Rules of Pesticide Use for Food Crop Diseases and Pests: A Concise Diagram

Xiaojie YANG Shaohua ZHANG Weizheng LI Huaisen XING Guohui YUAN

Abstract In order to ensure food security, the selection of pesticide must take into account the controlling effect and economy in the chemical control of food crop diseases and pests.  Although China has registered a large number of pesticides and fungicides on food crops, traditional statistical methods cannot treat the complex relationships between crops, pests and diseases, as well as pesticides. To this end, the correspondence analysis was used to mine the data  of pesticides being registered after January 1, 2010 for the prevention and control of pests and diseases on wheat, corn, and rice in China, in order to reveal the general rules of pesticide use for the prevention and control of grain crop pests and diseases and to provide scientifically recommended pesticide regimens. ① Recommended pesticides for wheat diseases and pests: For underground pest control, phoxim, diazinon or clothianidin is recommended; imidacloprid is recommended for aphid control; avermectin is recommended for mite control; chlorpyrifos is recommended for midge control; the pesticides for rust control could be selected according to the order of epoxiconazole>flutriafol>kresoxim-methyl>cyproconazole≈hexaconazole; tebuconazole is recommended for the prevention and control of smut; difenoconazole is recommended for the simultaneous occurrence of take-all, smut and sheath blight; it is recommended to use carbendazim or thiophanate-methyl for the control of scab; and triazolone or pyraclostrobin is recommended for the control of powdery mildew. ② Recommended pesticides for corn diseases and pests: For the control of underground pests, fipronil, carbosulfan or chlorpyrifos are recommended; thiamethoxam is recommended for control of planthoppers and corn aphids; Bacillus thuringiensis or phoxim is recommended for control of corn borers; pyraclostrobin is recommended for the control of northern leaf blight; tebuconazole is recommended for the control of head smut; and fludioxonil+metalaxyl-M is recommended for the control of corn stalk rot. ③ Recommended pesticides for rice diseases and pests: It is recommended to use triazophos, avermectin analogs or methoxyfenozide for the control of borers; for rice planthoppers, the pesticides could be selected in order of pymetrozine>imidacloprid>thiamethoxam>thiazide>nitenpyram; avermectin analogs are recommended for the control of rice leaf rollers; tebuconazole is recommended for the control of rice false smut; prochloraz and its manganese salts are recommended for the control of bakanae disease; hymexazol is recommended for the control of wilt disease; thifluzamide or hexaconazole is recommended for the control of sheath blight; and the pesticides for the control of rice blast are in order of tricyclazole>kasugamycin>fenoxanil>isoprothiolane>Bacillus subtilis. In addition, potential applications of correspondence analysis in other fields of plant protection are discussed.

Key words Food crop; Wheat; Corn; Rice; Fungicide; Pesticide; Correspondence analysis; Pesticide use rules

Received: April 28, 2022  Accepted: June 25, 2022

Supported by National Science and Technology Innovation Project for High Grain Yield and Efficiency; Integration and Demonstration of Maize Disaster Reduction and Quality and Efficiency Improvement in Rain-fed Areas of Southern Henan (2018YFD0300706).

Xiaojie YANG (1996-), female, P. R. China, master, devoted to research about agricultural insects and pest control.

*Corresponding author. E-mail: hnndygh@126.com.

