I have just finished a rough draft of a manuscript describing the population decline of loons in my study area. That effort forced me to count and calculate, estimate and project. I like math, so the work was not unpleasant. I thought that I would pause and share some of the numbers I produced. First, let me share a complex graph!

Screen Shot 2020-01-28 at 1.28.13 PM

This graph shows two sets of values over an 18-year period. The blue line shows the proportion of all adults in the population that are floaters — that is, adults that lack a territory. This group includes both adults too young to have settled and old adults that have been evicted from a territory. The grey bars show the number of chicks banded in the study area three to four years before, adjusted for the number of territories covered those years. In short, the bars show how good the breeding years were in the study area three to four years before. I plotted things this way because three- and four-year-old loons make up the bulk of the new floaters each year, so by comparing the grey bar to the blue line, you can see what impact the breeding success three to four years ago had on the floating population. We expect that lots of chicks produced three to four years before will produce a surge in floater numbers. So the blue line should track the grey bars.

Okay, now let’s see what we can learn from the graph. Notice, first, that the blue line descends overall. This means that the proportion of the adult population made up of floaters declined during the past 18 years. Notice, next, that the grey bars are roughly the same height throughout the graph, which means that chick production was relatively stable over this interval. Already you should be thinking that something odd is going on. The blue line is NOT tracking the grey bars as closely as we expected. The bars and line DO track each other pretty closely at many points, however. For example, the stable chick production from 2002 to 2006 is paralleled by the stable floater population. 2008 went as we expected: good chick production 3-4 years before meant a positive bounce in the floater population. 2013 saw a loss of floaters, as we expected from low chick numbers 3-4 years before.

Now, focus on the last four years. Floater number had fallen to less than 40% of the adult population by 2015. Huge chick production from 2012 and 2013 should have “rescued” floater numbers in 2016 (caused them to spike upwards), but we only see a small bump up in floater numbers that year. And 2017 is worse as, despite big crops of 3- and 4-year-olds, there is a sharp downward turn in floater numbers! Likewise,  2019 should have seen an uptick, but floater numbers actually declined to an all-time low.

I know it is messy to look at this plot. If you find my fine-scaled analysis too picky, forget about the trees I have been discussing and look at the forest. Chick production during the past 18 years has been okay, yet consistently we see fewer floaters than we expect. This is the main puzzling finding of the paper I have written.

It is the mysterious loss of floaters, in fact, that seems to imperil the population of Oneida County loons. We estimate that there are about half as many floaters now as in the late 1990s and that the entire adult population has fallen by 22%. We do not know what is happening to the missing floaters — whether they are dying in their first fall, on migration, or perhaps during winter. But those floaters, which are  future breeders, will have to stage a comeback to get the population back on track.



Having read that northern Wisconsin loons are reproducing poorly and returning to the breeding grounds in very low numbers, many of you are probably wondering, “How widespread is the problem?”. Alas, most efforts to mark and monitor loon populations in other parts of the Upper Midwest have been fragmentary, short-term, and limited in scope. Lacking longitudinal data from other studies of marked individuals — the only kind of data that will permit a reliable assessment —  we cannot say whether other populations in the Upper Midwest have experienced the same downturn as our study population.

Two points are worth making here. First, the loons that we study in Oneida County do not exist in an isolated pocket. Rather, they are part of a continuous swath of loons that stretches from central Wisconsin to the Great Lakes, and northwards across most of Canada. Moreover, loons exhibit the sex-specific natal dispersal pattern characteristic of birds generally: males settle to breed close to where they were hatched and reared; females disperse much greater distances. So the female breeder on your lake is likely to be tens or even hundreds of miles from where she grew up, like the current female from Two Sisters Lake, who was reared on Crab Lake in Vilas County, or the female on Manson, who grew up on Rock Lake, also in Vilas County — or the female that dispersed over 200 miles east and wound up in Antrim County, Michigan. Hence, the loons in northern Wisconsin are part of a vast interdependent network that stretches to adjacent counties, states, and provinces. Females raised in Oneida County breed in Price County, Wisconsin, Michigan, and even Minnesota, while females from those distant places provide breeding females back to Oneida County. The whole system relies upon a dynamic exchange of females across great distances. In short, the downturn in chick production in northern Wisconsin does not spell trouble merely for local loons, it means fewer females are available to breed in outlying counties and adjacent states.

