WU Qiu-lin ,SHEN Xiu-jing, ,HE Li-mei ,JIANG Yu-ying ,LIU Jie ,HU Gao,WU Kong-ming

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 College of Plant Protection,Fujian Agriculture and Forestry University,Fuzhou 350002,P.R.China

3 National Agro-Tech Extension and Service Center,Beijing 100125,P.R.China

4 Department of Entomology,Nanjing Agricultural University,Nanjing 210095,P.R.China

Abstract The fall armyworm (FAW,Lepidoptera:Noctuidae),Spodoptera frugiperda (J.E.Smith),invaded China in mid-December 2018; since then,it has become a great threat to Chinese agricultural production.Qinling Mountains–Huaihe River region(QM–HRR) is the transitional zone between northern and southern China,an important region for both corn and wheat production.Based on the actual occurrence of QM–HRR invaded by FAW in 2019,daily mean surface air temperature and nocturnal wind conditions at 925 hPa were examined,and migratory routes of FAW moths originated in QM–HRR were modeled by a forward-trajectory-analysis approach.The results indicated that migratory activities of FAW adults emerged in QM–HRR were initiated from late June.The moths from western QM–HRR,where has complex topographic terrain,mainly flied to Ningxia and Inner Mongolia before mid September.However,FAW moths from the eastern QM–HRR primarily engaged in high-altitude northward transport assisted by the prevailing southerly winds before mid August,and the North China Plain was identified as the main destination of FAW.Meanwhile,the migration trajectories of FAW moths had a possibility to reach the Northeast China Plain.From mid August,FAW moths in eastern QM–HRR largely migrated southward and returned to the Yangtze River Valley.This study provides detailed information on the occurrence and migration routes of FAW moths from QM–HRR and will be helpful for early warning and development of integrated pest management strategies for the control of this exotic insect pest.

Keywords:Spodoptera frugiperda,invasive pest,Qinling Mountains–Huaihe River region,atmospheric circulation,windborne migration trajectory

1.Introduction

The fall armyworm (FAW),Spodopterafrugiperda(J.E.Smith) (Lepidoptera:Noctuidae),is a native of tropical and sub-tropical America that does not undergo diapause(Luginbill 1928; Sparks 1979).However,taking advantage of wind currents for transport and its own migratory nature,the FAW undertakes seasonal multi-generational migration to expand its distribution range throughout most of North America from spring,reaching Québec and Ontario provinces in Canada by the end of summer (Luginbill 1928;Roseet al.1975; Westbrook and Sparks 1986; Mitchellet al.1991).FAW moths can travel over hundreds of kilometers per night under favorable synoptic weather conditions(Westbrook and Sparks 1986; Westbrooket al.2016),whereby the insect has become the dominant agricultural pest species in America,infesting important economic crops such as maize,cotton,soybean,and potato,in addition to various vegetable hosts (Dayet al.2017; Montezanoet al.2018).As was the case with the cotton bollworm(Helicoverpaarmigera) introduced into the New World (Tayet al.2013; Joneset al.2019),the increasingly frequent agricultural and horticultural trade between Americas and other continents,together with long-distance migration,might cause the highly polyphagous FAW to become a serious risk to many crops in the rest of the world (Cocket al.2017; Earlyet al.2018).

The presence of FAW was suddenly reported in Nigeria and Ghana in January 2016 (Goergenet al.2016),from where it quickly spread throughout sub-Saharan Africa(Stokstad 2017).In May 2018,FAW was officially reported in India (Sharanabasappaet al.2018),after which Yemen,Sri Lanka,Bangladesh,and countries of the Indo-China Peninsula confirmed the infestation of maize plants with FAW (Wuet al.2019a).Using a searchlight trap,on 11 December 2018,the insect pest was spotted for the first time in Yunnan Province of China,while trying to immigrate into Lancang County (Sunet al.2021),and it has been spreading within China since that time (Jianget al.2019).

A major factor contributing to widespread crop damage by FAW throughout the American continent is thought to be the capacity of the species for long-distance migration at high altitude (Westbrooket al.2016; Earlyet al.2018).Myanmar is the purported origin of FAW currently invading southwestern China (Liet al.2020; Wuet al.2019c),from which point,it has already infested maize crops throughout 26 provinces (autonomous regions,municipalities) since early in 2019,thereby posing a severe threat to food security in the country (Jianget al.2019).

