Analysis on the migration of first-generation Mythimna separata(Walker) in China in 2013

ZHANG Zhi , ZHANG Yun-hui WANG Jian LIU Jie, TANG Qing-bo, Ll Xiang-rui CHENG Deng-faZHU Xun

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 Beijing Plant Protection Station, Beijing 100029, P.R.China

3 National Agro-tech Extension and Service Centre, Beijing 100025, P.R.China

4 College of Plant Protection, Henan Agricultural University, Zhengzhou 450002, P.R.China

1. lntroduction

Mythimna separata (Walker) (Lepidoptera: Noctuidae) is an important insect pest in China. Damages are caused by larvae on rice and wheat in southern China, wheat in central China, and corn, millet, sorghum and wheat in northern and northeastern China (Li 1993). Before the 1960s, the migratory capabilities of M. separata were unknown and it was very difficult to provide satisfactory forecasting and early warning information to farmers. During the 1960s, a research group led by Prof. Li Guangbo from the Institute of Plant Protection of the Chinese Academy of Agricultural Sciences established the National Research Collaboration Group for M. separata. Via a large-scale mark-releaserecapture experiment, the group clarified the migration and infestation pattern of M. separata in eastern China and established a novel forecasting method for this insect(Li et al. 1964; Lin et al. 1964; Chen and Bao 1987; Chen et al. 1989). Since then, the occurrence of M. separata has been greatly reduced due to significant reductions of wheat planting area in southern China. Over the past two decades, large-scale outbreaks of M. separata have rarely occurred with the exception of heavy infestations of the third-generation in northern China in 2002 (Pan and Zhai 2009), and in northern and northeastern China in 2012(Zhang et al. 2012).

Insect pest migration and outbreaks are closely related to local meteorological factors such as temperature,humidity and precipitation (Wang and Zhang 2001; Jiang et al. 2003; Yin et al. 2003). Previous studies have confirmed the general M. separata seasonal north-south migration patterns in China and its migration course, source area and immigration events can be largely predicted with the application of searchlight traps, radar observation and trajectory analysis model (Feng et al. 2008; Zhang Y H et al. 2013). However, it is very difficult to accurately forecast outbreaks at the meso- and micro-scale (Lin 1990). To accurately predict M. separata population dynamics, it is necessary to monitor and analyze its migration course at the meso- and micro-scale. More specifically, to make control decisions it is particularly vital to predict the probability of M. separata landing on a crop and subsequent re-migration. In this paper, M. separata migration events were analyzed to determine the influence of weather and terrain factors on its migration and landing.Trajectory simulation was used to determine the source area where these pests originated. These results offer aid in fine-scale forecasting and the prevention and control of M. separata in the future.

基层反腐打黑已取得明显的阶段性成效，广大农村的民主自治和经济社会发展开始步入正常轨道。但数千年传统习俗惯性强大，以宗法制度及其观念为主的封建意识在基层大量残留，特别是基层腐败与黑恶势力造成的消极影响，基层民众伦理观念受到了严重侵蚀。因此，乡村道德体系亟待修复，并通过这种修复，赋予乡村道德体系新的时代意涵，进而推动反腐倡廉成为基层官员和民众的一种内在追求。

2. Materials and methods

2.1. Monitoring locations

Monitoring sites for aerial insect pest populations were located in Yanqing District, Beijing (40°31´N, 116°5´E) and the Luancheng Ecological Experiment Station of the Chinese Academy of Sciences (37°53´N, 114°41´E). Ground monitoring sites were located in Henan, Hebei, Shandong,Liaoning, Inner Mongolia, Jilin and Heilongjiang.

2.2. Population data collection

Daily interval sampling data were collected by searchlight and ground traps; parameters for the setup of searchlight traps have been previously described (Zhang 2013). The search light trap was turned on at sunset and switched off at sunrise the following day. Whole night samples were divided into nine intervals, and each interval lasted approximately 1 h. Quantities and sex ratios of captured M. separata were recorded from each interval during the 2013 migration season. Ovarian developmental stage is important in judging migration events as female migrators are often of underdeveloped ovarian stages (Lin 1990;Zhang Y H et al. 2013). On each migration day, 50 randomly sampled females were dissected to determine ovarian developmental stage and were divided into five categories(Lin 1990; Jiang et al. 2009). On dates when female numbers were less than 50, all available females were dissected. For local ground data, the automatic light traps were setup to capture daily catch sizes of M. separata by the local plant protection stations in the provinces mentioned above.