The 2020 Central "No.1 Document", Opinions on Doing a Good Job in the Field of "agriculture, rural areas and farmers" for Ensuring the Achievement of a Well-off Society in an All-round Way as Scheduled, states that ensuring food security is always the top priority in governing the country. With the change of climate, the change of cultivation methods and the increase of crop multi-cropping index, crop diseases and pests in China are more frequent and recurring, especially for Spodoptera frugiperda, wheat stripe rust, rice planthoppers, Chilo suppressalis, wheat aphids, Helicoverpa armigera, and other major pests and diseases［1-3］. Irrational pesticide application will lead to unsatisfactory effect, excessive residues, and deterioration of agricultural product quality, which seriously threatens national food security［4-5］. Efficient plant protection measures can recover about 100 million t/year of food loss, accounting for 1/6 of the total output, which is equivalent to increasing 20 million hm2/year of invisible arable land［6］. During the "14th Five-Year Plan" period,  National Agricultural Sustainable Development Plan (2015-2030) requires that the comprehensive mechanization rate of China’s grain crop ploughing, planting and harvesting reach more than 85%, but agricultural control methods such as crop rotation and intercropping will cause hybridization of planting patterns, which is not conducive to the realization of the goal, while chemical control methods are most compatible with the general trend of agricultural mechanization. Regulations on the Prevention and Control of Crop Diseases and Pests (National Order No.725) requires firmly establishing the concept of "public plant protection, green plant protection, and scientific plant protection", continuing to implement the policy of "prevention first and comprehensive prevention and control", further promoting the integrated development of green prevention and control and unified prevention and control, and vigorously advocating scientific and safe pesticide use［1］. The key to efficient prevention and control of many major pests and diseases is still how to scientifically select pesticides under the premise of taking into account the controlling effect and economy［7］.

The Fifth Plenary Session of the 19th Central Committee of the Communist Party of China comprehensively outlined the "14th Five-Year Plan" and the long-term goals for 2035, and clearly pointed out that "One-map" of disease and pest information should be formed through meta-analysis of plant protection, according to the idea of "Top design, systematic integration, data integration, integrative display", by gathering existing information system data resources and strengthening data mining and comprehensive application［8］. As of June 7, 2021, China Pesticide Information Network (www.chinapesticide.org.cn) has registered 1 010 pesticides for wheat, 1 179 fungicides for wheat, 454 pesticides for corn, 276 fungicides for corn, 4 055 insecticides for rice, and 2 470 fungicides for rice, providing a rich source for big data meta-analysis. Since pesticides must pass a large number of field efficacy verifications when they are registered, and the dynamic fine-tuning of production, sales and dosage is constantly being carried out in the market operation, it can be regarded as an information base for pesticide application experience. However, for specific targets, which pesticide is scientific and reasonable to use? One pesticide can treat many targets simultaneously, but there are many pesticide to choose from for one pest. Traditional statistical methods cannot deal with the intricate relationship between crops, diseases and pests, and pesticides.

Correspondence analysis developed by Beh et al. is a powerful tool for analyzing the complex relationships among multivariates and multiclasses［9］. The basic idea is to first perform principal component analysis on the contingency table to reduce the dimension, and then display the similarity or concomitant of each category point in the row variable and column variable in the form of a scatter plot in the low-dimensional space［10-12］. It has been widely used in the fields of paleontology, sociology, economics, linguistics, ecology, medicine, and psychology［13］, but its significance has not been fully recognized in the field of plant protection. To this end, we analyzed the pesticide for pests and diseases in the three major food crops registered after January 1, 2010, and constructed "One-map" of pest-pesticide and disease-fungicide for each food crop, aiming to solve the difficult problem of scientific selection of pesticides.

Information Sources and Data Processing

Information sources

After logging on China Pesticide Information Network, "Data Center", "Registration Information", and "Pesticide Registration Data" were clicked in sequence. "Insecticide" or "Fungicide" was entered in "Pesticide Category", and "Wheat", "Corn" or "Rice" was entered in "Crop/Location", forming a total of 6 pesticide type-crop combinations. Single preparations registered after January 1, 2010 were checked, and then the hyperlink of the registration certificate was clicked one by one. The registration certificate numbers, pesticide names, crops and targets were copied into Excel. If the same registration certificate involved two or more targets on the same crop, each target should occupy a separate line.

