Dispatch

R183

photography condensing the entire

Pleistocene into a YouTube clip.
Suppose it takes a long time for a wasp
lineage to reach the Iberian glacial
refuge from the Italian one, but once it
gets there it prospers. In the time-lapse
clip, Iberia will be free of the lineage one
instant, and entirely colonised in the
next. The discrete event model is
a good approximation to the extent that
the details skipped during such
eye-blinks are irrelevant to the
evolutionary outcome.
Rejecting the strawman null would
allow us to state that the European
Pleistocene history of oaks and their
fellow travellers is not one of
spatiotemporal equilibrium. Yawn.
Next generation inference allows us
not only to estimate the extent to which
the fellow travellers have been
associated over the dynamic history of
the Pleistocene, but also, and of key
importance, to quantify our uncertainty
regarding those estimates. This is the
big leap forward, because opinionated
criticism of the conclusions can be
replaced by informed scientific debate
regarding the approximations in the
model and the degree of uncertainty
attached to the estimates. I judge that
the multi-taxon co-estimation model
used by Stone et al. [1] can reasonably
approximate the evolutionary/
ecological history that resulted in the
Oak wasp dataset. A match between
model and data lends credibility to their
results, which indicate associations
between gall wasps and their
parasitoids broke down during the
dynamic range changes of the
Pleistocene. Here is a surprise that
captures the scope of what we learn
from the study: the potentially rich
coevolution between these oak fellow

travellers is, on examination, highly

unlikely to have been elaborated
during the last 2.6 million years
because the close associations
necessary for coevolution to occur
were disrupted by the Pleistocene
glacial cycles.
As more and more evolution
researchers realise the inference
models we use are an integral part of
our work [19], rather than an
inconvenient requirement for
publication, the evolution of next
generation inference will most likely
outstrip the technological
breakthroughs which have driven the
pace of much recent research. Of all
the sciences it appears ours may be the
one that will advance most in the
information age.
References
1. Stone, G.N., Lohse, K., Nicholls, J.A.,
Fuentes-Utrilla, P., Sinclair, F.,
ka, G.r., Melika, G.,
Schonrogge, K., Csoc
Nieves-Aldrey, J.-L., Pujade-Villar, J., et al.
(2012). Reconstructing community assembly
in time and space reveals enemy escape in
a western palaearctic insect community. Curr.
Biol. 22, 532537.
2. Feynman, R.P. (1999). The Value of Science,
address to the National Academy of
Sciences (Autumn 1955). In The Pleasure
of Finding Things Out: The Best Short Works
of Richard, P. Feynman and J. Robbins, eds.
(Cambridge, Massachusetts: Perseus
Publishing).
3. Fisher, R.A. (1925). Statistical Methods for
Research Workers (Edinburgh: Oliver & Boyd).
4. Kuhn, T.S. (1962). The Structure of Scientific
Revolutions (Chicago: University of Chicago
Press).
5. Knowles, L.L. (2008). Why does a method
that fails continue to be used? Evolution 62,
27132717.
6. Edwards, A.W.F. (1972). Likelihood
(Cambridge: Cambridge University Press).
7. Dawson, K.J., and Belkhir, K. (2001). A
Bayesian approach to the identification of
panmictic populations and the assignment of
individuals. Genet. Res. 78, 5977.
8. Guillot, G., Mortier, F., and Estoup, A. (2005).
Geneland: a computer package for landscape
genetics. Mol. Ecol. Notes 5, 712716.

Pheromones: Fish Fear Factor

Fish, like many other animals, panic when another individual is injured. Now, the
chemical nature of a substance that mediates this reaction has been
uncovered.
Marcus C. Stensmyr1
and Florian Maderspacher2
Karl Ritter von Frisch the
Austro-German biologist best known
for his work on the senses and
communication of honeybees was

passionate about biology. So

passionate that he once allegedly
knocked a blueberry cake out of his
wifes hands because she had dared to
interrupt a scientific discussion at the
coffee table; so passionate that he even
used his summer holidays spent at the