2.3. Meteorological and map data collection

The global meteorological data from the Final Operational Global Analysis Data (FNL) at 1°×1° was collected every 6 h and was provided by the National Center for Environmental Prediction (NCEP) and the National Center for Atmospheric Research (NCAR), China (http://rda.ucar.edu). Precipitation and topographic data were taken from the National Meteorological Information Center, China. Terrain data were downloaded from the official website of the Weather Research and Forecast (WRF) Center (http://www.wrfmodel.org/). Chinese provincial administrative maps at a scale of 1:4 000 000 were downloaded from the website of the National Basic Geographic Information Center (http://nfgis.nsdi.gov.cn). Precipitation data were collected at 30-min intervals using the portable weather station deployed at the monitoring site in Yanqing.

2.4. Data analysis

In the past, trajectory analysis models were an important tool in atmospheric transport and diffusion analysis.These models were introduced into the study of plant disease epidemics and pest migration, because migratory insects are windborne (Chapman et al. 2010). However,trajectory analysis in early research was only conducted with meteorological data with low spatial and temporal resolution, so it was difficult to guarantee the accuracy of the trajectory simulation output. Fortunately, new models such as WRF-FLEXPART were built in recent years. The Hybrid Single-Particle Lagrangian Integrated Trajectory model (HYSPLIT) is often run when doing trajectory analysis(Zhang et al. 2012; Zhang Y H et al. 2013). Compared with HYSPLIT, FLEXPART is designed with terrain factors in mind and more accurately simulates migration trajectory(Hu et al. 2013, 2014; Bao et al. 2016). In this study, WRFFLEXPART was used instead of HYSPLIT. We found that trapped M. separata moths may originate from Shiyan in Hubei, Nanyang in Henan, northern Jiangsu and Fuyang and Suqian in Anhui, and Jining in Shandong. Furthermore,according to the bulletins from the National Agriculture Technique Extension Service Centre, first-generation M. separata occurred in large numbers in Shanghai,northwest of Anhui, as well as in areas along the Yangtze River in Jiangsu, northern Hubei and in moderate numbers in the middle and lower reaches of the Yangtze River and most parts of Jianghuai. From south to north, migrating routes were aggregated into a narrow channel that was consistent with the ground light trap data. The trap data showed that there were a large number of M. separata moths in the northern areas compared to that in southern areas during the trajectory simulation periods. When the M. separata flew over Yanqing, although they could continue to migrate into some areas in northeastern China, only a few migrators were able to successfully arrive at these monitoring sites.Perhaps during the migration, many migrators landed prematurely. This is why we should deploy additional monitoring sites along the migration routes. Finally, highdensity second-generation larvae occurred in many areas of Shandong in 2013, demonstrating that the WRF-FLEXPART model trajectory simulation offered high accuracy.

Background weather analysis using the WRF modelThe WRF model is a numerical forecasting simulator used for weather research and real-time forecasting, developed jointly by the National Center for Atmospheric Research, the National Center for Atmospheric Environment Research and the University of Oklahoma. It can carry out nested analyses,where each nested area can be operated independently (Skamarock et al. 2008). In this paper, the WRF_3.4.1 model was used to simulate the background weather patterns from the FNL data. The simulation was carried out using double nests. The meteorological element field was at a scale of 36 km×36 km and 12 km×12 km (Table 1) and the final output data were scaled to 20 inches. Output data were then converted into a Grid Analysis and Display System (GrADS) organization using the ARWpost Program. With the NCEP/NCAR re-analysis data and the output data mentioned above, GrADS was used to extract hourly background weather patterns to acquire the general atmospheric circulation characteristics and weather conditions over northern China.

From June 1st to 18th, there were five rainy days and three rainy nights (Fig. 4). During these days, downdrafts were popular on 13 of the nights except for the night of June 12–13th. On these nights, downdrafts were considered to be pressure gradient forces below 925 hPa while updraft winds were measured above 925 hPa. When there were southern winds with downdrafts, the catch number increased. When the vertical velocity of the downdraft declined, the catch number decreased (Figs. 5 and 6).On the night of June 3rd–4th, the first peak in catch sizes appeared in Yanqing where southern winds and updrafts benefited the migration of M. separata. However, on the following night, a heavy rainfall with a strong downdraft developed among the Beijing-Tianjin-Hebei region and this heavy rainfall would block the departure of the migratory M. separata. This downdraft accompanied by rainfall resulted in passive landing and therefore the trapped number(324) was much smaller than that of the previous night (1 062). During the second peak in catch sizes, downdrafts took place over all the nights and vertical velocity averages of most nights (21:00–05:00) were greater than 0.002 m s–1at 925 hPa, with the maximum being 0.008 m s–1(Fig. 6).The vertical velocity measured during these nights could force M. separata to passively land. To our surprise, on the night of June 12–13th, although the wind direction was northward, the vertical fields were at an updraft, and peak catch numbers still transpired (Figs. 5 and 6).

