This post is a relatively recent (May 24, 2006) critique of a PLoS paper.
There is a new study on PLoS – Biology that is getting some traction in the media and which caught my attention because it was supposed to be about circadian rhythms. So, I downloaded the paper and read it through to see what it is really about.
Well, it is a decent study, but, unfortunately, it has nothing to do with circadian rhythms. Many examples of tritrophic relationships involve parasitoids (usually small wasps) being attracted by plant volatiles which are released in response to herbivory by insect (usually moth) larvae. So, if a caterpillar munches on a plant, that plant releases chemicals which attract the wasps. When a wasp arrives, she injects her eggs into the caterpillar, often together with a cocktail of toxins or other chemicals that alter the development of the caterpillar, keeping it in the larval stage longer than normal, thus giving wasp eggs sufficient time to hatch and the little wasp larvae to eat their way out (and in the process devouring and killing the caterpillar).
It has also been known for a while now that wasps are strictly diurnal, i.e., they fly only during the day. It has also been known for a couple of years now that plants release the alarm chemicals mostly during the day. Production of this odors takes energy which would be wasted at night when the wasps are not active.
Activity of caterpillars is much harder to assess, and many people in the field swear that there is no diurnal rhythm of their activity, i.e., they are as likely to feed on leaves at night as during the day, though some mild rhtyhms were found in some species.
What this paper addresses is the problem with the previous studies of the caterpillar rhythms. Those were assessed on larvae placed on their host plants. Thus, those were not well-controlled experiments because apart from a light-dark cycle, the larvae were simultaneously exposed to signals generated by the plants.
So, in this paper, the larve were kept in cups and fed synthetic food. They were assayed in light and in darkness in a series of experiments, first in the absence of plants, then in the presence of uninjured plants, and finally in the presence of day-time and night-time volatiles released by either uninjured or insect-injured plants.
Result: the activity of caterpillars was affected by the presence of plants.
Larvae were more likely to hide in the presence of plants than in their absence, even more in the presence of day-time emissions than night-time emission from uninjured plants, and even more in the presence of day-time emissions from the injured plants, suggesting that plant volatiles, especially those produced during the day, and especially those produced by grazed plants, inhibit foraging activity of larvae and promote hiding activity of larvae. The statistics are nice and strong and the conclusion drawn from the data is correct.
If they framed it in this way, the study would be fine. But, for some unkown reason, they decided to frame the study within the context of “sexier” circadian research.
“Although many organisms show daily rhythms in their activity patterns, the mechanistic causes of these patterns are poorly understood.”
is the first sentence of the Abstract of the paper that contains the statement even in the title: Plant Volatiles, Rather than Light, Determine the Nocturnal Behavior of a Caterpillar (PDF).
Their first reference is to the Saunders’ book on Insect Clocks and most of the Introduction and Discussion treats the results of the paper in the circadian context.
“The caterpillars are believed to have evolved a nocturnal lifestyle in order to avoid predatory wasps that maraud throughout the day, but why they don’t use light cues like most other organisms remains a mystery, Takabayashi says” in a press release.
Yet, their experimental methods cannot say anything about response of circadian rhythms to light in these caterpillars. Why?
First, there is nothing said about the pre-treatment. Were the insects kept in light-dark cycle, constant dark or constant light prior to the onset of the experiment? Were they kept on plants or on artifical diet prior to the onset of the experiment? This information is essential to evaluate how entrained their circadian rhythms were prior to the experiment.
Was the onset of the experiment at the lights-off, lights-on or some other phase of the pre-treatment cycle? Just swithing on or off the lights on them at just any time of day or night will not shift their clocks so fast, or even at all, depending on their phase response.
Even if they have a fast-resetting Type O Phase Response Curve (and there is no reason to believe they do – those are rare in the animal kingdom), monitoring the response for just eight hours is not enough – the clocks take much longer to reset.
There is a reason why circadian rhythms are monitored over many days, weeks, months or even years and why the data collected over the first 2-5 days after any kind of treatment (light transition, light pulse, injections of chemicals, etc.) are not used in statistical analysis – the researcher waits that long until the post-treatment rhythm stabilizes.
So, from their data, we cannot say if plant volatiles affect the circadian clock. We also cannot say if the caterpillar clock is or is not responsive to light. The data are consistent with the hypothesis that their clock is light-blind, but is equally consistent with the hypothesis that it is not. The data are consistent with the hypothesis that plant volatiles entrain the clock, but also consistent with the hypothesis that plant odors exert only a masking effect on the overt rhythm of activity – the hands of the clock – without affecting the underlying gears of the clock.
Carl Zimmer wisely avoids any discussion of circadian clocks in his excellent description of what the paper really shows on Loom. The behavior is affected by plant volatiles. Period. Excellent demonstration of the effect. No need to bring in the stuff that was not really addressed by the research, no matter how much the authors wish it may be so.