The second point to make is that the reproductive downturn we are seeing is not a short-term pattern that seems likely to reverse course. The inexorable nature of the decline — the fact that the numbers have been slipping downwards steadily for the past two decades — implies that some relentless, slowly-worsening environmental factor has been at work that reduces the abundance of small fishes in northern lakes and will continue to do so in the coming decades.

I am sorry for all of my gloomy forecasts of late. I know: I have only made it worse here by stating that I think loons might be in trouble throughout the entire Upper Midwest. In truth, I am deeply worried. But I am also thinking of strategies that we might use to learn what is hurting the loons and even possibly turn things around. First, of course, we must understand the problem. If it is food, that is not entirely bad news, because humans have been altering fish populations in myriad ways for hundreds of years. By targeted manipulations of small fish populations in certain lakes that we observe closely, we might be able to pinpoint the cause of loons’ reproductive decline, design a strategy for reversing it, and put loons on the comeback trail.




In the wake of my litany of negative findings related to loon breeding success, one burning question presents itself. What is the cause? I am sure that this obvious question occurred to readers of my blog more than once. I feel now as if I have been dodging it. Although we are still gathering information, I will tell you what we currently know. I think I owe you that.

First of all, we must remember that this is a longitudinal decline. By this I mean that the declines are not based data from one, two, or even five years. Rather, these declines have been occurring since the beginning of my study in 1993. Look, for example at the increase in the number of singleton (1-chick) broods over the course of the study.

Screen Shot 2019-12-16 at 10.37.45 AM

It is a noisy pattern, but the pattern marches steadily upwards. There have been good years and bad years for 2-chick broods, but the overall pattern is clear. Progressively fewer pairs have been able to rear 2-chick broods as the years have passed. In the mid-1990s, about half of all pairs had 1- and 2-chick. Now almost all broods are singletons. Even more striking is the decline in returning chicks (below). In the mid-90s, we could expect about half of all chicks that we banded to come back to the study area as adults a few years later. Now, only about 1/6 of all banded chicks are ever seen again as adults.

Screen Shot 2019-12-16 at 10.21.13 AM

The longitudinal nature of the decline makes it difficult to dismiss the results as a blip resulting from a bad winter here or a severe outbreak of black flies there.

Secondly, the patterns connect logically in a worrisome way. There are not only fewer chicks these days but lighter chicks, leading to the obvious conclusion that low chick mass is a reflection of harder conditions for chicks that leads to lower survival. Furthermore, fewer chicks surviving to fledge should result in fewer young returning as adults, and indeed there are now far fewer young adults in the population than was the case two decades ago. Finally, the smaller young adult population should mean less pressure on territorial breeders to defend their territories. Consistent with this expectation, we now see significantly fewer intruders into territories than we did years ago, and, as I pointed out recently, territorial breeders are evicted much less frequently than they used to be. In short, the entire set of measurements paint a stark and coherent picture.

Third, the lower fledging rate of chicks together with their decreased mass over time points rather strongly to food as the likely cause. Starvation of chicks, if it is occurring, would by itself explain all of the “downstream” patterns (i.e. fewer returning adults, lower intrusion and eviction rates in territories) we have detected.

Pulling threads together, we have a consistent picture of long-term decline in loon breeding success across a broad swath of lakes, probably owing to a fall in food levels. This summary leads to the most vexing question: Why should food levels fall? To be more specific, what factor or factors might result in populations of small fishes falling gradually over the course of a 27-year study? The answer, it is clear, has to be some environmental factor that acts like a slow, incessant march, not a lightning strike.