3. Results

3.1. Aerial population dynamics

In 2013, the earliest appearance of adult M. separata was on the night of May 17–18th in searchlight traps and on June 2nd–3rd in ground light traps in Yanqing. By June 22nd –23rd,the cumulative number of trapped adult M. separata in searchlight and ground light traps was around 250 000 and 5 300, respectively. The majority of moths (99%) were trapped in June. There were obvious fluctuations in the number of first-generation trapped adults in searchlight and ground light traps and when plotted, four peaks in captured populations could be observed. The largest peak appeared on the night of June 7–8th in the searchlight traps and on June 9–10th in ground light traps. Within searchlight and ground light traps,the maximum number of trapped moths was 86 000 and 1 600, respectively, and the cumulative number of trapped moths during the three nights accounted for 88.4 and 87.2%of the total migration process, respectively (Fig. 1). During peak periods, most moths were trapped after 00:00 in searchlight traps. On the night of June 6–7th and 7–8th, trapped moths in searchlight traps before 00:00 accounted for 18.13 and 22.92% of the total, but accounted for 81.87 and 77.08% after 00:00, respectively.

?

3.2. Ground light trap population dynamics

From the end of May to the beginning of June, many peaks in moth catch amounts appeared in ground light traps in Shandong, Henan, and Hebei provinces. The earliest peak appeared during May 30th to June 3rd in Dancheng,Yongcheng, Mengzhou, and Qinyang, Henan Province(Table 2). The earliest peak period was postponed by 1–3 d in the Shandong and Hebei monitoring sites that were located north of the Henan monitoring site. The Hebei Province peak periods appeared on June 5–9th. Being located even more north, the peak period duration in Inner Mongolia and northeastern China was much longer, appearing after June 5th. Generally, sites in southern Beijing trapped more moths. The cumulative trapped moths by a single light trap was 3 748, 3 380, and 3 225 at Yongcheng, Mengzhou, and Dancheng in Henan Province, respectively; 597 and 541 at Qufu and Wenshang in Shandong, respectively; and 6 164 at Daming in Hebei Province. The largest catch size trapped by searchlights was in Luancheng County with 8 736 moths trapped and occurred on the night of June 5th. Numbers of trapped moths from Inner Mongolia and northeastern China were generally much smaller and the number of cumulative trapped moths at these locations during the whole migration period was less than 100 (Table 2).

3.3. Physiological characteristics

Fig. 1 Nightly catch amounts of adult Mythimna separata in searchlight traps and ground light traps at Yanqing, China in 2013.

Table 2 Migration peaks for Mythimna separata captured in China in 2013

Up to June 22nd, sex identification was conducted on 31 683 moths (♀:♂=17 966:13 717) in total. We found that the sex ratio was significantly higher than 1 (χ2=569.83,df=1, P＜0.001). From the May 3rd through June 22nd,there were 25 d when identified moth numbers were more than 30 and statistical analysis was executed. Females were captured significantly more than males on 20 of the 25 days tested. In the remaining days when males were captured more than females, the highest ratio was 8.86 (♀:♂=390:44). Sex ratios changed with fluctuating population dynamics. As the peak period approached, the ratio largely fluctuated. Prior to May 30th, ratios were greater than or approaching 1, but after May 31st ratios increased rapidly and reached 8.86 on the night of June 1st–2nd. After this, the ratio declined to 1.39 on the night of June 5–6th.During the greatest peak period from June 6–9th, sex ratios fluctuated less dramatically than population sizes. In the late peak period from June 19th–21st, the ratio was less than 1. A chi-squared test showed that there were eight nights where captured females were significantly greater than males and twelve nights where the ratios were not significantly different from 1, including five nights with sex ratios less than 1, and seven nights with sex ratio greater than 1 (Fig. 2). In addition, all sex ratios determined from ground light traps were also greater than 1 (Table 2). More than 80% of trapped females were at grade II–III of ovarian development. During the peak period, ovarian stages did not change significantly for several days, following the physiological characteristics of typical migratory pests.Similar data were recorded in ground light traps (Table 2).