Data preprocessing

The problem of upper and lower concepts of prevention and treatment objects

Some registration certificates have vague descriptions of targets, such as "underground pests", and it is not clear whether they are mainly aimed at grubs, crickets, wireworms or cutworms. Therefore, there are two ways to treat it. One is to delete the unspecific registration information items, but a certain amount of information will loss, and the other is to integrate into a higher-level concept, but it cannot reflect the characteristics of special harmful organisms. This study adopted the second method. Such cases included: ① wheat "seedling stage aphids" and "wheat aphids" were integrated into "aphids", and "grubs", "crickets" and "wireworms" were integrated into "underground pests"; ② wheat "leaf rust" and "stripe rust" were integrated into "rust", and "dwarf bunt of wheat" and "loose kernel smut" were integrated into "smut"; ③ "wireworms", "Agrotis ypsilon", "grubs" and "mole crickets" in corn were integrated into "underground pests"; ④ there were three types of registration objects in corn: "rice planthopper (PD20182887)", "planthopper (PD20140496)" and "Laodelphax striatellus", but rice Sogatella furcifera and Nilaparvata lugens rarely harm corn, while L. striatellus can transmit the maize rough dwarf virus, so they were also all classified as L. striatellus; ⑤ rice "brown planthoppers", "planthoppers" and "white planthoppers" were integrated into "rice planthoppers", and "weevils" and "Lissorhoptrus oryzophilus" were integrated into "rice water weevils", and "thrips" and "rice thrips" were integrated into "rice thrips"; ⑥ "rice sheath blight (PD20201073)" and "sheath blight" were integrated into "sheath blight", and "rice bacterial streak (PD20142170)" was verified to be exactly the same as that of "bacterial streak disease", and was revised to the latter; and ⑦ rice "sheath blight" and "rice seedling rot" are caused by a variety of pathogens that are still controversial［14-15］, and they were regarded as two different diseases in this study.

Combination of similar pesticide

The active ingredients of some pesticide differ only for isomer differences, a small amount of functional group modification or microbial physiological races, so they were integrated. Such cases included: ① "beta-cypermethrin" and "cypermethrin" in wheat were integrated into "cypermethrin";  ② in corn, "Bacillus thuringiensis G033A (PD20171726)" was integrated into "B. thuringiensis", and "avermectin" and "emamectin benzoate" were called "avermectin analogs", and "cypermethrin" and "cis-cypermethrin" were integrated into "cypermethrin"; ③ "avermectin", "emamectin benzoate", "methylamino abamectin benzoate" and "5% avermectin microemulsion (PD20180988)" in rice were integrated into "avermectin analogs", and "niclosamide ethanolamine salt" and "niclosamide" were integrated into "niclosamide class"; and ④ "prochloraz-manganese chloride complex " and "prochloraz" in rice were referred to as "prochloraz", and "Bacillus amyloliquefaciens LX-11 (Bacillus amyloliquens) (PD20190018)" was incorporated into "B. amyloliquefaciens".

Amendments to some registration information

Some registration information does not match the control target, and must be judged based on general knowledges of plant protection. Such cases included: ① "niclosamide ethanolamine salt (PD20110164)" and "calcium cyanamide (PD20110256)" were registered as "fungicides", but the target on rice is "Pomacea canaliculata", which was amended to the "pesticides" category; ②both "matrine (PD20132710)" and "osthole (PD20182169)" were listed as "pesticides/fungicides", but the targets on rice are only "rice planthoppers" and "sheath blight" respectively, and they were amended to the "pesticides" category and the "fungicides" category, respectively; ③"bensulfuron-methyl (PD20101253)" was registered as "fungicides", but the target on rice is "broadleaf weeds and sedge weeds", so it is eliminated; and ④ "sex pheromone of Chilo suppressalis (PD20200312 and PD20190003)" was registered as a mixture (cis-9-hexadecenal, cis-13-octadecenal, cis-11-hexadecenal), but the component and ratio of sex pheromone are very specific and should be regarded as a "single agent", and the information of these two registration certificates should also be supplemented to the original data for analysis.

Corresponding analysis operations

A new column of "weight" was created on the right side of the registration information, and the weight factor was assigned a value of 1 (the same registration certificate controls multiple objects, and each pesticide-target corresponding information was assigned a weight factor of 1). Then, the pesticide name and target were replaced with continuous numerical codes respectively.

