doi:10.1144/gsjgs.154.2.0265 1997; v. 154; p. 265-292 Journal of the Geological Society
A. R. MILNER, N. MORRIS, E. OWEN, B. R. ROSEN, A. B. SMITH, P. D. TAYLOR, E. URQUHART and J. R.
BURNETT, P. CHAMBERS, S. CULVER, S. E. EVANS, C. JEFFERY, M. A. KAMINSKI, A. R. LORD, A. C. MILNER,
N. MACLEOD, P. F. RAWSON, P. L. FOREY, F. T. BANNER, M. K. BOUDAGHER-FADEL, P. R. BOWN, J. A.
The Cretaceous-Tertiary biotic transition
Journal of the Geological Society
Email alerting to receive free email alerts when new articles cite this article click here
Permission to seek permission to re-use all or part of this article click here
Subscribe to subscribe to Journal of the Geological Society or the Lyell Collection click here
Downloaded by U C L Library on 24 August 2008
© 1997 Geological Society of London
Journal of the Geological Society,London, Vol. 154, 1997, pp. 265–292, 11 ﬁgs, 1 table. Printed in Great Britain
The Cretaceous–Tertiary biotic transition
, P. F. RAWSON
, P. L. FOREY
, F. T. BANNER
M. K. BOUDAGHER-FADEL
, P. R. BOWN
, J. A. BURNETT
, P. CHAMBERS
, S. CULVER
S. E. EVANS
, C. JEFFERY
, M. A. KAMINSKI
, A. R. LORD
, A. C. MILNER
A. R. MILNER
, N. MORRIS
, E. OWEN
, B. R. ROSEN
, A. B. SMITH
, P. D. TAYLOR
& J. R. YOUNG
Department of Palaeontology, The Natural History Museum, Cromwell Road, London, SW7 5BD, UK
Department of Geological Sciences, University College, London, Gower Street, London, WC1E 6BT, UK
Department of Anatomy and Developmental Biology, University College, London, Gower Street, London,
WC1E 6BT, UK
Department of Biology, Birkbeck College, Malet Street, London, WC1E 7HX, UK
Abstract: Mass extinctions are recognized through the study of fossil groups across event horizons, and
from analyses of long-term trends in taxonomic richness and diversity. Both approaches have inherent
ﬂaws, and data that once seemed reliable can be readily superseded by the discovery of new fossils and/or
the application of new analytical techniques. Herein the current state of the Cretaceous–Tertiary (K–T)
biostratigraphical record is reviewed for most major fossil clades, including: calcareous nannoplankton,
dinoﬂagellates, diatoms, radiolaria, foraminifera, ostracodes, scleractinian corals, bryozoans, brachio-
pods, molluscs, echinoderms, ﬁsh, amphibians, reptiles and terrestrial plants (macrofossils and
palynomorphs). These reviews take account of possible biasing factors in the fossil record in order to
extract the most comprehensive picture of the K–T biotic crisis available. Results suggest that many
faunal and ﬂoral groups (ostracodes, bryozoa, ammonite cephalopods, bivalves, archosaurs) were in
decline throughout the latest Maastrichtian while others (diatoms, radiolaria, benthic foraminifera,
brachiopods, gastropods, ﬁsh, amphibians, lepidosaurs, terrestrial plants) passed through the K–T event
horizon with only minor taxonomic richness and/or diversity changes. A few microfossil groups
(calcareous nannoplankton, dinoﬂagellates, planktonic foraminifera) did experience a turnover of
varying magnitudes in the latest Maastrichtian–earliest Danian. However, many of these turnovers,
along with changes in ecological dominance patterns among benthic foraminifera, began in the latest
Maastrichtian. Improved taxonomic estimates of the overall pattern and magnitude of the K–T
extinction event must await the development of more reliable systematic and phylogenetic data for all
Upper Cretaceous clades.
Keywords: K–T boundary, mass extinctions, taxonomy, phylogeny.
Turnover patterns among dinosaurs, ammonites, calcareous
nannofossils and planktonic foraminifera during the end
Cretaceous extinction event are causes célèbres for advocates
of various mass extinction scenarios (Fig. 1). Rarely, however,
has there been a concerted attempt to survey these as well as
other animal and plant groups to determine whether the
extinction/survival pattern of one group is matched by that of
another. This paper provides a new look at the broader
palaeontological record on either side of the Cretaceous–
Tertiary (K–T) boundary.
Magnitudes of extinction and origination are usually
measured by plotting the ranges of species, genera and families
against time (Raup & Sepkoski 1984). The validity of such
analyses depends on the sample size, completeness of the fossil
record, and the nature of the taxa themselves (that is, whether
they are real evolutionary entities or taxonomic artefacts). If
the record were complete the only problem would be to
recognize species and species lineages. The K–T fossil record is
not complete, however, and although many analytical tech-
niques can be employed to estimate the taxonomic complete-
ness of particular groups (see Schoch 1986), the extent to
which the record is incomplete is often not known. Some of the
more obvious phenomena that may lead to incorrect identiﬁ-
cation of extinction/origination events are listed below (see
also Smith 1994).
Hiatuses are ubiquitous in the stratigraphical record at all
temporal scales and in all depositional settings (Hedberg 1976;
Ager 1973, 1993). Hiatuses can occur without any lithological
expression (e.g. in condensed sequences, by variable sediment
accumulation rate patterns, or by coarse sampling schemes, see
MacLeod & Keller 1991; MacLeod 1995a,b). A temporary
cessation of deposition in any one area will give the impression
of multiple contemporaneous extinctions followed by multiple
contemporaneous originations (Birkelund & Hakansson 1982),
because true ranges will be truncated at both ends (Fig. 2).
This is particularly pertinent for the K–T boundary contro-
versy since many of the classic sequences contain missing
intervals (MacLeod & Keller 1991; Kennedy 1993; MacLeod
The presence of unusually good preservation with many iden-
tiﬁable species will enhance the species diversity for brief
time intervals, giving the impression of sudden origins at the
beginning of the Lagerstätten interval, followed by severe
extinctions at its end.
Reworking of fossils
Fossils and/or clasts containing fossils that are reworked from
older into younger sediments, and burrows ﬁlled with younger
sediments, will give a false impression of an extended range by
a ‘delayed’ time of extinction or ‘early’ time of origination.
Examples of both have been reported from K–T successions.
Non-biotic vertical mixing (Berger & Heath 1968) can have
the same eﬀect, though mixing models indicate that it would be
conﬁned to a few centimeters and recognizable by a smooth
and exponentially declining decay curve above the true last
Circular arguments for dating
If rocks are dated using particular index fossils then the
validity of the age assignment for the rock unit cannot be used
to date unambiguously origination or extinction events for
those fossils. For example, if dinosaurs or ammonites are
regarded as being conﬁned to Mesozoic strata, then any
dinosaur-bearing or ammonite-bearing unit will necessarily be
assigned a Mesozoic age regardless of its true (e.g. possible
Tertiary) chronological age.
Emigration–immigration and facies change
The movement of animals into or out of an area will give a
signal of only local signiﬁcance. Such movements may be
brought about by facies changes reﬂecting changing local
conditions. In theory this might be detected by ﬁnding the
fossil in younger deposits elsewhere, but this depends on a high
degree of temporal completeness.
If the fossil record of any particular species or group is poor
and patchy, or if a good fossil record is coarsely sampled, we
can have little conﬁdence that observed occurrences truly
reﬂect past events. Furthermore, data reproducibility exper-
iments, carried out by distributing similar rock samples con-
taining planktonic foraminifera to a variety of scientists, have
shown large degrees of variation in reported faunal compo-
sition (see Keller et al. 1995; MacLeod 1996a), suggesting that
either the distributions of microfossils may be more spatially
heterogeneous than previously suspected, or that reports of
Fig. 1. Theoretical patterns of mass
extinction in the fossil record. Three
end-member patterns indicate diﬀerent
causal mechanisms (e.g. progressive
means relatively long term, moderate
intensity environmental change;
catastrophic is a single, short-term, high
intensity environmental change; stepwise
is multiple short-term moderate-high
intensity episodes of environmental
change). However, these three models
are not mutually exclusive. A
combination of end-member models may
more accurately describe K–T extinction
patterns in the stratigraphical record (see
Figs 3, 10) and, in turn, may imply the
operation of multiple causal mechanisms.
Fig. 2. Potential bias of biostratigraphical data towards a
catastrophic faunal turnover pattern via the presence of an
unrecognized hiatus. (a) Hypothetical distribution of
biostratigraphical ranges for taxa A–N across an event horizon;
y-axis=time. (b) Distribution of the same datums as they would
appear in the stratigraphical record; y-axis=stratigraphic position. If
the event horizon lies within an interval of non-deposition or
post-event erosion, the stratigraphical distribution of relatively
short-ranging taxa may be disrupted so as to appear to terminate
(and originate) abruptly. Compare this diagram to ﬁg. 1 of
Bramlette (1965). All of the sections used by Bramlette (1965) to
infer an abrupt K–T turnover pattern for marine plankton are now
known to contain boundary hiatuses of up to tens of thousands of
years in duration (see MacLeod & Keller 1991, and references
266 N. MLEOD ET AL.
faunal compositions may be strongly biased by diﬀerent
processing methods, species concepts, or other factors.
Species named solely because they occur at a certain level
(biostratigraphical species) or in a certain area (biogeographi-
cal species) will distort true species ranges. There are instances
where the same morphotaxon is given diﬀerent names simply
because individuals lived on either side of a major stratigraphi-
cal boundary. One example, among sharks, is the common
practice of changing the generic name Cretolamna to Otodus
according to which side of the K–T boundary the specimen
The use of genera and families as units of analysis can distort
the fossil record since unnatural groups may be treated as
though they had the same evolutionary signiﬁcance as natural
(monophyletic) groups (Fig. 3). This has been demonstrated
most clearly for ﬁshes and echinoderms by Patterson & Smith
(1987, 1989) and Smith & Patterson (1988). These authors were
concerned with the taxonomic quality of the data used by
Raup & Sepkoski (1984) which had been claimed to support
extinction periodicity. Patterson & Smith (1987, 1989) showed
that 58% of the ﬁsh and echinoderm families which had been
claimed to show extinction periodicity were either polyphyletic
(the members of the family did not share an immediate
common ancestor), paraphyletic (the group shared a unique
common ancestor but not all members were included), or
monospeciﬁc. These non-monophyletic groups are phylo-
genetic mistakes and cannot be used as evidence of extinction
(see Discussion). Moreover, Patterson & Smith (1987, 1989)
showed more pseudoextinctions in their genus-level analysis.
The following are brief reports of turnover patterns in a
variety of microfossil, invertebrate macrofossil and vertebrate
macrofossil groups across the K–T boundary, in which the
possible biasing mechanisms described above have been given
due consideration. These summaries represent the state of the
K–T fossil record for the groups discussed. Any hypothesis or
scenario seeking to explain or otherwise account for the
K–T mass extinction should provide mechanisms consistent
with these patterns of long and short-term extinction and
The ﬂoral turnover across the K–T boundary is the most
dramatic feature in the fossil record of calcareous nannoplank-
ton (comprising coccoliths, nannoliths and calcispheres). These
groups have been studied in detail from virtually all known
marine boundary successions. Moreover, the record of nanno-
ﬂoral change across the boundary is well-established at the
species level (Fig. 4).
