Long-term seasonal forecasting of a major migrant insect pest: the brown planthopper in the Lower Yangtze River Valley.

Rice planthoppers and associated virus diseases have become the most important pests threatening food security in China and other Asian countries, incurring costs of hundreds of millions of US dollars annually in rice losses, and in expensive, environmentally harmful, and often futile control efforts. The most economically damaging species, the brown planthopper, Nilaparvata lugens (Hemiptera: Delphacidae), cannot overwinter in temperate East Asia, and infestations there are initiated by several waves of windborne spring or summer migrants originating from tropical areas in Indochina. The interaction of these waves of migrants and synoptic weather patterns, driven by the semi-permanent western Pacific subtropical high-pressure (WPSH) system, is of critical importance in forecasting the timing and intensity of immigration events and determining the seriousness of subsequent planthopper build-up in the rice crop. We analysed a 26-year data set from a standardised light trap network in Southern China, showing that planthopper aerial transport and concentration processes are associated with the characteristics (strength and position) of the WPSH in the year concerned. Then, using N. lugens abundance in source areas and indices of WPSH intensity or related sea surface temperature anomalies, we developed a model to predict planthopper numbers immigrating into the key rice-growing area of the Lower Yangtze Valley. We also demonstrate that these WPSH-related climatic indices combined with early-season planthopper catches can be used to forecast, several months in advance, the severity of that season's N. lugens infestations (the correlation between model predictions and outcomes was 0.59), thus allowing time for effective control measures to be implemented.

Determining the migration duration of rice leaf folder (Cnaphalocrocis medinalis (Guenée)) moths using a trajectory analytical approach.

Many moths finish their long distance migration after consecutive nights, but little is known about migration duration and distance. This information is key to predicting migration pathways and understanding their evolution. Tethered flight experiments have shown that ovarian development of rice leaf folder (Cnaphalocrocis medinalis [Guenée]) moths was accelerated and synchronized by flight in the first three nights, whereby most females were then matured for mating and reproduction. Thus, it was supposed that this moth might fly three nights to complete its migration. To test this hypothesis, 9 year's field data for C. medinalis was collected from Nanning, Guangxi Autonomous Region in China. Forward trajectories indicated that most moths arrived at suitable breeding areas after three nights' flight. Thus, for C. medinalis this migration duration and distance was a reasonable adaptation to the geographic distribution of suitable habitat. The development of female moth ovaries after three consecutive night flights appears to be a well-balanced survival strategy for this species to strike between migration and reproduction benefits. Hence, an optimum solution of migration-reproduction trade-offs in energy allocation evolved in response to the natural selection on migration route and physiological traits.

Wind-Related Orientation Patterns in Diurnal, Crepuscular and Nocturnal High-Altitude Insect Migrants.

Most insect migrants fly at considerable altitudes (hundreds of meters above the ground) where they utilize fast-flowing winds to achieve rapid and comparatively long-distance transport. The nocturnal aerial migrant fauna has been well studied with entomological radars, and many studies have demonstrated that flight orientations are frequently grouped around a common direction in a range of nocturnal insect migrants. Common orientation typically occurs close to the downwind direction (thus ensuring that a large component of the insects' self-powered speed is directed downstream), and in nocturnal insects at least, the downwind headings are seemingly maintained by direct detection of wind-related turbulent cues. Despite being far more abundant and speciose, the day-flying windborne migrant fauna has been much less studied by radar; thus the frequency of wind-related common orientation patterns and the sensory mechanisms involved in their formation remain to be established. Here, we analyze a large dataset of >600,000 radar-detected "medium-sized" windborne insect migrants (body mass from 10 to 70 mg), flying hundreds of meters above southern UK, during the afternoon, in the period around sunset, and in the middle of the night. We found that wind-related common orientation was almost ubiquitous during the day (present in 97% of all "migration events" analyzed), and was also frequent at sunset (85%) and at night (81%). Headings were systematically offset to the right of the flow at night-time (as predicted from the use of turbulence cues for flow assessment), but there was no directional bias in the offsets during the day or at sunset. Orientation "performance" significantly increased with increasing flight altitude throughout the day and night. We conclude by discussing sensory mechanisms which most likely play a role in the selection and maintenance of wind-related flight headings.

Predicting insect migration density and speed in the daytime convective boundary layer.

Insect migration needs to be quantified if spatial and temporal patterns in populations are to be resolved. Yet so little ecology is understood above the flight boundary layer (i.e. >10 m) where in north-west Europe an estimated 3 billion insects km(-1) month(-1) comprising pests, beneficial insects and other species that contribute to biodiversity use the atmosphere to migrate. Consequently, we elucidate meteorological mechanisms principally related to wind speed and temperature that drive variation in daytime aerial density and insect displacements speeds with increasing altitude (150-1200 m above ground level). We derived average aerial densities and displacement speeds of 1.7 million insects in the daytime convective atmospheric boundary layer using vertical-looking entomological radars. We first studied patterns of insect aerial densities and displacements speeds over a decade and linked these with average temperatures and wind velocities from a numerical weather prediction model. Generalized linear mixed models showed that average insect densities decline with increasing wind speed and increase with increasing temperatures and that the relationship between displacement speed and density was negative. We then sought to derive how general these patterns were over space using a paired site approach in which the relationship between sites was examined using simple linear regression. Both average speeds and densities were predicted remotely from a site over 100 km away, although insect densities were much noisier due to local 'spiking'. By late morning and afternoon when insects are migrating in a well-developed convective atmosphere at high altitude, they become much more difficult to predict remotely than during the early morning and at lower altitudes. Overall, our findings suggest that predicting migrating insects at altitude at distances of ≈ 100 km is promising, but additional radars are needed to parameterise spatial covariance.