Landing is an important process of insect migration. How and where migrators land is considered to be vital for pest insect forecasting. The landing process can be affected by many factors, including passive effects from circadian rhythms, rainfalls and downdrafts, and initiative effects from suitable breeding habitats or pheromones from the female.Landing effects from topographic stress can be sorted into two sides. The first is the obstructing effect of rising elevation, where a migratory route can be blocked directly or M. separata could be forced to fly at greater heights and thus their flight blocked due to air temperature reduction with the increased elevation (Chapman et al. 2008, 2010;Nowinszky et al. 2012; Zhang Y H et al. 2013). The second is that rising terrain could induce warm and humid airflow and bring rainfall, which can result in M. separata landing(Hu et al. 2013, 2014). In 2013, although large numbers of M. separata were captured in the searchlight trap, we could not ascertain whether or not they landed passively due to the lack of radar observations. We made an assumption based on the weather pattern being beneficial for landing instead of declaring a single factor that could be important during the landing process. Yanqing is located in northern China and surrounded by the Taihang Mountains and Yanshan Mountains which could also have an effect on M. separata migration as mentioned above. We found that southern wind was necessary for migrants and this finding was consistent with a previous study (Lin et al. 1990). Rainfall and downdraft winds were also very important factors. On some nights with heavy rainfall, migration was blocked. On other nights when wind direction was suitable but vertical velocity was not suitable for landing, the captured numbers also decreased(Figs. 4–6). In the future, the landing process would be better understood if additional migration event data were collected using multiple tools especially entomological radars.

3.4. Relationship between catch sizes and wind direction

脱靶量位于天线2监视的区域，表示偏靶方向为船首方向。而靶船长度约为50 m，根据测量的脱靶量(相对于脱靶量测量系统)和偏靶方向，导弹应击中靶船中部位置。靶船回岸以后，对靶船进行检靶，脱靶量测量系统与导弹命中位置如图11所示。导弹击中靶船中部位置，而且用皮尺测量脱靶量测量系统与命中位置的实际距离为20 m，测量值与真实值相符，且测量误差≤1 m。实际偏靶方向(相对于脱靶量测量系统)与系统检测偏靶方向相同。坐标转换后可直接转换为导弹相对于靶船中心的脱靶量。试验结果表明，该系统切实可行，可用于导弹脱靶量测量。

3.5. Rainfall and downdraft frequency during migration period

Fig. 2 The Mythimna separata sex ratio (female:male) and numbers caught in searchlight traps at Yanqing, China in 2013. According to Fisher’s theory, the sex ratio should be 1, so sex value in theory in this figure was supposed to be 1.

Fig. 3 Time-latitude profiles of wind fields along 116°E at 925 hPa in 2013 (m s–1). A, May 31st–June 4th. B, June 5–9th. C, June 10–14th. D, June 15–19th. Transverse lines indicate latitude where monitor site was located. UTC, coordinated universal time.

Fig. 4 Nightly precipitation and Mythimna separata numbers caught at monitoring site in Yanqing, China in 2013.

Trajectory analysisFLEXPART is a Lagrangian transport and dispersion model developed by the Norwegian Institute of Atmospheric Research (Favaretto et al. 2004).It is suitable for the macro- and meso-scale transportation, diffusion, wet and dry sedimentation, and radiation attenuation of tracers in the atmosphere by calculating the trajectories of large amounts of particles released by point, line, area or volumetric sources (Stohl et al. 2005). Trajectory analysis was carried out in FLEXPART using the output data from the WRF_3.4.1 model and monitoring sites were set up as starting point. For searchlight trap samples, the intervals with significantly larger catch numbers compared to the average (determined by a chi-square test) were considered the main pest migration periods, while the median time within the interval was defined as the starting time for trajectory analysis (Lin et al. 1963; Zhang and Li 1985; Cao et al. 1995;Chen et al. 1995; Feng et al. 2008). When catch numbers from each interval were not significantly different, the start time was designated at sunset for forward simulations and at sunrise for backward simulations. All backward simulations were terminated at sunset and forward simulations were terminated at sunrise. All simulations were performed over five continuous nights. The hourly migration trajectory data were imported into ArcGIS10.1 (ESRI, Redlands CA) for modeling. Trajectories that were not of biological significance were removed.

3.6. Trajectory analysis

Fig. 5 Relationship between wind velocity at 925 hPa and Mythimna separata numbers caught at monitoring site in Yanqing,China in 2013. Letter near hollow circle were wind directions, S, W, E and N represent south, west, east and north, respectively.

Fig. 6 Relationship between vertical velocity at 925 hPa and Mythimna separata numbers caught at monitoring site in Yanqing,China in 2013. For vertical velocity, negative values indicate updraft and positive values indicate downdraft. The vertical velocity values were very small and units were multiplied by 100.

