Low-level jet streams associated with spring aphid migration and current season spread of potato viruses in the U.S. northern Great Plains

Green peach aphid, Myzus persicae (Sulzer), is the principal and possibly sole vector of Potato leafroll virus (PLRV) and a key vector of Potato virus Y (PVY) in the U.S. northern Great Plains. Current season spread of PLRV and PVY in seed potato depends primarily on flight activity of winged aphids and the availability of virus inoculum. M. persicae does not overwinter outdoors in the northern Great Plains, populations are reestablished each year by spring immigrants from the south. The objective of this research was to relate spring low-level jet (LLJ) streams to intensity of M. persicae flight activity and current season spread of PLRV and PVY. Synoptic weather maps were used to identify LLJ events that could have brought aphids into the northern Great Plains and to record the timing and duration of each event from 1 May to 30 June. HYSPLIT (Hybrid Single-particle Lagrangian Integrated Trajectory), a trajectory analysis model developed by the National Oceanic and Atmospheric Administration and Australian Bureau of Meteorology, was used to track pathways of air particles to their source. Captures of M. persicae and other potential vector species by a regional aphid trapping network (Aphid Alert) from 1992 to 1994 and 1998 to 2003 were used as measures of aphid flight activity. Results from winter grow-outs of the Minnesota Seed Potato Certification Program were used as measures of available PLRV and PVY inoculum. As a possible surrogate for aphid flight activity we used data on spring wind events to predict current season seed lot rejections due to PLRV and PVY. Statistical models were developed relating frequency and duration of spring wind events to subsequent M. persicae abundance and severity of PLRV and PVY spread in the northern Great Plains. Results showed that the cumulative LLJ duration fit best with cumulative M. persicae capture through the first week of August (R² ranging from 0.597 to 0.883), and the current season spread of PLRV fit best with inoculum and cumulative M. persicae capture through the first week of August (R² = 0.75, P = 0.015). LLJ duration did not reliably predict seasonal abundance of PVY vectors other than M. persicae. The model for prediction of PVY using LLJ duration and PVY inoculum was considerably less reliable (R² = 0.30, P = 0.34) although using both M. persicae capture through the first week of August and PVY inoculum fit better (R² = 0.39, P = 0.231).