18 THE SCIENTIST | the-scientist.com
ANDRZEJ KRAUZENOTEBOOKresult in a hostile environment that
encourages the herbivorous insect to
move away from the area.
Finding evidence to distinguish
between these scenarios hasn’t been
easy, especially because plant communi-
cation is a small field. But a long-running
project offers new clues. In 1996, a team
at Cornell University started an elabo-
rate experiment on one goldenrod spe-
cies, S. altissima, regularly spraying
rows and rows of the plant with the
insecticide fenvalerate, while leaving
other rows untreated. After 12 years,
the researchers collected plants from
each of the rows, brought them to the
lab, snipped the stems, and grew clones.
Then, the team set up collections of the
clones in pots at a nearby farm, let gold-
enrod beetle larvae munch on some of
the plants, and measured the emission
of VOCs.
“This research was really mostly
curiosity driven,” says Aino Kalske, apostdoctoral researcher in ecology and
evolution biology at the University of
Turku in Finland and a former gradu-
ate student at Cornell who helped with
the experiment. She and her colleagues
were particularly interested to see if the
goldenrod’s chemical messages would
evolve differently depending on whether
the plants had been treated with insec-
ticide and were protected from insect
attacks or had been left untreated and
experienced higher levels of herbivory.
Differences in signaling between the
treated and untreated plants’ descen-
dants might be a small step toward
determining which hypothesis about
plant communication was correct.
The team found that VOCs emitted
by goldenrod plants whose predeces-
sors had been sprayed with the insecti-
cide only induced genetically identical
plants to mount preemptive chemical
defenses to insect invasion—consis-
tent with the kin selection hypothesis.But VOCs emitted by goldenrod whose
predecessors hadn’t been sprayed with
the insecticide induced the preemptive
defense from all the other goldenrod
plants around them, even plants that
weren’t their kin—consistent with the
mutual benefit hypothesis.
Additionally, the plants exposed to
herbivory converged on a shared VOC sig-
nal over the course of the study—with all
of the goldenrod plants eventually emit-
ting the same chemical signals whether
they were genetically identical to the
emitter plant or not. Plants treated with
insecticide showed no such signal conver-
gence, the researchers reported in Cur-
rent Biology last September (29:3128–33.
e3). This sort of convergence on a single
chemical signal is thought to benefit
plants exposed to herbivory by provid-
ing a stronger deterrent against invad-
ing insects or a stronger attraction for
the herbivores’ natural enemies. Kalske
says the study provides the first concrete