When backward and forward trajectories were stimulated where obvious migration events of M. separata appeared from local sites such as Daming County, Ningjin, Renqiu and Luancheng in Hebei Province, we found that first-generation moths were mainly originating from Shandong, Jiangsu,Anhui and Hubei (Fig. 7-A). The migratory trajectories seemed to be converged before migratory moths reached the Beijing-Tianjin-Hebei region. When they passed the Beijing-Tianjin-Hebei region, the migratory trajectories became dispersed (Fig. 7-B). Trapped moths from Yanqing mainly derived from Shiyan in Hubei, Nanyang in Henan,northern Jiangsu, Fuyang and Suqian in Anhui, and Jining in Shandong. The simulation of forward trajectories from Yanqing showed that M. separata moths could fly across Beijing or continue to migrate to western Liaoning,northeastern Heilongjiang and the junction of Inner Mongolia, Liaoning, and Jilin after a stopover in Beijing(Fig. 7-A). The simulated routes were in accordance with the trapped M. separata peak periods that appeared in the light traps at the plant protection stations in northeastern China (Table 2 and Fig. 7-A).

From late May to early June, winds in northern China were impacted by a subtropical anticyclone that gradually extended to the western region. From June 1st to 6th,there was a cyclone above the North China Plain and an anticyclone above the sea. As a result of the great difference between the winds in the center of anticyclone and cyclone, a southern low level jet was formed that benefited the migration of M. separata. From June 8th to 10th, a southwestern low level jet also accompanied the withdrawal of the subtropical anticyclone and benefited migration. During the peak period, frequent strong southern winds impacted northern China except during the nights of June 9–13th (Fig. 3). Captured M. separata population sizes were closely related to the direction of the wind. Trapped moths increased when southern winds occurred but decreased when the wind direction changed northward. On the night of June 3rd–4th, the first peak in captured moths appeared when the southern winds prevailed. On the night of June 4–5th, despite the southern winds, the captured population size decreased dramatically compared to the size of the previous night, possible due to heavy rainfall (Fig. 4). On the nights from June 6–9th,the heaviest migration night appeared when high speed southern winds (more than 10 m s–1) prevailed. From the night of June 7–8th when the convergence of the wind field appeared along the latitude of 40.5°N, the captured population size gradually decreased.

4. Discussion

Early studies have shown that first-generation M. separata moths have numerous bursts of high occurrence in early spring instead of one large peak occurrence. Furthermore,peak occurrence has also been shown to be consistent with strong southwest winds (Lin et al. 1963). Our monitoring data are in agreement with previous studies. In 2013, besides the main immigration peak found in the searchlight traps,there were also many small peaks before and after the main migration period. Data from ground light traps of local plant protection stations were similar to data from the searchlight trap in Yanqing, both indicating the presence of many occurrence peaks in the migration process. Furthermore,more northern regions seemingly had more peaks. Previous work showed that first-generation M. separata moths arrived in Liaoning, Jilin and Inner Mongolia from late May to early June, and in some places arrival date was delayed to late June (Chen et al. 1965, 1989). Peak occurrence dates derived from ground light traps were different, even at similar latitudes. Yanqing is located at the most northern part of China. In 2013, migratory first-generation moths above Yanqing was captured mainly during nights in early June (Fig. 1). This finding was consistent with the previous report. However, such an extensive migration of the firstgeneration M. separata moths has rarely been recorded prior to 2013. Extensive migration of this pest could cause serious damages to corn crops at this seasonal time because the highly sensitive corn seeding stage coincided with this migration. A future recommendation would be for the local plant protection stations to pay more attention to M. separata migration events in late June.

Fig. 7 The trajectories of Mythimna separata in north and northeastern China in 2013. A, the starting trajectory point began at the monitoring site in Yanqing, China, the black dotted line indicates backward trajectory while the solid line indicates forward trajectory. B, the starting trajectory point began at the monitoring sites from the plant protection stations presented in the legend,all were forward trajectories.

现如今，大部分建筑企业所面临的非常明显的问题之一就是其相关的管理人员和领导层不重视对工程的造价进行有效的控制。工程造价观念的缺失便会造成大量的资金和成本的浪费。尤其是现如今我国市场经济不断发展，工程造价预算的管理和控制也越来越重要。如果建筑企业内部一直无法提起对工程造价的管理的重视，便会阻碍整个建筑行业在市场中的发展。

M. separata moths do not fly long distances during the daytime. To travel over long distances, they must fly over several nights. Simulation trajectory analysis illustrated that moths passing by Yanqing could continue to fly to Inner Mongolia and the northeastern region. During the migration,some would stay for breeding and others would continue to fly, forming a progressive migration pattern. This could be the reason why total caught moths were fewer in traps from sites further north compared to those located more southern (Table 2). In addition, peak occurrences from the sites closer to the south appeared earlier and were delayed in northern sites.

