Region:
Categorized: Biology, monitoring, surveillance and scouting
Tagged: #realtime #technology
Research updates on in-field monitoring and predictive models: how far have we got?
Hi All,
I am CABI’s newly appointed Research Champion, specialising in the biology, scouting, monitoring and surveillance of fall armyworm.
Following a pilot project where we tested different technologies for monitoring FAW in realtime, I wanted to ask how the different research groups present here are approaching the challenge of population monitoring/ modelling. This seems especially relevant to our colleagues in Australia given the recent invasion, although I think the reality is the majority of countries are struggling to develop good predictive models given the huge areas over which the insect is present and can disperse.
The most accurate large scale models I have come across are those from Trapview (https://www.trapview.com/v2/en/) who rely on digital pheromone traps to ground truth their biological models. However, I have only seen fully functioning version of the software for Helicoverpa armigera, where they have about 50 traps across a large coastal area of Spain. When we were using them to monitor FAW in Kenya as part of the Smart Armyworm project, we found the digital pheromone traps worked really well on large farms with huge acreages of maize, but were at all well suited to small scale agriculture. On small scale farms “analogue” bucket traps were far better at capturing adult moths at low population densities, but the challenge with bucket traps is (1) identification, as the insects lose all their scales and become very hard to ID (2) it’s almost impossible to get accurate, timely data back from the field.
I’d love to hear of any research updates you might have, as well as honest opinions of any new technologies you have tried, what you believe works and what doesn’t. If you know of any digital bucket traps, I’d be especially keen to hear more.
Best,
Aislinn
A very quick tag for of those who seem to be specifically interested in this area of the science, @MARYLUCY ORONJE @Giorgio @Dan McGrath @Adam Kleczkowski @Marcellin Cuma Cokola @Robert Meagher @Obert Maminimini @Lorenzo Meschiari @Hazel Parry @Jessica Lehmann @Olivia Reynolds @John Wightman @Lorenzo Meschiari @Greg Chandler @Ken Young @Robert Meagher
We are indeed working on field detection, monitoring and modelling of FAW, with colleagues at ICRISAT-IN. It would be great to touch base, so please contact me on a.kleczkowski@strath.ac.uk
Hi Aislinn, Congrats on the position at CABI! We have used bucket traps to monitor large areas of the U.S. for fall armyworm and several points were important. One was having a sponge or toweling, or drilling small holes in the bottom of the trap so as to reduce moths getting wet. Two was having a 2-component lure so that a high percentage of moths were FAW and not another species. Third, using a strong kill agent (dichlorvos if possible) that hung along the side of the trap to kill moths quickly. Finally, each of our cooperators serviced the traps… Read more »
HI Rob, I think this is incredibly useful information, thank you. Also perhaps a more manageable way of monitoring FAW under smallholder conditions if perhaps the operators were paid a small stipend to encourage regular counting and trap updates. Although I assume your traps are run by institutional operators? May I ask how many traps you have across the US, and if you have done any work on validating them at different scales? If there is an online or published map of the monitoring locations that would also be incredibly useful, as well as interesting comparison with the way aphids… Read more »
Hi Aislinn, The traps were run by either county extension agents or state university researchers or extension agents. Trapping sites were either state experiment stations or grower fields. We supplied cooperators with bucket traps, pheromone lures, and vaportape and instructions on how trapping should be carried out. We also used numbers in the models from PestWatch (Penn State) that had some of the same cooperators. At the peak we had 47 locations in the Midwest and eastern U.S. Check out Rod Nagoshi’s pubs for the genetic results and John Westbrook’s pubs for the scale(s) that were modeled. (Westbrook, J. et… Read more »