Massive seasonal high-altitude migrations of nocturnal insects above the agricultural plains of East China

Significance High-altitude, windborne movements of insects occur on an enormous scale, and have significant impacts on ecosystem function, provision of beneficial services, disease spread, and agricultural productivity. We used a combination of insect monitoring radar, balloon-borne nets, and searchlight traps to characterize the intensity, taxonomic composition, direction, and geographical extent of nocturnal insect “bioflows” occurring at heights to ~1 km above the agricultural lands of East China during spring, summer, and fall. We demonstrate seasonal northward and southward flows and show that the transport of insect biomass is considerably greater above this globally important food-production region than above the United Kingdom (the only other region where it has been quantified to date) and is dominated by species that are agricultural pests.

), in 2015-2017.A total of 176 species, comprising 435,532 individuals, were trapped from 1 April to 31 October, with most being caught between June and September.Lepidoptera, Coleoptera and Orthoptera were the most abundant orders.Large insects (70-500 mg) were represented by 79 species of Lepidoptera (including 53 Noctuidae), and another 17 species from 8 other insect orders.Medium insects (10-70 mg) were represented by 43 species of Lepidoptera (including 17 Crambidae and 10 Noctuidae), 15 species of Coleoptera, and 13 species belonging to five other orders.All these species were large enough to be readily detected by the Insect Monitoring Radar.Also see Tables S2 and S3 for a complete taxonomic breakdown of the various insect orders sampled in the searchlight traps, and a complete list of the Lepidoptera species in the samples.Wind speeds during "mass migrations" (yellow) and "non-mass migrations" (gray) between 100-1800 m above ground.In the violin plots, solid black lines represent the median, dashed black lines represent the inter-quartile range (IQR).The mean wind speeds during mass migrations were not significantly different to wind speeds during non-mass migrations in all three seasons for large insects (mean speed ± SD; Spring: mass migrations, 6.5 ± 2.2 m s −1 vs non-mass, 6.9 ± 1.9 m s −1 , t = 0.31, P = 0.76; Summer: mass migrations, 5.3 ± 2.2 m s −1 vs non-mass, 5.4 ± 2.6 m s −1 , t = 0.58, P = 0.56; Fall: mass migrations, 5.2 ± 1.8 m s −1 vs non-mass, 5.2 ± 2.2 m s −1 , t = -0.04,P = 0.96).They were also no different for medium insects during mass vs nonmass migration in fall (mass migrations, 5.1 ± 2.5 m s −1 vs non-mass, 5.3 ± 1.6 m s −1 , t = -0.84,P = 0.40), but there were small but significant differences in spring (mass migrations, 6.0 ± 1.9 m s −1 vs non-mass, 7.0 ± 2.1 m s −1 , t = -2.12,P = 0.04) and summer (mass migrations, 5.9 ± 1.7 m s −1 vs non-mass, 5.0 ± 1.6 m s −1 , t = 2.61, P = 0.01).Wind data from every night in the study period (2015)(2016)(2017) from 800 to 975 hPa in 25 hPa intervals and averaged over 1-hour intervals.Wind dataset as in Fig. S4, but partitioned according to category of migration event, first for large insects and then for medium insects.S1.Total number of migration occasions, number of "mass migrations", and numbers of individual larger insects detected by the IMR.The proportion of the mass migrations which exhibited a significant degree of common orientation (see Methods) is also presented.Mass migrations are defined as the subset of occasions which cumulatively account for 75% of the total number of individuals in each size category across the year.Note. a From Tables S4 and S5.Note: a Scaled from columns 2-5 for representative 600 km width of the ECP; proportions (%) not shown as same as in columns 2-4.b Ranges from difference of small:larger ratio values in 2009 and 2017 aerial-trapping samples.c Estimated small insect proportion is the same for all rows as it is determined from the same small/larger ratio (Table S6).