In the SPSS19.0 variable view, four new variables, i.e., pesticide name (numerical nominal variable), crop (string nominal variable), target (numerical nominal variable) and weight (measurement variable) were created, and the registration information was imported into the data view. After weighting the weights, "Analyze", "Dimensionality Reduction" and "Correspondence Analysis" were clicked in turn. "Pesticide Name" and "Target" were set as row and column variables, respectively, and the " Range of Definition " of the row and column variables were clicked, respectively, to define their minimum value (input "1") and maximum value (input the maximum value of the continuous coding of the category), and the preliminary correspondence analysis was completed by the Euclidean distance metric method.   When the effective margin of row in the corresponding table was viewed in the results window to be 1, it meant that only one manufacturer believed that a certain pesticide is effective for a certain target, which was not representative, and the corresponding registration certificate entry was excluded from the corresponding analysis. However, the number of registration certificates is related to the registration year, and pesticide registered later tend to have fewer registration certificates, so the registration frequency cannot reflect the controlling effect of a certain pesticide in the strict sense. In addition, some targets appear only once, and there is only one registered pesticide. Such targets would be deleted together with the above-mentioned elimination operation, such as sterenol and wheat mosaic virus disease in registration certificate PD20181615, fenpyroximate and corn mites in PD20183984, and oligosaccharides and maize rough dwarf disease in PD20121813, resulting in a condition that very important targets in local areas cannot be expressed on the corresponding analysis plot. In order to take into account the generality and forward looking of the conclusions, these relatively rare cases of registration certificate items were listed separately.

After removing the relevant registration entries, other categories of various variables were re-coded consecutively, and the corresponding analysis was performed again. The coordinate values (X, Y) of each pesticide and each target were found in the result output window, and a bivariate composite scatterplot was made in Origin2021.

Xiaojie YANG et al. Rules of Pesticide Use for Food Crop Diseases and Pests

Interpretation of the picture

Now, a virtual correspondence analysis diagram (Fig. 1) will be used to illustrate the interpretation method. Assuming that the scatter point A is a certain target, and the scatter points B to I are several pesticide, a series of forward vectors (directed rays starting from the coordinate origin and passing through a scatter point) are drawn to assist in interpretation.

Recommended pesticides

The B, C, D, and E on the positive vector side of the target A are regarded as optional pesticides, and the three pesticide F, G, and H on the negative vector side of the target A are not recommended.

Acceptable degree of controlling effect

Among B, C, D, and E, the pesticide corresponding to the scattered point farther away from the coordinate origin (0, 0) will have more registration certificates, and the acceptable degree of the controlling effect is higher. For example, C is recognized by manufacturers with a higher degree than B, D, and E.

Registered target specificity

Among B, C, D, and E, a smaller angle between the forward vector of a certain pesticide and the forward vector of the target A indicates stronger registered target specificity of the pesticide, that is to say, more manufacturers believe that the pesticide can only control target A. For example, the registered target specificity of pesticide B for target A is stronger than those of C, D, and E.

Potential control targets and pesticide with similar control spectra

In general, the smaller the angle between the forward vectors of two targets is, the easier it is to apply the same pesticide to both. Correspondingly, the smaller the angle between the forward vectors of two pesticides, the more similar the control spectra.

Targets with no obvious pesticide application characteristics

Pests and diseases closer to the origin of coordinates have fewer pesticide registration certificates, and it is appropriate to choose a broad-spectrum pesticide for use in combination.

Rules of Pesticide Application for Food Crop Pests and Diseases

Rules of pesticide application in wheat

Wheat aphids, underground pests, and mites have more prominent pesticide application characteristics, while midges and L. striatellus have less obvious pesticide application characteristics. In terms of pesticides, there are many registration certificates for imidacloprid, phoxim, diazinon, and avermectin.