Syntheses are provided by Thierstein (1981), Gartner (1996)
and Pospichal (in press). Percival & Fischer (1977) divided the
Maastrichtian and Danian nannofossil ﬂora into vanishing
species, persistent species and incoming species. Precise num-
bers of species in these categories depends on subjective species
concepts, but using mid-range concepts, and excluding calci-
spheres, we recognize 80 vanishing Maastrichtian species, 10
K–T survivors, and about 30 incoming Danian species.
The vanishing species are those that occur in the topmost
Maastrichtian but do not occur consistently in Danian strata
and do not give rise to obvious Tertiary descendants. Members
of this group may form >90% of Maastrichtian assemblages in
most deep-sea cores from this interval and include numerous
long-ranging species. There does not appear to be any fore-
shadowing of the K–T event by nannofossil extinctions in
the Maastrichtian. Species-level Maastrichtian nannofossil
extinctions do occur, but there is no evidence of an increase in
Fig. 3. Distinctions between monophyletic and paraphyletic taxonomic groupings in the study of mass extinction as seen on a cladogram (a) and
in an equivalent phylogenetic tree (b). A monophyletic group is a set of taxa that uniquely share a homologous, derived character (Smith 1994).
Taxa C–E all share a unique character (the grey symbol) or its derived homologue (the black symbol). A subsidiary monophyletic group
composed of taxa D and E can also be recognized. A monophyletic group includes the ancestral taxon of a lineage and all its descendents. A
paraphyletic group contains an ancestral taxon and some, but not all of its descendants. Paraphyletic groups are usually deﬁned by a
combination of present and absent (=negative evidence) characteristics. Monophyletic taxa are part of the evolutionary process and possess a
unique history that transcends taxonomic convention. Paraphyletic taxa are artiﬁcial in the sense that their deﬁning attributes are matters of
systematic convention and not an intrinsic part of evolutionary history. Stratigraphical ranges, taxonomic richness values, and extinction rate
estimates for higher taxonomic groups should be constructed on the basis of monophyletic sets of taxa in order to avoid bias due to an
imperfectly observed fossil record and/or inappropriate taxonomic practice.
THE K–T BOUNDARY TRANSITION 267
268 N. MLEOD ET AL.
the extinction rate towards the K–T boundary (Pospichal
1994), nor is there any strong evidence of anomalous assem-
blages immediately prior to the boundary (Pospichal 1994;
Ehrendorfer 1993; Gartner 1996).
Representatives of the vanishing ﬂora (the relative number
of which depends upon one’s interpretation, see below) occur
in the lowermost Danian and may dominate particular assem-
blages. In most sections these vanishing species show a rapid
decline in abundance over a lowermost Danian interval of a
few tens of centimetres to a few metres, typically correspond-
ing to the ﬁrst 20 000–80 000 years of the Danian (Fig. 4;
Pospichal 1995). During their decline the relative abundances
of each species within the assemblage of vanishing species
remains essentially constant and, despite extensive eﬀorts,
consistent last appearance datums within the Danian have not
been identiﬁed for any of these taxa.
Various explanations for these Danian occurrences of
Maastrichtian nannofossils have been oﬀered. Survivorship of
these taxa into the Paleocene is an obvious and simple expla-
nation. The main problems with this scenario are (1) if the
vanishing group as a whole survived the apparently anomalous
conditions of the earliest Danian ocean, why did they then fail
to recover when more normal conditions were resumed (by
Zone NP2; Gartner 1996), (2) why did members of this large
group of species not have widely varying, biogeographically
related distributions in the (presumably) highly variable early
Danian ocean (as the surviving and incoming species did), and
(3) why do they not exhibit clear last appearance data?
These questions have led many nannofossil palaeontologists
to entertain the alternative scenario of a global catastrophic
event, which rapidly (perhaps over 1–10 years) killed oﬀmost
of the nannoﬂora, leaving only a few survivor species. This
alternative scenario requires an explanation for the presence
of Cretaceous specimens above the boundary. Possibilities
include bioturbation, reworking, and mixing of remobilized
material. Such mechanisms are entirely consistent with knowl-
edge of and interpretations of other parts of the nannofossil
record. However, deﬁnitive evidence may never be found
to prove whether or not the Cretaceous specimens in the
lowermost Danian are in situ.
Persistent (or survivor) species are Upper Cretaceous species
regarded as occurring in situ in the Danian, because they
exhibit high relative abundances, consistent occurrences in
Danian strata, and/or represent the inferred ancestors of
Danian species. These species dominate nannofossil assem-
blages between the decline of the Cretaceous ﬂoras and evolu-
tion of incoming Danian species. Most of these taxa do not
belong to the major coccolith or nannolith families but to
evolutionarily conservative taxa of uncertain aﬃnities (miscel-
laneous taxa of Fig. 4). These groups are usually rare (indeed
only sporadically present), but become much more abundant
in the lowest Danian. By contrast only a very few of the more
common Cretaceous taxa persist. Calcispheres are exception-
ally abundant in the survivor assemblages. Kienel (1994)
suggested that a major turnover occurs within this group, but
in the lower Danian rather than at the K–T boundary.
The ﬁrst incoming species occur shortly above the boundary
in expanded sections (e.g. El Kef, Tunisia) and then give rise
to easily traceable lineages (Fig. 4). The origins of these
new Paleocene lineages are not easy to determine. This is partly
due to a lack of careful phylogenetic research in this interval,
but also reﬂects the facts that the lowest Danian species are
all very small and that the new lineages have structures that
diﬀer very signiﬁcantly from those of any obvious Mesozoic
The K–T dinoﬂagellate record is diﬃcult to interpret with
conﬁdence because only a small proportion of modern dino-
ﬂagellate species, and by inference K–T species, produce
fossilizable organic-walled cysts. While ancient cyst-forming
species have provided palynologists with a fossil record that
can be used for correlation and palaeoenvironmental analysis,
the K–T dinoﬂagellate fossil record most likely remains an
underestimate of the total number dinoﬂagellate species that
actually existed during K–T time. Consequently, it may be
highly misleading to infer turnover patterns among ancient
ﬂoras on the basis of data from cyst-forming species alone.
Details of the K–T dinoﬂagellate record are best known
from sections at El Kef, Tunisia (Brinkhuis & Zachariasse
1988) and Seymour Island, Antarctica (Elliot et al. 1994; Askin
& Jacobson 1996). In both successions, ﬁrst and last appear-
ances of dinoﬂagellate cysts are not conﬁned to single levels or
restricted intervals near the boundary, but occur throughout
the uppermost Cretaceous and lowermost Tertiary in a broadly
progressive pattern. The turnover rate does increase in the
uppermost Maastrichtian (close to the K–T boundary), but
this interval precedes the emplacement of impact debris in all
studied sections and is probably better correlated with the
latest Maastrichtian eustatic sea-level rise (see Haq 1991). A
pronounced decline in marine palynomorph species in this
uppermost Maastrichtian interval occurs on Seymour Island
(Askin & Jacobson 1996).
Diatoms are microscopical unicellular algae whose valves are
made of opaline silica. Because of the instability of this form of
silica, diatoms are highly susceptible to dissolution, recrystal-
lization, and/or other diagenetic processes. The presence of
diatoms diminishes down the geological column, and in
pre-Eocene sediments they are known from only a handful of
localities worldwide (see Harwood & Gersonde 1990).
Living diatoms are found in a wide variety of freshwater and
marine settings, but there are no known occurrences of dia-
toms in Mesozoic freshwater facies. Until recently, the lack of
diatomaceous K–T successions led to wildly varying estimates
of diatom survival across the K–T boundary. These estimates
ranged from 13% to 100% survivorship, and were based
Fig. 4. Calcareous nannofossil record across the K–T boundary. The
diagram includes most common low-latitude species. High latitude
assemblages show similar patterns, but with some diﬀerent species.
Numerous rare species, particularly disappearing Cretaceous species,
are not shown. Species are grouped into families. For each
Cretaceous family, estimates are given of the number of species in
the topmost Maastrichtian interval and the number of these taxa
that deﬁnitely survive into the Danian (e.g. 10 spp., 0 survivors). The
pattern of decline in abundance of questionably reworked
disappearing species in the Danian varies between sections.
Gradualistic evolutionary transitions within some lineages are
indicated by placing the species very close together. Arrows indicate
other, less certain, evolutionary transitions. Magnetostratigraphy
and biostratigraphy from Berggren et al. (1995). Nannofossil data
from Heck & Prins (1987), Pospichal (1995, in press), Perch-Nielsen
(1979), Romein (1979), Varol (1989) and personal observations.
THE K–T BOUNDARY TRANSITION 269
primarily on presence/absence of Upper Cretaceous species in
Paleocene ﬂoras (Wornardt 1972; Strelnikova 1975; Jouse
1978; Fenner 1985). However, there was general agreement
that the characteristic Upper Cretaceous genus Gladiopsis
became extinct at or just below the K–T boundary. Harwood
(1988) described a diatom ﬂora from the K–T succession
on Seymour Island, Antarctica. This high-latitude locality
contains the only diatomaceous K–T boundary succession
presently known. From that succession, Harwood (1988)
reported 84% of Upper Cretaceous diatom species surviving
into Tertiary strata.
High diatom survivorship rates across the K–T boundary
event horizon were predicted by Milne & McKay (1982) and
Kitchell et al. (1986) who suggested that the ability of some
diatom species to form dormant resting spores during times
of environmental stress (especially during low nutrient/light
conditions) enabled these species to survive a short-term
catastrophic event. Diatom resting stages are readily recogniz-
able in the fossil record (Hargraves 1986; Kitchell et al. 1986),
and Harwood (1988) noted a great increase in their relative
abundance at and above the K–T boundary on Seymour
Island. However, such increases in resting spore frequencies
are known from other stratigraphical levels. Accordingly, the
hypothesis that an increase in diatom resting spores in the
Seymour Island section might have resulted from local (i.e.
non-impact related) environmental factors cannot be ruled
On a broader scale, a survey of Santonian to lower Eocene
diatom assemblages, undertaken by one of the authors (P.C.),
indicates that out of 447 valid Upper Cretaceous species, only
203 species (46%) survive to the Upper Palaeocene and Eocene.
Since only a single diatomaceous K–T boundary succession is
known, this ﬁgure cannot be taken as indicative of the magni-
tude or pattern of diatom extinctions across the K–T bound-
ary. However, it does suggest that a substantial, though by no
means catastrophic, turnover among diatom species took place
within the Late Cretaceous–early Palaeogene.
A preliminary analysis of these Santonian–lower Eocene
data also shows strong discrepancies in survivorship rates
among planktonic and benthic diatom taxa. Planktonic genera
exhibit substantially higher survivorship rates (50–90%) rela-
tive to benthic genera (10–40% survivorship). Since the ability
to form resting stages is more common among modern plank-
tonic (but not benthic) diatom species (Hargraves & French
1983), this might explain the high survivorship rates reported
by Harwood (1988) for the predominantly planktonic
Seymour Island ﬂoras (see also Kitchell et al. 1986). However,
modern marine diatom resting stages have not been known to
remain viable for more than two years (Hargraves & French
1983). This suggests that if resting stages were the mechanism
used by the Seymour Island diatoms to survive the K–T event,
that event must have had a very short duration. Regardless, a
short-event scenario is inconsistent with the high relative
frequency of diatom resting stages throughout the lowermost
Danian on Seymour Island (Harwood 1988).
