I have pointed out many times how science proceeds not in simple linear fashion but haphazardly. A finding may appear rock solid but subsequent findings modify our understanding, even forcing us to discard earlier conclusions in some cases. Viewed from space — and over large stretches of time — we might seem to advance steadily in understanding a phenomenon. From ground level, progress in comprehending a topic is herky-jerky; we make clear progress for awhile, then reach a dead end, back up, and find a new path forward.

So it has been with loon conservation. Thirty years ago, we feared that methylmercury was a major health hazard for breeding loons and their offspring. At the time, our fears seemed well grounded. Now, having spent millions of dollars measuring mercury levels in loons and looking for its impacts, we must reassess. We now see that, at worst, mercury has the potential to harm loons in the eastern 1/3 of the breeding range — and even there only in acidic lakes. Consequently, where loon populations appear to be declining, we can cross mercury off the list of threats and turn our attention to other potential causes. By the way, it was no mean feat to cross mercury off the list of dangers to loons. Thoughtful, rigorous work by dozens of biologists across North America has made this conclusion possible.

Our study of a much narrower topic — the loon yodel — also seems to be moving forward by fits and starts. (Linda’s gorgeous photo above shows a male yodelling in typical crouched position.) A really cool paper by Jay Mager — a collaborator with the Loon Project for several years — showed fifteen years ago that the frequency of the yodel was strongly dependent upon body mass. Large males, Jay found, had low-pitched yodels. This was fascinating to us, because it meant that male loons were revealing their size to territorial competitors! Now it is well and good to advertise your size if you are a big loon. In that case, the fact that you are yodeling tells opponents that you are motivated to defend your territory, and the low frequency of your yodel informs them that you can back up that determination with beef! But consider the mixed message that a small male sends by yodeling: “I am small but feisty”. Hmmmm. Not such a deterrent to a territorial challenger, it would seem. We are still coming to grips with this so-called “honest-signalling” of body size and trying to learn how small yodelers might benefit from yodeling.

Among males whose ages are known exactly (blue) and males whose ages are estimated (red), older birds have higher-pitched yodels.

But wait! Lately we have had cause to wonder whether body size really does have a strong impact on yodel pitch. Now that a good many males that we marked as chicks have settled to breed, we know the ages of many of our breeding males precisely. Brian Hoover, a postdoc here at Chapman, took these new age data — data unavailable to Jay Mager — and looked to see whether age is correlated with yodel frequency. As you can see from the figure, age is strongly correlated with yodel frequency: old males have high-pitched yodels. Reassuringly, the better our age information is, the stronger the correlation. That is, the blue points show a tighter age/frequency relationship than the red points.

Where does this leave us? I just explained that small males have high-pitched yodels. Now I am telling you that old males have high-pitched yodels. Can it be both? It is possible that both mass and age influence yodel frequency and could exert additive effects. (Imagine the extreme falsetto of a male that was both old and small!) However, Brian examined mass in his more complete analysis and found no evidence that it affects yodel frequency after all. In other words, we might have found a mass/frequency pattern back in 2006 simply because we did not have age data that would have shown us what was really an age/frequency pattern. Statisticians are quite familiar with this frustrating phenomenon; conclusions can change dramatically when a variable that was left out of a previous analysis (age, in this case) is added.

The finding that it might be age, not mass, that affects yodel pitch would be more palatable if high-pitched yodels by old males made more sense than high-pitched yodels by small males. But, intuitively, the new finding makes less sense! This is the delightful pickle we now find ourselves in. “Delightful” because the age pattern is robust and clear, as the graph shows. A “pickle” because we have no idea, at present, why or how old males have high-pitched yodels.

Well…so it goes in science!

Science is, by nature, cumulative. Theories put forward centuries, decades, or years ago form the foundation of ideas we test today. If those theories fail to explain patterns we see in nature, they are refined or discarded and replaced by new theories that themselves must be tested ceaselessly and revised or rejected.

For our part, we scientists spend years learning our field, which means achieving a deep understanding of the sweeping theories that have withstood the test of time. We also must have an intimate knowledge of recent findings of colleagues in the sub-discipline that forms the context of our own research.