Phoxim and diazinon are the most acceptable pesticides for controlling underground pests in the wheat fields, and clothianidin is less frequently registered. Among them, the registered target specificity of phoxim and clothianidin is slightly higher than diazinon. Imidacloprid has the highest acceptable degree for the controlling effect of wheat aphids and has stronger registered target specificity, and a total of 150 registration certificates (referred to as "150 times", the same below) have been handled; other suitable pesticides are thiamethoxam, pymetrozine and cyhalothrin, but the registered target specificity is not strong, while dinotefuran, acetamiprid, cypermethrin and cyfluthrin have low registration frequency and very similar control spectra; and pirimicarb, dichlorvos, omethoate, deltamethrin and matrine are densely clustered near the origin of coordinates, and only two units have registered for each insecticide. In terms of the acceptable degree for wheat spider mites, avermectin is slightly higher than bifenthrin. Chlorpyrifos is the most frequently registered insecticide for midges, but chlorpyrifos is also an important underground pest control agent, so its scatter point is located in the angle between the forward vectors of midges and underground pests, and it can treat both. The pesticide application characteristics of wheat L. striatellus are not obvious, and only pymetrozine is currently registered as a single preparation (Fig. 2).

Rules of fungicide application in wheat

Wheat rust has the most obvious pesticide application characteristics, followed by head blight, take-all, smut, and powdery mildew. Those with the least obvious pesticide application characteristics are sheath blight and root rot. The four most frequently registered fungicides for wheat are difenoconazole, epoxiconazole, and carbendazim and thiophanate-methyl with similar mechanisms of action. The acceptable degrees of pesticides for wheat rust prevention and control rank as epoxiconazole>flutriafol>kresoxim-methyl>cyproconazole≈hexaconazole, and pyrimidine nucleoside antibiotics are used less frequently. When wheat take-all, smut and sheath blight occur at the same time, difenoconazole in the cross region of the forward vectors can be selected for concurrent treatment. Tebuconazole is registered in smut with a high frequency (22 times), although the registered target specificity is not strong. Silthiopham, thifluzamide and Trichoderma are less used to control these three diseases. Carbendazim and thiophanate-methyl are the most frequently registered pesticides for the prevention and control of wheat scab. From the perspective of plant protection, these two fungicides have similar mechanisms of action and tend to produce cross-resistance. The pesticides for the control of wheat powdery mildew with strong registered target specificity and high degree of recognition are triazolone (19 times) and pyraclostrobin (8 times). Although tebuconazole and flutriafol are registered with higher frequency (23 and 9, respectively) in powdery mildew than the above two fungicides, tebuconazole was registered in smut (22 times) and flutriafol in rust (13 times) with a high registration frequency. Therefore, although these broad-spectrum fungicides can also be used to control powdery mildew, they are not recommended as control agents when powdery mildew occurs alone, in order to delay the development of pathogen resistance (Fig. 3).

Laws of pesticide application in corn

The corn pests with most obvious pesticide use characteristics are L. striatellus, underground pests, and corn borers, and fipronil and thiamethoxam have the most registration certificates. Although Athetis lepigone is an underground pest, its pesticide use characteristics are similar to those of ear borers and leaf-eating pests.

Among them, Spodoptera frugiperda has only been introduced to China in recent years, and there has not many pesticides been registered so far. In addition, there are not many registered pesticides for occasional pest insects such as A. lepigone and armyworms, which reflects the trade-off between benefits and risks by manufacturers. The acceptable degrees of the controlling effects on corn underground pests are shown as follows: fipronil>chlorpyrifos>carbosulfan>cypermethrin.

Chlorpyrifos is applied by soil treatment, and other three are used as coating agents. Among the pest insects with piercing-sucking mouthparts (L. striatellus, aphids), L. striatellus is more important, probably because it can also indirectly transmit maize rough dwarf disease. Thiamethoxam is the most acceptable and has the highest registered target specificity for the control of these two pests, and imidacloprid and clothianidin are used less frequently. There are many pest species involved in the fourth quadrant, and the proportion of biological pesticides is relatively large. The pesticide with the highest registration frequency for corn borer is phoxim (8 times), and the registration frequencies of Bacillus thuringiensis and Beauveria bassiana are roughly the same (6 and 5 times, respectively), and chlorantraniliprole also has a relatively high registration frequency (4 times). There are fewer registered pesticides for other three pests in the fourth quadrant. Among them, the insecticides registered for armyworms are chlorantraniliprole (PD20100677, PD20171109), acephate (PD20101948), B. bassiana (PD20180788) and cyhalothrin (PD20110291); the registered insecticides for the control of S. frugiperda are Mamestra brassciae multiple nucleopolyhedrovirus (PD20150817), Metarhizium anisopliae CQMa421 (PD20171744), and B. bassiana (PD20190002); and the insecticides registered for the control of A. lepigone are emamectin benzoate (PD20132150, PD20141273) and chlorantraniliprole (PD20100677) (Fig. 4).

