The other three bushes were left untouched and monitored as controls.
The team followed up by counting ants periodically on specific sections of the coffee bushes during the days that followed.
To decide whether the newly established string highways were indeed granting protection,
three days after building the connections Dr Jimenez-Soto and Dr Morris
attached white cards bearing ten dead adult female borer beetles to the trunks of all the coffee bushes in the research sites.
They monitored these cards for half an hour, noting ant activity on them, and also recording how many beetles were removed during that period.
The string highways proved popular with the ants. Three-quarters of them turned into ant trails,
and at least some were used in this way in every one of the study sites.
Presumably as a consequence, ant activity on bushes connected by strings to cuaniquil trees more than doubled,
while that on unconnected neighbours saw no statistically significant change.
This extra activity resulted in more beetle-scavenging.
Coffee bushes connected by string to a cuaniquil had an average of three of the carded insects removed by the ants during the 30-minute window,
triple the rate for unconnected bushes. Dead beetles on cards are clearly easier prey than live ones hidden in coffee berries.
Nevertheless, this is an encouraging result for plantation owners.
If further experiments back these results up
it may be that the coffee-borer problem can be alleviated by a combination of planting the right sorts of shade trees and the wise deployment of some balls of string.