The Virtue of Visuals

Sometimes, as a scientist, I get tunnel vision. I get so locked-in while running statistical tests to verify simple behavioral patterns that I cannot see beyond those patterns.

This past week, I got another illustration of the problem. I was asking a basic question: “Do territorial loons show stronger territory defense when they have chicks than at other times?”. We might expect such a pattern for two reasons: 1) young chicks are sometimes killed by “rogue” intruders, and 2) intruders learn that a territory is of high quality — and possibly worth fighting for later — from seeing a chick or chicks in it; so territorial pairs should hide their chicks from intruders. As stated, this question is binary; I am just asking if the chick-rearing phase is characterized by more intense territorial behavior than other phases of the breeding cycle (like the pre-nesting and and incubation periods). And there is nothing unsound about asking that question. It is just a bit narrow.

I didn’t see the limitations of the question until I plotted the data on territorial defense against stage of the breeding cycle. Here are those data.

At first glance, I suppose, the graph looks a little busy. “Why did he have to plot TWO lines on a single graph?”, you might ask. My goal was to allow the viewer to compare two kinds of territorial behavior towards intruders at once: 1) territorial yodels by males and 2) outright attacks of intruders and other forms of physical aggression — and to look at how those behaviors vary throughout the breeding cycle. The graph allows us to see not merely how territory defense varies when territory owners have chicks or do not have them, but how territory defense changes throughout the breeding season — from 20 days before the eggs hatch to 50 days after.

What do we see? Whereas the statistical analysis I did simply told me that both yodels and aggression are more likely during chick-rearing than at other times, the graph paints a more nuanced picture. First, we see that yodels spike sharply at hatching and are rather infrequent at other times (red line). That is, males yodel with surgical precision during the period when their chicks are less than two weeks old and seldom at any other time. In contrast, aggression (blue line) by male and female parents peaks much later — when chicks are three to six weeks old. In short, territorial pairs seem to employ yodels and aggression for different purposes.

Here is my interpretation. As a grad student of mine showed experimentally, yodels are effective tools for discouraging landing by intruders that have entered the airspace above a territory. By yodelling, a male can cause an intruder bent on making a territorial visit to change its mind and visit elsewhere. Frequent yodels by males with tiny chicks, then, keep intruders away from chicks when they are small and most vulnerable to being killed by territorial intruders. But if intruders are so dangerous to young chicks, shouldn’t territorial aggression also be very frequent at this time? No, it should not. Yodels are so effective at driving intruders away at this time that few intruders approach pairs with young chicks closely enough for aggression to be necessary! Instead, it seems, parents only need to defend their chicks with physical aggression when this critical stage has passed, fathers have stopped yodeling their heads off, and intruders are comfortable enough to land in a defended territory and engage in social behavior with territorial pairs.

As I was strutting about the house and congratulating myself for solving this small puzzle, I presented the idea to my wife. She inspected the graph and asked a very reasonable question: “Why do males stop yodelling?”. After making a mental note never to share my ideas in the future, I puzzled over my wife’s vexing but insightful observation.

Here is my tentative response. We know from Jay Mager’s work that yodels are costly. Why? Because, by yodeling, a male is telling young nonbreeders about its identity, size, and body condition. Such information might allow those young adults to decide whether or not to try to evict the male from his territory immediately, to do so at some point in the future, or never to try. So yodels betray valuable information about the yodeller that might best be hidden. In particular, a small old male that has fallen into poor condition is placing a great big target on its back by yodeling to protect his chicks. Since yodeling males keep intruders at bay — a short-term benefit — but are also giving away valuable information — a long-term cost — males should use this potent vocal weapon only in time of greatest need.

Now……that is just so much arm-waving. I have conveniently tailored my hypothesis about the laser-targeting of yodels to the observations we have made. But at least, by developing a testable hypothesis, I have laid the groundwork for future progress. One day experimental playbacks of yodels at different stages of the breeding phase — not just at the young chick stage, where they normally occur — will determine whether the hypothesis has merit. So it goes with science.