Birds are getting smaller

March 15, 2010 • 8:15 am

This may be a portent of global warming: a new paper in the bird journal Oikos, based on 46 (!) years of research, shows that migratory birds in the U.S. are getting smaller over time.

One of the generalizations about biogeography that I teach my students is “Bergmann’s Rule,” the observation that within species of mammals and birds, populations from more northerly locations have larger body sizes than those from further south.  This has classically been attributed to selection to conserve heat: if you’re twice as large (in terms of doubling every linear dimension), you increase your body volume by a factor of 23, or 8.  Thus the amount of heat you generate, which is based on body mass, increases eightfold.  Body surface area, however, is proportional to the square of linear dimension, and doubling that would increase surface area by 4.  Thus, by doubling body size, the ratio of heat lost/heat produced would be halved (4/8). In other words, by getting bigger, you conserve heat more efficiently.

This, at least, is the explanation we give students when describing Bergmann’s rule.

But there’s a problem with this explanation.  Bergmann’s rule holds not only for endotherms (warm-blooded animals like birds and mammals), but also for ectotherms—cold-blooded creatures like insects and amphibians, who don’t generate their own body heat.  In one of my old papers, for instance, I that found Bergmann’s rule was scrupulously obeyed by fruit flies.

Another problem is that we don’t know whether a pattern of more northerly populations being larger reflects true evolutionary change (evolved body-size differences attributed to genetic differences), a developmental response to temperature (you get bigger if you’re born and grow up in a cold climate), or both.  In fruit flies, I found that both factors were at work, but certainly there had been some evolutionary change within the species I studied (Drosophila melanogaster).  But it’s unclear why, for a fly, it’s good to evolve a bigger body in a colder climate. They don’t have to conserve body heat.

In the Oikos paper, van Burkirk et al. combined bird data collected since 1961 at the Powdermill Nature Reserve in Pennsylvania.  Birds were trapped during spring and fall migrations (local residents were trapped, too) and were measured for weight, “fat score,” and “wing chord,” a measure of wing size.  In toto, they measured nearly half a million birds from 102 species.

Upshot: most of the species got smaller over the 46 years of study.

Of 65 species breeding in the study area, 51 got smaller as measured by body mass  Of 83 species caught migrating north during the spring, 60 got smaller.  And of 75 species caught migrating south in fall, 66 got smaller.  All of these trends over time were statistically significant. For the birds that bred locally, and hence for which local temperature could be measured, the decline in mass was significantly correlated with an increase in temperature. This of course reflects the fact that temperature has been going up since 1961.

The decline in body mass wasn’t large: for spring migrants, for example, mass decreased only 1.3% over 46 years.  However, this is a fairly large change over evolutionary time.  The thing is, we don’t know if this change, even if related to temperature, is due to evolution of the birds (genes for smaller body size have replaced those for larger size), developmental change (birds simply grew up smaller as their environment got hotter) or both.  This could be tested by rearing the offspring of birds under constant laboratory conditions, but that would be onerous. The rather small decrease in mass means that very large laboratory samples would be needed to detect such a small change.  And it’s no picnic to rear wild birds of even a single species in captivity, much less the dozens and dozens it would take to see if the change of mass over time is an evolutionary change.

While these differences might reflect declining “conditions” in the birds’ habitats (i.e., less food), the authors address this by looking at population densities at the birds’ breeding grounds.  Presumably bad conditions should be reflected in lower densities. But they found no association between bird density and bird body size.