9. Pritchard, J.K., Stephens, M., and Donnelly, P.

(2000). Inference of population structure using
multilocus genotype data. Genetics 155,
945959.
10. Hey, J., and Nielsen, R. (2007). Integration
within the Felsenstein equation for improved
Markov chain Monte Carlo methods in
population genetics. Proc. Natl. Acad. Sci. USA
104, 27852790.
11. Wilson, I.J., Weale, M.E., and Balding, D.J.
(2003). Inferences from DNA data: population
histories, evolutionary processes, and forensic
match probabilities. J. Roy. Stat. Soc. A 166.
12. De Beaulieu, J.L., Miras, Y., Andrieu-Ponel, V.,
and Guiter, F. (2005). Vegetation dynamics in
north-western Mediterranean regions:
Instability of the Mediterranean bioclimate.
Plant Biosystems 139, 114126.
13. Cook, J.E. (1996). Implications of modern
successional theory for habitat typing: A
review. For. Sci. 42, 6775.
14. Hickerson, M.J., Stahl, E., and Takebayashi, N.
(2007). ms Bayes: pipeline for comparative
phylogeographic inference using approximate
Bayesian computation. BMC Bioinformatics 8,
268.
15. Lohse, K., Harrison, R.J., and Barton, N.H.
(2011). A general method for calculating
likelihoods under the coalescent process.
Genetics 189, 977U398.
16. Hewitt, G.M. (2004). Genetic consequences of
climatic oscillations in the Quaternary. Philos.
Trans. R. Soc. Lond. B Biol. Sci. 359, 183195.
17. Baird, S.J.E., and Santos, F. (2010). Monte
Carlo integration over stepping stone models
for spatial genetic inference using approximate
Bayesian computation. Mol. Ecol. Resources
10, 873885.
18. Estoup, A., Baird, S.J.E., Ray, N., Currat, M.,
Cornuet, J.M., Santos, F., Beaumont, M.A., and
Excoffier, L. (2010). Combining genetic,
historical and geographical data to reconstruct
the dynamics of bioinvasions: application to the
cane toad Bufo marinus. Mol. Ecol. Resources
10, 886901.
19. Philippe, H., and Roure, B. (2011). Difficult
phylogenetic questions: more data, maybe;
better methods, certainly. BMC Biol. 9, 91.

Research Centre in Biodiversity and Genetic

Resources (CIBIO), University of Porto,
Campus Agrario de Vairao, Rua Armando
Quintas 7, 4485-661 Vairao, Portugal; and
French National Institute for Agricultural
Research (INRA), 147 rue de lUniversite,
75338 Paris Cedex 07, France.
E-mail: StuartJ.E.Baird@gmail.com

DOI: 10.1016/j.cub.2012.02.011

idyllic Wolfgangsee to do experiments

(Figure 1). In the early 1930s, von Frisch
was interested in the question of
whether fish can hear. He had
conditioned a swarm of minnows,
a common small freshwater fish, to
sounds and wanted to label the
conditioned fish: To label another
member of the shoal, I one day caught
a minnow, severed its sympathicus
nerve with a thin knife near the end of
the tail, which causes a darkening of the
skin caudal of the incision site, and set
it free. There something unexpected

Current Biology Vol 22 No 6

R184

Figure 1. Lake of Fear.

The Wolfgangsee in Austria on a postcard from 1927, the kind von Frisch would have come
across during the time he spent there investigating fish behaviour and discovering the fish
alarm pheromone.

happened: some of the fish became

interested in the slightly bleeding
wound of the injured comrade and
snapped at it; the fish itself retired to
the depths and wasnt seen again; our
swarm, however, until then very tame,
was now visibly scared and only
reluctantly approached the food. About
a month later [.] one of the minnows
was trapped under the edge of the
feeding tube. The others regarded the
struggling comrade until I liberated it.
Now the bad news started spreading
through the whole shoal. An increasing
unrest took hold, and after a while, half
a minute may have passed, they
all fled [1].
Von Frisch was intrigued and,
passionate enquirer that he was, spent
the summer holidays of 1937 further
investigating the curious phenomenon.
Four years later, he had assembled
a 100-page paper [2] concluding: the
skin of minnows and other fish
harbours a substance von Frisch
called it Schreckstoff (fear or fright
stuff) that is released upon injury and
triggers fearful behaviour and escape in
conspecifics (Figure 2); the substance
is not present in other organs, and is
sensed through the nose; skin extracts
from one species can sometimes, but
not always, induce the fright response
in others; a predator attack releases
sufficient amounts of Schreckstoff
in an injured individual to scare off
the swarm.