Radiolaria have a geological record dating from at least
Ordovician times. The mineral skeletons or tests of polycystine
radiolarians are secreted by the organism following uptake of
monosilic acid from seawater and its transformation into
amorphous silica (Opal-A). After death, the skeletal test can be
fossilized in a series of mineral phases from Opal-A through
Opal-CT (cristoballite/trydimite) to quartz, depending on
diagenetic history (mainly time, temperature and pressure
conditions). Due to the high susceptibility to dissolution of
the initial and transitional silica phases in both marine and
connate waters, the fossil record of radiolaria is extremely
patchy, both stratigraphically and geographically.
Very few studies of radiolarian faunal transitions across
the K–T boundary, or of highest Maastrichtian and lowest
Paleocene radiolarians, have been published. These include:
Frizzel & Middour (1951), Borisenko (1958, 1960), Foreman
(1968), Dumitrica (1973), Kozlova (1983), Nishimura (1986,
1992), Goltman (1988), Ling (1991), Hollis & Hanson (1991),
Hollis (1993, 1996) and Strong et al. (1995). Of all these publi-
cations, only two authors claim to have studied radiolarian
assemblages from either questionably complete (Foreman
1968) or complete (Hollis 1993; Hollis in Strong et al. 1995)
Foreman (1968) concentrated on the analysis of upper
Maastrichtian Nassellaria from four localities in the Moreno
Formation in California. She considered ﬁve identiﬁable
species to have survived from the late Maastrichtian to the
early Paleocene: Amphipyndax stocki,Cornutella californica,
Dictyomitra cf. multicostata,?Staurodictya fresnoensis and
Stichomitra alamedaensis. These species were recovered
throughout a core section of c. 6 m in association with other
undescribed forms of deﬁnite Cretaceous character, together
with a predominantly ?Palaeocene radiolarian assemblage.
Foreman (1968) did not address the problems of possible
reworking or sedimentary gaps. Biostratigraphical control was
also based on a marked change in dinoﬂagellates and spores at
the same level as the radiolarians in this series of samples as
recorded by Drugg (1967).
Hollis (1991, 1993) investigated ﬁve onshore sections from
New Zealand, and re-examined faunas from the ﬂanks of Lord
Howe Rise previously examined by Dumitrica (1973) from
coeval strata at DSDP Site 208. The New Zealand rocks are
siliceous limestones, mudstones and cherts whereas lithologies
from DSDP Site 208 are primarily chalks and cherts. An
iridium-rich boundary clay layer is reported in three of the
onshore New Zealand sections. Biostratigraphical age control
is provided by foraminifera, dinoﬂagellates and calcareous
nannofossils, and Hollis (pers. comm. 1996) discounts the
possibility of reworking on the grounds of assemblage compo-
sition and preservational state of individuals together with
sedimentological evidence. Hollis (1993) found no evidence for
mass extinctions of radiolarians at the K–T boundary. Forty-
two of the 45 taxa identiﬁed by Hollis (1991) from Cretaceous
rocks had reliable Palaeocene occurrences. Reference to ranges
of Cretaceous–Tertiary taxa in Hollis (1991, 1993) suggests
that a faunal turnover in radiolarians occurred in the mid
Palaeocene (at least in southern high latitudes), in contrast to
the earlier major ﬂoral turnovers experienced by calcareous
nannoplankton at the K–T boundary.
Although the K–T boundary in these sections does not
mark a faunal discontinuity, it was a time of changing compo-
sition among siliceous plankton. In particular the frequency
of radiolarian-rich samples, radiolarian/diatom ratios,
nassellarian/spumellarian ratios and Si/Ca ratios all increase
across the K–T boundary (Hollis 1991, 1993). These compo-
sitional data suggest that marine productivity increased in
the southern high latitudes and contradict predictions of the
Strangelove Ocean model (Hsü & MacKenzie 1985). Ad-
ditional support for high phytoplankton productivities in the
southern high latitudes across the K–T boundary comes from
270 N. MLEOD ET AL.
the isotopic work of Stott & Kennett (1990), and show that
they may have been related to climatic cooling (Stott &
Kennett 1990; Hollis 1996).
The idea that foraminifera suﬀered a pronounced extinction
across the K–T boundary dates back to the 1930s (e.g.
Glaessner 1937). Abrupt turnover in planktonic foraminiferal
faunas at the K–T boundary began to be mentioned in the
late 1950s (e.g. Troelson 1957; Reiss 1957). In introducing
Schindewolf’s (1962) concept of neocatastrophism (‘Neo-
katastrophismus’), Newell (1962) identiﬁed the K–T boundary
as one locus of mass extinction in Earth history and identiﬁed
planktonic foraminifera as one of the principal victims of this
The abrupt turnover of the planktonic foraminifera across
the K–T boundary in many sections was cited by Bramlette
(1965) as evidence that this event took place on a much shorter
time scale than the Permo-Triassic mass extinction. Bramlette
(1965) acknowledged that a stratigraphical hiatus was present
in most K–T boundary sections, thus imparting the appear-
ance of an instantaneous biotic turnover to those records, but
speculated that the K–T boundary hiatus ‘may involve many
thousands of years but probably [represented] much less than a
million . . .’ (Bramlette 1965, p. 1697). Interestingly, much of
the early literature on the K–T planktonic foraminiferal mass
extinction also contains anecdotal accounts of Cretaceous
species in Danian strata (e.g. Troelson 1957).
The 1980 proposal of the asteroid impact scenario (Alvarez
et al. 1980) resulted in renewed interest in the planktonic
foraminiferal record across the K–T boundary. Detailed
sampling of several classic K–T boundary localities (e.g.
Gubbio, Italy; El Kef, Tunisia; Brazos River, Texas), and the
discovery of new K–T boundary successions (e.g. Agost and
Caravaca in Spain), has forced a re-evaluation of the K–T
planktonic foraminiferal record. In particular, Keller (1988b,
1989, 1993; Canudo et al. 1991; Keller & Benjamini 1991)
reported many species previously thought to be conﬁned to the
Upper Cretaceous, associated with undoubtedly Tertiary
faunas above the K–T boundary impact debris layer.
Subsequent work by independent investigators (e.g. Liu &
Olsson 1992) and a heretofore unprecedented biostratigraphi-
cal blind test of the El Kef stratotype succession (see Keller
et al. 1995; MacLeod 1996a,c; Ginsburg et al. in press) have
conﬁrmed Keller’s observations. It appears that planktonic
foraminiferal extinctions occurred prior to, at, and well after
the K–T boundary event on a global scale. Olsson & Liu (1993)
and Liu & Olsson (1992) have argued that the presence of
Cretaceous morphotypes in Danian rocks is the result of
reworking. To assess this interpretation, Keller’s data have
been extensively tested for both survivorship, reworking
(Barrera & Keller 1990; Schmitz et al. 1992; MacLeod & Keller
1994; MacLeod 1995a,b,c, 1996a,b,d) and the Signor-Lipps
eﬀect (MacLeod 1996a,b,d; see Discussion). Results of these
tests have revealed patterns of isotopic, geographical, morpho-
logical and stratigraphical variation that are inconsistent with
reworking or a signiﬁcant Signor-Lipps eﬀect within these
data, but consistent with Keller’s original interpretation of
widespread planktonic foraminiferal survivorship into the
MacLeod & Keller (1991) combined planktonic foramini-
feral and calcareous nannoplankton data from thirteen bio-
stratigraphically complete marine K–T boundary successions
to determine which of these sections and cores contained
boundary hiatuses, and if possible, to estimate the location
and duration of these hiatuses. Results showed that previous
data used to support models of catastrophic faunal and ﬂoral
turnover across the K–T boundary, and the physical evi-
dence used to associate a bolide impact with that turnover,
have often come from boundary successions that contained
unrecognized hiatuses (e.g. Gubbio, Italy; Stevns Klint,
Denmark). Other sections (e.g. Brazos River, Texas; Nye
Kløv, Denmark), exhibiting a more progressive turnover
pattern, were shown to be chronostratigraphically complete
across the boundary. Comparative stratigraphical analysis of
these sections and cores (MacLeod & Keller 1991; MacLeod
1995a,b, 1996a, b, d) indicates that the temporal and environ-
mental distribution of marine boundary hiatuses are consistent
with predictions of sequence stratigraphical models for the
rapid eustatic sea-level rise that is known to have taken place
during the Cretaceous–Tertiary transition (Haq et al. 1987;
Haq 1991). A summary of global biostratigraphical data
from MacLeod & Keller (1991) and MacLeod (1995a,b,
1996a) reveals a broadly progressive pattern of planktonic
foraminiferal extinction and origination through the latest
Maastrichtian and earliest Danian (Fig. 5).
From a biogeographical point of view (MacLeod & Keller
1994), these data suggest that the K–T planktonic foramini-
feral turnover took place in a series of waves that removed
taxonomically and geographically uniﬁed components of the
Late Cretaceous fauna (Fig. 6). The ﬁrst of these waves began
in the latest Maastrichtian (Plummerita hantkeninoides Zone)
and primarily aﬀected large, highly ornamented, intermediate
and deep-dwelling globotruncanoid species in the low and
middle latitudes. Survivors of this extinction pulse appear to
have crossed the K–T boundary horizon without suﬀering
signiﬁcant additional loss in species richness. Local species
extinctions do occur at the K–T boundary for many
Cretaceous species; however, when appropriate, high-
resolution, global correlations are made, and the composite
record of K–T planktonic foraminiferal turnover assembled
(MacLeod & Keller 1991; MacLeod 1996a), no Cretaceous
species can be assigned unambiguously to an extinction
horizon coincident with the K–T boundary.
The planktonic foraminiferal K–T survivor fauna persisted
some 40 000–100 000 years into the Tertiary after which it
suﬀered a second wave of extinctions that removed Cretaceous
survivors in the low and middle latitudes. Throughout these
ﬁrst two extinction waves high-latitude planktonic foramini-
feral faunas appear to have suﬀered relatively little extinction.
However, all but a few of these high latitude Cretaceous
survivor populations disappear from the stratigraphical record
in overlying Danian biozones. The incoming fauna of Tertiary
planktonic foraminiferal species occupied a very small portion
of overall Tertiary species diversity in the earliest Danian, but
expanded to replace Cretaceous survivor species as these
became extinct in the low and middle latitudes in the early
Danian and later in the high latitudes. This temporally ex-
panded and geographically structured interval of planktonic
foraminiferal extinctions seems strongly suggestive of a long
period of environmental instability across the K–T boundary,
perhaps driven by relatively long-term eustatic sea-level/
climatic ﬂuctuations, with some contribution from relatively
short-term volcanic and impact-related event(s) at or near
the boundary horizon itself, coupled with post-boundary
competition from radiating Danian lineages.
THE K–T BOUNDARY TRANSITION 271
Agglutinated benthic foraminifera
The record of agglutinated benthic foraminifera exhibits few
species extinctions in the Late Cretaceous, though deep-water
agglutinated benthic foraminiferal assemblages do appear to
exhibit a local shift to dominance by opportunistic species
during the ﬁrst few hundred-thousand years of the Palaeocene.
In acid residues from the Gubbio sections (Central Italy), a
signiﬁcant decrease in agglutinated foraminiferal abundance
and diversity, and a shift in community structure, are observed
across the K–T boundary. Upper Maastrichtian assemblages
are also dominated by epifaunal suspension and detritus
feeders whereas the lower Danian boundary clay is dominated
by infaunal species that were probably bacteriovores. These
changes may have been a consequence of the signiﬁcant
decrease in the ﬂux of organic matter to the sea ﬂoor associ-
ated with the K–T boundary event.
