Intercontinental Migration Facilitates Continuous Occurrence of the Desert Locust Schistocerca gregaria (Forsk., 1775) in Africa and Asia

Abstract

The desert locust, Schistocerca gregaria (Forsk., 1775), stands as one of the most pervasive pests globally, inflicting extensive damage across Asia and Africa. Facilitated by intercontinental migration, the desert locust engages in population exchange between different source areas, perpetuating its widespread proliferation. Despite the wind being recognized as a key factor during migration events, elucidating its precise influence on intercontinental migration has remained elusive. In this study, we scrutinized monitoring data sourced from the FAO monitoring system, pinpointing 13 desert locust events featuring intercontinental migrations since 1967. From these events, four migration routes were summarized, traversing the Red Sea (RS-WE and RS-EW) and the northern Indian Ocean (IO-WE and IO-EW). Typically, RS-WE and IO-EW migrations occurred between December and March, whereas RS-EW and IO-WE migrations were observed from May to June and April to July, respectively. Our examination of wind field data spanning the past 15 years revealed that wind direction and speed facilitated intercontinental migrations. Furthermore, migration trajectory modeling indicated that desert locusts might exhibit migratory behavior both during the day and at night in the cases of RS-WE and RS-EW, with cross-oceanic migration potentially lasting for a week for IO-WE and IO-EW. In summary, our study identifies four migration routes for the intercontinental migration of the desert locust, providing crucial support for the scientific prediction of its occurrence and contributing to international food security efforts.

1. Introduction

The desert locust, Schistocerca gregaria, stands out as one of the most destructive migratory pests, with related studies focusing on its basic biology, distribution and migration, monitoring, and management [1,2]. Over the span of nearly a century, drawing from extensive research and observations (Figure 1), it has been established that the desert locust predominantly thrives in desert shrub habitats, spanning North Africa, the Horn of Africa, West Asia, and South Asia [3,4]. These areas, usually characterized by annual rainfall levels below 200 mm, collectively span an estimated 16 million km² [5]. Moreover, even during phases of decline, known as recession, the desert locust can still inflict localized damage across approximately 30 countries in Africa and Asia [6].

Migration of the desert locust stands as a pivotal negative factor precipitating recurrent outbreaks across expansive territories. Recent outbreaks negatively affected the livelihoods and sustenance of farmers and herdsmen throughout Africa and Asia [7,8]. Desert locust swarms exhibit the remarkable capacity to traverse hundreds to thousands of kilometers, propelled by sustained winds, which was exemplified by their journey from northwest Africa to the British Isles in 1954 and by desert locust swarms moving across the Atlantic ocean from West Africa to the Caribbean in 1988 [9]. When migrating at considerable altitudes, there is no specific directional preference for migration direction. Upon reaching the swarm’s periphery, individual locusts instinctively veer towards the center to uphold group cohesion. Consequently, the velocity and trajectory of collective movement followed the prevailing wind patterns at their aerial elevation [9,10,11]. Studies have revealed that the desert locust can fly at 1500 m above ground, with long-distance population migrations more prevalent before sexual maturity [12].

Desert locust swarms, like other nomadic arid zone animals, typically migrate from regions where rainfall has waned to areas where it has burgeoned [13,14]. At elevated population densities, gregarious phase locusts assembles into highly mobile swarms, traversing during the day and resting at night [15,16,17]. For solitarious-phase individuals, they usually migrate during the night, when wind conditions often differ from daytime patterns [18]. The desert locust exhibits the capacity to cover hundreds of kilometers along its migration trajectories, harboring a significant potential for initiating new invasions [19]. In favorable environmental conditions, desert locusts can sustain flight for several days or even weeks [20]. When conditions permit, the desert locusts settle on trees or shrubs for the night, resuming their journey at sunrise.

Climate factors play a pivotal role in triggering desert locust occurrence [21]. The seasonal distribution of desert locusts in East Africa follows a regular pattern, largely influenced by the seasonal migration trends and the geographical positioning of breeding sites [22]. This predictability hinges on the physiological response of the desert locust to climatic variables, which exhibit regular fluctuations across different seasons, rather than solely on the intrinsic migration or reproductive cycles of its population [23,24]. The wind has been proven to be a key factor during the migration process, especially for intercontinental migrations [22]. Consequently, understanding the mechanisms by which desert locusts navigate between continents, leveraging seasonal wind patterns, has emerged as a new question.