What environmental factor influences fish communities broadly but gradually? Could a recent increase in recreational fishing be the cause, which has led to fewer large fish capable of producing the small ones on which loons depend? Possibly. If overfishing is to blame, it must be true that the problem has gotten especially severe in the last few decades and was not a great problem in the 1980s and 1990s. Yet small fish (“panfish”) populations apparently declined most during the 1960s through 1990s and have somewhat rebounded during the past 20 years. Could the problem be instead that anglers are now more apt to practice “catch and release” of large fishes, whereas they caught and kept their fish in the old days? If so, these released large predatory fishes might be competing with loons for small fishes and driving small fish populations down so much that loon chicks suffer. Again, the “catch and release” explanation would only work if this practice has intensified across all kinds of lakes in Oneida County over the past quarter century. Yet catch and release increased most sharply in the 1980s and had hit a plateau by the time I began my loon study in 1993. So, upon quick inspection, neither obvious explanation for a reduced small fish population over the past quarter century passes muster.

Our search will continue until we learn what factor — probably related to a decrease in small fish populations — explains loons’ recent reproductive slide. I feel confident that we will ultimately learn what is causing our loons to struggle. We must all hope that the cause is reversible.



By now, most of you are aware that the loon population in northern Wisconsin is falling. Since my last report on this topic, we have made two separate formal calculations of λ (“lambda”), which estimates the number of adults in the population in year 2 divided by the number in year 1. Lambda is convenient and intuitive; if λ equals one, there are as many loons in the population this year as there were last year, and we are okay. λ greater than one tells us that the population is growing; λ less than one tells us that it is in decline. Our two separate calculations generated λ values of 0.96 and 0.94, which indicate that the loon population in Oneida County is currently falling at a rate of 4% to 6% per year. The picture is somewhat worse, it seems, than we had thought a few months ago.

This rate of decline — if it is correct, and if it persists — is grave news for humans who love loons. If these numbers are accurate, we will notice the effects of the decline within the next several years. Territory vacancies will go unfilled. Pair members that lose their mates will struggle to re-pair with new ones. Still fewer surviving young will fledge than do now. And our loons will have entered the dizzying downward vortex of a dwindling population.

In the short term, though, one cohort of the loon population benefits from falling floater numbers. The sharp downturn in floater abundance has territorial pairs breathing a sigh of relief. For breeding males and females, you see, fewer intruders — fewer scenes like that depicted in Linda Grenzer’s photo above — means fewer young upstarts seeking to evict them from their territories and a higher rate of territory tenure. How much better off are breeders? As the plot below shows, they are a good deal better off.

Screen Shot 2019-11-01 at 3.47.25 PM

The threat of being evicted from a territory in a given year is now only 1/4 to 1/5 what it was only two decades ago. Small and/or old loons that were lucky to hold a good territory for a year or two in 1998 can kick back and relax nowadays, because the eviction rate is trivial. The decreasing floater population is making the prospect of lifelong breeding on a single territory look like a reasonable expectation for both sexes. People who have become familiar with the breeding pair on their lake might feel better off in the short term. They can be much more confident now that the birds they greet each April are the same two from the previous year.

Though I feel that I know several dozen of my study animals reasonably well and look forward to seeing them each spring, I cannot celebrate the fact that I now stand an even better chance than before of doing so. To me, the dynamism of the system — the likelihood that a breeding female or male might have to accept eviction, lick its wounds, and find a new territory with a new mate nearby — was part of its beauty. Knowing what I do now, each reunion with a familiar breeder for me will be a reminder of the new normal: unnaturally long breeding tenure made possible by the drastic decrease in territorial challengers.

Many of you have e-mailed me to ask, “What became of the duckling reared by loons?” It is a reasonable question. Each passing day during the summer revealed startling new behavioral quirks in the peculiar, touching relationship between these inseparable misfits. Having witnessed well over a thousand loon families — and by this I mean those consisting entirely of loons — I found each of my visits to the Long Lake pair a revelation. Each time I watched male and female loons feed their precious adoptee a fish or warn it about a passing eagle, I involuntarily shook my head. How could two species separated by 70 millions of years of evolution come together into such a tight and successful makeshift family? Every day the family remained together seemed to defy logic.

Yet their familial bond persisted. Following my most recent post on the loon-duckling family, the duckling grew and grew some more. By the end of July (as Linda Grenzer’s photo shows), the duckling was close to adult size, and the only uncertainty was whether or not it would sink its parents by continuing to ride about on their backs. By mid-August, the pair and duckling were spending more time apart. On multiple occasions, we saw the duckling take off and fly around the lake a few times before landing near its anxious guardians. By now fully capable of feeding itself and weeks beyond the normal fledging date for mallards reared by their own species, the duckling seemed to cling to its parents more for their sake than for its own.