Laws of fungicide application in corn

Pyraclostrobin (12 times) is the most acceptable for the prevention and control of corn northern leaf blight, and validamycin is used less frequently. Tebuconazole (32 times) is the most acceptable for the prevention and control of corn head smut, and triticonazole and difenoconazole are used less frequently and both registered by five units. Far more fungicides have been registered for the control of corn stalk rot than smut, southern leaf blight and sheath blight. Among them, fludioxonil (6 times) is the most frequently registered fungicide for the prevention and control of corn stalk rot, which is mainly used as a seed dressing agent. Metalaxyl-M is less used, and there are only a handful of fungicides registered for other three diseases. The registered pesticides for the prevention and control of smut are difenoconazole (PD20141612) and sedaxane (PD20150321), while for the control of corn sheath blight and corn southern leaf blight, only validamycin has been registered (PD20150331, 1 registration certificate involves two disease targets at the same time) (Fig. 5).

Laws of pesticide application in rice

Rice planthoppers, Cnaphalocrocis medinalis and borers (including Scirpophaga incertulas) have the most obvious pesticide application characteristics. Minor pests are aphids, leaf miners, Sesamia inferens (belonging to the boring moths), leafhoppers, Orseoia oryzae and rice water weevils. Avermectin analogs, pymetrozine and triazophos are the most frequently registered. The pesticide application characteristics of piercing-sucking pest rice planthoppers are more obvious than those of aphids.

For the rice planthopper insecticides, their acceptable degrees rank as: pymetrozine>imidacloprid>thiamethoxam>thiazide>nitenpyram.

Avermectin analogs are most acceptable for controlling rice leaf rollers, followed by chlorpyrifos and indoxacarb, while Bacillus thuringiensis is the least used.

In the third quadrant, stem-boring moths (C. suppressalis, T. incertulas and S. inferens) are the most harmful, and their damage ways and pesticides are similar. Specifically, triazophos (33 times) is the most frequently registered pesticide for the control of stem-boring moths, followed by avermectin analogs (17 times) and methoxyfenozide. In addition, the two registered pesticides for the control of Pomacea canaliculat are niclosamide (29 times) and metaldehyde (8 times). When rice planthoppers and rice thrips occur at the same time, thiamethoxam can be selected for the treatment of both (Fig. 6).

Laws of fungicide application in rice

Rice blast, rice false smut, and wilt disease have the most obvious pesticide application characteristics, followed by bakanae diseasee and wilt disease. Tebuconazole, thifluzamide, hexaconazole and tricyclazole are the most frequently registered. Rice smut, bakanae and wilt disease are expected to be treated with the same fungicide. Tebuconazole (39 times) is the most acceptable for the controlling effect of rice false smut and had the strongest registered target specificity. Prochloraz (25 times) is the most acceptable for the prevention and controlling effect of bakanae disease, and fludioxonil and hymexazol are also used more frequently. Hymexazol (18 times) is the most acceptable for the controlling effect of wilt disease and have the strongest registered target specificity.

Thifluzamide (104 times) and hexaconazole (108 times) are the most acceptable for the controlling effect of sheath blight. Tricyclazole (44 times) is the most acceptable for the controlling effect of rice blast, and other fungicides rank as kasugamycin>fenoxanil>isoprothiolane>Bacillus subtilis according to the acceptable degree.