Tubular agglutinated species (e.g. Rhizammina) alone show
an 80% reduction in numbers beginning 5 cm below the
boundary (Fig. 8). The ﬁrst genus to increase in numbers
above the boundary is Reophax (Event 1 in Figs 7 and 8),
a form known to colonize modern deep-sea environments
following disturbances by strong currents (Kaminski 1985;
Kaminski et al. 1988) and ash falls (Kuhnt & Hess 1994). It is
possible that organic matter scavenged from the water column
by settling ﬁne-grained particles may have been able to sustain
a bloom of small infaunal Reophax.
Following the event, successive maxima of two other in-
faunal taxa occur within Zone P1a; Spiroplectammina ex. gr.
dentata and S. ex. gr. spectabilis, corresponding to events 2 and
3 in Figs 7 and 8). These assemblages may represent periods of
increased organic ﬂux associated with erratic phytoplankton
blooms within an early Palaeocene period of generally low
productivity. Similar patterns of abundance increase have
recently been observed at two other K–T boundary sections in
Spain (Kuhnt & Kaminski 1993; Coccioni & Galeotti 1994).
An increase in Spiroplectammina above the K–T boundary
has also been reported from the ﬂysch sediments of the
Alpine Gossau in central Asia (Peyrt et al. in press). The
Spiroplectammina biofacies, which appears to characterize the
upper part of Zone P0 and the lower part of Zone P1a, seems
an especially useful lowermost Danian marker horizon.
The timing and structure of the agglutinated benthic
foraminiferal recovery undoubtedly depended on deposition
of organic matter from the mixed layer. The succession of
biofacies across the K–T boundary is consistent with a
model of strongly reduced primary productivity in the earliest
Palaeocene. In an early Palaeocene lower and middle latitude
Strangelove Ocean, suﬃcient nutrients may have been avail-
able in the mixed layer to sustain large blooms once the
phytoplankton ﬂora re-established itself after the K–T bound-
ary event. It is likely that export production from occasional
phytoplankton blooms provided the nutrients needed to sup-
port agglutinated benthic foraminiferal assemblages. Once the
ﬂux of organic matter became more substantial, a high-
productivity adapted community of r-selected infaunal forms
began to thrive. The initial drop in agglutinated benthic
foraminiferal abundance and diversity appears to have taken
place in the late Maastrichtian (Fig. 8). Ultimate recovery of
agglutinated benthic foraminiferal community structure during
the early Palaeocene occurred in several stages lasting several
hundred thousand years (Kuhnt & Kaminski 1993). This
Fig. 5. Spindle diagrams of observed
changes in Cretaceous and Tertiary
planktonic foraminiferal species richness
across the K–T boundary (based on a
synthesis of biostratigraphical data from
El Kef, Tunisia; Agost, Spain; and
Brazos Core, Texas; see MacLeod
1996a). Positions in the sequence
represent binned (10 cm) intervals along
the K–T Composite Standard Reference
Section (K–T CSRS) for these three
successions. Note that in terms of the
observed data, the Cretaceous planktonic
foraminiferal fauna loses approximately
one-third of its Maastrichtian species
richness in the uppermost Maastrichtian
P. hantkeninoides Zone, and that the
decline in species richness beginning in
this interval carries through the K–T
boundary horizon without any signiﬁcant
deﬂection. The most dramatic reductions
in Cretaceous species richness in these
sections occur at the top of Tertiary
Zone P1a almost 300 k.y. after the K–T
boundary. Analysis of the original
biostratigraphical data (MacLeod 1996a)
suggests that, in these three sections, the
pre-boundary decline in planktonic
foraminiferal species richness and the
post-boundary occurrence of Cretaceous
morphotypes cannot be accounted for by
the Signor–Lipps eﬀect and reworking
272 N. MLEOD ET AL.
perturbation of a component of the benthic foraminiferal
fauna may have been linked to the evolution and recovery
of the phytoplankton that are the source of nutrients for
Calcareous benthic foraminifera
Lipps & Hickman (1982) considered that shallow water ben-
thic foraminifera underwent a mass extinction at the K–T
boundary whilst deep-sea foraminifera were generally un-
aﬀected. Miller (1982), however, concluded that neither shallow
nor deep-dwelling species exhibit a pattern of mass extinction.
More recent investigations of Tethyan faunas (Keller 1988a;
Coccioni & Galeotti 1994) and faunas from the southern
Atlantic (Widmark & Malmgren 1992) and high southern
latitudes (Thomas 1990) agree with the latter interpretation,
though Speijer & Van der Zwann (1996) have argued for an
explicitly impact-related mass extinction.
Keller & Lindinger (1989), working in the relatively shallow
water deposits of El Kef, Tunisia, described relatively warm
temperatures followed by sudden global cooling that trig-
gered reduced surface water productivity during the late
Maastrichtian, and led to reduced calcareous benthic fora-
miniferal diversity. These authors considered this productivity
low to continue (along with associated unstable environmental
conditions) some 300 000 to 400 000 years into the Tertiary.
Fig. 6. Schematic diagram of diﬀerential geographical responses of the global planktonic foraminiferal fauna to the K–T mass extinction event.
There is no scale on the x-axis and the widths of the bars are not meant to imply that either Cretaceous or Tertiary planktonic foraminiferal
species richnesses were the same in all biogeographical provinces. Figure based on results of MacLeod & Keller (1994).
Fig. 7. Abundance of important groups
of deep-water agglutinated foraminifera
across the K–T boundary in the
Bottaccione section near Gubbio, Italy.
The scale bar along the vertical axis is
divided into 10 cm intervals. Numbers
alongside the section represent bed
numbers. Position of the K–T boundary
is at the level of 0.0 on the right-hand
scale. All abundance values are given as
individuals >63 ìm per gram of
THE K–T BOUNDARY TRANSITION 273
Ortiz and Keller (pers. comm.) have also suggested that a
short-term episode of global warming, sea-level rise, and
expanded oxygen minimum zone conditions on the continental
shelf, resulted in decreased abundances of species in many
sections, along with a switch in ecological composition that
favoured infaunal, low oxygen-tolerant species. Epifaunal and
low oxygen intolerant species presumably migrated to refugia
during this time because these Cretaceous species reappear in
overlying Danian sediments. Speijer & Van der Zwann (1996),
however, speculated that these changes in the El Kef assem-
blage were the result of climatic feedback mechanisms trig-
gered by bolide impact. In an analogous study of the Caravaca
K–T boundary section in southern Spain, Coccioni & Galeotti
(1994) argued that the reduction in diversity coincided with a
decrease in the rate of organic ﬂux to the ocean ﬂoor via the
Strangelove Ocean model.
Most likely, both oxygen minimum zone expansion and
productivity decrease played a role in this benthic foramini-
feral ecological event. It should be noted though, that while
ﬂuctuations in the oxygen minimum zone are commonly
associated with sea-level changes and have been known to
exert a long-term eﬀect on benthic foraminiferal populations
throughout the last 100 million years, there is no known
mechanism for reducing the amount of organic carbon reach-
ing the sea ﬂoor by eliminating a large part of the marine
phytoplankton population and then keeping that population
depressed for geologically signiﬁcant time periods (75 000–
80 000 years), the Strangelove Ocean model. Indeed, phyto-
plankton doubling times are on the order of hours to days and
the direct eﬀects of a bolide impact (e.g. darkness, acid rain)
are estimated to last for only a few months to (at most) a
few years (see D’Hondt 1995). While interpretations of the
causes of the faunal dominance changes observed in upper
Maastrichtian benthic foraminiferal assemblages will no doubt
continue, the bulk of the evidence indicates that the latest
Cretaceous was not a period of widespread calcareous benthic
Larger benthic foraminifera
The so-called larger foraminifera are benthic taxa that develop
complex and characteristic internal skeletons. Skeletally, they
are the most elaborate foraminifera and their phylogenies are
usually well established. Modern larger benthic foraminifera
occur world-wide, but in the tropical–subtropical belts only. It
is in regions that had palaeoclimates similar to these that the
larger foraminifera provide evidence of variations in species
richness or faunal composition across the K–T boundary.
Because the larger foraminifera are benthic and conﬁned to
the shallow seas of the inner continental or island shelves, few
managed to cross the widening Atlantic Ocean. Thus, the
larger foraminiferal faunas of the Americas and Euro-Afro-
Asia were diﬀerent in both the Late Cretaceous and the Early
Tertiary. Many suprageneric groups of larger foraminifera
became extinct in the Late Cretaceous, and others appeared
for the ﬁrst time during the Palaeocene. However, few if any
species made an evolutionary ﬁrst appearance in the Danian or
its equivalents. Many workers believe that the extinctions of
Tethyan lineages occurred virtually simultaneously, though
the chronostratigraphical control necessary to evaluate this
hypothesis is lacking. No obvious environmental, sedimento-
logical, geochemical or isotopic anomalies coincide with the
extinction of any Maastrichtian larger foraminiferal taxon.
There are few analyses of ostracode occurrence patterns across
the K–T boundary that provide reliable data. Perhaps the most
important single K–T ostracode record is that of the El Kef
(Tunisia) boundary stratotype section. Donze et al. (1982) and
Peypouquet (1983) recorded presence/absence patterns of 50
upper Maastrichtian and lower Danian species from this
succession. These data show an interruption in ostracode
occurrence near the boundary horizon itself, but many Lazarus
taxa are re-established subsequently. Peypouquet considered
this pattern to reﬂect expansion of the oxygen minimum zone
at the time of the K–T boundary event. Evidence for oxygen
based on the morphology of Krithe is considered questionable,
especially insofar as Peypouquet made no comment on the
possibility that the rising carbonate compensation depth across
the boundary interval may have inﬂuenced these faunas.
Marine, inner to middle shelf ostracodes from the Brazos
River, Walkers Creek, and Littig Quarry, Texas have been
described by Maddocks (1985). Although these sections may
not be entirely complete across the boundary (e.g. the Brazos
River sections contain an obviously disconformable coarse
Fig. 8. Abundance of important groups
of deep-water agglutinated foraminifera
across the K–T boundary in the
Contessa section near Gubbio, Italy. The
scale bar along the vertical axis is
divided into 10 cm intervals. Numbers
alongside the section represent bed
numbers. Position of the K–T boundary
is at the level of 0.0 on the right-hand
scale. All abundance values are given as
individuals >125 ìm per gram of
274 N. MLEOD ET AL.
clastic unit that has been interpreted by some as a ‘tsunami
deposit’), a progressive pattern of faunal replacement was
An ostracode fauna from sections along the Colville River,
northern Alaska has been described by Marincovich et al.
(1990) and Brouwers & De Deckker (1993). The Arctic Ocean
was largely isolated in the Late Cretaceous–Early Tertiary, and
this is reﬂected in the preponderance of endemic ostracode
taxa in these sections. However, the older beds do contain
Cypridea, a freshwater genus usually regarded as indicative of
the Cretaceous. Nonetheless, a combination of facies changes
reﬂecting the uppermost Maastrichtian marine transgression
(see Haq 1991), along with the endemic nature of this fauna,
make the Colville River ostracode record diﬃcult to evaluate
in terms of its bearing on the ostracode record across the K–T
In the Far East, Ye (1993) reported the presence of an
assemblage of marine shelf ostracode species that implied the
presence of a K–T boundary sequence, but provided no details
as to the nature of the faunal transition. Pang & Whatley
(1990) noted the need to deﬁne the Chinese K–T boundary
in a paper that provides references to ostracode assemblages
and relevant Chinese literature. Musacchio (1990) described
Neuquenocypris, a non-marine genus thought to range from
Aptian to Paleocene from the Neuquen basin. The associated
ostracode fauna was qualitatively described as resembling that
of the Maastrichtian, but with reduced diversity.