The first question von Frischs

exhaustive description left behind
was obvious: what is Schreckstoff?
Soon after his initial description people
began to chemically hunt for the
substance, but it took several decades
and many kilos of fish skin until a good
candidate substance, hypoxanthine3-N-oxide (H3NO), had been isolated.
Now, writing in this issue of Current
Biology, Ajay Mathuru, Suresh
Jesuthasan and colleagues add a new,
potent ingredient to the Schreckstoff
cocktail [3].
The Alchemy of Alarm
H3NO had been shown to be able to
elicit alarm responses in fish of various
species [4,5], both in laboratory and
natural settings. It had also become
clear that it is not so much the purine
skeleton of H3NO that is crucial in
eliciting the response, but rather its
nitrogen oxide group, a group found
in many other compounds [5]. Early
biochemists had already noticed that
the Schreckstoff activity also has a high
molecular weight component, and
in zebrafish (Danio rerio), H3NO could
elicit some, but not all aspects of
the fright behaviour [3]. It thus looked
as though H3NO was not the only
Schreckstoff there had to be other
components.
Mathuru and colleagues [3] tested
fractions of zebrafish skin extract for
whether they elicited fright responses.

Two fractions of different molecular

weight together elicited all aspects
of the alarm response affirming the
notion that Schreckstoff is a cocktail
rather than a single substance. While
the low molecular weight fraction
proved tricky to identify, the high
molecular weight fraction indicated
long polymers, possibly
polysaccharides (polymers of
carbohydrate units joined by glycosidic
bonds) as mediators of the alarm
response.
Then, the authors noticed something
unexpected: when zebrafish were
vigorously shaken, but otherwise not
injured, the resulting mucus or slough
also elicited a mild alarm response.
This was curious because the dogma
had always been that Schreckstoff
is only released upon injury. But this
proved to be a fortunate finding, as
the mucus samples turned out to be
chemically much less complex than
the skin extracts, thus greatly reducing
the number of potential leads to
investigate. Even more curiously,
boiling the mucus for 2 hours made it
a much more potent inducer of the
alarm response. A major component
of mucus is glycosaminoglycans
(GAGs), unbranched polysaccharides
composed of multiple repeated
disaccharide units. Upon heating, the
GAGs are fragmented and released,
and could hence be the high molecular
weight alarm component hinted at.
And indeed skin extracts contain
GAGs, and their amount correlates
with their potency of eliciting an alarm
response. This was suggestive,
but what was the precise identity
of the alarm component? Only
after mass spectroscopy,
fluorescence-assisted carbohydrate
gel electrophoresis, immuno-depletion,
enzymatic digestion and comparisons
with synthetic standards did
Jesuthasan and his modern day
alchemists manage to turn fish goo into
scientific gold [3].
If GAGs are a component of the
Schreckstoff, they should also be
detected through the olfactory system,
like the original Schreckstoff.
Repeating von Frischs simple
approach, Jesuthasan and his team [3]
sealed the nostrils of zebrafish, and
indeed, such fish no longer showed
alarm responses. Adding a more
contemporary twist, imaging of neural
activity revealed that the new
Schreckstoff component and skin
extract both activate the same region