The way that scientists carry out the scientific process should sound robust and logical. It is. This approach has led to steady progress in our understanding of the world and a guarantee that — although we may occasionally take a wrong turn in understanding some process or phenomenon — we shall not stray too far and for too long.

But the innate teamwork that typifies the scientific process has a major drawback. So desperate are scientists to keep up with discoveries and hypotheses of others in our own discipline that we expect to replicate their findings in our own work. Of course, such replication is vital to the scientific process; repeated similar findings confirm for scientists that we are seeing consistent patterns and have a solid understanding of nature. In our tendency to look for and find what others have found, though, we are often blind to what is novel. Indeed, if we discover some oddity, we are more likely than not to try to reconcile it with current theory by treating it as an aberration or an artifact of our procedures, rather than a truly new pattern that we do not yet understand.

So it was with spotlighting by loons. For years, I had observed the visits of territorial loons to their neighbors’ lakes. This behavior was curious, to be sure, but my training convinced me that these visits must have an explanation within the fabric already woven by other scientists. No, we would not expect territorial pairs with chicks ever to leave them at home and visit their neighbors with chicks. It made no sense. But until I took a long, hard, robust look at our data, I simply shrugged and trusted that someday we would be able to make sense of it based on what my scientific colleagues had found in other species.

On the other hand, all scientists are aware of this bit of wisdom familiar to fans of Sherlock Holmes:

When you have eliminated the impossible, whatever remains, however improbable, must be the truth. —Arthur Conan Doyle

Most ecologists encounter this situation seldom. It is, as you must imagine, a most unsettling outcome. However, in trying to test hypotheses to explain intrusions by loons with chicks into neighboring lakes, I encountered this situation exactly. Territorial loons, I thought, might be visiting their neighbor’s territories to look for food. That possibility did not stand up to scrutiny; intruders rarely forage during intrusions. Neighboring pair members, I reasoned, might intrude to learn about nearby territories, so that they would be positioned to “trade up” to a new one, given the opportunity. This possibility works for neighbors that failed to produce chicks, which sometimes trade up to the territory next door, but not for neighbors that hatched chicks themselves, which do not. Neighbors with chicks might intrude because — if they wish to draw attention away from their own chicks — they must go somewhere. This explanation fails because parents would be better off, in that case, visiting one of the many uninhabited lakes throughout the study area, where they could forage without interference from other loons and replenish their energy reserves. The fact that parents target other territories with chicks in precise, laser-like fashion rather than studiously avoiding them indicates that they are visiting specific territories with a specific goal. Thus, the improbable explanation that remains after all impossibilities fall away is spotlighting.

To conclude that loons are reciprocally spotlighting each other’s chicks is unsettling. No one has ever proposed such a convoluted mechanism of territory defense before. Our ability to develop the spotlighting hypothesis depended upon knowing loon behavior intimately. When you consider that: 1) nonbreeding floaters are obsessed with finding chicks to gauge territory quality for eviction attempts, 2) pairs with chicks are desperate to hide them from floaters, and 3) floaters are strongly attracted to other adults already intruding in a territory, it is not a great conceptual leap to suppose that adults eager to hide their own chicks would visit the neighbors to draw floaters to the neighbors’ territory and the neighbors’ chicks.

Since I am on a roll, I will add that the form of eavesdropping that loon pairs appear to do on each other’s yodels to keep track of each other’s breeding success is rather novel. That is, behavioral ecologists have long known the animals listen to each other and intercept each other’s messages in order to boost their own reproductive opportunities. Nightingales, for example, use their neighbors’ songs to determine where intruders are, so that they can defend their territories more effectively. But the kind of eavesdropping that I propose in loons — whereby loons use intercepted yodels to learn about neighbors’ chicks, spotlight neighbors’ chicks, and thus cause nonbreeders to evict neighbors — is far more sinister. The fact that eavesdroppers are causing harm to the loons whose yodels they intercept sets the loon system apart from other forms of eavesdropping that have been described in animals.

