby as moderately susceptible (MS);KRIMA and FD 4 as susceptible (S); and Alfaking and SOCA as highly susceptible (HS) (Table 5, Appendix D).

Table 3 Standard error (SE) associated with the aphid quantity ratio (AQR) and improved AQR (αAQR) in three different stages of alfalfa harvest during over one growing season in 2014 northern China

Fig. 1 Estimated cultivar resistance to aphids by aphid quantity ratio (AQR) model. Twenty-eight alfalfa cultivar varieties are as follows:var1, FD4; var2, SARDI 10; var3, SARDI 7; var4, 53HR; var5, Zhongmu 1; var6, Alfaking; var7, Farmers Treasure; var8, WL319HQ;var9, Algonquin; var10, Sanditi; var11, Zhongmu 3; var12, WL354HQ; var13, WL343HQ; var14, KRIMA; var15, Gongnong 1;var16, WL363HQ; var17, WL440HQ; var18, Sitel; var19, WL168HQ; var20, Zhongmu 2; var21, Derby; var22, Cangzhou; var23,WL323; var24, Apex; var25, Queen; var26, SARDI 5; var27, Gongnong 2; var28, SOCA.

Fig. 2 Improved aphid quantity ratio (AQR) model to estimate cultivar resistance to aphids. Resistant class, tolerant class, and susceptible class include 9, 9, and 10 cultivars, respectively by median clustering method. Twenty-eight alfalfa cultivar varieties are as follows: var1, FD4; var2, SARDI 10; var3, SARDI 7; var4, 53HR; var5, Zhongmu 1; var6, Alfaking; var7, Farmers Treasure;var8, WL319HQ; var9, Algonquin; var10, Sanditi; var11, Zhongmu 3; var12, WL354HQ; var13, WL343HQ; var14, KRIMA; var15,Gongnong 1; var16, WL363HQ; var17, WL440HQ; var18, Sitel; var19, WL168HQ; var20, Zhongmu 2; var21, Derby; var22,Cangzhou; var23, WL323; var24, Apex; var25, Queen; var26, SARDI 5; var27, Gongnong 2; var28, SOCA.

Table 4 Contributions of each natural enemy in the cultivar resistance to aphids by gray correlation analysis

4. Discussion

Previous methods assessingM.sativaresistance to aphids only considered aphid population density as a reference parameter to evaluate resistance (Rana 1999; He and Zhang 2006). In contrast, the method described in this paper considered not only the aphid density but also aphid predator density (Table 4). We considered several key factors in this study. The first was how to select the appropriate period to estimate aphid resistance. Because the aphid population numbers were more uniform across all cultivars early in the season (May to June), this was deemed a suitable time for assessing cultivar resistance differences (Wuet al.2007). It is possible that different developmental stages(e.g., three-leaf, six-leaf and flowering) may vary for each cultivar, which in turn could influence aphid colonisation, and therefore predator aggregation (Wuet al.2007, 2012). In this paper, we focused on cultivar development dynamics from the three-leaf to the flowering stage, but it may be helpful to eliminate differences among developmental stages(Geet al.2011).

Table 5 Grade division of 28 cultivars by the new standard of median value (1±0.1) and equal difference (0.1)1)