Dispatch
R185

of the primary olfactory processing

centre the medio-dorsal posterior
region of the olfactory bulb, which has
also been shown to detect fear signals
in other fish species [6]. What is
interesting about this particular region
is that it is innervated by so-called
crypt cells, a morphologically distinct
type of olfactory sensory neuron
unique to fish and thought to mediate
reproductive behaviours [7]. Crypt cells
may thus constitute a special sensory
channel for signals related to vital
behaviours such as sex and fear.
Of course, alarm pheromones are
not only found in fish; they are found
everywhere: in plants, for example,
specific volatiles, such as the green leaf
volatile (3)-hex-3-enyl acetate, are
released into the air following herbivore
attack and trigger upon detection
defensive measures in neighbouring
plants [8]. In the predatory sea slug
Navanax inermis, a blend of alarm
substances, including 10-(30 -pyridyl)3E,5E,7E,9E-decatetraen-2-one
(or Navenone A for short), is produced
in a dedicated gland and is secreted
into the slime trail upon molestation,
as the technical term goes, which
triggers avoidance and alarm
behaviour in trail-following
conspecifics [9]. Alarm pheromones
have also been characterized from
a number of insects; a well-known
example is the alarm pheromone
(E)-b-farnesene of the pea aphid
(Acyrthosiphon pisum), which not only
elicits avoidance in other aphids, but
also triggers the production of winged
offspring capable of deserting the
dangerous birthplace [10]. Among
social insects, such as ants and bees,
sophisticated alarm signals are
particularly prevalent, and serve
a number of critical functions in the
colony [11].
Everybody Panic!
In terms of behaviour, the alarm
response is largely innate and varies
between and within different species of
fish [12,13]: when facing alarm signals,
some fish skitter, dart or dash, while
others freeze; shoals disperse or
huddle together; some species swim
nose down on the ground stirring up
mud, or go into hiding, yet others rise
to the surface, even jumping out of the
water. So what is lumped together
under the label fright or alarm reaction
is actually a whole suite of behaviours
that vary within and between species.
Despite this variation, the fear program

is conserved across wide evolutionary

distances, even outside fish, a notion
betrayed by the observation that fish
alarm responses can be dampened
by human anxiolytic drugs [14].
But what are the benefits of these
behaviours? To those that respond to
the Schreckstoff emitted by an injured
fish the benefits seem rather obvious:
they can escape, hide, conceal
themselves, play dead or confuse
predators. So, even if the initial victim
should perish, the other members of the
group can survive. In the good old days
of group selection, such apparent group
benefits seemed to be sufficient to
explain the evolutionary origin of the
Schreckstoff system: they were good for
the benefit of the group even if one fish
got seriously injured or eaten. But with
the advance of the gene-centric view of
evolution this became a problem. If the
sender of the alarm signal did not gain
a fitness advantage, how could genes
encoding the machinery required to
make and receive the Schreckstoff
persist in evolution?
Kin selection the idea that a trait
can evolve when it confers a benefit
also in genetically related
individuals looked like the obvious
answer. But the problem is that shoal
members, of minnows or other fish
species, are not necessarily close
relatives [15]. Another, even simpler
explanation might be that there is no
signal at all. In order to be stable in
evolution, a signal needs to confer
a fitness benefit, to offset the costs
of making, storing and releasing the
signal. But, what if the signal carries
no cost and is just a by-product? After
all, the release of Schreckstoff appears
to be a passive process caused by
rupture of the skin.
At some point, doubts arose whether
an alarm system existed at all. Most of
the experiments following von Frischs
initial serendipitous discovery had
been made in aquaria or semi-enclosed
spaces, where chemicals can spread
easily. Indeed, when wild Eurasian
minnows in a stream were confronted
with skin extract they failed to show the
expected escape response [16] a
finding that caused itself considerable
alarm in the research community, but
was later countered with evidence
favouring presence of the response
in wild North American minnows [5].
Fright Club
There is evidence that the alarm signal
does in fact come at a cost for the

Figure 2. Schreckstoff in action.