A clear understanding of movement ecology of this invasive pest is crucial for effective,integrated pestmanagement practice.Atmospheric conditions greatly regulate flight initiation,displacement,and termination of high-flying insect migrations (Drake and Farrow 1988).Thus,insect migration trajectories and patterns have been identified based on a series of atmospheric simulation models (Westbrook and Sparks 1986; Chapmanet al.2012;Huet al.2013; Westbrooket al.2016); similar studies have been conducted to investigate the seasonal displacement of migratory flights of FAW moths in China (Liet al.2020; Wuet al.2019a,b,c).The results indicate that the Yunnan and Guangxi were the first to bear damage by FAW migrating from Myanmar before April (Liet al.2020; Wuet al.2019c).Once the FAW quickly infested the tropical and southern subtropical zones of China in the spring,the population provided sufficient FAW migrants to colonize the Yangtze River Valleyviatransport support by the prevailing southerly wind currents (Wuet al.2019b).Taking advantage of the summer monsoon generating over the whole of eastern China,southwesterly airflows enabled the Yangtze River Valley population to continue migrating in a northward direction to occupy the North China Plain (Wuet al.2019a),which is the major corn-growing region in China.These studies have placed the whole country on the alert for the incursion of this exotic insect pest and,more importantly,they are remarkably consistent with the timing of FAW appearances in different regions (Jianget al.2019).Qinling Mountains–Huaihe River region (QM–HRR) consists of six studied provinces:Jiangsu,Shandong,Henan,Shanxi,Shaanxi,and Gansu,is an important region for both corn and wheat production,as well as the transitional zone between northern and southern China.As predicted,the host crops planted in QM–HRR were initially damaged by FAW since June 2019 (Jianget al.2019; Wuet al.2019a).Meanwhile,QM–HRR has been on high alert and relevant response plans have been immediately undertaken against the invasion of this new pest.However,given the great mobility and rapid development of FAW,the establishment of this invasive pest in QM–HRR has been assumed to pose an increasingly severe threat to host plants in other key cropping areas.Furthermore,the infestation status and migratory movement of FAW offspring produced in QM–HRR remain unclear.

To provide more information for developing proper management strategies against this pest both regionally and nationally,it is important to understand the dynamics of incidence and redistribution of the migratory FAW populations originated in QM–HRR.Therefore,based on the dynamics of invasion and the local status of FAW incidence in 2019 in QM–HRR,the migration periods (i.e.,adult moth emergence dates) of the first and second generations of adult FAW from QM–HRR were determined based on the national field investigations and on the actual day–degree method proposed for predicting FAW development (Taylor 1981; Westbrook and Sparks 1986).Then,we used the Weather Research and Forecasting (WRF) model(Skamarocket al.2008) to conduct a numerical simulation of high spatiotemporal synoptic weather conditions during the migration periods.Subsequently,migratory flights of FAW moths from QM–HRR were modeled by inputting the atmospheric background to a three-dimensional trajectory analysis program (Hu 2015) that considered self-propelled flight behavior variables (such as departure time,flight altitude,flight duration,common orientation,airspeed,etc.)and unfavorable cold weather conditions leading to forced landing of FAW mothsenroute.Lastly,main landing,sweeping areas,and potential infestation risk of FAW migrants from QM–HRR were quantified.

2.Materials and methods

2.1.Study site and field survey of newly-invasive FAW in QM–HRR

The area of QM–HRR under study shows a complex topography.Specifically,the eastern plain consisting of Jiangsu,Shandong,and Henan provinces is located at an altitude below 1 000 m above mean sea level (a.m.s.l.); the western region,including,Shanxi,Shaanxi,and Gansu provinces averages 500–2 000 m a.m.s.l.All map data were derived from the National Catalogue Service for Geographic Information (http://www.webmap.cn/).During the multiprovincial systematic field surveys organized by the China National Agro-Tec Extension and Service Centre,the initial crop infestation in QM–HRR by FAW larvae was recorded for a total of 98 sites from 1 June to 10 July 2019.For a better understanding of the biology of FAW in the field,which can be shared and used for early warning,a standardized field scouting at each site was conducted every 2 days (Jianget al.2019).In the corn field,5 plots were sampled using‘W’ pattern to cover the entire field.At each plot,10 plants in a row were inspected.Pheromone traps were also built up to monitor the presence and establishment of FAW in a particular site.FAW-specific lures were routinely replaced in traps for the whole cropping season.To accurately monitor the invasion and occurrence of FAW adults in QM–HRR,traps were also checked daily by counting the number of FAW moths inside.Thus,the date of the first invasion of FAW and the developmental stage of FAW larvae were recorded for each site.