In the most comprehensive synoptic study of K–T ostra-
codes presently available, Whatley (1990) recorded 70 generic
originations (representing 678 species originations) and 230
generic extinctions (representing 1372 species extinctions) dur-
ing the Maastrichtian. Coles (1990) reviewed Cenozoic North
Atlantic deep-water and adjacent shelf ostracode faunas, and
found diversity was lower in the Palaeocene than at any other
time in the Tertiary. These reviews lack the precise chrono-
stratigraphical control needed to determine the timing, magni-
tude, and geographical scope of extinctions occurring within
the boundary interval itself. Based on these studies, the entire
Late Cretaceous interval appears to have been a time of major
change for ostracode faunas. Much more work needs to be
done to establish the precise character, ecology and timing of
About half the number of extant scleractinian corals genera
are zooxanthellate (exhibit intracellular symbiosis with dino-
ﬂagellate algae zooxanthellae). These ‘z-corals’ are commonly
colonial and restricted to warm, shallow, photic-zone tropical
waters. The remaining azooxanthellate (‘az-corals’) are com-
monly solitary and collectively eurytopic. This z/az distinction
cuts across formal scleractinian classiﬁcations. Thus, the algal
symbiosis has been lost in some groups and evolved more than
once. Colonial forms of both z- and az-corals contribute to a
variety of tropical reefs and deeper or colder water carbonate
buildups (Stanley & Cairns 1988). The functional, ecological
and biogeographical importance of scleractinian symbiosis has
meant that the z/az distinction pervades most coral studies
explicitly or otherwise, including those pertaining to mass
Unfortunately, there is still no universally applicable way of
knowing whether extinct scleractinian taxa were zooxanthel-
late. Consequently, there is considerable ambiguity in the use
of most terms related to this aspect of fossil coral biology in
the systematic literature. For present purposes, however, these
conceptual distinctions can be conveniently gathered into two
groups. Group I includes Tethyan (in both strict and broad
senses) tropical, reef-dwelling, reef-building, hermatypic, z-like
corals. Group II includes boreal, temperate, non-reefal, deep-
and cold-water, ahermatypic, az-like corals that are geographi-
cally widespread and, in certain instances, overlapping in
distribution with Group I.
Rosen & Turnsek’s (1989) analysis remains the most exten-
sive global study of K–T extinctions in scleractinian genera
and species based on actual taxonomic records. This study
determined that about 60% of Late Cretaceous scleractinian
genera, failed to survive into the Palaeocene. To place this
ﬁgure in context, approximately the same number of sclerac-
tinian coral taxa failed to survive from the Early to Mid-
Cretaceous, and 20–40% of Mid-Cretaceous genera failed to
persist into the Late Cretaceous. All patterns appear statisti-
cally signiﬁcant given current data. Partitioning of these data
into Groups I and II shows the same relative extinction
patterns for both groups through the Cretaceous. However,
extinction intensities are generally greater for Group I than
Group II, especially in the Upper Cretaceous. Of the total
number of Maastrichtian taxa (about 160 species), 97–98%
species and 83% genera are unknown in the Tertiary. These
ﬁgures may also partly reﬂect stratigraphical ‘chauvinotypy’
(Rosen 1988, p. 447). Palaeobiogeographical analysis of
Maastrichtian species yielded a Tethyan cluster dominated by
Group I corals, but no similar diﬀerentiation emerged for the
The widely held belief that post-Cretaceous Group I faunas
were rare, if not absent, until the Eocene is inconsistent with
the long-known Thanetian Ranikot fauna (Duncan 1880).
Recent research (Bryan 1991; Moussavian & Vecsei 1995)
suggests that the Group I hiatus may have lasted only until the
late Danian. In addition, Late Cretaceous Group I corals are
typically rather small (usually <0.5 m), rarely branching and
occur in bedded deposits (e.g. Coates & Jackson 1987; Smith
et al. 1995; Skelton et al. in press), often downslope relative to
rudist facies (e.g. Matteucci et al. 1982; Skelton et al. in press).
Group I corals of the Tertiary to Recent are much larger,
branching forms dominate some assemblages, and buildups
span a range of depths and energy conditions (e.g. Bryan
Although the stratigraphical resolution in the data described
above is coarse, it suggests that K–T extinctions took place at
a progressively increasing rate through the last stages of the
Cretaceous. Recent work in the Oman Mountains has revealed
a rich ‘Gosau-type’ (broadly Santonian) Group I fauna of
Maastrichtian age (Smith et al. 1995) indicating that extinction
rates prior to the Maastrichtian may have been lower, taxo-
nomic richness may have remained higher, and extinctions
may have been more concentrated in the later Maastrichtian
than previously realized. The foregoing patterns also suggest
that Group I corals were more susceptible to K–T extinction
factors than Group II corals (see Rosen & Turnsek 1989).
While this conclusion may seem at odds with the analyses
of Sheehan (1985) and Raup & Boyajian (1988) who found
that ‘reefs [cf. Group I] are not so diﬀerent’, these authors
were addressing relative intensity of extinctions through
time. Present patterns (cf. point 2) accord with theirs, whilst
additionally showing that in terms of the K–T extinctions,
there were also signiﬁcant diﬀerences between Groups I and II.
THE K–T BOUNDARY TRANSITION 275
Explanations compatible with the apparently greater extinc-
tion susceptibility of Group I corals range from a collapse of
algal symbiosis (due to volcanic or meteorite-impact dust, high
nutrient levels, reduced global irradiance), large-scale tectoni-
cally and/or eustatically generated loss of shelf seas, increased
terrigenous input in shallow marine Tethyan habitats, climatic
cooling, climatic ‘over-warming’ (cf. coral bleaching), and the
ever-popular, but surely mistaken (Skelton et al. in press),
notion of competition from rudists (Kauﬀman & Johnson
1988). Algal symbiosis collapse may also explain how the
characteristic might have evolved independently in diﬀerent
coral groups (see above). Perhaps zooxanthellae survived the
K–T boundary in a non-symbiotic state, or in hosts other than
corals. Alternatively, the algal symbionts of the Cretaceous
may have belonged to one or more species that became extinct,
with new symbiotic species arising in the Tertiary. Unfortu-
nately, it is diﬃcult to test such ideas as it is still diﬃcult to
estimate how many Group I corals of Late Cretaceous age
were speciﬁcally zooxanthellate (see above), because there
remains insuﬃcient understanding of living zooxanthellae, and
because zooxanthellae are not preserved as fossils. Neverthe-
less, use of coral data to support the widespread concepts of
ecosystem collapse and post-extinction recovery (in this case of
reef ecosystems) must be tempered by recognition of the
considerable changes that took place in the respective eco-
systems of which Group I corals were a part, on either side of
the K–T boundary.
Species belonging to two orders of bryozoans (Cheilostomata
and Cyclostomata) are present in Cretaceous and Palaeocene
marine shelf sediments. However, poor knowledge of their
systematics and stratigraphical distributions constrain inter-
pretations of bryozoan extinction and survival across the K–T
Maastrichtian bryozoan faunas are known from several
parts of the world, but Palaeocene faunas are much rarer. The
only regions where bryozoans are found on both sides of the
K–T boundary are Denmark and possibly the Majunga basin
in Madagascar (Brood 1976). Bryozoans have not been re-
corded from such classic K–T boundary sections as Brazos
River (Texas), Gubbio (Italy), El Kef (Tunisia), and Zumaya
Bryozoans are the dominant macrofossil group in the
Danish Maastrichtian and Danian. They occur in chalks,
including the grey Maastrichtian chalk of Stevns Klint where
their skeletons are carbon stained and rich in iridium (Hansen
et al. 1987), and in bryozoan limestones that often accumu-
lated as mounds on the sea-bed (Thomsen 1976). More than
500 species are present (Hakansson & Thomsen 1979), but
nothing has yet been published on species ranges across the
Hakansson & Thomsen (1979) have summarized unpub-
lished data from the Nye Kløv section. These authors found
species diversity to be high in the white chalk at the top of the
Maastrichtian (where c. 70 cheilostome species were found),
dropping to four species in the basal Danian marl, and
recovering to a peak of more than 40 species in a bryozoan
limestone approximately 6 m above the K–T boundary
before declining again in the overlying Danian pelagic chalk.
Of 115 cheilostome species, only 11 are found in both the
Maastrichtian and Danian parts of the sequence at Nye Kløv.
However, this very low survivorship ratio (9.6%) is a local
feature because a number of the Maastrichtian species re-
appear in the Danian elsewhere in the basin. The four species
from the basal Danian marl at Nye Kløv form a specialized
community of rooted and free-living colonies very diﬀerent
from the faunas above and below this unit (see also Håkansson
et al. 1996).
Within Denmark there is a strong facies-related diﬀer-
ence between Maastrichtian and Danian bryozoan faunas
(Hakansson & Thomsen 1979). Whereas pelagic chalks contain
high diversity faunas in the Maastrichtian, bryozoans are
almost lacking in Danian chalks. Bryozoan limestones have
similar diversities in the Maastrichtian and Danian, but
cheilostomes dominate in biomass in the Maastrichtian and
cyclostomes in the Danian. Selective extinction of pelagic
chalk species and of cheilostomes is implied by these data. The
low K–T extinction rate of cyclostomes is supported by data in
Brood’s (1972) monograph of Scandinavian cyclostome species
which shows a much greater extinction across the Campanian–
Maastrichtian boundary (23 of 51 species) than across the
Maastrichtian-Danian boundary (2 of 56 species).
The relative fates of cheilostomes and cyclostomes can be
compared on a coarser scale by looking at the compositions of
Maastrichtian and Palaeocene bryozoan assemblages world-
wide. Taylor & Larwood (1988, ﬁg. 5.7) showed that the K–T
event did little to disturb the long-term temporal trend in
the proportion of these two groups, and a more complete
database analysed by Lidgard et al. (1993) revealed no signiﬁ-
cant diﬀerences between within-assemblage percentages of
Maastrichtian and Palaeocene cheilostomes.
Tentative global genus and family data are available at
stage-level for the Cretaceous and Palaeocene. On the basis of
his extensive taxonomic experience, Ehrhard Voigt recognized
the extinction of 38 cheilostome genera and 32 cyclostome
genera during the Maastrichtian (reported in Taylor &
Larwood 1988). Viskova (1994) compiled data on Late
Cretaceous and Palaeogene bryozoan genera, ﬁnding that the
Maastrichtian contained high diversities of cyclostomes (178
genera) and of cheilostomes (172 genera), but that the generic
diversity of both groups declined equally and by more than a
half by the upper Palaeocene. Therefore, the selective extinc-
tion of cheilostomes mentioned above is not apparent in
Viskova’s analysis based on her global generic database. At a
higher taxonomic level still, the diversity of cyclostome and
cheilostome families shows no more than a slight fall across the
K–T boundary (Taylor & Larwood 1988; ﬁg. 6 of Lidgard
et al. 1993). The database of Taylor (1993) shows the presence
of 20 cyclostome families in the Maastrichtian of which 2
(10%) became extinct during or at the end of the stage, and 28
cheilostome families with 4 (14%) extinctions.