Second, there have been some deviations in evaluating resistance in the same cultivar, even within the same region.For example, Sanditi was a susceptible variety according to He and Zhang (2006), but research by Huang (2007) and our results indicated that the cultivar could be classified as resistant. Conversely, Gongnong 1, an important cultivar in high-latitude region of northern China, has been classified as resistant (Wuet al.2007, 2012), but it was a susceptible cultivar (Fig. 1) based on the AQR value (Table 5). While the AQR is useful, there were questions concerning validity of the results because of the confounding effects caused by the presence of predators. In other words, plants that were attacked by low numbers of aphids may be classified as resistant, but the low aphid counts may simply be a factor of abundance of predators that suppress the population.As a consequence, a seemingly resistant plant may be susceptible. Natural enemies, especially spiders, may play an active role in decreasing aphid populations (Table 4),and these diverse species would provide a valuable role in aphid biocontrol (Zhuet al.2000). These results may be related to crop resistance or pest-induced resistance.Studies have demonstrated that aggregation of predators and parasitoids can be attributed to volatiles from the crop itself and from pest damage to host plants (Kapperset al.2005; Geet al.2011), as well as semio-chemicals from pests such as kairomones (Vet and Dicke 1992), allomones and synomones (Lin and Chen 2009). Thus, we introduced a natural enemy ratio, ‘α’, to construct a new model and reevaluate cultivar resistance to aphids (Fig. 2; Table 4).Compared to the AQR model, the αAQR model implied an interesting grade division, as differences between grades were all 0.1. The format of the new model was similar to the aphid damage index (ADI), another important method used to evaluate cultivar resistance. However, the ADI requires investigation of the leaf damage index and damage grades due to aphid infestation by artificial tests, including more uncertain factors and requiring more time to collect data(Wuet al.2007, 2012). The new model considered the natural ecosystem effect and allowed data to be obtained more easily than the ADI, and it is also more convincing than the AQR model. Hence, the new model, αAQR, with its tolerant value of 1±0.1 and equal difference of 0.1 would be a suitable method for examining cultivar resistance to aphids in field studies. However, the αAQR model has been used to evaluate alfalfa resistance only in northern China, and more attention should be paid when using it in different locations in future work. Also environmental factors such as temperature, precipitation, etc. vary greatly in different locations. Although these factors would not change the intrinsic resistance of crops, they play a vital role in decreasing aphid population density (Luoet al.2014). In addition, we did not evaluate parasitoids such as parasitoid wasps in this study, even though we recognized that the parasitoids may be also important as the predators on aphid control (Liet al.2013). We also have not displayed the different roles of all natural enemies in this paper.For example, some other invertebrate predators such as syrphid flies may decrease aphid population dramatically(Nelsonet al.2012). For the invertebrate predators we did include, we only displayed the correlation coefficient and rank ordering of four natural enemies, but the role of each species has not been illustrated. For example, if we counted the same number of ladybugs of different species, ladybugs with larger body sizes would affect the aphid population more than ladybugs with smaller body sizes (Jaro?íket al.2003).These considerations should be included when discussing crop resistance to pests in future studies.

5. Conclusion

In this report, we studied three periods during the alfalfa growing season in the Cangzhou, Hebei Province, China.Results showed that the earlyM.sativagrowing season(May to June) was the most suitable stage for assessing cultivar resistance because insect populations were at their most uniform across all plots. Moreover, we found that predator population density had a significant negative correlation with aphid population density during May to June of each year, although we had not revealed the roles of all predators in this study. Comparing the AQR model with the αAQR model, we found that 17 of the 28 cultivars (61%) were reassigned to resistant, tolerant, and susceptible groups.The αAQR model enhanced the accuracy of the resistance comparison based on the median clustering method. A standard graduated scale to rank cultivar resistance to aphids is suggested based on the median value of 1±0.1.The new standard is consistent with the αAQR model, and retrospective allocation of the cultivars produced consistent outcomes.

Acknowledgements

The study was funded by the earmarked fund for China Agriculture Research System (CARS-34-07) and the National Department of Public Benefit Research Foundation,China (201303057).

Appendices associated with this paper can be available on http://www.ChinaAgriSci.com/V2/En/appendix.htm

Chen G M. 2005. Evaluation ofAegilopsresources for resistance toSitobion avenaeand biochemical mechanism of aphid resistance. MSc thesis, Sichuan Agricultural University, Chengdu. (in Chinese)

Chen J L, Sun J R, Ding H J, Ni H X, Li X F. 1997. The resistant patterns and mechanism of biochemical resistance in various wheat cultivars.Acta Entomologica Sinica, 40,190–195. (in Chinese)

Cheng L, He C G, Hu G X, Wang S S, Zhu Y L. 2009. The effects ofTherioaphis trifoliion the activities of PAL, POD and PPO in five alfalfa varieties.Plant Protection, 35, 87–90.(in Chinese)

Cuperus G W, Radcliffe E B. 1982. Economic injury levels and economic thresholds for pea aphid,Acyrthosiphon pisum(Harris), on alfalfa.Crop Protection, 1, 453–463.

Fang Q, Wang L, Zhu J Y, Li Y M, Song Q S, Stanley D W,Akhtar Z R, Ye G Y. 2010. Expression of immune reponse genes in lepidopteran host is suppressed by venom from an endoparasitoid,Pteromalus puparum.BMC Genomics,11, 484.