In this study, our focus centered on the wind-assisted intercontinental migration dynamics between Africa and Asia, specifically examining the desert locusts’ traversal across the Red Sea and the northern Indian Ocean. We directed our attention to three aspects: (1) intercontinental migration routes inferred from historical monitoring data, (2) the impact of the wind field during desert locust migration, and (3) migration timing (day or night) and duration.

2. Materials and Methods

2.1. Desert Locust Migration Routes Inference

The reports of desert locust swarms were retrieved from FAO Locust Hub (https://locust-hub-hqfao.hub.arcgis.com/, access date 3 April 2024) and are shown in Figure 1. The occurrence data of S. gregaria since 1967 were collected from FAO Locust Watch (https://www.fao.org/ag/locusts/en/info/info/index.html, access date 15 April 2024), where 13 desert locust events were reported to have intercontinental migration activities (Table S1). We focused on the 4 intercontinental migration routes between (1) the east and west coastlines of the Red Sea (RS-WE and RS-EW) and (2) the Pakistan/India border and the Horn of Africa (IO-WE and IO-EW). After checking the occurrence of these routes by month, we obtained their occurrence periods (

Table 1. The observed occurrence of 4 intercontinental migration routes in 13 desert locust events since 1967 from the FAO Locust Watch Database.

Time	Level	Red Sea West–East	Red Sea East–West	Indian Ocean West–East	Indian Ocean East–West
		December to March	May to June	April to July	December to March
1967~1969	Plague		1967.12		1967.12–1968.6
1972~1974	Upsurge	1973.12–1974.3			1972.10–1973.3
1986~1989	Plague	1988.11–1989.3
1992~1994	Upsurge	1992.12–1993.3	1993.5–6
1994~1996	Upsurge	1994.12–1995.2
1996~1998	Upsurge	1996.12–1997.3	1997.5–7
2003~2006	Upsurge	2003.12–2004.3
2006~2008	Outbreak	2007.1–2007.2	2007.4–5	2007.4–6
2009	Outbreak			2009.5
2010~2011	Caution	2011.1–2011.3	2011.4–2011.6
2012~2014	Outbreak	①2013.2–2013.3	2013.5–2013.6	2014.5
		②2014.1–2014.3
2015~2017	Outbreak	2017.1–2017.3		2016.5
2018~2021	Plague	2019.1–2017.3	2020.4–2020.5	①2019.5	2020.1
				②2020.5–2020.7

). Specifically, the 4 migration routes were retrieved by reading the monthly FAO Locust Watch reports and finding the arrow labels representing migration activities. Repeated migrations across these waterbodies were observed and lasted for one to three months. These results were summarized and shown by red pie charts in Figure 2.

2.2. Wind Field Analyses

The time resolution of data collection was once every 6 h, and the spatial resolution was 1.0° × 1.0°. Their monthly and daily global-gridded data, including the geopotential height, U-component, and V-component of the wind, come from the National Centers for Environmental Prediction (NCEP) and the National Center for Atmospheric Research (NCAR) global reanalysis data (Final Operational Global Analysis, FNL) from 1967 to 2020. The flying height of the desert locust is about 1500 m, so the wind field height was set to 850 hPa. Wind field figures in the month were produced, and we checked the occurrence of the 4 routes manually with the summarized data shown by blue pie charts in Figure 2 and Table S2. In total, 648 pictures from 1967 to 2020 were drawn and deposited in figshare: 10.6084/m9.figshare.24648444. FNL data from 2008 to 2022 were fed into GrADS 2.1 for the wind field figures for each route (Figure 3A–C). The wind roses represent the direction where winds were coming from, and the rose figures were drawn with R 4.2.1 using the ggplot2 package (Figure 3D–G). Statistical analyses were conducted using the RColorBrewer, CircStats, and Circular packages in R 4.2.1. The map data were from the Global Aviation Data Management (GADM) database (https://gadm.org/download_country_v3.html, access date 13 July 2024).