By September 4th, the duckling and its loon parents were gone from the lake. We will not ever know where the duckling went or how it lived after leaving Long Lake. Although we could have attempted to mark it in July, as we do loon chicks, I could not bring myself to do so. Even as a scientist fascinated by the behavioral outcome, I was too transfixed by the beauty of the family to capture them and risk disrupting it.


Juvenile loons are in a race against time. While their parents seem to relax following the breeding season — wandering from lake to lake as if on a goodwill tour — juveniles, like the three-month-old in Linda Grenzer’s photo, face a ticking clock. After hatching in June or July, juvies must reach near-adult size by ten weeks of age, practice takeoffs and landings, and become strong enough to make flights of hundreds of miles on their southward migration in early November.

They are racing the ice. Temperatures cool in September, become unpleasantly chilly in October, and truly plummet in November — and lake temperatures follow suit. Ice-up can occur anytime between mid-November and mid-December in northern Wisconsin, and ice-up is the end of the line for juveniles. Opportunistic bald eagles await juveniles that are not prepared to migrate and become trapped in the ice. Apparently sensing the desperate task that will confront their offspring in the fall, parents stuff them with fish for eight long weeks in July and August. Chicks grow explosively during mid-summer. But they face their most challenging task in autumn, when parental support wanes and they must learn to feed themselves, improve their body condition, and prepare for their southward journey.

In general, scientists have paid little attention to the juvenile period in birds. Our neglect is natural enough. The breeding season is chock full of interesting behavioral and ecological events: pairing of mates, defense of breeding territories, selection of nest sites, and relentless territorial intrusions by nonbreeding adults seeking to settle. Perhaps ecologists can be forgiven for focusing their attention on breeding behavior and trusting that juveniles will take care of themselves.

But we wondered. If young adults settle on breeding lakes that closely resemble their natal lakes, might juveniles — which must fight for their lives just to become adults — also exhibit clear preferences for certain kinds of lakes over others? Constrained by flightlessness to forage only within the lake where they hatched, we might expect juveniles to become highly specialized to hunt and consume the species of prey found on the natal lake. So once they become capable of flight, we might expect them to visit and forage on other lakes very similar to their natal one. That is, juveniles reared on a diet of bluegill sunfish and used to hunting that species should spend most of the pre-migratory period visiting lakes full of bluegill that they can catch and consume efficiently. And juveniles accustomed to eating snails and leeches should find lakes full of those invertebrates on which they can feast.

Our interest in lake visitation patterns of juveniles during fall inspired us to plot the local movements of youngsters between lakes in the fall of 2012, 2013, and 2014. Kristin, Gabby, and Nathan used their band-spotting skills to locate juvies in September and October of these years. They found close to 200 cases where a juvenile we had marked had flown to forage on a lake other than its own. Using these data, Brian, who joined us this summer, asked, “Do juveniles forage on lakes at random, or do they prefer to forage on lakes like the one that hatched them?”. As the figure below shows, the mean difference in pH between a juvenile’s natal lake and the lake where we spotted it foraging (red vertical line) was far less than the distribution of differences we would have expected, if juvies had foraged randomly (grey bell-shaped curve).


Although Brian has a few statistical checks to complete, the pattern seems clear. Juveniles exhibit strong preference for lakes that resemble their natal one in two respects: 1) pH and 2) water clarity (data not shown). Brian’s analysis is ongoing, and he is trying to learn how closely these chemical and physical attributes predict the food available to loons in a lake. But we are betting that the stark preference of three-month-old juveniles for lakes that remind them of home occurs for a simple reason. Juvies try to spend their time hunting prey in familiar conditions to build themselves up for their most dangerous first southward journey.

I have put off writing those words for some weeks now. The patterns in my data were clear; every measure of breeding success was pointing downwards. If you have been following the blog, you might recall that number of chicks per pair has fallen sharply since I began studying loons in 1993. Although I had not reported it yet, loss of chicks after hatching has also increased significantly since I began my work. That is, many pairs hatch two young but lose one or both of them nowadays. Furthermore, even chicks that survive to five weeks of age are now in poorer condition (as measured by body mass) than in 1998 or 2006 or 2013. In short, breeding pairs in northern Wisconsin now raise fewer and less robust chicks than they did 25 years ago.