Farmer E E. 2001. Surface-to-air signals.Nature, 411, 854–856.

Feng Z Q. 1990. Color handbook of spiders in China. Hunan Science and Technology Press, Changsha. (in Chinese).

Ge F, Wu K M, Chen X X. 2011. Major advance on the interaction mechanism among plants, pest insects and natural enemies in China.Chinese Journal of Applied Entomology, 48, 1–6. (in Chinese)

Girousse C, Faucher M, Kleinpeter C, Bonnermain J L. 2003.Dissection of the effects of the pea aphidAcyrthosiphon pisumfeeding on assimilate partitioning inMedicago sativa.New Phytologist, 157, 83–92.

Gutierrez A P, Ponti L. 2013. Deconstructing the control of the spotted alfalfa aphidTherioaphis maculate.Agricultural and Forest Entomology, 15, 272–284.

Havlí?ková H. 1997. Differences in level of tolerance to cereal aphids in five winter wheat cultivars.Rostlinná Vyroba, 43,593–596.

He C G, Zhang X G. 2006. Field evaluation of Lucerne(Medicago sativaL.) for resistance to aphids in northern China.Australian Journal of Agricultural Research, 57,471–475.

Hesler L S, Riedell W E, Kieckhefer R W, Haley S D, Collins R D. 1999. Resistance toRhopalosiphum padi(Homoptera,Aphididae) in wheat germplasm accessions.Journal of Economic Entomology, 92, 1234–1238.

Huang F, Shi M, Chen Y F, Cao T T, Chen X X. 2008. Oogenesis ofDiadegma secularism(Hymenoptera: Ichneumonidae)and its associated polydnavirus.Microscopy Research and Technique, 71, 676–683. (in Chinese)

Huang W. 2007. Evalution of the resistance to aphid of alfalfa varieties and preliminary studies on the resistance mechanism. MSc thesis, Northwest A&F University,Yangling. (in Chinese)

Jaro?ík V, Honěk A, Dixon A F G. 2003. Natural enemy ravine revisited, the importance of sample size for determining population growth.Ecological Entomology, 28, 85–91.

Kappers I F, Aharoni A, van Herpen T W, Luckerhoff L L,Dicke M, Bouwmeester H J. 2005. Genetic engineering of terpenoid metabolism attracts bodyguards toArabidopsis.Science, 309, 2070–2072.

Li Q, Ye H Z. 2002. Studies on resistance of wild relatives in triticeae to oat bird-cherry aphids (Homoptera, Aphididae).Scientia Agricultura Sinica, 35, 719–723. (in Chinese)

Li S J, Zhang Z Y, Wang X Y, Ding J H, Ni H X, Sun J R, Cheng D F, Chen J L. 1998. Resistance identification of wheat varieties (lines) to aphid with fuzzy recognition technology.Plant Protection, 24, 15–16. (in Chinese)

Li Y J, Wang L M, Wen Z Z. 2013. Influence factor analysis on control efficiency of parasitoid wasp to aphid.Hubei Agricultural Sciences, 52, 3478–3481. (in Chinese)

Lin H Q, Chen S B. 2009. Advance in research on the orientation mechanism of herbivorous insects and natural enemies.Fujian Journal of Agricultural Sciences, 24, 191–196. (in Chinese)

Liu D, Jiang H X, Wang Z F, Cao Y, Zhang S E, Zhai G Y. 2012.The prevention and control of alfalfa aphid.Shangdong Journal of Animal Science and Veterinary Medicine, 33,94–96. (in Chinese)

Luginbill P. 1969. Developing resistant plants-the ideal method of controlling insects.USDA-ARS Product Research Report,11, 1–14.

Luo J H, Huang W J, Zhao J L, Zhang J C, Ma R H, Huang M Y.2014. Predicting the probability of wheat aphid occurrence using satellite remote sensing and meteorological data.Optik, 125, 5660–5665.

McLaren N W, Craven M. 2008. Evaluation of soybean cultivars for resistance to sclerotinia stalk rot in South Africa.Crop Protection, 27, 231–235.

Nelson E H, Hogg B N, Mills N J, Daane K M. 2012. Syrphidflies suppress lettuce aphids.BioControl, 57, 819–826.