2.2.Calculation of cumulative degree–days for estimation of phenological growth stage of FAW

Temperature directly affects phenological development from the egg stage through eclosion of the adult moth (Hogget al.1982; Barfield and Ashley 1987; Heet al.2019).To calculate cumulative degree–days for estimating the initial emergence date of FAW adults for the first and second generations,daily mean surface air temperature (DMSAT)at each site from June to September in 2019 was derived from the SURF_CLI_CHN_MUL_DAY dataset (version 3.0)which was downloaded from the National Meteorological Information Center (http://data.cma.cn/).In the present study,FAW development was calculated based on heat accumulation of degree-day (DD) values above a threshold temperature of 9.16°C (DD9.16°C,same thereafter; Heet al.2019).All calculations described here were conducted at daily intervals.Therefore,the emergence dates of the first and second generations of FAW adults in the newly invaded QM–HRR in 2019 were computed.Specifically,adult moth emergence of first FAW generation was estimated to take place on the day associated with a certain cumulative value,for example,254.18 DD9.16°Cafter 4th-instar FAW larvae were reported during the field investigation (Heet al.2019); emergence of the second adult moth generation was also estimated to occur on dates with a cumulative value of 680.02 DD9.16°C(Heet al.2019),which is the effective accumulated temperature for the full life cycle of FAW.

2.3.Wind transportation and circular statistics

To investigate the transporting wind conditions,on which long-distance migration of FAW largely depends,we displayed wind speed and direction at 925 hPa (the altitude at which FAW typically fly; Wolfet al.1990; Westbrooket al.2016,2019).Long-term atmospheric weather data were extracted and calculated from the National Centers for Environmental Prediction (NCEP; http://www.ncep.noaa.gov/) final analysis data (FNL),which were produced at 6 h intervals (0000,0600,1200 and 1800 UTC) on a spatial resolution of 1.0°×1.0° global grids.Wind variables from 20:00 to 08:00 in the China Standard Time (CST; 1200 to 0000 UTC) were then obtained for the analysis of nocturnal FAW migration processes by using the Grid Analysis Display System (GrADS) ver.2.0.1 (COLA,Fairfax,VA,USA).

To illustrate the wind conditions for departure (i.e.,under which FAW emigrated from QM–HRR),we analyzed wind speeds and directions at 925 hPa during the migration periods of two generations of FAW moths at all detected sites in this region.Specifically,mean wind directions at 925 hPa were calculated by adopting the Rayleigh test of uniformity for circular data (Fisher 1993) and associated circular statistics.For the distribution of wind directions,we present three parameters:(i) mean downwind direction(to which the wind is blowing); (ii) mean vector lengthr(a measure of the clustering of the angular distribution of headings ranging from 0 to 1,with higher values indicating tighter clustering around the mean) for each distribution; and(iii) probability of the distribution of wind directions differing from a uniform distribution.Thus,if the Rayleigh testP-value is ＜0.05,it means that the distribution of wind directions is significantly unimodal; hence,there is a significant bias during that migration period for wind to blow toward a particular compass sector.

In this study,the first and second migration periods were defined as the dates on which the first and second generations of FAW adults departed from QM–HRR,respectively.Before further analysis of the migration pattern of FAW moths bred in QM–HRR,non-parametric tests for significant differences between circular data of daily wind directions during the first and second migration periods were performed using the Mardia-Watson-Wheeler test(Batschelet 1981).All circular statistics were conducted using the ‘circular’ package in R (R ver.3.6.1) (R Foundation for Statistical Computing,Vienna,Austria).

2.4.Atmospheric modeling and simulation of migration pathways of newly-emerged FAW adults

The 1°×1° and 6 h FNL data were also input to the WRF model,which is a state-of-the-art atmospheric modeling system that has been successfully implemented to study meso-scale migration of high-flying insects (Wuet al.2018a,b).The detailed WRF model scheme and parameterizations designed for FAW migration were previously described(Wuet al.2019a,b,c).In our experiment,the outputs from the WRF model were meteorological variable values obtained at 1 h intervals and at a grid resolution of 30 km×30 km.Subsequently,the outputs were applied to run a three-dimensional trajectory analysis program written in Fortran to simulate nocturnal migration pathways of FAW adults.These simulation procedures were run by the Linux Operating System.

