Possible evolution of hummingbird beaks since WWII

The report below may represent a case of rapid adaptive evolution of a trait: the beaks of Anna’s hummingbirds (Calypte anna) in California, though there are sufficient confounding factors that, were I teaching evolution, I would still use Peter and Rosemary Grant’s work on the beaks of medium ground finches in the Galápagos as my paradigm. (The Galápagos incident occurred over a single year on one small island and confounding factors are virtually nil).

First the species: a male Anna’s Hummingbird flying:

and a female hovering:

Click below to read the article, and find the pdf here.

The authors posited that the increasing use of hummingbird feeders after WWII would select for changes in the bill length of this species because individuals who could reach and consume more nectar from newfangled feeders (which reward copious nectar swilling) would have a reproductive advantage. Their predictions were met, but there are complications.

Here’s a hummingbird feeder:

That’s a very common design, with the feeder filled with sweet liquid: often sugar water, which is okay but commercial nectar containing other nutritive substances is better. The paper describes the spread of feeders and the morphology of AH beaks over time, using about 400 museum specimens gathered since 1860. Feeders, though, were introduced mostly after WWII (from the paper):

Although it likely existed earlier, we report that the widespread recreational hummingbird feeding can be traced back to an article published in National Geographic in 1928 documenting how to ‘tame’ hummingbirds by making bottles of sweet liquid masquerading as flowers (Bodine 1928); this method is thought to have directly influenced the first patented hummingbird feeder in 1947 (True 1995). As a result of this newly popularized feeder, terms associated with hummingbird feeders in local newspapers increased rapidly from southern to northern California, where feeder density began its increase in the historic range accompanied by an increase of ANHU populations as they moved north.

Based on the spread of hummingbird feeders, the authors posited an evolutionary change in beak shape (remember, this is over 80 years):

We therefore expect feeders to select for increased volume with each lick resulting from increased bill length and thickness. In feeders, unlike flowers, nectar pools are not quickly depleted and therefore the short distance between the bill tip and the nectar surface remains relatively constant, such that minimizing the bill-nectar gap allows higher licking rates and extraction efficiency (Kingsolver and Daniel 1983; Rico-Guevara et al. 2015; Rico-Guevara and Rubega 2011; Kingsolver and Daniel 1983).

“Minimizing the bill-nectar gap” involves evolving longer bills. And getting more capacious bills allows you to take in more nectar in one slurp.

And this is what they found.  First, though, there are quite a few confounding factors that the authors had to consider:

• Eucalyptus trees, an invasive species and also a source of food for Anna’s Hummingbird (called AH in this post), also spread over that period
• Humans also spread, and urbanization spread from southern to northern California, so there is a climatic factor to consider, too. Since bills are a source of heat loss, we expect birds in colder climate sin the north to have shorter bills (and they did indeed find this)
• Feeders could have a secondary effect by promoting fights between males, who try to monopolize the “nectar” source. It could be this fighting that would select for changes in bill shape, since bills are used in fighting. Attendant changes in female shape could simply be a byproduct of selection in males.
• Increased urbanization itself could change beak shape, perhaps because it leads to planting of flowers that select for longer bills

Data analysis was done (this is above my pay grade) using a multivariate analysis, taking into account year, location, temperature, beak measurements, and the abundance of feeders and Eucalyptus trees. The latter two factors were estimated—not very satisfactorily—using newspaper mentions since 1880. The results were these:

• The abundance of eucalyptus trees had a small effect on increasing bill length and thickness, but it was much smaller than. . . .
• The density of feeders, which had a highly significant effect, increasing both bill length and thickness (bill dorsal area) in the predicted way
• However, bill size was smaller in colder climates, representing a presumed tradeoff between acquiring nectar from feeders and conserving heat when it’s cold
• Human population size and year also had strong effects, changing the trait in the expected direction, as one would expect if natural selection were causing evolution of bill size and shape over time
• Feeder density had a stronger effect on population size of AHs in northern rather than southern California. From the paper:

We find that feeders and human population size are both strongly positively associated with ANHU [Anna’s Hummingbird] counts (Figure S9) and each appear to have facilitated population growth differently throughout California (Figure 1B,C). Specifically, feeder availability appears to have facilitated population growth at northern latitudes, whereas human population size appears to have contributed more strongly to population growth in ANHU’s native range in southern California. These findings corroborate work conducted by Greig et al. (2017) suggesting that hummingbirds at northern latitudes are more reliant on feeders in winter than those at southern latitudes, while ANHU population growth is supported by urbanized human environments.

Why urbanized environments select for higher hummingbird populations independently of feeders is a bit counterintuitive, but perhaps it has to do with planted gardens.

The upshot:  So, do we have an example of evolution by natural selection here, one based on the proliferation of feeders causing evolution in beak length and shape? It’s possible, but there are a lot of problems. They include a rather small sample size for a model with many covarying factors, the use of newspapers to estimate feeder and Eucalyptus density, an unexplained change in beak shape with feeder density (a constriction appears in the middle of the beak), and no solid evidence that the change is really genetic rather than a change in beak shape induced environmentally by the use of feeders.  (I’ll add, though, that increasing change in time suggests genetic evolution rather than a one-time environmental modification by using feeders.) But the Grants’ work had pretty strong evidence that the change in beak size in the Medium Ground Finch on Daphne Island was genetically based. (They did a heritability analysis.)

One way to test this hypothesis would be to take an area lacking many feeders, but having Anna’s Hummingbirds, and then saturate it with feeders (best to use commercial nectar). If you monitor the birds over a number of years, one should expect to see, in that one small area, a change in beak shape. But nobody is going to do this experiment, because they’d probably expire before it was done. The Grant’s experiment documented change in beak shape over just a single year, and is, to me, far more convincing.