2.3. Windborne Migration at High-Altitude

We calculated the potential forward migration trajectories for the 4 routes during their occurrence periods to assess the potential destination within the locusts’ flight capacity. For the desert locust, gregarious locusts migrate during the day, while solitarious individuals migrate during the night. To determine the migration period for the desert locust, we calculated migration trajectories for RS-WE and RS-EW during the day and the night, respectively. Specifically, we selected February and May in 2019 as representative months for RS-EW and RS-WE in both the day and night (Figure 4A,B). The program for trajectory modelling was designed in Fortran (64) and run under CentOS 7.4 on the server in Professor Gao Hu’s lab in Nanjing Agricultural University. The Weather Research and Forecasting (WRF) model (version 3.8, https://www.mmm.ucar.edu/wrf-model-general, access date 10 March 2024) was used to produce a high-resolution atmospheric background for the trajectory calculations (https://www.mmm.ucar.edu/weather-research-and-forecasting-model, access date 10 March 2024). Like the wind field data analysis, WRF used FNL data from NCEP as the meteorological data for the model input. The modelled flight durations are in two modes: during the day (9:00 to 19:00) or the night (19:00 to 5:00), and the locusts migrate 7 times. The heights were set to 1500 m, 2000 m, and 3000 m, respectively. Other key WRF parameters included wrf_core = ‘ARW’ and map_proj = ‘lambert’. The start points for RS-WE and RS-EW are (37.16 E, 18.61 N) and (37.71 E, 24.46 N), which are among the key take-off areas.

The desert locust can travel thousands of kilometers by riding suitable winds. However, few scenarios were available to test the long-distance migration of the desert locust. Here, we calculated migration trajectories for IO-WE and IO-EW, which required continuous flight to pass through the Indian Ocean (Figure 4C). Specifically, we selected the most recent observations, January and July in 2020, as representative months for IO-WE and IO-EW, respectively. The flight duration was set as 168 h or 96 h and migrated for a single time. We chose the duration when most trajectories just arrived at the termination as the final settings. We found that 96 hours of flight was a suitable value for IO-WE, with 168 h working well for IO-EW. The start points for IO-WE and IO-EW are (50.18 E, 10.16 N) and (67.62 E, 24.26 N). Other parameters followed the RS-WE/RS-EW simulations.

3. Results

3.1. Historical Monitoring Data Supports 4 Intercontinental Migration Routes of the Desert Locust

Based on desert locust observation data from FAO, intercontinental migrations mainly happened between Africa and Asia (Figure 1), providing supplementary sources for the continuous occurrence of desert locust around these routes. Since 1967, 13 desert locust events with large-scale intercontinental migration have been recorded by the FAO Locust Watch Database (Table S1). These data support four intercontinental migration routes, including flying across the Red Sea and Indian Ocean in a west–east (RS-WE, IO-WE)/east–west (RS-EW, IO-EW) direction. Of all 13 historical events, RS-WE and RS-EW covered 11 and 7 events, respectively, indicating that the Red Sea coastline is a hotspot of desert locust activities (Figure 2, red pie charts). RS-WE occurred from December to March, and the desert locust migrated from Africa to the west and south Asian regions, while RS-EW was mostly observed from May to June. The locusts can fly across the Red Sea (about 300 km in width) within two days. The frequent migration between the Red Sea coasts directly promotes the intercontinental movement. Nonetheless, monsoon air currents in the northern Indian Ocean can carry locust swarms from the Horn of Africa to South Asia in summer (IO-WE) and take the swarms back in winter (IO-EW). IO-WE and IO-EW were, respectively, supported by six and three events historically. IO-WE happened from April to July, while IO-EW occurred, like RS-WE, from December to next March. It is noteworthy that all four routes were present in the latest desert locust plague from 2018 to 2021, with IO-WE present both in 2019 and 2020, and all four routes were used for the flight duration calculations. These routes are relatively stable based on FAO monitoring data, which provide good evidence for the guidance of the desert locust forecast.

3.2. Long-Term Wind Field Data Support the Four Migration Routes

A stable migration route is usually supported by a stable wind field across the years during migration route study [25]. Hence, we analyzed the wind field data from 1967 to 2020 by month and checked if the wind was present during the occurrence date of each route over the 54 years. The wind field appeared in all 54 years for RS-EW, IO-WE, and IO-EW, while the wind direction along RS-WE appeared in 52 years of the 54 years (Table S2 and Figure 2 blue pie charts). Overall, the historical wind field indicated stable wind fields along the routes during all of the months.

All four routes occurred during the last desert locust plague during 2018–2021, indicating that recent climate conditions have facilitated the intercontinental migration of the desert locust. Subsequently, we analyzed the wind field data from the last 15 years (2008–2022) and checked if the wind speed and direction are suitable for migration. We first analyzed the migration across the Red Sea. We observed a predominance of northwestern winds in both the RS-WE route (December-March, Figure 3A,D) and the RS-EW route (May-June, Figure 3C,G), suggesting that winds blowing along the Red Sea from the northwest to the southeast constitute a significant portion of the wind patterns in this region (RS-WE 21%, RS-EW 25%). During RS-WE, the predominant wind direction was towards the east (Rayleigh test, mean value −53°, r = 0.13, p < 0.0001), accounting for 34% of all wind units (including west, southwest, and south winds, with an average speed of 5.8 m/s). Conversely, for RS-EW, the prevailing winds blew towards the west (Rayleigh test, mean value 1°, r = 0.35, p < 0.0001), comprising 47% of all wind units (including north, northeast, and east winds, with an average speed of 6.0 m/s).