The flight pathways of FAW adults at high altitudes were calculated using the forward trajectory analysis program combining high-resolution atmospheric backgrounds simulated by the WRF Model and flight behavior parameters related to migratory FAW.The known parameters of flight behavior of FAW,which has been well employed in the trajectory program (Wuet al.2019a,b,c),were also introduced in the present analysis.Specifically,the departure time of FAW migrants was set to 1800 GST the local time (Westbrooket al.2016); moths were simulated to migrate at a model altitude of 500 m above ground level(a.g.l.),which is nearly synonymous with 500 m a.m.s.l.(Mitchellet al.1991,Westbrooket al.2016); 4.5 m s–1of the similar-sized migrant’s self-powered airspeed was added to the wind vector in the trajectory modelling (Wolfet al.1995; Westbrook 2008); FAW migrants exhibit common orientation to the right of the downwind vector,and the crab angle (difference between insect heading and wind transport direction) is 30° (Wolfet al.1995); thus,the flight trajectory was calculated for up to 12 h on each night but terminated when the air temperature at the flight height dropped below 10°C (Westbrooket al.1998,2016; Geet al.2021).Departure altitudes of FAW adults in the eastern plain of QM–HRR were set to 500,750,1 000,1 250,and 1 500 m a.m.s.l.In particular,flight heights of FAW moths over sites with much higher and more complex topography in the western part of QM–HRR were set to 1 500,1 750,2 000,2 250,and 2 500 m a.m.s.l.After 1–3 successive nights of flight,those FAW migrants computed to land into large bodies of water (e.g.,the Bohai Sea)were removed from the overall endpoints of the forward trajectories.Thus,the probability of immigration (%) was obtained as the ratio of the number of valid endpoints within the target province to the total number of valid endpoints.

3.Results

3.1.Infestation in QM–HRR with FAW in 2019

The first invasion event by FAW in QM–HRR occurred in June 2019; in all,98 sites confirmed an initial invasion by FAW between 1 June and 10 July 2019 (Fig.1).Among all these sites,97 cases of infestation occurred in maize plants and only one site,called Tancheng County,in Shandong Province,registered one adult moth trapped in a wheat field with thin rice stubble by using a pheromone trap on 20 June 2019.Three counties in Shaanxi Province,four in Jiangsu Province,and one in Henan Province reported the first appearance of FAW in early June.In all,37 counties in Jiangsu,Henan,Shaanxi,and Shandong provinces reported new invasions by FAW in mid-June.No counties in Shandong Province reported FAW between 21 June and 10 July 2019.However,there were 28 and 25 counties in QM–HRR where the first FAW infestations on corn plants were recorded in late June and early July,respectively.Observations indicated that FAW populations continued spreading further northward.On 5 July 2019,Shanxi,in the northernmost part of QM–HRR,documented the first arrival of FAW.The invasive FAW populations in QM–HRR were found distributed at altitudes ＜2 000 m a.m.s.l.Despite the numerous FAW adult moths trapped,all damage found in corn plants was due to 4th-instar caterpillars,which predominated in the region.

3.2.Estimated peak migration dates of FAW adult moths from QM–HRR

Fig.1 Newly fall armyworm (FAW)-invaded field sites and occurrence in Qinling Mountains–Huaihe River region and locations of other principal regions in China.AH,Anhui; BJ,Beijing; CQ,Chongqing; FJ,Fujian; GD,Guangdong; GS,Gansu; GX,Guangxi;GZ,Guizhou; HA,Henan; HB,Hubei; HE,Hebei; HI,Hainan; HK,Hongkong; HN,Hunan; HL,Heilongjiang; JL,Jilin; JS,Jiangsu;JX,Jiangxi; LN,Liaoning; MO,Macao; NM,Inner Mongolia; NX,Ningxia; SC,Sichuan; SD,Shandong; SH,Shanghai; SX,Shanxi;SN,Shaanxi; TJ,Tianjin; TW,Taiwan; XJ,Xinjiang; XZ,Xizang; YN,Yunnan; ZJ,Zhejiang.