2009 video from Spider Lk, Oneida County.

I know. Web sites, books, “loon experts” — and our own hearts — tell us that loons are fundamentally good. Even when a pair with chicks hides them along a shoreline while confronting a raft of intruders, all of the attendees of the social gathering seem so respectful to each other, so congenial. But I am going to ask you to take the same sort of journey that I did as a scientist recently. Despite having been steeped for decades in narrow, Pollyannaish loon lore that holds that all loons are friends and are looking out for each other and the population as a whole, open your mind to the possibility that loons, like humans, do not always see eye to eye. Maybe, like me, you will find that this brings you closer to them.

After I explained that breeding loons must learn about their neighbors’ chicks through yodeling of neighboring males, my mother-in-law shot back:

Wouldn’t that behavior be counterproductive for the yodelling male and hence, be modified in his progeny……….unless it’s paired with positive results. What would that be? What’s the point of the yodel once he has a mate?

There are a couple of points to address here. First, a small one. Joanne implies that male loons — like most songbirds — use the yodel to attract a mate. The concept that bird song is a signal to potential mates is so firmly entrenched in our brains that we almost take it for granted. We know that most males that yodel are already paired with a female (like the Manson Lake male in Linda’s awesome photo, above). So, at best, mate advertisement could only be one of multiple functions of the call. Alas, though, we have no data on this question. I know — the yodel as a signal to potential mates seems a simple idea — but no one has yet played yodels from a territory in April and early May and counted the ratio of males and females that hear the yodels and visit the territory. If yodels function as advertisement for a mate, of course, we would expect a high proportion of all visitors to a territory from which yodels are emanating to be unpaired females. Sounds like a good Masters’ project!

Now on to Joanne’s main point. She is absolutely correct that behavior, like other biological traits of animals, should not occur unless it increases fitness — that is, unless it increases the number of offspring produced by the individual showing the behavior — and, hence, the number of individuals that possess that trait in the next generation. So yodels by males should not have evolved if all they do is bring the neighbors in from next door to spotlight the chicks! However, as Joanne suggests, we might sometimes expect that a behavior could have two (or more) impacts on evolutionary fitness. In other words, behavior might affect fitness negatively in one way, yet provide a benefit in another way that more than offsets the negative impact.

Now let’s apply this concept of “multiple impacts” to the loon case. Based on published work, we know that yodels are useful to loons in that they: 1) prevent landings of intruders that might evict a territory owner, and 2) keep intruders that have entered a territory from approaching the chicks closely. Therefore, I surmise that these two fitness-enhancing consequences of yodels are so beneficial to yodelers that they more than offset the negative impact of increased spotlighting of chicks by neighbors. In the interest of full disclosure, I must admit that it is very difficult to measure the impact of behaviors on fitness precisely, so we must be content, at this point, to know that yodels are beneficial in some ways and costly in others.

In thinking about Joanne’s question some more, let me provide a bit of context for the idea of spotlighting. While it is an exciting advance for loon biology to learn that loons are probably spotlighting each other’s chicks, “eavesdropping” on each other’s yodels is not surprising to behavioral ecologists. Eavesdropping, defined as intercepting of signals directed at a specific receiver by a third party that is not the intended target of the signal, turns out to be widespread in animals and makes a lot of sense. If you are a young loon floater trying to obtain a territory — especially a male floater — you would benefit immensely from listening to the territorial defense signals of established males in your neighborhood, since yodels are known to convey information about the identity, body size, body condition, age, and motivation to attack of the yodeler. If you hear from his yodel that a certain male is large and likely to be aggressive to intruders, you would do well to avoid landing on his territory!

Eavesdropping is one of several phenomena that behavioral ecologists have discovered by taking a second look at animal behaviors that seemed not to square completely with known behavior patterns. Another such phenomenon is “prospecting” by young animals looking to settle in a region. Young adults of many species that are in search of breeding territories explore their habitat extensively in one year and then settle the next year in specific areas where they had detected cues indicating successful reproduction (such as the sight or sound of offspring) the year before. Loons prospect for territories too.