During the same timeframe as RS-WE (Figure 3A), IO-EW occurred, characterized by dominant northeastern winds in the northern Indian Ocean. These winds provided favorable conditions for desert locust migration from the Pakistan–India border to the Horn of Africa. The prevailing wind direction, from the land to the ocean, predominated in the wind field (Rayleigh test, mean value −41°, r = 0.26, p < 0.0001), encompassing 66% of all wind units (including western, northwestern, northern, and northeastern winds). This migration direction aligns well with the winter monsoon air currents in the northern Indian Ocean, which blow southwestward (Figure 3E). During the summer, the monsoon air currents reverse direction, blowing northeastward, and IO-WE occurs from April to July, with winds from the southwest constituting 55% of all wind units (Rayleigh test, mean value −142°, r = 0.56, p < 0.0001). High wind speeds were observed around the Horn of Africa, with an average wind speed reaching 14 m/s (Figure 3F), facilitating the take-off activities of the desert locust.

3.3. Modelling Windborne Migration along the 4 Routes

Desert locusts exhibit two distinct phases: gregarious desert locusts form migrating swarms and typically migrate during the day, whereas solitarious individuals tend to migrate during the night. In our study, we modeled the windborne migration of RS-WE and RS-EW during both day (Figure 4A) and night (Figure 4B) periods throughout the latest plague (2018–2021) to investigate these migration phases. A total of 168 and 186 forward trajectories (half during the day and the remainder during the night) were simulated for RS-WE (February 2019) and RS-EW (May 2019). Our results revealed strikingly similar trajectories between the day and night simulations for both RS-WE and RS-EW cases, suggesting that desert locusts may migrate during either the day or night to reach similar destinations. During RS-WE (Figure 4A,B), desert locusts could traverse the Red Sea and reach the Persian Gulf states, including Saudi Arabia, Iran, Pakistan, and India, within 7 days. Some trajectories also veered in the opposite direction, reaching the Central Africa area, corroborating the wind field data (Figure 3D). For RS-EW (Figure 4A,B), desert locusts from the east coast of the Red Sea flew westward, reaching the Central Africa area, while other trajectories extended to northern Saudi Arabia, western Iran, Iraq, and the Caspian Sea region.

There have been reports of desert locusts undertaking prolonged flights lasting an entire week, a capability crucial for transoceanic migration. In this study, we simulated forward trajectories during January 2020, originating from the Horn of Africa and heading towards South Asia (IO-WE), with a flight duration of 96 h (Figure 4C). Ninety-three trajectories were generated, and the endpoints covered regions in middle and southern India. In addition, backward simulations were initiated from the border of Pakistan and India, employing wind field data from July 2020 and simulating flight durations of 168 h (Figure 4C). Out of the 93 simulations conducted, 11 trajectories reached the Horn of Africa, while others migrated eastward, posing a significant threat to regions including Nepal, Bangladesh, Myanmar, and China. Our findings unequivocally demonstrate the potential for the transoceanic migration of desert locusts within a timeframe of 7 days.

4. Discussion

In this study, we have elucidated the pivotal role of wind in facilitating the intercontinental migration of the desert locust, drawing insights from analyses of 13 historical events spanning from 1967. The wind patterns along the four desert locust migration routes serve as crucial routes for overcoming natural barriers between Asia and Africa (Red Sea and the Indian Ocean). These migration routes were summarized from data from the FAO monitoring system, which has proven invaluable in dissecting the migration patterns of the desert locust. Through the comprehensive analysis of wind fields and modeling of forward trajectories, we have determined that both day and night periods are conducive to locust migration across the Red Sea. Additionally, the robust northern Indian Ocean monsoon air currents play a key role in facilitating locust movements between the Horn of Africa and South Asia. These findings underscore the complicated interplay between environmental factors and locust migration dynamics, providing essential insights for understanding and managing the impact of desert locust outbreaks on a global scale.