Daily DD values at each site under study in QM–HRR were accumulated to estimate the dates on which FAW larvae developed to eclosion of moths.Cohorts of 4th-instar caterpillars colonizing QM–HRR from 1 June to 10 July 2019 were calculated to have developed into adults (i.e.,first generation) from 20 June to 30 June (A1),29 June to 7 July(B1),9 July to 17 July (C1),and 20 July to 28 July (D1),after accumulating 254.18 DD9.16°C.Moreover,emergence of the second generation of adult moths was calculated to occur from 1 August to 7 August (A2),5 August to 13 August (B2),15 August to 23 August (C2),and 30 August to 9 September(D2),based on a cumulative value of 680.02 DD9.16°C.Examination of the detailed daily surface temperature in QM–HRR shows that the average DMSATs in the region were near or slightly above 25°C (Fig.2).The environment was sufficiently favorable to support fast development of FAW and hence,encourage quick colonization of QM–HRR by new invading FAW adults.

3.3.Nocturnal windborne-transport conditions for FAW migrants en route

Fig.2 Daily mean surface air temperature for the estimated migration dates of the first (marked as 1) and second (2)generations of newly-emerged adults in Qinling Mountains–Huaihe River region,China.The bottom and top of the black box indicate the lower and upper quartile values,respectively.The horizonal solid black line shows the median for each category,while the red dashed line represents the mean.Whiskers indicate the 5th and 95th percentiles; the black circle represents the outlier for the data set; A1 and A2,B1 and B2,C1 and C2,and D1 and D2 are the first and second generations of newly emerged adults for counties in provinces undergoing invasion by FAW in early,mid,and late June,and in early July,respectively.

To identify the migration patterns of the first and second generations of FAW moths,we conducted a detailed examination of the nocturnal wind conditions at 925 hPa above the study sites during two FAW migration periods each day (Fig.3).Wind patterns across eastern China at the synoptic scale were complex and varied.On average,southerly winds blew towards the north over much of eastern China during the peak emergence period of the first generation of FAW bred in QM–HRR,while southwesterly winds prevailed in the Northeast China Plain (Fig.3-A).Nocturnal wind directions at 925 hPa above each study site in QM–HRR were generally suitable for migration of FAW from QM–HRR to the north,and the breeze increased the travel distance over which FAW moths may fly by their own self-powered airspeed (Fig.4-A; Rayleigh test:n=3 528,mean downwind direction (DWD)=(140.98±35.80)°,r=0.458,P＜0.0001,mean wind speed (MMS)=(4.60±0.05) m s–1).The wind conditions during the migration periods for the first generation of FAW from QM–HRR were significantly different from those registered for the second generation(Mardia-Watson-Wheeler test,W=700.95,P＜0.0001).Specifically,analysis of wind fields for the migration of the second generation of FAW from QM–HRR showed that there was a tendency for winds to blow to the west; however,there was a relatively high proportion of southeasterly winds which carried FAW moths to northern regions (Figs.3-B and 4-B,Rayleigh test:n=3 234,DWD=(92.94±33.12)°,r=0.512,P＜0.0001,MMS=(4.94±0.06) m s–1).

3.4.Simulated migration of FAW adults from QM–HRR

The suitability of the wind conditions for windborne transport of FAW moths was confirmed by the simulated forward trajectories over the eclosion periods of the first and second generations of FAW moths from QM–HRR.To better clarify emigratory pattern of FAW originating from QM–HRR,only flight trajectories at a model flight altitude of 500 m a.g.l.,i.e.,500 m a.m.s.l.in eastern QM–HRR,and of 2 000 m a.m.s.l.in western QM–HRR,and falling into provinces where exposed the top three immigration risk,were presented(Fig.5).Specifically,the computed endpoints of emigration trajectories of FAW initiated from the western QM–HRR,where has complex topographic terrain,mainly fell into the north of 35.2°N of latitude (i.e.,the northern boundary of QM–HRR) during the whole simulated study period.While FAW moths from the eastern QM–HRR were simulated to primarily engage in northward transport before mid August,after which their migration endpoints were mostly positioned below 31.3°N of latitude (i.e.,the southern boundary of QM–HRR) and landed into the Yangtze River Valley.The overall trajectories showed a relatively high proportion of endpoints reaching northern China (Fig.6).The importance of QM–HRR as the source area of migrant FAW moths was further suggested during the migration period.The results showed that both the first and second generations of FAW moths from the source QM–HRR breeding area had spread mainly across corn-growing regions in eastern China by 9 September,2019.The corresponding migrations from the first and second generations from QM–HRR indicated the same pattern,which showed that QM–HRR populations regularly migrated northeastward or southwestward covering the main corn-belt in China (Figs.5 and 6).