Previous light trapping experiments showed that the numbers of captured females were different from that of captured males (Lin et al. 1963). The possible reason could be their flying ability. Migratory insects generally begin to migrate during the immature period after eclosion, so the ovarian development of the migratory females are often at lower stage, e.g., grades I and II (Lin 1990). In this study,we found that more than 80% of trapped females were at a developmental grades II–III, following the physiological characteristics of obligate migratory pests. In an early report, the sex ratio of different insect pest species trapped at various times were found to be different, and the sex ratio of a M. separata population enumerated from a light trap could be divided into three stages (Li et al. 1987). The first stage was 1–4 d after emergence, and gave a sex ratio close to 1. The second stage was 5–10 d after emergence, when ovaries were developing and the flight capacity of females weakened as a result of increasing weight. Therefore, the percentage of captured females reduced gradually. The third stage was 10 d after emergence when male moths accounted for the majority (Li et al. 1987; Lin 1990). Wang and Zhang (2001) found that the proportion of migratory females (52.47%) was significantly lower than that of migratory males (97.14%). For this season, our results were distinct from the previous report from Wang and Zhang(2001) but consistent with Lin (1990). When considering total examined M. separata moths, the trapped females were found at higher numbers compared to the males, with similar findings observed from the ground light trap moths.Furthermore, there were more days with a female to male ratio＞1 than days with ratios＜1. This ratio greatly fluctuated with the population dynamics. For other species such as Scotogramma trifolii and Athetis lepigone, sex ratios were similar to those of M. separata observed in this study. When the population rises, the sex ratio was significantly higher than 1, and when the population reaches a stable period, the sex ratio was close to or significantly less than 1 (Zhang et al.2007; Zhang Z et al. 2013). The sex ratio pattern may be affected by different flying abilities. Further study is needed to test whether there are other factors affecting the sex ratio.

分析结果显示，有限元的排水量计算结果较大，剩余水头高度较小；而方法一的计算结果正好相反。因此，出于安全的角度考虑，在计算盲沟排水量时，建议参考有限元计算结果，而在计算剩余水头高度时，建议参考方法一的计算结果。在3个主要影响因素中，水位降深对双排水盲沟渗流的影响最大，且接近正比例关系；排水盲沟间距对排水量影响的敏感度相对较小，而对剩余水头高度影响相对较大；渗透系数则只对排水量有影响，对剩余水头高度无影响。

设计原则包括：（1）该区排水体制采用雨、污分流制。雨水分散排放，污水集中收集处理达标后排放。（2）采用重力流模式，雨水与排洪系统统一考虑，就近分散排放。合理布置管道标高，预留适量的支管接口，便于周边雨、污水的接入，并且可以避免与其他管线发生冲突。（3）海绵城市设计依据《武汉市海绵城市规划设计导则》（以下简称“导则”）相关内容执行。

首先，当客户不再拥有自己的车队和仓储库存而是全部依赖于“第三方”为它和它的客户提供部分或全部的物流服务时用户和第三方物流公司之间就会形成共享信息、共担风险、共同获利的关系[14]。

Migration event assessmentsM. separata cannot overwinter north of the 0°C isotherm (also referred to as the area along 33–34°N) in January (Li et al. 1964). According to the law of effective accumulative temperature, native M. separata moths do not emerge until early July. This means that all moths trapped in Yanqing before July were immigrants of the first generation. In order to finely assess migration events, the capture data from the searchlight traps and ground traps were analyzed in tandem with the ovarian developmental data as previously described by Feng et al. (2004) and Zhang Y H et al.(2013), respectively. When M. separata migration was detected on any given night, we concurrently performed a more extensive analysis using the meteorological data.

Insect migration is a very complex progress that can be affected by many factors such as wind fields, rainfall,temperature and humidity. In this study we reported the migration pattern of first-generation M. separata appearing in China in 2013. The number of trapped M. separata moths was astonishing compared to previous records.With our analysis, we found that several features were related to the migratory population. The first was that populations occur in waves and can suddenly increase as we saw with our captured moth data. The second was that peak occurrence was consistent with strong southwestern winds. Lastly, we found that the number of trapped females was significantly higher than males and the captured female moths were at a lower grade developmental stage, consistent with other migratory species. The statistics on the relationship among catch size,rainfall and downdraft showed that rainfall and downdraft have the greatest influence on migratory M. separata landing and that WRF-FLEXPART was able to provide high accuracy trajectory stimulations. Unfortunately, without an entomological radar, we have not observed the landing progress directly. M. separata forecasting would benefit from surveying when heavy populations occur in a source area with southern winds ensuing concurrently.