Qiao G X. 2009.Hebei Fauna, Aphids. Hebei Science and Technology Press, Shijiazhuang. (in Chinese)

Qu F, Dang J Y, Cheng M F, Lian J, Xie X S. 2012. Resistance identification of new variety wheat toMacrosiphum avenae.Journal of Shanxi Agricultural Sciences, 40, 386–388. (in Chinese)

Rana J S. 1999. Sceening of wheat (Triticum aestivum) varieties against wheat aphidSitobion avenae(F.).Annals of Biology(Ludhiana), 15, 267–269.

Schnee C, K?llner T G, Heid M, Turlings T C J, Gershenzon J, Degenhardt J. 2006. The products of a single maize sesquiterpene synthase form a volatile defense signal that attracts natural enemies of maize herbivores.Proceedings of the National Academy of Sciences of the United States of America, 103, 1129–1134.

Shiojiri K, Kishimoto K, Ozawa R, Kugimiya S, Urashimo S,Arimura G, Horiuchi J, Nishioka T, Matsui K, Takabayashi J. 2006. Changing green leaf volatile biosynthesis in plants,An approach for improving plant resistance against both herbivores and pathogens.Proceedings of the National Academy of Sciences of the United States of America,103, 16672–16676.

Smith C M, Clement S L. 2012. Molecular bases of plant resistance to arthtopods.Annual Review of Entomology,57, 309–328.

Souza E, Smith C M, Schotzko D J, Zemetra R S. 1991.Greenhouse evaluation of red winter wheats for resistance to the Russian wheat aphid (Diuraphis noxia, Mordvilko).Euphytica, 57, 221–225.

Sun D X. 2006. The roles of several kinds biochemical and enzymes in the spring wheat resistance toSitobion avenge Fabricius. MSc thesis, Gansu Agricultural University,Lanzhou. (in Chinese)

Temuer B H, Wu R T, Jin X L, Shuang Q. 2005. The preliminary studies on the injury of alfalfa by aphids.Inner Mongolia Prataculture, 17, 56–59. (in Chinese)

Thackray D J, Wrattent S D, Edwards P J, Niemeyer H M.1990. Resistance to the aphidsSitobion avenaeandRhopalosiphum padiin Gramineae in relation to hydroxamic acid levels.Annals of Applied Biology, 116, 573–582.

Tong H P, Zhu X M, Cao J Z, Guo Y Y, Hu Y, Wu Y Q, Zhou D H. 1991. The preliminary studies on the resistance identification of winter variety wheat to aphid.Crop Variety Resource, 2, 29–30. (in Chinese)

Tu X B, Fan Y L, Ji M S, Liu Z K, Xie N, Liu Z Y, Zhang Z H. 2016.Improving a method for evaluating alfalfa cultivar resistance to thrips.Journal of Integrative Agriculture, 15, 600–607.

Vet L E M, Dicke M. 1992. Ecology of infochemi use by natural enemies in a tritrophic context.Annual Review of Entomology, 37, 141–172.

Wang X Q, Ding X Y, Huang F. 1993. Analysing the key factor in insect population dynamics with grey correlation coefficient.Journal of Shenyang Agricultural University, 24, 120–124.(in Chinese)

Wu D G, Du J L, Wang S S, Hu G X, He C G. 2012. Evaluation on resistance of 4 alfalfa (Medicago sativa) cultivars to pea aphid (Acyrthosiphon pisum).Pratacultural Science, 29,101–104. (in Chinese)

Wu D G, He C G, Wu T J, Tang S R, Jia B. 2007. Resistance comparison of eleven alfalfa varieties to aphid.Grassland Turf, 4, 54–57. (in Chinese)

Yang W G, Chai H, Huang X Y, Yang Z, Gao H J, Li H. 2013.The main pests and occurrence regulation of alfalfa in Qiqihar of Heilongjiang Province.Grass Culture, 12, 26–30.(in Chinese)

Zeng R S, Su Y J, Ye M, Xie L J, Chen M, Song Y Y. 2008. Plant induced defense and biochemical mechanisms.Journal of South China Agricultural University, 29, 1–6. (in Chinese)

ZhuY Y, Chen H R, Fan J H, Wang Y Y, Li Y, Chen J B, Fan J X, Yang S S, Hu L P, Leung H, Mew T W, Teng P S, Wang Z H, Mundt C C. 2000. Genetic diversity and disease control in rice.Nature, 406, 718–722.