Fig.3 Average nocturnal wind conditions at 925 hPa over East Asia,responsible for the migratory flights of the first (A) and second(B) generations of the newly-emerged fall armyworm (FAW) moths from Qinling Mountains–Huaihe River region,China.

Fig.4 Daily nocturnal downwind directions at 925 hPa,responsible for the emigratory flights of the first (A) and second (B)generations of the newly-emerged fall armyworm (FAW) moths at sites in Qinling Mountains–Huaihe River region,China.For easy comparison with FAW displacements,the downwind direction is the direction to which the wind is blowing.The area of the dark grey segments is proportional to the number of occasions when downwind directions fell within each 22.5° sector.The bearing of the red arrow indicates the mean downwind direction,while its length is proportional to the clustering of the dataset around the mean direction (the r-value shown on the y-axis).

In all,116 291 and 124 310 valid endpoints of the forward trajectories of FAW moths during the two migration periods,respectively,were calculated and marked in red (Fig.6).Of all the endpoints identified during two migration waves,61.52 and 57.70% were distributed to the north of the departure sites.For those flyers which migrated northward,Shaanxi Province showed a 25.71%probability of receiving the incoming population from 20 June to 28 July,and a 27.06% probability of receiving it from 1 August to 9 September; Henan Province followed with the corresponding probabilities at 23.42 and 27.66%,and Jiangsu and Shandong provinces with the same probabilities at around 12 and 11%,respectively.Additionally,subordinate regions that might be invaded by FAW from QM–HRR were located in Shanxi,Inner Mongolia and Gansu,where endpoints accounted for 4 to 6% of the total probability of receiving the migrant population,respectively.Hebei Province was also under the threat of FAW invasion at rates of 5.91 and 2.05%.Liaoning Province,where corn is cultivated,was affected by the long-distance migration of FAW initiating flight from QM–HRR.Meanwhile,the simulated migration of a FAW population from QM–HRR also showed massive southward displacement (Fig.6),with 38.48 and 42.30% of the total valid endpoints covering much of the Yangtze River Valley during the first and second migration periods,respectively,thus allowing for the migrant population to mix with the local population breeding there.

4.Discussion

We confirmed that a total of 98 sites were initially invaded by the FAW in QM–HRR.Most of the affected locations were located south of the North China Plain.Compared with the rapid spread of invasive FAW in southern China before June 2019,the incidence of this migratory insect pest in QM–HRR thereafter was much lower (Jianget al.2019);this reduction is largely a benefit of the implementation of a FAW regional management and control strategy in early 2019,which greatly depressed the source populations colonizing the Yangtze River Valley and the tropical and southern subtropical zones of China (Jianget al.2019).Another reason for the deceleration of the spreading of the insect may be the large scale and continuously-heavy rainfall processes in southern China,especially in the Yangtze River Valley,in May and June of 2019 (Guan and Zhang 2019; Liu and He 2019),during which,presumably,the FAW source population initially invaded QM–HRR.As documented,heavy rainfall was one of the major causes inhibiting the departure of migrants (Gatehouse 1997) and inducing flight termination of insectsenroute(Dickisonet al.1983,1986;Rileyet al.1983).As recently reported,QM–HRR was the major landing area for FAW populations from the Yangtze River Valley (Wuet al.2019a).Thus,a series of rainy nights might also hamper the flight of FAW moths engaged in northward migration from the Yangtze River Valley into QM–HRR.However,more observations and studies using radars are required to increase our understanding of the correlation between atmospheric factors and migration processes,and the underlying mechanisms.

Fig.5 Typical migration trajectories of the first (left) and second (right) generations of the newly-emerged fall armyworm (FAW)adults from Qinling Mountains–Huaihe River region,China.A1 and A2,B1 and B2,C1 and C2,and D1 and D2 are the first and second generations of newly emerged adults for counties undergoing invasion by FAW in early,mid,and late June,and in early July,respectively.The grey,red and blue lines represent migration at the first,second and third nights,respectively.

Fig.6 Endpoint distribution of the forward migration trajectories for the first (A–C) and second (D–F) generations of the newlyemerged fall armyworm (FAW) adults from Qinling Mountains–Huaihe River region,China.