5. Conclusion

According to our observation data, heavy migration of first-generation of M. separata appeared in northern China in 2013. In Yanqing District, Beijing, the cumulative number of captured adults in searchlight traps was around 250 000 and the daily maximum for trapped moths was 86 000. The sex ratio at each monitoring site was greater than 1 and greatly fluctuated with population dynamics. With the assistance of WRF model and related outputs, we found that prevailing downdraft winds benefited M. separata moths to land passively during the migration peak and trajectory simulation showed that immigrants were from Anhui, Jiangsu and Hubei provinces and most of them could continue to fly into the northeastern regions of China. Migration is a very complicate process, includes takeoff, transmission and landing. Whether migrants could land or not is vital to judge their damages at defined place. In the future, we could provide more accurate pest information to farmers at mesoand micro-scale with this method mentioned in this article.

Acknowledgements

We are grateful to many students from Henan Agricultural University and Shandong Agricultural University, China, for their assistance in field tests. This work was supported by the Special Fund for Agro-scientific Research in the Public Interest, China (201403031), the National Natural Science Foundation of China (31571998), the earmarked fund for China Agriculture Research System (CARS-03), and the National Key Research and Development Program of China(2016YFD0300702).

两组的总有效率为93.33%和76.67%,观察组的总有效率高于对照组,统计学有意义(P<0.05)。见表1。

Bao Y X, Sun M Q, Yan M L, Lu M H, Liu W C. 2016. Comparative study of migration trajectories of the brown plant hopper,Nilaparvata lugens (Stål), in China based on two trajectory models. Acta Ecologica Sinica, 36, 6122–6138. (in Chinese)

Cao Y Z, Huang K, Li G B. 1995. The effect of relative humidity on flight activity of adult oriental armyworm. Acta Phytophylacica Sinica, 22, 134–138. (in Chinese)

Chapman J W, Nesbit R L, Burgin L E, Reynolds D R, Smith A D, Woiwod I P, Middleton D R, Hill J K. 2010. Flight orientation behaviors promote optimal migration trajectories in high-flying insects. Science, 327, 682–685.

Chapman J W, Reynolds D R, Mouritsen H, Hill J K, Riley J R,Sivell D, Smith A D, Woiwod I P. 2008. Wind selection and drift compensation optimize migratory pathways in highflying moth. Current Biology, 18, 514–518.

Chen R L, Bao X Z. 1987. Research on the migration of oriental armyworm in China and a discussion of management strategy. Insect Science and Its Application, 8, 571–572.

Chen R L, Bao X Z, Drake V A, Farrow R A, Wang S Y, Sun Y J, Zhai B P. 1989. Radar observations of the spring migration into northeastern China of the oriental armyworm moth, Mythimna separata, and other insects. Ecological Entomology, 14, 149–162.

Chen R L, Leo C S, Liu T Y, Li M C, Sien N Y. 1965. Study on the source of the early spring generation of the armyworm Leucania separata Walker in the kirin province studies on the original ground of the immigrant moths. Acta Phytophylacica Sinica, 4, 49–58. (in Chinese)

Chen R L, Sun Y J, Wang S Y, Zhai B P, Bao X Z. 1995.Migration of the oriental armyworm Mythimna separata in East Asia in relation to weather and climate. I. Northeastern China. In: Drake V A, Gatehouse A G, eds., Insect Migration:Tracking Resource through Space and Time. Cambridge University Press, Cambridge, United Kingdom. pp. 93–104.

Favaretto A, Ceresoli G, Paccagnella A, Barbieri F, Bearz A,Ghiotto C, Oniga F, Schiavon S, Frustaci S, Villa E. 2004.Source-receptor matrix calculation with a Lagrangian particle dispersion model in backward mode. Atmospheric Chemistry & Physics, 4, 51–63.

Feng H Q, Zhao X C, Wu X F, Wu B, Wu K M, Cheng D F,Guo Y Y. 2008. Autumn migration of Mythimna separata(Lepidoptera: Noctuidae) over the Bohai Sea in northern China. Environment Entomology, 37, 774–781.

Hu G, Lu F, Lu M H, Liu W C, Xu W G, Jiang X H, Zhai B P. 2013. The influence of typhoon Khanun on the return migration of Nilaparvata lugens (Stål) in eastern China.PLoS ONE, 8, e57277.