Mechanical injury such as that inflicted by
the teeth of predators causes the release
of an alarm pheromone (Schreckstoff) from
the skin, which triggers avoidance and
escape behaviour in conspecifics. Drawing:
Rakel Stensmyr.

releaser: in the skin there are special

secretory cells, so-called club cells
(Kolbenzellen), that were thought to be
dedicated to the alarm system via guilt
by association the more club cells in
a piece of skin, the more potent the
extract is in causing a fright response,
and nearly all species that have club
cells also have a fright response [17].
In fact, the presence of club cells and
alarm response neatly matches
a taxonomic unit within the bony
fishes the super-order Ostariophysi
that comprises nearly a third of all fish
species and over two-thirds of all
freshwater fish. These club cells are, so
to say, the embodiment of the cost of
the alarm system and their presence
is what requires an evolutionary
explanation.
The best ecological evidence so far
for a benefit to the sender of the alarm
signals comes from a paradoxical
effect of the alarm pheromone. As the
Schreckstoff can also be perceived
by members of other fish species,
also some predators, such as pike,
are attracted to it [18]. It can even
attract aquatic beetles that also prey
on smaller fish. More predators would
seem to mean more trouble, but there
is evidence that predators can get
in each others way and thus in effect
increase the fishs chance of getting
away after an attack [19]. But whether

Current Biology Vol 22 No 6

R186

this is a general mechanism in the many

fishes that presumably have the
Schreckstoff system let alone
in zebrafish about whose ecology
deplorably little is known remains
an open question.
The chemical nature of the
Schreckstoff component identified by
Mathuru and colleagues [3] and its
being a constituent of mucus do
suggest that the alarming function
could be a secondary by-product.
But then, what about the club cells?
Surprisingly, there is evidence that
their alarm function might also only
be secondary. When fathead
minnows are exposed to parasites,
pathogens or UV light, they increase
the number of club cells in the skin,
an effect that is inhibited by
immuno-suppresion [20]. So, after all,
the Schreckstoff could just be
a by-product of other protective
functions that enhance fitness of
the bearer much like its
serendipitous discovery was the
by-product of a summer holiday
and the study of hearing in fish. And
much like that first minnow that von
Frisch cut, the presumed evolutionary
enigma of Schreckstoff might just
disappear. Oh, by the way, did we
mention what the new Schreckstoff
component was? Oligosaccharides
of chondroitin-4-sulfate and
chondroitin-6-sulfate.

References
1. von Frisch, K. (1938). Zur Psychologie des
Fisch-Schwarmes. Naturwiss. 26, 601606.
2. von Frisch, K. (1941). Uber einen schreckstoff
der Fischhaut und seine biologische
bedeutung. Z. Vergl. Physiol. 29, 46149.
3. Mathuru, A.S., Kibat, C., Cheong, W.F.,
Shui, G., Wenk, M.R., Friedrich, R.W., and
Jesuthasan, S. (2012). Chondroitin fragments
are odorants that trigger fear behavior in fish.
Curr. Biol. 22, 538544.
4. Pfeiffer, W., Riegelbauer, G., Meir, G., and
Scheibler, B. (1985). Effect of
hypoxanthine-3(N)-oxide and hypoxanthine- 1
(N)-oxide on central nervous excitation of the
black tetra Gymnocorymbus ternetzi
(Characidae, Ostariophysi, Pisces) indicated
by dorsal light response. J. Chem. Ecol. 11,
507524.
5. Brown, G.E., Adrian, J.C., Jr., Smyth, E.,
Leet, H., and Brennan, S. (2000). Ostariophysan
alarm pheromones: laboratory and field tests of
the functional significance of nitrogen oxides.
J. Chem. Ecol. 26, 139154.
6. Hamdani, E.H., and Dving, K.B. (2003).
Sensitivity and selectivity of neurons in the
medial region of the olfactory bulb to skin
extract from conspecifics in crucian carp,
Carassius carassius. Chem. Senses 28,
181189.
7. Hamdani, E.H., and Dving, K.B. (2006).
Specific projection of the sensory crypt cells
in the olfactory system in crucian carp,
Carassius carassius. Chem. Senses 31,
6367.
8. Kost, C., and Heil, M. (2006). Herbivore-induced
plant volatiles induce an indirect defence
in neighbouring plants. J. Ecol. 94,
619628.
9. Sleeper, H.L., Paul, V.J., and Fenical, W. (1980).
Alarm pheromones from the marine
opisthobranch Navanax inermis. J. Ecol. 6,
5770.
10. Kunert, G., Otto, S., Rose, U.S.R.,
Gershenzon, J., and Weisser, W.W. (2005).
Alarm pheromone mediates production of
winged dispersal morphs in aphids. Ecol. Lett.
8, 596603.
11. Blum, M.S. (1969). Alarm pheromones. Annu.
Rev. Entomol. 14, 5780.

