[Coral-List] Misleading ABH Baker al 2004 Nature paper

Martin Pêcheux martin-pecheux at wanadoo.fr
Fri Nov 19 17:09:09 EST 2004

Dear  all,

I know that there was excitation in Okinawa about Symbiodimium clades
and possible adaptation to coral reef mass bleaching, as professed by
the somewhat ecologically irresponsible "Adaptative Bleaching Hypothesis
(ABH)" (Buddemeier and Fautin, 1993, Buddemeier et al., in press). As
for "cyclones increase reef biodiversity" (which is a short term view,
think of Amazon forest, and nobody knows why tropics are diverse), I
feel that this kind of theory has its roots in Reaganic-Bush thinking,
namely that poor people will adapt anyway, so better not to do anything,
particularly helping them. This would be a classical relationship
between science and society, as Victorian darwinism or "quantic
Picasso". Alternatively, reef scientists may be refusing the evidence in
front of their eyes, namely that reefs are collapsing. At the 1998
Europ. Meet. Coral Reefs, Perpignan, when I saw once more a picture of
mass bleaching, I joked with few colleagues: "in 20 years, there will be
no more reef biologists, only geologists". Latter, two people told me it
back, sign of its teasing truth. May be it is natural "compassion", or
reef biologists are afraid to loose their job. This is even worse for
reef managers, who call for more reef protections in time of mass
bleaching (I don't say that protection must not be reinforced, but just
that CO2 mitigation is the efficient way to save reefs). But let me be
very serious, and sorry for my bad English.

Previous works have been unable to prove clearly that mass bleaching
induces change in symbiont composition to high-temperature tolerant ones
in hard corals (synthesis in Coles and Brown, 2003), although every body
can just hope it. Now two new papers on this topic appeared in Brief
Communications in the 12th August issue of Nature (Baker et al., 2004,
Rowan, 2004).

Baker, Starger, McClanahan and Glynn (2004), "Corals' adaptative
response to climate change. Shifting to new algal symbionts may
safeguard devastated reefs from extinction", provide proportions of
Symbiodinium clades A, C, D in corals in several areas (Uva Island,
Kenya, Mauritius, Red Sea, Persian Gulf). Apart from the former, these
analyses were conducted just once, in 2000-2002. It just indicates that
different clades inhabit preferentially different localities, likely due
to different maximum temperatures, as already known (see their
references, and the excellent review of Baker, 2003), and not bleaching
effects. For the time serie data of Uva Island (Chiriqui Gulf, Panama,
Pacific side), "colonies containing [the resistant] Symbiodinium in
clade D were already common (43%) in 1995" and "by 2001
had become
dominant (63%)", after the two bleaching bouts in late 1997 and early
1998. In 1995, there was in fact 50% corals bearing clade D (n=17/34, in
Supplementary Information). This is because they counted the corals
hosting both clades C and D for an half (A. Baker, pers. com.). So the
change is twice smaller than indicated. They do not gave the
significance of the change between 1995 to 2001, from 50.0% to 63.4%,
but stated "Here we show that corals containing UNUSUAL algal symbionts
that are thermally tolerant [
] are MUCH MORE ABUNDANT on reefs that
have been severely affected by recent climate change". The significance
is p=0.121 (unilateral t-test; Chi-2=1.348, p>0.1). Nature's referees
must apologise.
There was significant changes of C-symbiont-bearing corals (p=0.0075)
but mainly due to the disappearance of corals with both C and D
symbionts (from 14.7%, n=5/34, to 0%, n=0/41, p=0.0055), not due to less
corals with C only (50%, n=17/34, to 36.6%, n=15/41, p=0.0696, just not
significant). This is practically the contrary to the ABH. C symbionts
are eliminated when in competition inside a colony, and no gain of D
symbionts are evident.
The proportion of clade D (found only in non-bleaching corals) in
October 1997 during the mass bleaching appears to be 78.9% (n=33/42).
Perhaps there is a bias if they choose to sample more normal corals than
bleached ones (they draw two pie charts), but they would have to told
it. The proportion of bleached corals, although measured, is not given
in Glynn et al. (2001). The mortality was only 3.2%.
In addition, in Kenya, with the most severe bleaching (Wilkinson, 2002),
the proportion of the still resistant D-symbiont-bearing corals is only
mean 36.2%, with variations between two neighbouring localities from
21.7% in Kisite to 66.6% in Kanamai (different at p=0.001). In Mauritius
(with only 3% D-bearing corals, n=1/31), there was bleaching in early
1998, up to 39%, and 31% in Blue Bay where some sampling were done
(Turner report, in www.reefbase.org).
Concerning the proportions of clades A, C, D in other countries,
attributed to the magnitude of the 1997-1998 bleaching event, the
biogeography of the symbiont clades is still in its infancy. They could
be alternatively explained, at least in part, by thermal history
previously to the bleaching. The proportion of clade C in the 5
countries is inversely correlated with the mean of the annual maximum
temperature 1950-1996 (HadISST data) at r2=0.842, p=0.028. And perhaps,
to be confirmed or infirmed, the proportion of clade A with its variance
(r2=0.373, p=0.19), or with a threshold at standard deviation>0.4°C.
In brief, Baker et al. do not prove anything. Nature reply do not allow
much development, so this email. I must emphasise that I found in
general that the works of these authors are very good and of great
importance (and our exchange very interesting), but, for their
hypothesis to be established, facts are still needed.