The migration direction and speed of swarms of S. gregaria are mainly influenced by the wind [26]. The higher the altitude, the stronger the wind speed. The orientation of S. gregaria flying at low altitudes is usually upwind; when the wind speed reaches 4 m/s, the migratory direction of the desert locust will shift to be in alignment with the wind direction [27]. When the wind speed exceeds 4 m/s, the flight speed of the desert locust will increase correspondingly [28]. For example, the S. gregaria swarm migrating near Nakuru, Kenya in 1944 initially flew at a speed of 3 m/s [29]. Without any other weather changes, the wind blowing from the southeast suddenly increased to 6 m/s. The locust swarm was swept back and redirected downwind, its migration speed increasing to 10 m/s [11]. Riding the wind is a smart migration strategy for the desert locust, corn earworm moth, and other species [30]. During the analyses of IO-WE and IO-EW, we found a stronger wind for IO-WE, and the corresponding flight duration was 96 h, much shorter than the 168 h for IO-EW (Figure 4C). We can obtain this result from the flight duration requirement, where 96 h is much lower than 168 h. IO-EW is far less common than IO-WE based on the requirement of a higher continuous flight process. Nonetheless, the historical data gave the same conclusion, i.e., that IO-WE was recorded six times in the past 20 years but only once for IO-EW (Table 1).

Desert locusts exhibit distinct migration patterns: gregarious locusts, which form swarms, tend to migrate during the day, while the majority of solitarious individuals migrate during the night. In our study, we aimed to determine which type of desert locust migrates during these intercontinental migrations and at what times. However, we found no significant difference between RS-WE and RS-EW in day and night trajectories. Consequently, both types of locusts can migrate across the Red Sea, contributing to the enlargement of the population size on the opposite coastline. This makes the Red Sea coastlines one of the most critical hotspot areas for desert locust propagation. It is noteworthy that winds blowing to the east or west persist during both the summer (Figure 3D) and winter (Figure 3G), facilitating population exchange throughout the year.

Long-distance migration plays a crucial role in the distribution of the desert locust and facilitates its occurrence on different continents. A notable example occurred in 1988, when desert locusts migrated from West Africa to the Caribbean, covering 5000 km in about ten days [9]. In our present study, we identified six cases of IO-WE from 2007 to 2020 (Table 1), occurring notably during the three most recent desert locust events spanning from 2012 to 2021. Additionally, the IO-EW route was documented half a century ago, during 1967–1968 and 1972–1973, and was documented again during the most recent event in 2020. The active occurrence of both the IO-WE and IO-EW routes in recent years underscores the necessity for heightened attention towards desert locust monitoring and forecasting efforts. During IO-EW trajectory modelling, we noticed a big part of the trajectories distributed along the Himalayas, and the desert locust was intercepted even in Nyalam County, China, in July 2020 [31]. A warming climate may expand the habitat suitability range of the desert locust, necessitating the monitoring of areas where desert locusts have not previously been distributed [32].

We have observed that seasonal changes in wind speed and direction facilitate the migration and dispersal of the desert locust. However, a substantial locust population size is the underlying driver of large-scale movement [33]. Among various factors, precipitation stands out as a key element facilitating the breeding and expansion of the desert locust [34]. For instance, the desert locust plagues during 1986–1989 [35,36] and 2018–2020 caused immense global damage, coinciding with heavy rains in 1985 and 2018 that facilitated the proliferation of the desert locust (FAO of the United Nations, 2020). Massive populations emerged and spread widely, propelled by the wind. However, it is essential to note that wind speed and direction also influence rainfall patterns, indirectly contributing to the plagues of the desert locust, constituting a complex interactive relationship [37].

Under climate change, unpredictable weather phenomena, such as cyclones, are anticipated to occur more frequently, providing additional opportunities for desert locust propagation. Our model, coupled with climate data (mainly the wind) and monitoring data, will prove highly beneficial for predicting desert locust migration and offering crucial insights to policymakers. Nonetheless, establishing an early warning system to monitor the population dynamics of migratory insects and meteorological factors is imperative [19,38]. Although the FAO suggested that countries affected by the desert locusts should monitor and control desert locust populations, the response speed of local stakeholders remains slow, leading to relatively poor prevention efforts, especially during proliferation phases [37]. Given the relatively low agricultural productivity and social instability in many countries within this region, concerted efforts through various channels are necessary to mitigate the damages caused by desert locusts in the world.

5. Conclusions

In this study, we summarized four intercontinental migration routes and their corresponding occurrence time from the monitoring data. These routes were supported by long-term wind field data and trajectory modelling. Migrations crossing the Red Sea could occur during the day or the night, while the desert locust’s flight across the northern Indian Ocean might last for a whole week. Our work is useful to predict the movement of the desert locust in combination with the established monitoring systems, which provides important support for the management of this notorious insect pest and helps to ensure international food security.