Back to the butterfly release debate!

Patrick Foley patfoley at
Thu Oct 12 23:31:08 EDT 2000

Dear Leppers,

For those who haven’t seen past exchanges with Paul Cherubini and Bruce
Walsh, I offer these old remarks of mine. These are not the last words
on the subject, but they do suggest that there may be unresolved dangers
to butterfly releases, even in widespread species. Whatever I write,
Paul will want more convincing proof. But Paul seems unwilling to study
the literature on population viability, population genetics and
epidemiology. Less time poring over, more time on basic
population biology and Paul could answer his own questions. I write this
not to convince those who have no desire to learn, but for those who
would like to investigate for themselves. I received no reply from Bruce
or Paul to my last question below.

Patrick Foley
patfoley at

Here is my short list of problems to solve before accepting
unregulated butterfly release.

1)  Let us first decide how many species we are willing to open to
release. Every
species that is released is a species harder to study scientifically
with respect
to phylogeography (John Avise's term for phylogenetic-geographic studies
the species or in the process of speciation). Pro-releasers often state
that most
releases are of monarchs and painted ladies, so this issue is small.
into those species might not agree. The evolution of painted lady
species around
the world could be a great future revelation that I hate to give up.
monarch researchers are already in print against unregulated release.
But a case
could be made that scientific research should not reserve all butterfly
for themselves. Where are releasers willing to draw the line? How many
Which ones?
    If we can get past this problem, we reach the threat to the species
subspecies themselves. Here are some poorly solved problems that need
solutions before I could be happy with unregulated releases of

2) How important is the natural metapopulation structure of the species
future evolution? Sewall Wright argued, and such researchers as Ernst
Stephen Jay Gould and Alan Templeton, agree that local, semi-isolated to
populations are critical to the production of new adaptive types because
genetic drift is easily overwhelmed by even a little migration for
alleles. However, strongly selected alleles should not mind a little
Moreover Barton, Coyne and Turelli argue that Wright's shifting balance
and related speciational processes are not as important as local
selection in the
production of new types. The point here is that it is scientifically
controversial whether isolation is critical to future evolution.

3) What is the geographic distribution of infectious diseases in
butterflies and
how critical is isolation to avoid endemic (enzootic if you must) and
outbreak dangers? This is such a big problem, I will just let it sit
sunning itself.

4) What roles do various forms of (loosely-defined) selfish DNA play in
species? Included here are any forms of DNA that might spread to the
detriment of
most of the genome, e. g. transposons, sexually selected runaway traits.

    Bruce is probably right to argue that a little unregulated release
is usually
safe for a species, and in fact can add to the useful genes subject to
selection. But the four issues raised above are not just quibbles, they
go to the
heart of our ignorance and our hopes to overcome it before we destroy
most of the
natural world on Earth. Perhaps, in view of our ignorance and past
errors in the
release of diseases and pests, some humility is warranted.

    The measure Bruce Walsh  proposed (Sewall Wright’s Fst , a kind of
correlation of allele frequencies within populations compared to the
total species) is a good broad-stroke measure of genetic
differentiation over space. Low Fst species are plausible candidates for

release. But don't set the bar too low. Local populations of smallish
size or
unknown genetic structure might make relatively little contribution to
species Fst and yet be worth preserving intact for all of worries 1
through 4.
    Would you release mainland North American Red Admirals (_Vanessa
in Hawaii? Presumably not since they are not native there, and the
Hawaiian species (_Vanessa tameamea_ one of only two native
butterflies!) is
fully reproductively isolated from the Red Admiral. Or is it? And even
if it is
now, was it always? This is an extreme case, but consider that there are
species of _Vanessa_ in North America that arose despite the famous
tendencies of the clan and a presumably low Fst. Either this happened
constant migration (supporting some level of uncontrolled release) or
isolation (or at least attenuation of gene flow) was required. We don't
    Even in a poorly differentiated species, the increase of disease
flow can
have disproportionate effects. In a highly differentiated population,
potential for catastrophe seems high. But we don't know!
    We still don't have a generally accepted theory that calculates the
value of
peripherally isolated and semi-isolated populations for the evolutionary

potential, for disease resistance and for refuges against selfish DNA.
Let's not
smear a species before we know it.

    For the list as a whole, consider this hypothetical proposal that
might come
close to a consensus between Bruce and me. No butterfly species should
released in the wild without scientific reason until these criteria are
a) The species must have low population genetic differentiation (the
level of
Fst to be worked out). Such data must be gathered and made publically
By whom? Where?
b) The butterfly can only be released in its studied range of low Fst.
serious solution to the problem of geographically unsuitable releases
must be
developed before suitable releases are accepted because of the problems
of poor
education and malice.
c) Some species of high long term scientific research may still be off
Which are these?
d) The species considered for release need infectious disease
monitoring. The
cost of this should be borne by whom?

Any other requirements? Or a complete rejection of the attempt at

Paul Cherubini wrote:

> Patrick Foley wrote:
> > While there are legitimate arguments suggesting that
> > butterfly releases are _usually_  not harmful to the species
> > of butterflies involved, several of us have argued that in many
> > cases there are important dangers (see the archives).
> Pat, at no time on this list have you ever explained to us
> what you think are the "important dangers" involved with
> the interstate transport and release of monarch butterflies or
> painted ladies - the two species that are used for nearly all
> commercial releases.
> We would love to hear your ideas, from both a theoretical
> standpoint and from actual case histories that you might be aware
> involving these or other migratory insect species with worldwide
> distributions.
> Best regards,
> Paul Cherubini

More information about the Leps-l mailing list