Hu G, Wu Q L, Wu X W, Jiang Y Y, Zeng J, Zhai B P. 2014.Outbreak mechanism of second generation armyworms in northeastern China: A case study in 1978. Chinese Journal of Applied Entomology, 51, 927–942. (in Chinese)

Jiang X F, Cai B, Luo L Z, Cao Y Z, Liu Y Q. 2003. Influences of temperature and humidity synthesize on flight capacity in the moths of oriental armyworm, Mythimna separata(Walker). Acta Ecologica Sinica, 23, 738–743. (in Chinese)

Jiang Y Y, Qu X F, Xia B, Zeng J. 2009. Rules for Investigation and Forecast of the Armyworm [Pseudaletia (Mythimna)Separata Walker]. National Standard of the People’s Republic of China, Standards Press of China, Beijing. (in Chinese)

Li G B. 1993. The general situation and progress of the research on the armyworm in China. Plant Protection, 19, 2–4. (in Chinese)

Li G B, Wang H X, Hu W X. 1964. Route of the seasonal migration of the oriental armyworm moth in eastern China as indicated by a three year result of releasing and recapturing of marked moths. Acta Phytophylacica Sinica, 3, 101–109.(in Chinese)

Li R D, Wang J Q, Su D M. 1987. Insect Ovarian Developmental and Prediction of Their Population Dynamics. Fudan University Press, Shanghai. pp. 99–108. (in Chinese)

Lin C S. 1990. The Physiology and Ecology of the Oriental Armyworm. Peking University Press, Beijing. (in Chinese)

Lin C S, Sun J R, Chen R L, Zhang Z B. 1963. Studies on the regularity of the outbreak of the oriental armyworm,Leucania separata Walker. I. The early spring migration of the oriental armyworm moths and its relation to winds. Acta Entomologyica Sinica, 12, 243–261. (in Chinese)

Lin C S, Zhang Z B. 1964. Studies on the regularities of the outbreak of the oriental armyworm (Leucania separata Walker). V. A model for seasonal long-distance migration of the oriental armyworm. Acta Phytophylacica Sinica, 3,93–100. (in Chinese)

Nowinszky L, Kiss O, Szentkiralyi F, Puskás J, Ladányi M.2012. Influence of illumination and polarized moonlight on light-trap catch of caddisflies (Trichoptera). Research Journal of Biology, 2, 79–90.

Pan L, Zhai B P. 2009. Analysis of source population and immigration process of Mythimna separata (Walker) caused an outbreak in North China in 2002. Acta Ecologica Sinica,29, 6248–6256. (in Chinese)

Skamarock W C, Klemp J B, Dudhia J, Gill D O, Barker D M,Wang W, Powers J G. 2008. A description of the advanced research WRF version 3. In: NCAR Technical Note NCAR/TN-475+STR. National Center for Atmospheric Research,Colorado, USA.

Stohl A, Forster C, Frank A. 2005. The Lagrangian particle dispersion model FLEXPART version 6.2. Atmospheric Chemistry & Physics & Discussions, 5, 2461–2474.

Wang Y Z, Zhang X X. 2001. Studies on the migratory behaviours of oriental armyworm, Mythimna separata(Walker). Acta Ecologica Sinica, 21, 772–779. (in Chinese)

Yin J, Feng H Q, Cheng D F, Cao Y Z. 2003. Observations on the flight behavior of the oriental armyworm, Mythimna separata (Walker) in airflow. Acta Entomologia Sinica, 46,732–738. (in Chinese)

Zhang Y H, Chen L, Cheng D F, Zhang Y J, Jiang Y Y, Jiang J W. 2007. Radar observation and population analysis on the migration of the clover cutworm, Scotogramma trifolii R. (Lepidoptera: Noctuidae). Acta Entomologica Sinica, 50,494–500. (in Chinese)

Zhang Y H, Zhang Z, Jiang Y Y, Gao Y B, Cheng D F. 2012.Preliminary analysis of the outbreak of the third-generation armyworm Mythimna separata in China in 2012. Plant Protecion, 38, 1–8. (in Chinese)

Zhang Y H, Zhang Z, Li C, Jiang Y Y, Zeng J, Cheng D F.2013. Seasonal migratory behavior of Mythimna separata(Lepidoptera: Noctuidae) in Northeast China. Acta Entomologica Sinica, 56, 1418–1429. (in Chinese)

Zhang Z. 2013. Monitoring and population dynamics analysis of important migratory pest insects in northern China. Ph D thesis, Chinese Academy of Agricultural Sciences. (in Chinese)

Zhang Z, Zhang Y H, Jiang Y Y, Xie A T, Wei S J, Cheng D F, Jing J W, Zhang F M, Peng H. 2013. Monitoring of the population dynamics of Proxenus lepigone (Lepidoptera:Noctuidae ) in North China and analysis of the source of its populations in northern Beijing. Acta Entomologica Sinica,56, 1189–1202. (in Chinese)

Zhang Z T, Li G B. 1985. A study on the biological characteristics of the flight of the oriental armyworm (Mythimna separata Walker) moth. Acta Phytophylacica Sinica, 12, 93–100.(in Chinese)