Direct evidence from the large-scale field surveys of FAW in 2019 throughout China confirmed that FAW continuously explores new habitats.The DMSATs examined in this study were highly suitable for the establishment and reproduction of FAW.In addition,maize plants were recorded as the only host crop for the insect in QM–HRR from 1 June to 10 July,which is consistent with the results of molecular characterization analysis of FAW populations in China (Zhanget al.2019).Based on our examination of the surface air temperature,the emergence of the first generation of FAW adults in QM–HRR occurred from 20 June to 28 July 2019,and the emergence of the second generation of FAW adults was recorded from 1 August to 9 September 2019.

In the present study,a unique modeling approach linking the flight behavior of the target insect,atmospheric cues which lead to flight termination,and an advanced numerical model based on atmospheric data,was used for the accurate simulation of the spatiotemporal migration route and distribution of FAW moths.The newly-emerged FAWs from QM–HRR engaged in long-range migration to exploit new breeding sites hundreds to thousands of kilometers apart (Figs.5 and 6),which benefited from moth’s strong flight capacity (Geet al.2021).It was confirmed that FAW is able to fly over five consecutive nights (10 h per night) and the longest flight duration was tested at 164 km.Moreover,most sampled FAW adults engaged in flight at 1–3 successive nights (Geet al.2021).Hence,we assumed that FAW migrants will undertake three successive nocturnal flights,which ranges from one to three nights of flights,that are sufficient to describe the geographical redistribution and movement pattern of this pest.The present study demonstrated that QM–HRR is a key area for FAW migrants that may have invaded further north,consistent with new arrivals and establishments of FAW immigrants in the North China Plain and the rest of the Loess Plateau in the following months (Jianget al.2019).Our results suggested that Hebei Province (4 198 endpoints of the total northward landing during the first migration wave) would be infested by the first generation of FAW originating from QM–HRR as early as in early July,but before 29 July,which is in good agreement with occurrences of invasive caterpillars of FAW investigated in this province (Jianget al.2019).As reported,the initial appearance of FAW in Hebei Province was monitored to occur on 16 August 2019 in Weixian County,where 5th–6th instar FAW larvae were detected.Our results also identified that there was a high proportion of endpoints located in the North China Plain,one of the key corn producing regions in China.As recorded,there are 246 ha areas affected by FAW in Hebei Province (Jianget al.2019).Meanwhile,the Northeast China Plain was also under a threat (i.e.,of 2.66 and 0.98% FAW-invasion probabilities for the first and second migration periods,respectively) from the northward migration of the FAW population initiating their journey from QM–HRR during late June to early September.Nevertheless,there was no report of colonization by FAW in the Northeast China Plain in 2019,which might be also explained by a very low source population density in QM–HRR and other source regions of China (Wuet al.2019a).According to the regular field investigations throughout the whole China,especially in QM–HRR,average percentages of injured corn plants in Jiangsu,Henan,Shaanxi,Gansu,Shanxi and Shandong provinces ranged between 1.3 and 6.4% (Jianget al.2019).Nonetheless,from the initial sudden appearance of FAW in the North China Plain we can still conclude that both warm temperature and northward atmospheric-transport conditions were matched with the availability of host plants at the flight endpoint from QM–HRR for successful migration of the insect.

Examination of wind fields in the present study provided further evidence in support of a recognizable trend for the northward invasion from QM–HRR.As FAW moths are medium-size,high-flying insects,their long-distance migration patterns will be greatly determined by winds(Chapmanet al.2011).Our trajectories underline the importance of synoptic wind patterns at night during the migration periods of FAW moths.The average wind speed from June through September was approximately 5 m s–1,which is comparatively higher than the flight speed of FAW moths.Moreover,the advance of the summer monsoon over all eastern China provides highly favorable winds(Huet al.2019; Wuet al.2019b,c) for transporting FAWs northwards in late summer.Thus,the mobility of FAW moths,combined with prevailing southerly winds,enabled them to quickly invade much of China in spring and summer(Liet al.2020; Wuet al.2019b),whereas autumn winds showed a very dynamic situation,apparently providing FAW moths with less carrying opportunities for northward flight.Nevertheless,a few studies in the northern hemisphere have shown that mass-migration of medium-size insect migrants flying at hundreds of meters of altitude are likely to move along their seasonal ‘preferred direction’ (Chapmanet al.2015; Reynoldset al.2016).The orientation strategies that these high-flying moths follow were illustrated by radar studies (Chapmanet al.2011) which emphasized that the selection of favorable high-altitude winds promotes insect migrants to achieve long-distance displacement (Chapmanet al.2008,2010).Thus,it would be of the greatest benefit to the invasive FAW to exploit more available resources and colonize new habitats if moths can select favorable -in this case,northward -transport airflow.Our study confirmed that winds facilitated long-range migration of FAW moths from QM–HRR,and that the migration destination,i.e.,maize-growing regions in northern and northeastern China,received frequent favorable winds for invasion by FAW populations from QM–HRR during their peak emergence.However,what and how synoptic weather stimulates in FAW moths to initiate migratory flights remains unclear and warrants further detailed study to improve our knowledge of migratory movement of this insect pest.