Rowan "Thermal adaptation in reef coral symbionts" (2004) only shows
that his two strains of Symbiodinium clades C and D have different
tolerances to high temperature, which was already proposed. It is OK to
observe the decrease of Fv/Fm from 31.3°C thereafter 32°C during 5 days,
at 890µE.m-2.s-1, in clade C not D, however from only about 0.54 to 0.47
(from Fig. 1a, not in Supplementary Information), probably far away from
bleaching threshold. But he said that p=0.02 (n=7), and I know how
precise can be chlorophyll fluorescence measurements (Pêcheux, 1997,
Appendix III, 2002, Tsimilli-Michael et al., 1998). More interesting is
the 20% Fo increase, not Fm decrease, indicative of PS II thermal
denaturation. Thermal denaturation, probably at the oxygen evolving
complex, and Fo rise begin in land plants at quite higher temperatures
(38°C at least, up to more than 45°C). The standard procedure is to
increase slowly the temperature (1°C/minute) in darkness on a fixed
sample (Fo/T curves). With M. Havaux, I found that critical temperature
is also high, 36.5-38°C in Symbiodinium in hospite of Sorites orbiculus
from French Mediterranean (in situ max T=29-30°C), whereas they bleach
at 32°C (Pêcheux, 1997) (see also rise of Fo at temperature higher than
bleaching with different protocols, 35°C in Iglesias-Prieto, 1995, 34°C
in Warner at al., 1996, also crash of NPQ in 5 minutes at 38°C, not
37°C, Ralph et al., 2001; see Jones et al., 1998, Fig. 8; no Fo rise,
Warner, unpublished in Fitt et al., 2001;  also with same signification,
rise of K step at 300µs at 36°C, Iglesias, 1997; and see Fo rise in
Jones and Hoegh-Guldberg, 2001, at 21.5-27°C, with full sun light on
shade colonies, not temperature). And it is to be known if Fo rise is
not just associated to light-dark state transition induced by the
darkness incubation before measurement. The Fig. 1d of the
photosynthesis/respiration ratio is the same data than Fig. 1c. The
Fv/Fm down-regulation with light increase (Fig. 1b) is known for decades
(also now from several papers with diel cycles of corals; with a log
pattern in light-adapted state, not linear, Pêcheux, 2002), and saying
that p<0.0001 is clearly no information for Nature readers. Many of the
interesting points are in fact in Rowan submitted. So we are left with
two experiments with different coral species, under one stress
condition, with measures one by Fv/Fm, the other by O2
 Rowan is
relatively cautious when he concluded "symbiont recombination MAY BE one
mechanism by which corals adapt, IN PART, to global warming". Yes, may
be, but there is still no evidence of recombination. However he didn't
hesitate to state in the introduction that "if other coral species can
host these or similar Symbiodinium [D]", "they might adapt to warmer
habitats RELATIVELY EASILY". Thanks for them, but I prefer that USA sign
the Kyoto protocol at minimum.

Those works rather only indicate that ribosomal clades are partitioned
according to maximum temperatures and that their physiological responses
to temperature are different. This was already known (their references).
It has long been known that coral temperature threshold is close to
local maximum temperature (synthesis in Coles and Brown, 2003).
Thresholds have great range (from 23.9°C in Mediterranean, Perez et al.,
2000, to ~33.5°C° in Persian Gulf). This implies a very strong
selection, thus a very high cost of host/symbiont thermal adaptation
(which biochemical pathway?). So it is probable that this adaptation
took place on long time scale, with glacial oscillations or probably
during millions years. With "current controversy over time scale"
(Baker, 2003), "especially needed are data on the time frame required
for selective adaptation" (Coles and Brown, 2003). The correspondence of
temperature response with ribosomal DNA clades reinforces the deep root
of this adaptation. But there is a problem with contradictory data, at
the same phylogenetic resolution, not discussed nor even quoted, with
independence of clades and physiological responses to light (Savage et
al., 2002) or temperature (Kinzie et al., 2001, and now Tchernov et al.,
2004). And theoretically here counter-examples are proofs. Divergence
between two clades B from Aiptasia pulchella, one with thermal
tolerance, the other not, is as young as <1 million years (Tchernov et
al., 2004, Fig. 4).
Whatever the case, more important, the strong selection indicates that
even if ABH is true (Stat et al., 2004, abstract), its cost will be
high. This was shortly envisaged by Rowan et al. (1997) and is now
demonstrated for juvenile corals with 2.7 lower growth rate with D clade
than with C one (Little et al., 2004), and 1.8 greater mortality
(logarithmic) rate in Acropora millepora (not in A. tenuis) (op. cit.,
data in Fig. 2 legend). The consequence for competition are obvious, in
particular against algae (corals have to concrete fast against them).
Saying that clade D is thermally resistant is not all the truth. In a
chapter of its excellent review, Baker (2003) pointed out that clade D
appears "unusual, weedy, opportunistic". It is found in very shallow or
very deep waters, or at mid-depth at the transition between two other
clades, or in marginal environments with terrestrial impact or
temperature excursion. If it is indeed just profiting of what place is
left under various stresses, it is certainly a poor competitor with low
fit to normal environment, with probable (very) high cost of ABH if
In Panama in fall 1997 all bleached samples contained clade C (n=9) and
all non-bleached ones contained clade D (n=33). There was a second bout
of bleaching in early 1998, which included corals that did not bleach
earlier (Glynn et al., 2001). So seemingly corals with clade D bleach
also, though less. They are not safeguard. Given that mass bleaching
threshold is only 0.4°C above local mean annual maximum temperature
(subm.), thermal resistance of clade D might be around an half of a
degree higher than that of clade C. If all corals species adopt clade D,
this would only delay the reef extinction from 2015 to 2035, according
to calculation of Sheppard (2003) for Kenya as a typical example.
Extinction of reefs in next decades, total or not depending on future
CO2 level, must be very seriously envisaged.