Regardless of whether the birds’ change in body size reflects genetic change, developmental plasticity, or both, it does indicate that organisms have responded to a long-term increase in temperature. The authors don’t say a lot about global warming, but do raise the issue at the very end of the paper:

Of course, we have long known that evolved changes are an inevitable consequence of almost any human activity that modifies the environment and thereby influences the selective regime experienced by organisms. Classic examples include adaptation to urbanization and contaminated soils (Bradshaw and Jain 1966, Partecke and Gwinner 2007). Similar responses to climate change may be on-going and widespread;whether they will prove to be adequate remains to be seen. Particularly salient and sobering, however, should current trends continue unabated, is the immense biological scope and geographic scale of changes that are taking place compared with the limited information and resources we presently have for measuring, understanding and mitigating those changes.

h/t: Matthew Cobb


van Buskirk, J., R. S. Mulvihill and R. C. Leberman.  2010.  Declining body sizes in North American birds associated with climate change.  Oikos, early view doi: 10.1111/j.1600-0706.2009.18349.x

Coyne, J. A., and E. Beecham. 1987. Heritability of two morphological characters within and among natural populations of Drosophila melanogaster. Genetics 117:727-737 (couldn’t resist).

18 thoughts on “Birds are getting smaller

  1. It is remarkable what the process of science can show, and it is remarkable that scientists can be this dedicated for so long while still not jumping to any conclusions.

  2. It’s not really true that endotherms “don’t generate their own body heat.” They just don’t generate as much, or enough to maintain a constant internal body temperature. Lots of insects use metabolic heat to reach flight temperature faster. Is it possible that even though something like a fruit fly doesn’t have to maintain a constant internal temperature, it is still beneficial to be bigger in colder climates so there is a greater capacity to generate and conserve metabolic heat to warm up faster in the morning, or stay active longer?

  3. Was any attempt made to look at age and population size?

    I was immediately reminded on the shrinking fish stock, which is indeed shrinking the size of individuals due to selective predation (by us). (If I’m not misremembering completely.)

    If the birds measured today are for some reason younger on average than birds of yesteryear, I’d expect that to have a similar effect.

    I am, of course, just talking through my hat here.

  4. I think you overlooked one important aspect of this. The global climate is warming. Aerodynamic bodies (such as airplanes and birds, and as opposed to bouyant bodies like hot air balloons) get less lift from warmer air, and so have to be lighter to compensate.

    You may recall a few years when the temperature in Phoenix, Arizona, USA got up around 125 F and many flights had to be grounded.

  5. While these differences might reflect declining “conditions” in the birds’ habitats (i.e., less food),

    It could be more food too I guess, the first this layman heard of Bergmann’s rule was in connection with some work on bioproductivity.

    IIRC the idea goes something like “colder – less bioproductivity – less plant biomass – higher relative grazing – selective pressures for being less nutritious/more toxic – lower food quality starting with plants but throughout the food chain – selective pressures for having larger reserves to survive patchier resource – selective pressures for being large”.

    Btw, dunno if I read it here, but if not the week before this AGW passed 2 sigma significance in testing. This is considerably up from the IPCC 2007 review, where they discussed 80 % certainty in some cases.

    [It is interesting to predict when the usual physics theory significance (ie 3 sigma) is passed. A rudimentary analysis of the deviance between the two rather similar noise distributions (noise vs AGW signal + noise) is to look at the overlap as the tail of a gaussian.

    Wikipedia has a figure of probit estimates of normal distribution effects, and the same rate from IPCC -07 to today gives 98-99 % certainty. Given that the IPCC -07 review is gathered earliest from IPCC 2001, 3 sigma will happen latest ~ 2020 or so. The current AGW signal is ~ 0.8 K since 1850 IIRC, and increases at ~ 0.15 K/decade, so we will have ~ 1 K AGW then.

    Not bad resolution for observations on a single system.]

  6. Questions from the statistical curmudgeon reporting for duty:

    1. Yom-Tov et al. 2006b show clear examples of non-linear, non-monotonic response to temperature rise in body sizes of English birds. In a new study unreferenced by van Buskirk et al., Salewski et al. 2010 [doi: 10.1007/s00442-009-1446-2] show a tight temporal dependence of body mass and feather length on yearly temperature fluctuations in birds from central Europe. Clearly, the van Buskirk et al. study, by its sheer mass and duration of observation, is time-averaging fluctuations on a smaller spatiotemporal scale. To what extent would a statistical treatment which would take into account the non-linear nature of change reveal more of the dynamics of the the adaptive processes? What exactly would you consider ‘long-term’ variations in such a study?