Like other noctuid,medium-sized and relatively strongflying moth species,FAW flyers have adaptive migration altitudes at 200–700 m a.g.l.(Wolfet al.1990),especially at the height existing a low-level jet (wind speed ＞12 m s–1),and were generally simulated to migrate at an initial flight altitude of 500 m a.g.l.(Mitchellet al.1991; Westbrooket al.2016).Many nocturnal insect migrants at high altitudes can utilize fast-moving and stable airstreams,in which nighttime layers generally form,to increase the transport distance (Chapmanet al.2010,2011).Besides,the direct effect of flight height of FAW on their migratory movements,especially on the displacement distance in East Asian migration arena,largely depends on seasonal weather conditions and geography (Liet al.2020).Regarding the prior radar observation of insect migration in northern China during the autumn season,including the dominant pests,such asSpodoteraexigua(Hübner) andHelicoverpaarmigera(Hübner),the flight heights of these noctuid moths range between 300–2 000 m(Wuet al.2001).Thus in the present trajectory modelling,we adopted different flight altitude values to ensure that we would capture the most likely height of FAW under complicated meteorological and topographical conditions in the study area.Nonetheless,further quantitative information on the FAW migrant’s height of flight and their nighttime layering in the New World is still essential for ascertaining their likely flight pathways and re-distribution.Future studies on the migration trajectories of FAW moths,and other migratory insect species,would be enhanced greatly by application of multiple monitoring techniques including regionally and nationally systematic trapping,mark and recapture experiment,long-term-operated radars,and genetic analysis.Furthermore,the migration model (WRF model and three-dimensional trajectory analysis program)should be constrained by additional research of FAW flight behavior and its response to atmospheric factors.Such insights should not only lead to the improvement of an accurate numerical model for simulation and prediction of FAW migration,but it also should provide valuable scientific information for forecasting,monitoring,and managing of this and other major agricultural insect pests.

5.Conclusion

The migration periods of the first and second seasonal generations of FAW moths emerging from QM–HRR were estimated approximately from late June to early September.We simulated daily flight routes of the first and second generations of FAW adults from QM–HRR by employing the forward trajectory analysis approach in combination with flight behavior parameters of FAW moths and highresolution atmospheric variables simulated by the WRF model.Our results indicate that southerly winds were dominant during the nights of migration of FAW moths from QM–HRR.Under such synoptic conditions,the North China Plain was identified as the major destination of migrating FAW moths,thereby facing the serious threat posed by this invasive pest.Meanwhile,the Northeast China Plain was also at risk of colonization by FAW.Successively,nocturnal migrations of the first and second generations of FAW moths from QM–HRR spread southward and back into the Yangtze River Valley,which led to a mixture of the local and the migratory populations.Currently,FAW has become a regular and major target of pest control in China.This study will improve our knowledge of the occurrence and distribution of migratory FAW moths from QM–HRR,an essential requirement for monitoring and implementing effective control measures against this exotic and destructive pest.

Acknowledgements

We express our gratitude to the National Agro-Tech Extension and Service Centre,China for conducting field surveys.We are grateful to the staff of the Plant Protection Stations in Jiangsu,Shandong,Shanxi,Shaanxi,Gansu,and Henan provinces of China,for providing the field investigation data applied in this study.We thank the National Oceanic and Atmospheric Administration (NOAA),U.S.Department of Commerce,for providing NCEP/NCAR reanalysis data.This work was supported by the National Natural Science Foundation of China (31901873 and 31727901),the China Postdoctoral Science Foundation (2019M660896),the Central Public-interest Scientific Institution Basal Research Fund,China (CAAS-ZDRW202007),and the National Key R&D Program of China (2019YFD0300105).We are also grateful for the invaluable suggestions and pertinent comments from anonymous referees and journal editors.

Declaration of competing interest

The authors declare that they have no conflict of interest.