A combination of scientific caution and denial caused me to delay describing the implication of these trends — and others that I have detected more recently. As a scientist, I am used to finding one pattern, drawing a tentative conclusion based on that single finding, and then being proven wrong by the next finding. Having had an initial conclusion reversed many times during my career, I was unwilling to sound the alarm without a stronger, more consistent set of findings. But the data I now have are consistent and come from multiple independent analyses. Furthermore, the data are fundamentally simple. My conclusions are not derived from complex models based on measurements of invisible particles detected by finicky high-tech devices. We are just counting and weighing loons here!

Still, you might wonder why my earlier caution has turned so suddenly to alarm. Let me explain. I followed a simple line of reasoning. If loon pairs are producing fewer and weaker chicks, then fewer chicks must be able to migrate to the wintering ground. And if fewer juveniles make it to Florida, then fewer should survive long enough to return to northern Wisconsin (which happens at 2 to 4 years of age) and look for a breeding territory of their own. So, declining chick production should result in a reduced population of nonbreeders (or floaters) in our study area, which are young adults looking to settle on their first territory with a mate. Since we mark chicks and obsessively reobserve them as young adults, we can test the idea that lower chick production has resulted in fewer floaters. The results are stark. After one adjusts for number of observer-hours spent looking for floaters each year, a dramatic pattern

Screen Shot 2019-09-11 at 9.20.36 AM

emerges. The population of floaters has plummeted. Look at the scale of the graph carefully. This is not a small decline. Between 1998 and 2006 the number of floaters seen per observer hour fell from 0.020 to about 0.010. From 2006 to 2015, that number has fallen still farther — to about 0.006. In other words, we have seen roughly 1/3 as many floaters from the 2015 year class (which are 4 years old now) as we saw from the 1998 year class. In terms of percentages, we reobserved about 45% of all chicks banded in 1998 and 1999 much later as adults; we see only about 14% of banded chicks as adults these days.

So what? Floaters are nonbreeders. They contribute no offspring to the population. Do they really matter? There are not even enough territories to hold them. They are surplus individuals, in a sense. You might view the presence of any floaters at all as a positive sign that the population is bursting at the seams. Yes….but floaters are also the future. That is, floaters are loons not yet old enough to claim territories but waiting to fill in for dead breeders (or evict them forcefully). Without floaters, a breeding population cannot sustain itself, because, inevitably, breeders die and must be replaced.

Even now that I have a set of strongly suggestive patterns, I cannot be absolutely certain that loons are in trouble in Wisconsin. Perhaps the steep decline in floater survival simply means that the weak floaters now die off long before settlement, leaving the strong ones to replace dead breeders. Perhaps something very odd is happening in the north-central part of the state (more intense human recreational activity?) that is not happening elsewhere. So, let’s keep fingers crossed that this is a fluke of some kind. But I am convinced that we are seeing a worrisome pattern that is unlikely to be confined to my study area.

What does the future look like, if the floater population truly is declining markedly, as I suggest? Despite the strength of the pattern, the short-term effects might be subtle, because loons are long-lived. Imagine that families in your town or city suddenly began to have only 1/2 or 1/3 as many children as in years past. When would you notice a marked change in the community? You might — if you were very sharp — notice fewer large families at local playgrounds. Sometime later, you might notice the closing of a school here or there. But decades would pass before you would discern a loss of occupied homes in your neighborhood (presuming that kids in your town grow up to inhabit local homes).

So it will be with loons. On some lakes, the death of a pair member — especially a male, because males are in short supply — will leave a vacancy that is not filled by a floater. Some lakes that always supported a breeding pair will lose that pair and instead only see loons that come and go to forage. Large lakes with six perennial breeding pairs will, over time, see some territories “wink out”, leaving only three or four. In a few decades perhaps, even those territories will vanish. If I am correct, and unless this long-term pattern in the survival of young loons reverses itself, loons will ultimately disappear altogether from northern Wisconsin.