    2. Teplistky et al. 2009 conclude to the mean body mass decrease vs. negative temperature anomaly being an environmentally induced (metabolic?) response rather than already a genetic adaptive response. Statistically, I find their methodology rather appealing, independently of their conclusions. Your comment?

    3. Teplistky et al. 2009 also have a quite sinister warning in their conclusions:
    “However, in the context of global change, many other factors, such as changes in habit at quality and interspecific interactions, may account for this decline. If so, this also means that temporal Bergmann clines could perhaps be viewed as warning signals, rather than comforting examples of microevolutionary adaptation in response to climate change.” [my italics]
    What is known about the speed of genetic adaptation to climatic change in birds?
    What if the speed of genetic adaptation in an unknown number of species were non-commensurate to the dynamics of human-induced climate change?

  7. That’s the sort of thing that worries me much more than the scaremongering claims of monster storms. These are small changes which are extremely difficult to attribute to any one thing. One small change, then another, and after decades we finally see that things are very different from the past, and not in a good way. Numerous species of plants in temperate zones are also indicating that there is a significant (and not necessarily positive) response to the observed global warming. Well, so long as there remains enough water (but not too much) then as far as food goes the worst may be a change of diet (being forced to eat whatever will still grow in a region). Unfortunately there is still so much unknown – for example, how will the rising temperatures and acidification of surface waters affect the globe’s primary oxygen source?

  8. Isn’t it the case that as volume decreases by the cube, area decreases by the square? So a cube 10 meters on the side has an area of 600 square meters, and a volume of 10,000 cubic meters. But an object 5 meters on a side has an area of 150 square meters and a volume of 125 square meters. So the smaller cube has a higher amount of space compared to volume for both the intake of energy, and the radiation of energy. Which means that small objects tend to heat up and cool down faster than larger ones. It also explains while warm climate animals tend to be spindlier than cold climate ones, with longer limbs and more slender proportions.

    Why do warm clime animals tend to have smaller sizes than cold clime ones? Because small size means they shed heat faster and that increases their chances of survival.

    1. Now putting the above measurements in millimeters we get an area of 600,000,000 square millimeters, and 1,000,000,000,000 cubic millimeters for the larger object; while the smaller comes in at 150,000,000 square millimeters area and 125,000,000,000 cubic millimeters volume, which gives you a better idea of the ratios involved.

      1. “a cube 10 meters on the side has …a volume of 10,000 cubic meters.”
        No, 1000 cubic metres.
        “a volume of 125 square meters.” No, 125 cubic metres.
        “Now putting the above measurements in millimeters we get … a better idea of the ratios involved.” It does? Not me.

        If you want simplicity, how about cubes 1 and 2 metres on a side, 6 and 24 sq. metres in area, 1 and 8 cubic metres in volume? That’s about the volumes of a cow and a small whale.

        1. Birds tend to get most of their heat from their metabolism, which goes up and down by body mass which correlates with volume; not radiation (unlike basking reptiles), which goes by area. But they lose heat by radiation. Therefore larger animals stay warm (and can overheat) more easily than small ones.

          I think it was Konrad Lorenz who wrote vividly about the tiny shrews that have to eat nonstop to stay warm.

  9. ““Bergmann’s Rule,” the observation that within species of mammals and birds, populations from more northerly locations have larger body sizes than those from further south.”

    Oy, some of us mammals (and birds) live in the Southern Hemisphere!

    I doubt that Bergmann was so borealicentric, but would instead have said “populations further from the equator have larger body sizes than those closer to it.”

    In humans it’s true that equatorial Melanesians and Micronesians have slight builds, but Polynesians tend to have large frames: it’s suggested those who could build up fat reserves could survive long sea voyages. This is now proving a handicap in the form of more diabetes and the like.

Leave a Reply