The simple statistical results of pesticides for controling uncommon diseases of rice with less registration certificates are as follows: for rice bacterial streak, chlorobromoisocyanuric acid and bacteria octyl amine acetate are each registered 4 times; to control rice stripe disease, lentinan is registered 7 times; for rice bacterial blight, chlorobromoisocyanuric acid is registered three times; for rice seedling rot, ethalicin is registered 4 times and metalaxyl-M is registered 3 times; and for rice black-streaked dwarf disease, dufulin is registered twice (Fig. 7).

Summary of pesticides with low registration frequency

With the processes of climate change, trade exchanges, changes in farming systems, biological invasion, and development of resistance to pests and diseases, the pests and diseases of food crops will also change, and some minor pests and diseases have been recognized as major pests and diseases. In this case, the laws derived from correspondence analysis may not be suitable for a long time in the future.

Therefore, the information on the registration certificates that have been registered on the three major food crops but only recognized by one agricultural material production enterprise (called "pesticide-target single correspondence" in this study) was summarized, as shown in Table 1.

Conclusions

In this study, the laws of pesticide application of food crop diseases and pests were depicted as a concise "One-map", which contains richful information, and has good referred significance for growers, pesticide producers and sellers, technical guidances, and revision of plant protection textbooks. The conclusions were given below. ① Wheat pests: phoxim, diazinon and clothianidin are recommended for control underground pests; imidacloprid is preferred for controling aphids, while thiamethoxam, pymetrozine, and cyhalothrin also can be chosen for application; mites should be controlled with avermectin; wheat midges are controlled with chlorpyrifos; the pesticide for rust prevention can be selected and used in the order of "epoxiconazole>flutriafol>kresoxim-methyl>cyproconazole≈hexaconazole"; tebuconazole is recommended for smut that occurs alone, and difenoconazole can be used in the case of simultaneous occurrence of take-all, smut and sheath blight; carbendazim or thiophanate-methyl is recommended for control of gibberellic disease, but it is best to use them with other fungicides in rotation; and powdery mildew is recommended for control triazolone and pyraclostrobin. ② Corn pests: Seed dressing with fipronil or carbosulfan, or soil treatment with chlorpyrifos, is recommended for underground pests; L. striatellus and corn aphids can be controlled by thiamethoxam coating; biological pesticides, such as B. thuringiensis or phoxim, are recommended to control corn borers in the large flare stage; for other leaf-eating pests, it is recommended to spray with emamectin benzoate or chlorantraniliprole; pyraclostrobin is used for spray control of northern leaf blight; head smut is controlled with tebuconazole coating; and seed dressing with fludioxonil+metalaxyl-M is used for stem stalk rot. ③ Rice pests and diseases: Triazophos, avermectin analogs or methoxyfenozide can be used to control borers by spraying; for the control of rice planthoppers, the pesticides can be selected in the order of "pymetrozine>imidacloprid>thiamethoxam>buprofezin>nitenpyram"; rice leaf rollers are controlled by spraying avermectin analogs; rice false smut is controlled by spraying tebuconazole; prochloraz and its manganese salts are recommended for control bakanae disease; hymexazol is recommended for control bacterial blight; thifuramide or hexaconazole is recommended for control wilt disease; sheath blight is controlled by thifluzamide or hexaconazole; and rice blast can be controlled using pesticides in the order of "tricyclazole>kasugamycin>fenoxanil>isoprothiolane>B. subtilis".

In the practical application of pesticide use rules in this study, it is necessary to pay attention to the naming preference of the targets of registration certificates. For example, we should pay attention to various trivial names of "mites" and "mites", "grubs" and "chafers", and then look for the most suitable pesticide from the plot. If a certain pest has been accurately identified but no suitable pesticide can be found, a suitable pesticide can be found from the category of its superordinate concept. For example, a reasonable pesticide can be found for wheat aphids according to the scattered point "aphids" in Fig. 2. Because the correspondence analysis is difficult to deal with registration period, as well as the synergistic effect or relative contribution of the active ingredients in a mixture, this study only counted the newer pesticides registered after 2010, and only collected the registration information of single pesticides. And the pesticides with low registration frequency were listed separately, so as to take into account the generality and forward-looking of the conclusions.