Active Vision: Fixational Eye

Movements Help Seeing Space
in Time
The significance of the miniature eye movements that we make during visual
fixation has been intensely debated for the last 80 years. Recent studies have
revealed that these motions of the eyes fulfill an important functional role:
helping to extract useful information from natural scenes.
Igor Kagan
We live in a dynamic environment, in
which the visual scenery changes from
one instant to another. Some of these
changes are caused by external events,
such as the movement of trees in the
wind. But even when the world is
stationary, our own movements
constantly shift the projection of the
visual scene on the retina. Even when

we have no intention to look around,

but try to maintain a steady gaze
during fixation, our eyes are still in
constant motion because of the
instability caused by fixational eye
movements: slow ocular drifts and
fast abrupt shifts called fixational
saccades or microsaccades. For
decades, the role of these
self-generated retinal motions has
been an important and controversial

12. Smith, R.J.F. (1992). Alarm signals in fishes.

Rev. Fish Biol. Fish 2, 3363.
13. Pfeiffer, W. (1962). The fright reaction of fish.
Biol. Rev. 37, 495511.
14. Rehnberg, B.G., Bates, E.H., Smith, R.J.,
Sloley, B.D., and Richardson, J.S. (1989). Brain
benzodiazepine receptors in fathead minnows
and the behavioral response to alarm
pheromone. Pharmacol. Biochem. Behav. 33,
435442.
15. Bernhardt, B., Lampert, K.P., Leese, F.,
Mayer, C., and Tollrian, R. (2012). Are
shoals of minnow Phoxinus phoxinus formed
by close kin? J. Fish Biol. 10.1111/
j.1095-8649.2011.03198.x.
16. Magurran, A.E., Irving, P.W., and
Henderson, P.A. (1996). Is there a fish alarm
pheromone? A wild study and critique. Proc. R.
Soc. B 263, 15511556.
17. Pfeiffer, W. (1977). The distribution of fright
reaction and alarm substance cells in fishes.
Copeia 4, 653665.
18. Mathis, A., Chivers, D.P., and Smith, R.J.F.
(1995). Chemical alarm signals:
predator-deterrents or predator attractants?
Am. Nat. 145, 9941005.
19. Chivers, D.P., Brown, G.E., and Smith, R.J.F.
(1996). Evolution of chemical alarm signals:
attracting predators benefits alarm signal
senders. Am. Nat. 148, 649659.
20. Chivers, D.P., Wisenden, B.D., Hindman, C.J.,
Michalak, T.A., Kusch, R.C., Kaminskyj, S.G.W.,
Lack, K.L.M., Ferrari, C.O., Pollock, R.J.,
Halbgewachs, C.F., et al. (2007). Epidermal
alarm substance cells of fishes maintained
by non-alarm functions: possible defense
against pathogens, parasites and UVB
radiation. Proc R. Soc. B 274,
26112619.

1Max Planck Institute for Chemical Ecology in

Jena, Germany. 2Florian Maderspacher is
Current Biologys Senior Reviews Editor.
E-mail: mstensmyr@ice.mpg.de;
florian.maderspacher@current-biology.com

DOI: 10.1016/j.cub.2012.02.025

issue, fraught with methodological

complications [1]. Are these
movements useful, irrelevant,
or damaging to our vision? Despite
progress in elucidating the
perceptual and neuronal effects of
fixational instabilities in specific,
artificial laboratory settings, a much
needed conceptual and
computational framework for
understanding the role of fixational
eye movements in natural vision has
been lacking. A series of recent
advances by Rucci and colleagues,
including the most recent study by
Kuang et al. in this issue of Current
Biology [2], make a crucial step
towards this fundamental
understanding, demonstrating that
fixational eye movements are an
integral part of early visual processing
strategy to efficiently analyse and
encode natural scenes.