Gallagher (1991) regarded the greater abundance of bryo-
zoans in Palaeocene compared to Maastrichtian strata of the
Atlantic Coastal Plain (USA) as a function of the diﬀerential
survival of the mostly non-planktotrophic bryozoans relative
to planktotrophic molluscs. It would be interesting to see
whether there is any evidence for selective extinction of plank-
totrophs within the bryozoans, but this may be diﬃcult
because of the very small numbers of planktotrophic species
that existed in the Late Cretaceous.
Much more research needs to be done on the rich
Maastrichtian and Danian bryozoan faunas of Denmark to
establish or refute the existence of a signiﬁcant and geologi-
cally instantaneous K–T extinction, and to seek evidence for
selectivity related to biological traits, clade membership and
habitat distribution. The few data that are available suggest
276 N. MLEOD ET AL.
that cheilostomes were more strongly aﬀected than cyclo-
stomes, and pelagic chalk species more than bryozoan lime-
stone species. Ecological recovery, at least in the Danish area,
appears to have been rapid in geological terms, judging by the
presence of abundant and diverse bryozoan faunas in the lower
Danian type area. Whatever the short-term eﬀects of any K–T
mass extinction may have been, with no more than 13% of
families becoming extinct there seems to have been little eﬀect
on the long-term pattern of bryozoan evolution, at least not
when compared with the profound extinctions of the Permo-
The work of Surlyk & Johansen (1984) and Johansen (1987)
gives a clear and indisputable indication of a sudden diminu-
tion in brachiopod species at the K–T boundary in Denmark.
Johansen’s analysis (1987) of the brachiopod fauna in the
lowest 0.5 m of the basal Danian Fish Clay at Nye Kløv dealt
with a large collection of minute and immature specimens,
most of which she concluded had been derived from the upper
Maastrichtian chalk. Thirty-ﬁve species were named, of which
six were said to be restricted to the Fish Clay, six were
common in the Maastrichtian and not present in lower Danian
strata, and the remainder were claimed to be new to the
Danian. The weakness in this analysis lies in the identiﬁcation
of established taxa from a collection of immature specimens.
Many brachiopod genera and species assume diﬀerent mor-
phological shapes during their ontogenetic development prior
to emerging as mature specimens. This applies particularly
to species of Terebratulina,Gisilina and Rugia that form a
substantial part of the fauna examined by Johansen (1987).
Surlyk & Johansen (1984) believed the sudden extinction
of brachiopod species at the K–T boundary to coincide with
an equally sudden extinction of coccoliths and planktonic
foraminifera, and suggested a common cause for all three
events. However, restudy of the Nye Kløv planktonic fora-
miniferal fauna has shown that this succession is characterized
by very few extinctions at or near the boundary horizon
(Schmitz et al. 1992; Keller et al. 1993). Furthermore, it is not
uncommon to ﬁnd a paucity of brachiopod species within clay
facies. An obvious analogy to the basal Danian Fish Clay is
that of the Plenus Marls in Britain. This facies supports very
few brachiopod species, but is followed by a chalky facies of
early Turonian age that supports a diverse brachiopod fauna
related both generically and speciﬁcally to lower and upper
Cenomanian forms. These results and observations suggest
that the interpretations of Surlyk & Johansen (1984) and
Johansen (1987) may need to be revised.
Although global temperatures were rising at the end of the
Cretaceous (Schmitz et al. 1992; Keller et al. 1993), a variety of
data suggest that a short-term cooling pulse occurred just
before the K–T boundary. This was followed by a warm
interval where temperatures returned to pre-event ranges.
Bearing in mind the geographic variation characteristic of
Cretaceous brachiopod faunas and the time factor between the
late Maastrichtian and early Danian, it is not surprising that
slight morphological changes may have inspired new taxa.
Examples of this taxonomic practice in Danish faunas have
been recorded by Posselt (1894), Nielsen (1909, 1914) and
others. Such taxonomic splitting has tended to obscure the
direct relationship that certainly exists between faunas from
other European localities of late Maastrichtian and Danian
age (e.g. The Netherlands, Belgium, France, Poland, Russia).
Far more species than have been described by Danish authors
escaped the K–T event and regained their diversity in the
Danian, thus exemplifying the extraordinary ability of this
group to adapt and survive.
The Ammonoidea almost became extinct on several occasions
during their long evolutionary history (House 1993). Each
time, a progressive decline in diversity ceased just short of
extinction. In this sense, the only diﬀerence characterizing the
K–T event is that the ammonoids ﬁnally failed to pull through.
The terminal decline in ammonoid diversity began at least 30
million years before their ﬁnal extinction. From a Cretaceous
peak of 31 families in the Albian there was a progressive
decrease to 14 families in the Campanian and only 12 in the
Maastrichtian, although at the generic level this overall decline
was interrupted by a marked increase in numbers during the
Campanian (Wright 1986). The decline was also characterized
by an increasing patchiness in the areal distribution of the
faunas. No new families appeared during the last 20 million
years of the Cretaceous. On the other hand, some of the
longest-living ammonoid families existed almost to the end of
The best documented late Maastrichtian ammonoids occur
in Europe. Kennedy (1993) reviewed faunas from Denmark,
Poland, Austria, the Cotentin Peninsula (France) and the Bay
of Biscay. Some 14 or 15 ammonite species were present in
these areas during latest Maastrichtian time. These faunas
show considerable diversity, falling into 12 diﬀerent genera
that represent all four Cretaceous suborders: Phylloceratina,
Lytoceratina, Ancyloceratina and Ammonitina. Several
species are very long-ranging. However, the Stevns Klint
(Denmark), Zumaya (Bay of Biscay) and Sopelana (Bay of
Biscay) sections are all chronostratigraphically incomplete,
lacking strata that represent at least 50 000–100 000 years in
the earliest Danian and an undetermined interval in the
Maastrichtian (MacLeod & Keller 1991; Ward & Kennedy
Near Bjala, in eastern Bulgaria, about 45 m of upper
Maastrichtian strata are exposed beneath the Danian (Marin
& Stoykova 1994). Here the faunas are very similar to those of
other European sections, with 13 genera and 22 species repre-
senting all four suborders. There is a gradual decrease in
diversity up this sequence and the last (indeterminate) speci-
mens occur 40 cm below the K–T boundary. Further east, at
Tetrickaro in southern Georgia, the Maastrichtian has yielded
16 species belonging to 9 diﬀerent genera (Adamia et al. 1993).
Exact stratigraphical ranges of these Georgian species are not
One of the most interesting upper Maastrichtian sequences
outside Europe is that of Seymour Island, Antarctica. In a
thick sequence, Macellari (1986) recorded nine species termi-
nating in the top 100 m of his Unit 9. The top of this unit is
now known to lie about 9.5 m below an Ir anomaly that is
thought to mark the local K–T boundary. Subsequent investi-
gations (Elliot et al. 1994) have extended the last appearance
datum of some of these species upwards to a level only 2 m
below the Ir anomaly. This Antarctic succession has not
been collected as intensely as some of the European sections
and it remains uncertain how gradual the extinctions were
(Marshall 1995; see also Discussion below). However, these
data do indicate that high latitude faunas may exhibit an
THE K–T BOUNDARY TRANSITION 277
uppermost Maastrichtian extinction pattern that is similar to
that recorded from lower latitudes.
Among the Coleoidea, the Belemnitida are by far the most
abundant and familiar Mesozoic representatives. They too
declined throughout much of the Late Cretaceous to become
extinct near the end of the Maastrichtian (with the exception of
Bayanoteuthis which Doyle et al. 1994 provisionally excluded
from the Belemnitida). As they declined, they became re-
stricted to high latitudes in the southern and northern hemi-
spheres. Southern forms disappeared quite early in the
Maastrichtian, and the last remaining family (the Belemnitel-
lidae) became limited to northern Europe alone, surviving to
near the end of the Maastrichtian.
The Nautiloidea appear to be a much simpler cephalopod
group, much overshadowed by the ammonoids and belemnites
through the Mesozoic. Yet, they continued largely unchanged
across the K–T boundary. Kennedy (1993) pointed out that
nautiloids may have had a diﬀerent reproductive strategy,
producing few, large eggs with a large yolk and embryo. In
contrast, ammonites apparently produced numerous small
eggs. This distinction may have played a role in survival of the
nautiloid clade, especially if the end-Cretaceous event was
associated with a reduction in phytoplankton abundance.
Gastropods and bivalves
Inoceramid and rudist bivalves, together with nerineid and
acteonellid gastropods, were important components of many
Upper Cretaceous benthic faunas. Although the extinction of
these groups has traditionally been associated with the K–T
boundary (see Alvarez et al. 1980), it has been claimed that
they became extinct by the middle Maastrichtian (Kauﬀman
1984; Johnson & Kauﬀman 1996). Inoceramid prisms are
known to occur in the upper Maastrichtian (Ward et al. 1991),
but these occurrences do not reach the K–T boundary. Polar
extinction of inoceramids has been recognized by Crame (pers.
comm.) before the Maastrichtian, but movement into shal-
lower, Tethyan habitats is shown by this family. In Arabia,
several inoceramid species are associated with rudists and
acteonellids (Smith et al. 1995) in the upper Maastrichtian, but
these are of early late Maastrichtian age and occur below an
unconformity and deep water sediments with no macro-
benthos preserved. Upper Maastrichtian inoceramids also
occur in Madagascar (W. J. Kennedy pers. comm.) and in
Cotentin, North France, but, apart from Tenuipteria, not at
Maastricht. This rather strange pattern of late Maastrichtian
occurrence might be consistent with global cooling and in-
creased oceanic temperature stratiﬁcation. The pattern of local
rudist disappearance has been related to loss of habitat before
the K–T boundary (Swinburne 1990).
Extinctions are apparent in the Pectinacea and Ostreacea.
The pectinid family Neitheinae is represented by two species in
the lower Maastrichtian of the Middle East, and at least one of
these continues to be common in the upper Maastrichtian of
Arabia and at Maastricht and Cotentin (North West Europe).
No member of this subfamily has been found in post-
Cretaceous rocks. A major group of ribbed Mesozoic oysters
(the Palaeolophidae) and the majority of genera of the
Exogyrinae do not seem to have survived beyond the late
Maastrichtian (Malchus 1990).
The Trigoniidae are widespread in the Maastrichtian but
usually limited to one species in any assemblage. They seem to
have survived into the Tertiary only in Australasia.
A profound change also took place in the solenacean
bivalves. In addition to Cretaceous representatives of the
Solenidae and the Pharidae, elongate, near vertical burrowing
shells belonging to the Quenstedtiidae and the Veneridae have
not been reported from Tertiary rocks.
The gastropods as a whole (and the carnivorous forms in
particular) were undergoing a considerable increase in diver-
sity through the later Cretaceous and did not show any
reversal of this pattern through the K–T boundary event at the
taxonomic level of species within families (Taylor et al. 1980).
Some gastropods, however, did not survive from the Mesozoic
into the Tertiary.
The archaeogastropods certainly underwent considerable
redistribution at the end of the Cretaceous but their classiﬁ-
cation is diﬃcult. It is possible that some Late Cretaceous taxa
among the Trochoidea, superﬁcially resembling the Palaeozoic
family Euomphalidae but with nacreo-prismatic aragonite
shell structures, are related to the Pseudophoroidea. If so, this
group includes long-term K–T boundary survivors. On the
other hand, they may not belong to the Pseudophoroidea, in
which case the group may not have survived into the Tertiary.