Implications: ① When spraying pesticides, if two or more pest insects (or diseases) occur simultaneously, we can observe whether the angle between the forward vectors of these targets on the corresponding analysis plot is small. If so, an equivalent broad-spectrum pesticide scattered in the area between the forward vectors of the two targets can be used, thereby reducing the number of spraying frequency and reducing labor and costs. ② Pesticide with a very small forward vector angle should not be mixed together, because their control spectrum or mechanism of action is very similar, and cross-resistance is likely to occur. ③ The negative effects of pesticide sometimes need to be used for a long time to be highlighted, and then they are included in the list of prohibited and restricted use. When some pesticide are banned due to pesticide resistance or residue problems and new alternative products have not been successfully developed, the range of available pesticide can be determined first according to the principle that the angle with the forward vector of the target in the corresponding analysis figure is less than 90°, and then the emergency pesticide application system can be constructed according to the order of the distances between the scattered points of these available pesticide and the origin of the coordinates.

References

［1］ GUO ZP. Implementing the Regulations on the Prevention and Control of Crop Diseases and Pests and effectively protecting the safety of grain production［J］. Hubei Plant Protection, 2020(6): 3. (in Chinese).

［2］ XIA JY. Vigorously popularizing the achievements of agricultural science and technology and strive to ensure the national food security［J］. China Agricultural Technology Extension, 2011, 27(8): 4-8. (in Chinese).

［3］ LIU J, JIANG YY, HUANG C, et al. Forecast of the occurrence trend of major diseases and pests of grain crops in China in 2021［J］. China Plant Protection, 2021, 41(1): 37-39, 42. (in Chinese).

［4］ XIANG T, QI Y. Food security, food safety and trade: An analysis based on cross-country panel data of pesticide use intensity［J］. Journal of International Trade, 2014(7): 33-41. (in Chinese).

［5］ XU BJ, YOU Y, XUE HZ, et al. Thoughts and suggestions on science and technology supporting national food security［J］. Management of Agriculture Science and Technology, 2020, 39(4): 25-29. (in Chinese).

［6］ XU SR, LIU WC, ZENG J, et al. Retrospect and prospect of plant protection in China for 70 years［J］. China Plant Protection, 2021, 41(4): 29-32. (in Chinese).

［7］ ZHANG CQ, ZHEN XP. Research on the existing problems and countermeasures of reducing pesticide application by grain growers in China［J］. Liaoning Economy, 2019(11): 9-11. (in Chinese).

［8］ WANG YJ, DI F, XIN L. The status and problems of grain production in the main grain production areas of China and policy suggestions［J］. Research of Agricultural Modernization, 2018, 39(1): 37-47. (in Chinese).

［9］ BEH EJ, LOMBARDO R. Correspondence analysis: Theory, practice and new strategies［M］. Hoboken: John Wiley & Sons, 2014: 130-186.

［10］ CAO YR. Research on qualitative data analysis method based on SPSS correspondence analysis［J］. Fujian Computer 2018, 34(10): 4-6, 20. (in Chinese).

［11］ SUN DZ. Application of correspondence analysis to statistical analysis［J］. Journal of Huangshan University, 2006, 8(3): 13-16. (in Chinese).

［12］ TAO FM. On the mathematical model of correspondence analysis［D］. Changchun: Jilin University, 2005. (in Chinese).

［13］ FREUDENTHAL M, MARTN-SUREZ E, GALLARDO JA, et al. The application of correspondence analysis in palaeontology［J］. Comptes Rendus Palevol, 2009(8): 1-8.

［14］ YANG QF. Screening and identification of biocontrol bacteria to induce the resistance against rice seedling blight (Fusarium oxysporum) by seed coating in rice［D］. Shenyang: Shenyang Agricultural University, 2020. (in Chinese).

［15］ YU XD. Main diseases and control methods of rice［J］. Auhui Agricultural Science Bulletin, 2008, 14(10): 37, 102. (in Chinese).