Content uploaded by James Michael Carpenter
Author content
All content in this area was uploaded by James Michael Carpenter
Content may be subject to copyright.
Nature © Macmillan Publishers Ltd 1998
8
Here we report fossil ants, including a
new genus of Ponerinae, about 50 mil-
lion years (Myr) older than the previous old-
est specimens. These discoveries in amber
from the Turonian stage (92 Myr ago) of
New Jersey in the United States have impor-
tant implications for estimates dating the
origin of ants, and extend the age of an
extant ant subfamily back about 50 Myr.
Until now, a specimen of Sphecomyrma
freyi in a piece of amber from Cliffwood
Beach, New Jersey, was one of the few non-
compression fossil ants from the Cretaceous
period showing most morphological details1.
The metapleural gland is the only morpho-
logical trait unique within the Hymenoptera
that distinguishes ants, and can be seen in
most ants embedded in amber. This gland
produces antibiotic-like substances2, neces-
sary to maintain nests underground or in
humid pieces of wood, where bacteria and
fungi would otherwise invade immobile
broods. The gland is seen in modern species,
opening above the hind coxae. Development
of the gland and eusociality were probably
correlated and are involved in the great eco-
logical success of the ants. In an Amazonian
rainforest, for example, ants make up more
than 25% of the total animal biomass3.
Wilson et al.1described an almost per-
fectly preserved worker of S. freyi, but the
debate remained whether this fossil is an ant
because the presence of the metapleural
gland was uncertain. Sphecomyrma was thus
excluded from a recent phylogenetic analysis
of the Formicidae4, and its relatively short
first antennal segment in workers was
assumed to prevent the basic social behav-
iour of trophallaxis5.
Our new specimens include three worker
and four male ants. One complete, well-
preserved worker was the second discovery
of S. freyi. Among other well-preserved traits
in this specimen are the external and even
part of the internal anatomy of the meta-
pleural gland. The gland has a circular open-
ing and a very large subcuticular atrium (Fig.
1a). Sphecomyrma is thus an ant.
Two of the males are a new species of the
Cretaceous genus Baikuris, known previous-
ly only from Upper Cretaceous (Santonian)
amber from Taymyr, northern Siberia. The
third male belongs to a genus as yet undeter-
mined. The fourth male is tentatively
assigned to Sphecomyrma and would be the
first known male of that genus.
Another worker represents a new genus,
distinguished by its clubbed antennae, pro-
portions of the antennal articles, thin man-
dibles that lack teeth and cross extensively
when closed, and a girdling constriction of
the gaster typical of the large modern sub-
family, the Ponerinae (Fig. 1b). Furthermore,
the broad attachment of the third and fourth
abdominal segments, and the apical denticles
on the genae, indicate a relationship with
ants of the ponerine tribe Amblyoponini.
Addition of these two new taxa confirms
the basal position of Sphecomyrma6, but as
part of a quadritomy. The position of the
new ponerine within one clade of the Poner-
inae demonstrates that one major lineage of
extant ants was established well before the
formation of amber from Sakhalin Island
(probably Palaeocene)7and of Eocene Baltic
amber8, which are about 50 Myr younger
than the New Jersey amber.
A reasonable estimate would thus place
the origin of the ants into the lowermost
Cretaceous (about 130 Myr ago), but proba-
bly no older. This conclusion is consistent
with the relationship of the Formicidae to
the wasp families Vespidae and Scoliidae9
and the phylogenetic position of the Creta-
ceous Vespidae10. On this basis, a hypothesis
of a Lower Jurassic origin of ants11, based on
estimated divergence time of mitochondrial
cytochrome bsequences, is unlikely.
Although ants originated in the Creta-
ceous, it was not until the Tertiary7period
that they became so dominant and diverse in
terrestrial ecosystems, as documented in
Baltic8and Dominican amber12. We are still
examining why the radiations were delayed
until the Tertiary.
Donat Agosti, David Grimaldi,
James M. Carpenter
Department of Entomology, American Museum of
Natural History, Central Park West at 79th Street,
New York, New York 10024-5192, USA
e-mail: agosti@amnh.org
1. Wilson, E. O., Carpenter, F. M. & Brown, W. L. Jr. Psyche 74, 1–19
(1967).
2. Beattie, A. J., Turnbull, C. L., Hough, T. & Knox, R. B. Ann.
Entomol. Soc. Am. 79, 448–450.
3. Fittkau, E. J. & Klinge, H. Biotropica 5, 2–14 (1973).
4. Baroni Urbani, C., Bolton, B. & Ward, P. S. Syst. Entomol. 17,
301–329 (1992).
5. Dlussky, G. M. Paleont. J. 1983 (3), 65–78 (1983).
6. Grimaldi, D. et al. J.M. Am. Mus. Novit. (in the press).
7. Dlussky, G. M. Paleont. J. 1988 (1), 50–61 (1988).
8. Wheeler, W. M. Schr. Phys.-Ökon. Ges. Königsb. 55, 1–142 (1915).
9. Brothers, D. J. & Carpenter, J. M. J. Hymenopt. Res. 2, 227–304
(1993).
10. Carpenter, J. M. & Rasnitsyn, A. P. Psyche 97, 1–20 (1990).
11. Crozier, R., Jermiin, L. S. & Chiotis, M. Naturwissenschaften 84,
22–23 (1997).
12. Hölldobler, B. & Wilson, E. O. The Ants(Harvard Univ. Press,
Boston, 1990).
NATURE
|
VOL 391
|
29 JANUARY 1997 447
scientific correspondence
a
b
MGO
Posterior end
Atrium
(subcuticular)
0.5 mm
m
0.2mm
Figure 1 Photomicrographs of New Jersey ant
fossils. a, Alitrunk and portion of the gaster of
the new
S. freyi
worker with a drawing showing
the position and structure of its metapleural
gland (MGO). b, Lateral view of the new Creta-
ceous ponerine genus (m, mandible in frontal
view). A detailed description will be provided
elsewhere6(http://research.amnh.org/entomol-
ogy/social_insects).
Oldest known ant fossils discovered
tionship where a 1 °C fall in diurnal temper-
ature range increased Australian wheat yield
by 0.52 t ha11. This effect, taken with the
trend in diurnal range, accounted for 45% of
the yield increase between 1952 and 1992. In
an extended model with three climate vari-
ables, Nicholls found that changes in mini-
mum temperature had had most impact on
wheat yield and that rainfall change had con-
tributed little. In our view, Nicholls’s results
need qualification and should be interpreted
with caution. They are only estimates, and
do not include standard errors to indicate
their precision.
The analytical challenge is to separate the
effects of climate change from other factors
affecting Australian wheat yields. Non-
climate factors include total wheat area, the
Climate change and
Australian wheat yield
Nicholls1reported that 30–50% of the
increase in Australian wheat yields in the
period 1952–92 resulted from climate
change. He estimated a simple linear rela-