It is also uncertain whether the essentially Mesozoic family
Amberleyidae survived the K–T event, though it is possible
that some Recent deep water trochids are their descendents.
The Nerineoidea may also be opisthobranchs. They have no
authentic record in the Old World above the upper Campa-
nian. (Note: records from the Maastrichtian of Baluchistan
may be attributed to the Turritellidae or Campanilidae, see
Noetling 1897.) Sohl (1987) discussed their rapid pre-end
Cretaceous decline, but also mentioned abundant lenticular
rock masses, largely formed of the shells of single nerineid
species, in the Maastrichtian of Puerto Rico. Consequently,
it appears that the Cretaceous nerineid extinction had a
pronounced geographical as well as temporal element.
Among the opisthobranchs, the acteonellids persist in some
numbers into the upper Maastrichtian, but have never been
recorded from younger rocks. The family Pseudomelaniidae
(=Trajanellidae not Pseudomelaniidae auctt.) also does not
survive into the Tertiary but in addition is very rare in the
uppermost Cretaceous, and may not have survived up to
Of approximately 100 gastropod families known to occur in
the Late Cretaceous, only one, the Acteonellidae, is known
to have become extinct during the late Maastrichtian. Five
other families and some undeﬁned groups of trochids reached
the Maastrichtian, but are not known from the upper
Maastrichtian. Eighty-one bivalve families are recognized in
the later Cretaceous. Of these, 10 do not extend above the
Maastrichtian, but only ﬁve of these (plus one subfamily) are
known with certainty from the upper Maastrichtian. There is
presently no evidence to indicate that the extinction of any
of these molluscan families occurred at the K–T boundary,
although approximately 3% did become extinct during the late
Jablonski & Raup (1995) have recently published an analysis
of selectivity in the end-Cretaceous marine bivalve extinctions
in which they claim a 70–80% reduction in diversity at the
species level and a 50% reduction at the genus level coincident
with the K–T boundary. However their analysis does not
subdivide the latest Maastrichtian and, in their own words
(p. 389), ‘our results apply to extinction at the K–T boundary
event only to the extent that genus-level bivalve extinctions
were concentrated at the end of the Maastrichtian stage’. In
the absence of detailed collections of bivalve faunas from
278 N. MLEOD ET AL.
uppermost Maastrichtian strata, the statements made herein
with respect to the magnitude and timing of the K–T bivalve
and gastropod turnover are consistent with Jablonski &
Raup’s (1995) results.
Of the ﬁve echinoderm classes, holothurians have by far the
poorest fossil record, being based almost exclusively on
their isolated microscopic body-wall spiculation. The Late
Cretaceous record of holothurian spicules is entirely derived
from mid-latitude chalks and clays. Gilliland (1990, 1992)
distinguished 22 distinct Late Cretaceous spicule morphotypes,
all of which extend into the Tertiary. An additional 14 spicule
morphologies found in extant holothurians are known from
pre-Late Cretaceous deposits and must also presumably have
survived the transition. All ﬁve orders of holothurians and
all 13 of the families known to have existed by the Late
Cretaceous passed into the Tertiary. Indeed, at this coarse
level of analysis, Gilliland found that no distinctive spicule
morphologies were lost at the K–T boundary.
Almost as little is known about the fate of Late Cretaceous
asteroids and ophiuroids, because of the rarity of articulated
specimens and the lack of attention given to isolated ossicles.
Our entire knowledge comes from the work of Rasmussen
(1950, 1972) and Gale (1986, 1987: asteroids only) who docu-
mented Maastrichtian and Palaeocene faunas of Denmark
and northern Europe, working largely from isolated ossicles.
Although changes at species level occurred, all ﬁve upper
Maastrichtian asteroid genera and four ophiuroid genera
continued into the Danian in Denmark. An analysis of
asteroid and ophiuroid taxa across the K–T boundary in
the type Maastrichtian area is in progress (Jagt, personal
The K–T crinoid fauna is now reasonably well-known from
north-western Europe, thanks to the work of Rasmussen
(1961, 1972) and Jagt (1995). Rasmussen (1961) observed that
13 of the 15 Maastrichtian benthic genera were also known
from Danian deposits, and that the major decline in crinoid
diversity occurred after the Danian, when groups such as
bourgueticrinids, holopodinids and isocrinids largely dis-
appeared from shelf environments (except in high southern
latitudes: Stillwell et al. 1994). Representatives of these groups
survive today only in deep-sea environments. Jagt (1995)
found a major drop in diversity of crinoids at the end of the
Maastrichtian, with the diverse shallow-water comatulid-
dominated fauna of the latest Maastrichtian being replaced
by a lower diversity bourgueticrinid-dominated fauna in the
Danian. However, since most of the comatulid genera are
known to survive in other areas such as the Danian of
Denmark, this drop in diversity must to a large extent repre-
sent a local sampling problem associated with a change of
Two clades survived to the end of the Maastrichtian and
then disappeared from the fossil record, the saccocomids and
the microscopic roveacrinids (Jagt 1995). Roveacrinids are
unique amongst crinoids in being obligate planktotrophs and
saccocomids have also been interpreted as planktotrophs.
Together with the comatulids, these constitute about 20%
extinction at the generic level.
Of all echinoderm classes echinoids have by far the best
fossil record, yet this group has been surprisingly neglected in
K–T boundary analyses. In his pioneering study, Kier (1974)
found a substantial drop in the numbers of echinoids described
from the Palaeocene compared with the upper Cretaceous. He
identiﬁed only 108 Palaeocene species as opposed to 1137
Senonian species, based on the Lambert & Thiéry’s (1909–
1925) compilation. Roman (1984) subsequently estimated that
this change represented more than a ﬁve-fold decrease in the
numbers of species, though how much of this may simply be
accounted for by sampling biases was not considered.
Gravesen (1979) and Asgaard (1979) provided an account of
the echinoid changes across the classic Maastrichtian–Danian
sections in Denmark. Gravesen found that, although there
were changes at species level in regular echinoids, all genera
appeared to cross the boundary. Asgaard noted that the
changes in irregular echinoid fauna largely reﬂected changes in
facies, although she pointed out that infaunal holasteroids
were more aﬀected than ploughing forms. Stokes (1979)
analysed changes in spatangoids and concluded that the K–T
boundary event had little impact on this group. He found
that there were more profound changes in generic composition
at the Campanian–Maastrichtian and Palaeocene–Eocene
boundaries than at the end of the Maastrichtian. A detailed
cladistic analysis of the Maastrichtian and Paleocene species of
Cyclaster and related taxa (Jeﬀery in press a) shows that this
group of spatangoids was unaﬀected by the K–T boundary
Van der Ham et al. (1987) recently revised the echinoid
fauna from the Maastrichtian and Danian of the Maastricht
district. Their table summarizing stratigraphical distributions
implies that, at a generic level, about 35% of late Maastrichtian
taxa became extinct. Those disappearing are, in the main,
representatives of Tethyan clades that appear in the uppermost
units as a broad and shallow phytal platform began to develop
in the region. Only one fossiliferous horizon with Danian
fossils is known from this region (Van der Ham 1988) and this
is in chalk facies. This Danian fauna (26 species) is almost as
large as that known from the latest Maastrichtian Meerssen
Member of the Maastricht Formation (27 species), although
only six species are in common, implying at the very least, a
major facies change. A similar major faunal and facies change
has been noted across the K–T boundary at Kazachstan
(Jeﬀery in press b), with again extinction at a generic level of
A compilation of the species of the former Soviet Union and
genera worldwide from the Campanian to early Eocene was
published by Moskvin et al. (1980), and subsequently these
data were used to construct patterns of survivorship and
extinction by Shimanskii & Solovyev (1982) and Roman
(1984). Roman found that the K–T boundary coincided with a
major extinction in echinoids, with 83 out of 127 genera (70%)
supposedly present for the last time in the Maastrichtian.
However, a more recent phylogenetically based analysis of
world-wide data has found extinction levels of no more than
about 35% for clades at generic level (Jeﬀery & Smith in press),
and has shown earlier estimates to have been grossly inﬂated
because of incorrect stratigraphical ranges and inconsistent
taxonomic assignment. Furthermore, they found a clear sys-
tematic bias in the taxa most heavily aﬀected by extinction,
with irregular echinoids more strongly aﬀected than regular
Three groups were particularly badly hit by extinction;
stomopneustid regular echinoids and cassiduloid and hol-
asteroid irregular echinoids. Highest extinction rates are found
amongst shallow-water carbonate-platform faunas, but a
few groups in other environments were also strongly aﬀected.
For example, the specialist deposit-feeding echinoid clade
THE K–T BOUNDARY TRANSITION 279
composed of the holasteroid Stegaster and its relatives, found
in upper continental slope and deep-water basin clastic facies
of mid-latitudes, disappeared at the K–T boundary. Echino-
derms were clearly aﬀected by changes at the K–T boundary,
but the dearth of phylogenetically standardized taxonomic
studies, combined with sampling biases and facies changes
that occurred at this time, greatly complicate any direct
interpretation of the fossil record.
Of the ﬁve classes, holothurians show no evidence of extinc-
tion at a fairly coarse level (i.e. disappearance of characteristic
spicule morphologies, roughly equal to taxa at the family
level). Asteroids and ophiuroids are hardly aﬀected at the
generic level, but this is based almost entirely on the fossil
record in mid-latitude chalks: nothing whatsoever is known
about low-latitude, shallow carbonate platform faunas. The
same is true for crinoids, except that the two planktotrophic
clades, roveacrinids and saccocomids, extend into the top zone
of the Maastrichtian and then become extinct. Echinoids
overall have extinction levels of around 35% of clades at the
generic level, but show a strong taxonomic and geographic
bias (Jeﬀery & Smith in press). Taxa such as cassiduloids that
had their greatest diversity in low-latitude shallow-water car-
bonate platform environments, were most severely reduced in
diversity at the end of the Cretaceous, whereas mid-latitude
deeper-water chalk-facies faunas were much less aﬀected,
displaying much higher levels of extinction at the end of the
Danian than at the end of the Maastrichtian. This is reﬂected
in a latitudinal bias seen in the extinction of echinoids with mid
to high latitude faunas displaying lower extinction levels than
low-latitude faunas. Furthermore, the marked reduction in
diversity of crinoids at the end of the Danian mirrors the
pattern seen in echinoids from mid-latitude chalks.
The emerging pattern points to habitat loss, speciﬁcally the
drowning of carbonate platforms, as playing a major role
in restructuring the echinoderm fauna at the end of the
Cretaceous. Major transgressions have previously been shown
to be strongly correlated with periods of signiﬁcant change in
the composition of Jurassic echinoid faunas (Thierry &
Neraudeau 1994). However, clearly not all extinctions can be
attributed to this cause. Extinctions of the planktotrophic
crinoid clades as well as of the specialist deposit-feeding
deeper-water holasterid echinoids in middle latitudes, show
that a multiplicity of factors must have been involved in
determining K–T echinoderm extinction and survivorship.
The fossil record of jawed ﬁshes, which stretches back some
430 million years, is rich and varied and provides fossils from
a wide range of sediment types. Classiﬁcation of many groups
of ﬁshes is also more reﬁned than that of most other animal
groups. This means that patterns of extinction/origination may
be tied to phylogenies in order to provide some estimate of
what is real extinction and what is artiﬁcial (see Discussion).
