The Nature
Ramblers book group is currently reading The
Most Perfect Thing: Inside (and Outside) a Bird’s Egg, by Tim Birkhead. The
book was published in 2016 and, as usual, science has progressed since its
publication. A new hypothesis has been proposed for one of the subjects
summarized in the book: bird egg size and shape. Birkhead
does a great job explaining
the various ideas for the adaptive nature of bird egg shapes. But he admits
that there is no definitive answer. A year after the publication of the book a
paper1 appeared in Science that offered another hypothesis about egg
shapes. There is a wonderful visual presentation of the study at this link: https://vis.sciencemag.org/eggs/ It’s beautifully illustrated and tells the
paper’s story using animated illustrations and without technical language. If
you are at all interested in the puzzle of bird egg shapes you’ll be enchanted
with this presentation.
Another area of
study, based on the surprising effect the color of plastic leg bands have on
male Zebra Finch sexual attractiveness, has been questioned.
Starting on p.
158 Birkhead describes a study done with Zebra Finches in the 1980s. These birds
have become the fruit flies of ornithology because of their ease of handling
and adaptability to a caged environment. For recognition purposes each captive
bird is identified by a colored plastic band or bands that are place around
their legs. One of the studies reported that the color of the plastic leg bands
affected altered the attractiveness of male birds. Males with red leg bands
became popular mates with the female birds. That discovery made it possible to
manipulate matings in Zebra Finches by simply changing the color of a bird’s leg
bands. Subsequent studies, discussed by Birkhead on pages 158-165, used colored
bands to alter male bird desirability to see if or how females altered the
contents of their eggs.
Recently two groups
of researchers performed a meta-analysis of all the papers that used the leg
band effect to manipulate female Zebra Finch mating preferences2,3. A
meta-analysis, as the name implies, is not a new experimental study. Instead data
from all the preceding research publications is compiled to see if they
collectively support or fail to support a hypothesis, in this case the effect
of leg band color on male attractiveness.
One thing a
meta-analysis tests is publication bias. This is not a bias in the research,
but potential bias in the publication of research results. Research that
discovers a new, novel or surprising result is often featured in leading
journals. The attention given to the new findings stimulates similar research
by other investigators. But this type of confirmatory research is not as “sexy”
as the original discovery and is looked at more critically or is accepted by
second tier journals or is rejected as “just confirmatory” research. It may not
be accepted at all. Also, grants for research that simply replicates an
existing study are hard to get. There is intense competition of many fields for
a limited amount of grant money and work that seems more novel or “sexy” than
just replication of existing research. Researchers become discouraged and don’t
submit their work for publication, moving on to a subject that is more likely
to be published.
The sample size
of the study has an effect. Studies that use a smaller number of males are inherently more variable in outcome than
those that use more males. By averaging the effect size over all the studies
you can approach the real magnitude of male advantage. But that assumes that
all studies are equally likely to be published. If the only small studies that
get published are those that confirm an observation, the size of the effect will
be artificially inflated – because the studies that found a lower value never
got published. That is what the two papers found: the advantage of males with
red leg bands was highest in the smaller studies and much lower or non-existent
in the larger studies.
The significance
of these two studies is that all the research that depended on male birds with
red leg bands being more attractive to females is called into question.
Scientists are
aware of the publication bias problem and are actively discussing it at
present. Some have suggested a “journal of replicated studies” approach, but
this seems unlikely at present. Most granting agencies have tight budgets and
cannot support all the application they currently consider worthy of support.
They will probably be unlikely to fund a “replication study” and most
investigators seeking tenure and promotion will likely consider such research
as not creating new science. We live in interesting times.
1 Stoddard, M.C. et al., 2017, Avian egg shape:
Form, function, and evolution. Science 356, 1249–1254.)
2 Seguin, A., and
Forstmeier, W., 2012, No Band Color Effects on Male Courtship Rate or Body Mass
in the Zebra Finch: Four Experiments and a Meta-Analysis. PLOS ONE 7, e37785.)
3 Wang, D. et al.,2018,
Irreproducible text-book “knowledge”: The effects of color bands on zebra finch
fitness. Evolution 72, 961–976