Fig. S2 .
Fig. S2.The annual numbers of insects of each order caught in a searchlight trap at Yuanyang County, Henan (see Fig. 1A), in 2015-2017.A total of 176 species, comprising 435,532 individuals, were trapped from 1 April to 31 October, with most being caught between June and September.Lepidoptera, Coleoptera and Orthoptera were the most abundant orders.Large insects (70-500 mg) were represented by 79 species of Lepidoptera (including 53 Noctuidae), and another 17 species from 8 other insect orders.Medium insects (10-70 mg) were represented by 43 species of Lepidoptera (including 17 Crambidae and 10 Noctuidae), 15 species of Coleoptera, and 13 species belonging to five other orders.All these species were large enough to be readily detected by the Insect Monitoring Radar.Also see TablesS2 and S3for a complete taxonomic breakdown of the various insect orders sampled in the searchlight traps, and a complete list of the Lepidoptera species in the samples.

Fig. S3 .
Fig. S3.The proportions of insects from each order in night-time aerial netting samples taken at Jiangpu (32.01°N, 118.62°E),Jiangsu province in 2009 and at Yuanyang County, Henan province in 2017.The netting samplings were dominated by Hemiptera, Diptera and Hymenoptera, which cumulatively accounted for 84.5% and 55.4% of the totals in 2009 and 2017, respectively.See also TableS7for a taxonomic breakdown of abundances in the aerial samples.

Fig. S4 .
Fig. S4.Windspeeds (A) and air temperatures (B) at a range of altitudes above ground level during all migration events.In the violin plots, solid black lines represent the median, dashed black lines represent the inter-quartile range (IQR).Data from ECMWF reanalyzes for every night in the 2015-2017 study period, averaged over 1 h intervals; nominal heights correspond to 975 to 800 hPa pressure levels at 25 hPa intervals.

Fig. S6 .
Fig. S6.Vertical profiles of the ground speed recorded by the IMR during "mass migrations" of large (brown) and medium (blue) insects in each season and in each 200 m-deep height intervals.In the violin plots, wide vertical lines represent the median, thin vertical lines represent the inter-quartile range (IQR).Long upper tails, especially evident in large insects, are thought to represent defective or misinterpreted echo signals, and targets with ground speeds >17 m s −1 have been eliminated from the analyses.

Fig. S7 .
Fig. S7.Log-log linear regression analyses of nightly catches of migrant fauna in the searchlight trap at Yuanyang, and the migration intensity of larger insects detected by the accompanying IMR.(A) The entire three-year study period; (B) each of the three years analyzed independently.

Table S2 .
Total abundance and relative proportion of all insect taxa, and those considered to be high-flying migrants, in the searchlight trap catches at the radar site in Yuanyang County, Henan province of China from 1 April to 31 October in2015-2017.
Note: "Y" indicates this family are regular high-flying groups and "N" represents groups not suspected to be regular high-flying migrants.

Table S3 .
Total abundance and relative proportion of the 122 species of larger Lepidoptera (>10 mg) caught in the searchlight trap at the radar site in Yuanyang County, Henan, from 1 April to 31 October 2015-2017.The 10 most abundant species are in bold blue font; 80 pests of crops and ornamental plants are shaded grey.

Table S4 .
Taxonomic composition of insects sampled in aerial nets about 200 m above the ECP.Samples from nights of 15 August -25 September 2009 at Jiangpu, Jiangsu, China (32.01°N, 118.62°E).

Table S6 .
The ratio of small insects (<10 mg) to larger insects (10-500 mg) in night-time aerial netting samples taken at ~200 m above ground at two locations in the ECP.

Table S7 .
Abundance of key groups of small insects in aerial netting surveys carried out above Jiangpu, Jiangsu province,and Yuanyang, Henan province in ECP.

Table S8 .
Annual cumulative omnidirectional flows of nocturnal insect migrants above the ECP

Table S9 .
Annual cumulative omnidirectional biomass transfer of nocturnal insect migrants above the ECP Note.Notes a-c of TableS8also apply here.

Table S12 .
Comparison of abundance and biomass transfer and of nocturnal insects migrating above the southern UK and the ECP.