Prospecting?? Eavesdropping?? Wow…..loons sound more like us every day!

If you have been reading my posts, you are aware that we now have good evidence for spotlighting of chicks. That is, our data suggest that: 1) parents of chicks systematically visit neighbors that also have chicks; 2) the added presence of these visitors draws in more young adult “floaters” to those neighboring lakes; 3) some of the floaters induced to visit neighboring lakes spot the neighboring chicks; and 4) these added chick detections by floaters result in increased attempts to evict the neighboring pair during the following year. Thus, adult loons with chicks draw the attention of local floaters to neighboring lakes and chicks and away from their own lake and chicks, decreasing the risk of losing their own territory to eviction.

Like most newly discovered behavioral processes, spotlighting alters the landscape and forces us to inspect some aspects of loon behavioral ecology more closely. Here is one puzzle raised by spotlighting: how do loons with chicks find out about the chicks of neighbors? Put another way, how does a breeding pair that is in the midst of protecting, feeding, and hiding their own young have time to spy on the neighbors so that they know where to spotlight? Remember that floaters, unlike established breeders, have no territories or chicks to defend, so they can spend weeks and weeks doing nothing but obsessively intruding into lakes to spot chicks and thus planning future eviction attempts. Territorial breeders with chicks, in contrast, must defend their territory, incubate eggs, and forage for and guard chicks. At best, they only have an hour here or there when they might leave their chicks behind and search for the chicks of others.

The answer probably has to do with territorial yodels. Yodels — like the one I recorded above on Muskellunge Lake (Lincoln Co.) in 2008 — are quite rare, and they occur mostly in a few narrow contexts. Specifically, yodels are frequent in all territories during the first few weeks after territory resettlement in the spring, are quite infrequent throughout incubation, and then suddenly spike again right at the time of hatching. This very precise, predictable pattern of territorial yodels thus conveys reliable breeding information to all loons (and knowledgeable humans) within acoustic range. Imagine, for a moment, that you are a territorial loon with close territorial neighbors both north and south of you. You hear: 1) an early burst of yodels from the north in late April and early May, 2) few or no yodels from the north for a four-week period, 3) a sudden burst of yodels from the north over a two-week period, and 4) occasional yodels for a few more weeks after that. This yodel pattern almost certainly indicates that a loon pair settled on the territory north of you, incubated their eggs for four weeks, hatched chicks, and reared them successfully for at least several weeks. Hence, this yodel profile from the north territory gives you vital information about the presence or absence of chicks without you ever having to leave the safety of your own territory. In fact, you might even be able to infer whether the north pair has two chicks or only one, because males defending two chicks yodel about three times as often as males defending a singleton chick! The fate of the breeding efforts of the pair to the south of you will also be evident acoustically. If you hear the same pattern of yodels to the south as you heard from the north, you know that the south pair too has a chick or chicks. And if the south pair yodels often only in April and May but seldom during the remainder of the summer, then they have failed to hatch chicks. When we look at the system closely, therefore, we realize that the fact that a breeding pair can collect a wealth of information about the neighbors without ever leaving their territory makes it much easier for them to detect the chicks of neighbors than it is for floaters (which intrude only occasionally) to do so.

In summary, the greatest puzzle regarding spotlighting — “How do breeding pairs know where to do it?” — is easily solved. Furthermore, scientists salivate at a behavioral system of this kind. Why? Because we can do a simple experiment to confirm it. Specifically, we can record yodels from Lake A, which is adjacent to Lake B, play Lake A yodels back to the pair with chicks on Lake B so as to simulate chick production on Lake A, and see if the Lake B pair intrudes into Lake A to spotlight the chicks there. If, as we surmise, pairs with chicks are spotlighting neighbors’ chicks, we should be able to induce a pair with chicks to intrude into a neighboring lake without chicks by playing yodels to them in a seasonal pattern that simulates settlement, incubation, and hatching of chicks by the pair next door!