The Audubon report put me on notice. Loons are not immune from climate change. While I have wondered at times whether their aquatic habitat might somehow buffer them from the warming of the Earth and increased moisture in the atmosphere, this was a false musing. Recent changes in temperature and precipitation have myriad and complex effects on lakes and their inhabitants. Loons will have to confront the changing conditions just like all other organisms must.

I wondered whether my long-term data on loons might show climate-induced changes. I am not a climate scientist — nor even a hard-core ecologist who might routinely measure fish populations, water temperatures, or lake chemistry. But we do weigh all loons that we capture and band. Perhaps masses of adult loons or chicks have fluctuated in response to the changing climate.

Screen Shot 2019-08-21 at 7.15.28 AM

My findings are quite striking. Chicks have decreased in mass consistently since my team began capturing and marking loons. This finding alone is worthy of concern, but it is not the only one. Breeding males (see below) too show a decline in mass during our study. Breeding females, on the other hand, show no steady loss in mass.

Screen Shot 2019-08-21 at 8.19.14 AM

What are we to make of these patterns? Are populations of small fish down in the past few decades such that chicks and their male parents struggle to put on or maintain body mass? Or are lakes changing in ways (e.g. clarity) that might make fish more difficult to catch? Whatever the cause of these decreases in mass, why are female loons not affected similarly? These questions must remain unanswered for the time being. In fact, these results are so new that I must run some more double-checks before I fully trust them. Even if they are real effects, as it appears, it is much too early to attribute them to global warming. My worrying self, though, fears that these significant declines in body condition might be the leading edge of a changing climate’s impacts on Wisconsin loons.

On its face, it seems absurd. Why would loons ever communicate with eagles? Apart from raccoons, bald eagles pose a greater threat to loons than any other species. Eagles are opportunistic feeders, always looking for an easy meal. And they are large and well-armed enough to seize almost anything edible they find. Nesting loons provide a tempting stationary target for eagle attack. We have had two eagle kills of nesting loons during our study. The eggs themselves are vulnerable to eagles, because they do not flee and provide a nourishing snack. Loon chicks, whose diving skills are more limited than adults’, also draw unwanted attention from eagles. According to our own observations and those of others on our study lakes, eagles are a major cause of mortality among loon chicks older than two weeks. No doubt this explains why chicks learn to track the movements of nearby eagles obsessively, like the three Bass Lake chicks in Linda Grenzer’s photo.

Why would loons, which seem to spend most of their waking hours scanning the skies for eagles, ever speak to them? Could it ever be profitable to speak to your arch enemy? According to a number of studies by behavioral ecologists, it could, providing the information you pass along to your enemy increases your likelihood of surviving.

Deer and antelope engage in a behavior termed stotting in the presence of predators. Stotting means jumping upwards (often while flashing the tail upwards) in a way that makes the animal more conspicuous to its predator. But data collection and analysis on the occurrence and timing of stotting by gazelles has shown that they chiefly practice this behavior when they spot the predator at a distance and can easily outrun it. This and subsequent research suggests that prey often signal to predators to inform them of the unprofitability of an attack. That is, a prey species is saying to its predator, “I see you and am faster than you; save us both a lot of time and energy by looking elsewhere for food.” In fact, honest signals between prey and predators are not uncommon in nature. Many animals have evolved bright warning coloration to signal to potential predators that they are poisonous or dangerous to the predator in some other way that makes attacking them a bad idea. Colorful prey are, in effect, doing predators a favor by informing them that they should not attack! (Again, though, the prey are acting in their own best interests, not the predator’s.)

Could the mournful sounding wail that we often hear from loons be a signal to eagles that they have been spotted and that an attack would be fruitless?  If so, wails should: 1) occur often when eagles are passing overhead, but only when they are at a safe distance, and 2) be emitted by loons regardless of the presence or absence of other loons (like mates and chicks). The second prediction is crucial; if loons give wails to eagles only in the presence of their mates and chicks, it would seem as though they are simply warning their family about the eagle and not talking to the eagle itself. Our data clearly show that the wail is a long-distance signal given by loons when eagles are overhead. And loons wail to eagles whether they are alone or with mates and offspring. So loons certainly look as though they are speaking to eagles with their wails.

Strange to think that telling your arch-enemy anything could ever be a good idea!