I will preach for my pets, very easy to collect and cultivate. Large
foraminifers harbour also different symbionts (Symbiodinium "distinct
entities", Langer and Lipps, 1994, A, C, F, G, Baker, 2003; or several
diatom species, see Lee et al., 1995; chlorophytes) and bleach. They
have short life (few months to one year, apart exceptions) and are by
far more numerous than corals (above one million times, pers. eval.). So
indication of an adaptation to bleaching, by ABH symbiont shifting or
more simply by selection, is expected to be seen first in symbiotic
foraminifers, which seems is not the case. And now they show frequent
spectacular shell malformations (soon on Web), never seen in geological
time apart in planktonics at the Cretaceous/Tertiary boundary, and
recovery needed millions years (Pêcheux, 1999). They are the "canaris in
the coal mine" of the corals, themselves "canaris in the coal mine" of
Global Change. You know, it is very difficult to see their malformations
never seen in all my geological collections, bibliography and asked
contacts (I was a petroleum micropaleontologist 20 years ago), and read
so often: "they will adapt" and rarely "we have to envisage not".
The "Adaptative Bleaching Hypothesis" tends to suggest that symbiosis
exists to allow bleaching. It still exist in sponges with cyanobacteria
for surely more than 500 millions years without endosymbiosic process,
more than secondary and tertiary endosymbiosis time, as for example in
Euglena (which bleach at 33°C, refs in Pêcheux, 1997). I don't think
that symbiosis is for bleaching. It is generally assumed that it is for
nutrients supply by the host. Another hypothesis considers that
symbiosis provides "nested-dolls" for Carbon Concentration Mechanism
(Allemand et al., 1996, after discussions), also with carboxysomes in
Symbiodinium, or foraminifer shell, a more fundamental acquisition in
clear, oligotrophic, warm waters (allowed light cost of CCM and reduced
photorespiration which is CO2 and temperature-dependent - due to one
third to O2/CO2 solubility, two third to Rubisco enzymatic dynamic - and
nitrogen photorespiratory cycle loss).
I profit of this email to remember that CO2 rise and seawater
acidification is for me the causal factor of increased temperature
sensitivity of corals leading to bleaching, supposedly also of their
general weakening (Pêcheux, 1991, confidential, 1994, 1996, 1997, 2002,
in reefbase, and subm.). Aquariophilists take great care of
CO2/pH/sursaturation. Wilkens (1990) say that that "CO2 increased
bleaching in leather corals".  For a geochemist, it is an evidence that
CaCO3 ecosystems are beaten by CO2 (W. Berger, W. Broecker, pers. com.
exchanges, 1990). But the mechanisms are damned complex.
Rather than so numerous genetic works, indeed interesting but, I prompt
for more studies of the molecular mechanism of bleaching, unless we
won't be able to understand it, thus predict it. We may come to
interconnected multiple mechanisms (lipids?, oxygen radical, PS II,
Rubisco, CCM), all already very strongly selected at the limit of summer
synergies of high temperature, high irradiance, low water agitation, and
fourth now rising CO2/lowering pH. This would be typical of Global
Change. It is now time of both strong researches and moving for Global
Change of Civilisation.


Dr. Martin Pêcheux, Scientific Consultant/IPCC
Large Foraminifer Institute
3, allée des Elfes, 94260 Fresnes, France
Phone: +33 1 4237 4196
martin-pecheux at wanadoo.fr
Publications at  www.reefbase.org, Publication , Author=Pecheux (without

L'homme descend du singe, la preuve, c'est qu'il croit en dieu.
Human' origin is monkey, proof the belief in god.
Nosotros descendemos de los monos, prueba se cree en dios.
Mannschaft stammen von Affe, Beweis gott glaube.

More information about the Coral-List mailing list