Science is a cumulative pursuit. That is, the thirst for knowledge is never fully quenched. Rather, we answer one question, only to reveal another puzzle or two more. And thus begins another search for answers. That is certainly how my 24-year-old loon project has gone. But sometimes we reach a point where a vexing question is finally laid to rest, and it feels as though we have made real progress. I am at such a point now.

Let me back up. If you have been following this blog, you know that senescence in old loons is a phenomenon we have recently discovered. Senescence, loss of body condition and decline in survival rate in aging individuals within a species, is all too familiar to me and other humans. During the past twenty years, many studies have reported senescence in birds, mammals, fish, and reptiles. So what? Well, we expect that animals that lose condition as they grow old will change their behavior in response. In other words, scientists have long predicted that senescing individuals should start to behave so as to leave more to their offspring and care less for themselves. To put it another way, old individuals should be willing to take a hit to their survival if it allows them to pour more resources into their young and help their young survive. This makes sense, of course, because old individuals reach a point where they stand little chance of surviving longer, so they would do well to give whatever they can to their offspring, which DO have a bright future. Animals that behave this way should leave more and healthier offspring, and thus this behavior should spread in populations. This very logical idea is termed “terminal investment”. Again we can all probably think of human parallels.

Terminal investment, which I have mentioned before, has become a central theme of the loon project, ever since we published a paper 9 years ago on fatal fighting of males. Terminal investment became interesting to us because it was the most plausible explanation for such lethal contests. Our reasoning was as follows. If males are willing to die to defend their territories, then they must reach an age at which they have little to lose. And if males have little to lose, this must mean that senescence hits males (but not females, which seldom battle to the death) very hard to the point where old males have little future to look forward to. In this case, it might make sense for them to fight like crazy to hold a territory for another year or two, rather than give it up easily and leave themselves nowhere to breed during their last year or two of life. So we have two clear predictions here: 1) males, but not females, must start to die off at a certain age, and 2) males beyond this age must still fight like hell for their territories. It is this clash of body condition and behavior among old males that might cause fatal fighting.

At the time we started to consider the terminal investment hypothesis as a means to explain reckless battling by males, we had almost no solid information on the ages of males in our population. With patience and tireless field work by dozens of us, we have now turned things around. Analysis of loons of varying age has shown us that many males hit the wall at age 15. First, and most important, they start to die at a high rate. You can see from the figure below that males (blue bars) are suffering higher mortality than females (red bars), whether they are on territory (Terr) or without one (floaters: “Float”).


But males also lose mass at age 15, indicating loss in body condition, as shown here:


Finally (and predictably), males get evicted from their territories at a high rate at age 15:


Wow, males are really getting slammed after they pass the age of 15 years!

So all of these data tell us that the first prediction of the terminal investment hypothesis, abrupt senescence at a certain age in males but not females, is clearly met in loons. That age, surprisingly, is only fifteen. Females clearly remain strong, healthy, and vigorous well past age fifteen.

As hard as the first prediction of terminal investment was to test, the second prediction is even harder. You see, fights are common in loons if you take the perspective of a loon’s lifetime, but they are quite uncommon if viewed from the standpoint of human observers in canoes. In other words, most individual loons have engaged in several major battles during their long lives, but territorial battles are not common during day to day observations and often occur so quickly that we are not present to witness them.

Patience pays, however. Since we can draw upon 24 years’ worth of field observations, we now have a trove of observations that we can search for any evidence of aggression and territorial behavior. I made this search, looking for two kinds of evidence: 1) territorial yodels, which serve to communicate a male’s aggressive tendencies and willingness to battle, and 2) out and out aggression, in the form of battling, lunging, chasing and underwater attacks launched by territorial loons on intruders to their territories. I was simply asking “Do old male loons (above age 15) tend to maintain a high level of yodeling and aggressiveness towards intruders?”.  The answer is a resounding “Yes”:


As you can see from the figure above, old males actually increase their tendency to yodel (yodels per intruder), compared to young males. Similarly, old males step up their aggression (see below) and contrast in this way with females, who show no increase:


By the way, all of these patterns I have shown are “statistically significant” via tests that I have performed.

You cannot be as excited as I am about this set of results. No one is. But, as I mentioned, this is one of those rare cases where we have finally managed to answer a burning question to our satisfaction. Even better, the question is one that had been the foundation of my research funding from the National Science Foundation. So I can now report to them that I have found the holy grail! What makes this clear finding even more significant is that terminal investment is quite rare in vertebrates. Of the hundreds of species studied thus far, the only other one to show such a clear pattern of terminal investment is the California Gull. Appropriate, don’t you think?

(Photo by Woody Hagge.)

2015-05-23 05.38.41

Two mornings ago, I saw my very first chick of the year on Woodcock Lake. My visit to Woodcock was memorable — not just for that reason, nor because of the rarity with which the loon pair there hatches a chick. Rather, I chanced to witness with great clarity one strategy that adult loons employ to defend their offspring. It is moments of  this kind, when loons’ behaviors and motivations become visible suddenly and starkly,  that fuel much of my thinking and writing about the species.

If you have read my recent posts, you know that we have learned a good deal about how loons defend their young from opportunistic intruders, which on occasion find and kill chicks less than two weeks old. We know that males possess a acoustic tool that females do not — the yodel — which conveys aggressive motivation and therefore can be used to discourage intruders within earshot of the territory from landing there and imperiling the chicks. What was less clear was whether males yodeled at a high rate as a generalized strategy to inform would-be intruders that visits to the yodeler’s territory would likely provoke an attack or whether, instead, yodels were targeted at specific intruders as they passed over a territory or began to land there.

On Saturday morning I arrived at Woodcock to find the male with a tiny chick riding on his back. The female was on the nest 200 meters away, incubating the second egg, which by now might have produced a second chick. As the sun rose above the horizon that day, intruders criss-crossed the airspace above the territory. Two such “flyovers” were especially enlightening. At 535, two intruders began to descend as they crossed the lake, preparing to land next to the male, near the lake’s center. The male crouched down, dumping the chick into the water, and uncorked a deafening yodel at just this moment. In response, the two flyers checked their glide down towards the lake, flapped rapidly to regain lost altitude, and flew off to try their luck elsewhere. Seventeen minutes later, this pattern was repeated. This time a lone flyer crossed just above my canoe, descended to within three meters of the lake surface — the whistling of wind across its wings easily audible in the morning stillness — before hearing the male’s acoustic objection and beating its wings desperately to abort the landing and ascend.

Though we had statistical evidence to suggest that male loons used yodels to repel specific intruders, I had never observed as clearly the effective targeting of flying intruders by a yodeling male. While songbirds sing incessantly during the spring in nonspecific fashion — that is, they sing repeatedly over hours, days and weeks to communicate their readiness to mate to any female in the vicinity and/or their willingness to defend their territory to any male that happens to be nearby — male loons seem to use yodels with surgical precision. In other words, loon males yodel rarely, and when they do, they aim their yodel at a specific target with a specific goal in mind.

Why are male loons so stingy with their yodels? Two possibilities come to mind. First, yodels appear to have a relatively high metabolic cost compared to songs of other animals. Perhaps, then, mere energy conservation places a limit on the frequency of this call. Second, Jay Mager’s work has shown that males reveal both their physical size and their physical condition when they yodel. By yodeling, therefore, a male might convey information about himself that he would prefer to keep private. Of course, a large male in good physical condition should be more apt to yodel, one might argue, whereas a small, ailing male should keep his bill shut so as to avoid an eviction attempt by a rival passing overhead.

I look forward to testing the prediction that male loons yodel rarely — and vary systematically from one to another in their tendency to emit the vocalization — as a means to avoid sharing information about themselves. That is a clear, well-grounded prediction that might produce an important finding. But my morning on Woodcock reminded me of a great benefit that I earn from spending time in the field observing loons. One can spend countless hours entering data and churning through statistical analyses to reach a rock-solid conclusion about animal behavior (and I do). But the occasional “Eureka” moment spent with animals in the field is invaluable.