On the other hand it must be acknowledged that the fossil ﬁsh
record is very patchy and can be misleading to the extent that
many of our observations may be locality dependent.
Information on family-level extinction in ﬁshes can be
obtained from Gardiner (1993), Capetta et al. (1993) and
Patterson (1993). The modern ﬁsh world is dominated by
cartilaginous ﬁshes (846 species in 45 families) and teleost
actinopterygians (ray-ﬁnned ﬁshes) (23 637 species in 426
families). Other ﬁsh groups are represented by a handful
of agnathans (lampreys), non-teleostean actinopterygians
(bichirs, sturgeons, garpike and bowﬁn), and sarcopterygians
(coelacanth and lungﬁshes) (135 species, counts from Nelson
1994). Only cartilaginous ﬁshes and teleosts contribute signiﬁ-
cantly to extinction patterns of ﬁshes at the K–T boundary.
For the cartilaginous ﬁshes (sharks, rays and chimaeras), 35
families pass through the K–T boundary, seven became extinct
within the Maastrichtian, and one originated in the Danian.
The seven fatalities (Polyacrodontidae, Hybodontidae,
Anacoracidae, Sclerorhynchidae, Rhomdodontidae, Hypso-
batidae and Parapalaeobatidae) all appear to be monophyletic
and therefore are cases of real extinctions. Expressed another
way, the K–T boundary was survived by 80% of cartilaginous
ﬁsh families. The survival rate between the Campanian–
Maastrichtian is 95% and that between the Danian–Thanetian
is 94%. Therefore there is a reduced survival rate at the K–T
boundary, but it may be questioned whether this reduction is
signiﬁcantly acute to implicate a catastrophic mechanism.
Capetta (1987) has compiled the most comprehensive data
base for elasmobranchs (sharks and rays) by documenting gen-
eric occurrences throughout the Cretaceous and the Tertiary.
His counts show a signiﬁcant disappearance of genera within
the Maastrichtian: 23/53 (or 43.4%) became extinct and this
compares with 26.3% disappearance within the Campanian
and 8.1% in the Danian. However, while the extinction rate is
high, so is the origination rate: in the Maastrichtian 22.6% are
ﬁrst occurrences (comparative ﬁgures for the Campanian and
Danian are 21.4% and 18.9% respectively). It therefore appears
that the Maastrichtian was a time of high generic turnover and
not only a period of high extinction. One fact that needs to be
checked is how many of the Maastrichtian genera which
disappeared are monophyletic and represent real extinctions
and how many are paraphyletic, representing parts of species
lineages which crossed the K–T boundary and are therefore
not evidence of extinction (Smith & Patterson 1988).
For teleost ﬁshes the K–T record consists of 43 families
which pass through the K–T boundary, four became extinct
within the Maastrichtian and nine originated within the
Danian. The four extinctions were probably real events be-
cause they involved monophyletic families with a total of eight
genera and 13 species: Saurodontidae, Ichthyotringidae,
Cimolichthyidae and Dercetidae (taken here to include
Stratodus). Therefore, about 10% of teleost ﬁsh families
became extinct and 90% survived. Comparing this with stages
either side of the K–T boundary, there is an 81% survival from
the Campanian to the Maastrichtian, and an 85% survival
from the Danian to the Thanetian. At this crude level of
analysis there is no reason to believe that a K–T elasmobranch
mass extinction took place.
The teleost record shows clearly the eﬀects of Lagerstätten
as mentioned by Patterson (1993) and Patterson & Smith
(1989). There are at least two peaks of teleost diversity; one in
the Campanian and another in the Eocene. The ﬁrst is due to
a diverse fauna at Sedenhorst, Westphalia. The second is due
to the occurrence of the famous ﬁsh-bearing deposits of Monte
Bolca, Italy and the London Clay of the Anglo-Paris Basin.
This is set against the relative paucity of Palaeocene and upper
Eocene ﬁsh localities.
There is no evidence for amphibian extinctions at the K–T
boundary and tangible evidence of lineages passing through
280 N. MLEOD ET AL.
the event unaﬀected. The latest known temnospondyl amphib-
ian is an unnamed chigutisaurid from the Aptian–Albian of
Victoria, Australia (Warren et al. 1991). Its only signiﬁcance to
this discussion lies in the fact that there are no good Australian
continental vertebrate assemblages from the Cenomanian to
the Eocene. Thus, temnospondyls may have died out before,
at, or after the K–T event.
The albanerpetontids are an enigmatic and specialized small
amphibian family of two genera, initially assumed to be
salamanders, but now considered to represent a distinct lineage
(McGowan & Evans 1995). The family is known from the
Maastrichtian to the upper Palaeocene of North America,
from the Middle Jurassic to the Miocene of Europe, and from
the Cretaceous of Asia (Milner 1993). There is no evidence of
family or generic-level extinctions at or near the K–T bound-
ary, the genus Albanerpeton bracketing the K–T boundary in
both North America and Europe. All evidence suggests that
this family survived unscathed in these two regions at least.
About 160 species of living caecilians (Gymnophiona) have
been described from South America, Africa, the Seychelles,
India and Southeast Asia. The ﬁrst record is from the Lower
Jurassic of Arizona. The only other fossils are vertebrae of
modern type from the Upper Cretaceous of Sudan, and the
Palaeocene of South America. There is therefore no direct
fossil evidence of extinction or survival at the K–T boundary.
Some immunologically based cladistic analyses (Case & Wake
1977) show congruence between dichotomies producing mod-
ern families and the pattern of Gondwana rifting in the areas
where they occur. Mitochondrial DNA sequence data support
the interpretation that much of modern diversity was estab-
lished in the Mesozoic by vicariance during the break-up of
Gondwana (Hedges et al. 1993). There is no evidence for a
Palaeocene diversiﬁcation such as might have followed a K–T
mass extinction, with the survivors rediversifying.
About 350 species of living salamanders (Caudata) have
been described from the northern continents, but with pletho-
dontids extending into Central and South America and
salamandrids into North Africa. There are ten families with
living representatives and two extinct families, plus a scatter of
disparate primitive genera with no family assignment from the
Middle Jurassic to the Middle Cretaceous. Of these 12 families,
ﬁve bracket the K–T boundary as fossils: Sirenidae (Campa-
nian to Recent), Batrachosauroididae (Tithonian to Pliocene),
Scapherpetontidae (Albian to Eocene), Amphiumidae
(Maastrichtian to Recent) (Milner 1993) and Salamandridae
(Maastrichtian to Recent) (Astibia et al. 1990). Four families
have only post-K–T boundary records, but phylogenetic infer-
ence from cladogram structure (e.g. see ﬁg. 5 of Larson
& Dimmick 1993) imply that they either were present before
the K–T event: Proteidae (Palaeocene to Recent) and Pletho-
dontidae (Miocene to Recent); or that a stem-lineage shared
with one other family was present: Cryptobranchidae (with
Hynobiidae, Palaeocene to Recent) and Ambystomatidae
(with Dicamptodontidae, Oligocene to Recent) (Milner 1993).
Three families have no pre-Pliocene fossil records, but phylo-
genetic inferences from cladogram structure imply that they
either were present before the K–T event (Rhyacotritonidae),
or that a stem-lineage shared with one other family was present
(Hynobiidae and Dicamptodontidae). There is no evidence
among these groups for family-level extinctions.
The only high-resolution study of salamanders bracketing
the K–T event is that of Archibald & Bryant (1990) on seven
genera from Montana. These authors found that four genera
passed from the Lancian (=Maastrichtian) to the Torrejonian
(=upper Danian) of Montana apparently unchanged (the
sirenid Habrosaurus dilatus, the scapherpetontids Lisserpeton
bairdi and Scapherpeton tectum and the batrachosauroidid
Opisthotriton kayi), while one genus passed from the Lancian
to the Puercan (=lower Danian) unchanged (the batracho-
sauroidid Prodesmodon copei). The remaining two genera did
not cross the K–T boundary in Montana. However one, the
scapherpetontid Piceoerpeton sp., was very rare in the Lancian,
and is known elsewhere in North America to range into the
Eocene. The other, the amphiumid Proamphiuma cretacea,
makes its last appearance in the Judithian (=Campanian) of
Montana. As a genus, Proamphiuma may have become extinct
at or before the K–T boundary, but there is no morphological
evidence to preclude it being the lineal ancestor of the
Palaeocene–Recent amphiumid genus Amphiuma, in which
case this would be a pseudoextinction. Thus, of seven Lancian
genera of salamanders, six survived the K–T event unchanged
and one either became extinct or underwent morphological
Over 3500 species of frogs (Anura) have been described from
all continents except Greenland and Antarctica. There are 21
families with living representatives and two extinct families,
plus a few disparate primitive forms from the early Mesozoic.
The frog fossil record is geographically very uneven: fair for
North America, Europe and South America, but very poor
from other continents. Of these 23 families, seven bracket
the K–T boundary as fossils: Leiopelmatidae (Tithonian to
Recent); Discoglossidae (Bathonian to Recent); Pipidae
(Hauterivian to Recent); Palaeobatrachidae (Valanginian to
Pliocene); Rhynophrynidae (Tithonian (Henrici pers. comm.)
to Recent); Pelobatidae (Tithonian to Recent); and Lepto-
dactylidae (Santonian to Recent) (Milner 1993). Seven families
have a post-K–T fossil record only: Pelodytidae, Myobatrachi-
dae, Bufonidae, Hylidae, Ranidae, Rhacophoridae and
Microhylidae; and eight have no record at all: Hyperoliidae,
Heleophrynidae, Sooglossidae, Brachycephalidae, Rhino-
dermatidae, Pseudidae, Centrolenidae and Dendrobatidae
(Milner 1993). Only one family might represent a Cretaceous
extinction, namely the Gobiatidae known from the Albian to
Santonian of Mongolia and vicinity (Rocek & Nessov 1993).
However the latest gobiatid record occurs well below of the
K–T boundary and this family may only represent one genus
upgraded to family status non-cladistically.
As with the salamanders, the only high-resolution study of
frogs bracketing the K–T event is that of Archibald & Bryant
(1990) on three genera from Montana. These authors found
that the discoglossid Scotiophryne pustulosa passed from the
Lancian to the Puercan of Montana apparently unchanged.
The pelobatid genus Eopelobates sp. passed from the Lancian
to the Torrejonian of Montana apparently unchanged, but this
genus is currently something of a dump-genus, probably
polyphyletic and hence of uncertain value. The third genus is
the palaeobatrachid Palaeobatrachus sp. which makes its last
appearance in Montana at the K–T boundary at the Bug
Creek locality, but is known from the Torrejonian elsewhere.
As with Eopelobates sp., the species unity of the various
fragments is not established.
The inescapable conclusion from these data is that there is
no evidence for family-level extinctions of amphibians at or
near the K–T boundary, and, in terms of genera, no un-
ambiguous evidence for complete extinction of any Lancian
amphibian lineage in North America. The only possible
exception (Proamphiuma) is equally explicable as a pseudo-
THE K–T BOUNDARY TRANSITION 281
All three living reptile groups, the Crocodilia, the Chelonia,
and the Lepidosauria, are survivors from the Cretaceous. Only
one, Lepidosauria, represents animals that are predominantly
Living lepidosaurs are divided between the Rhyncho-
cephalia and Squamata. The Rhynchocephalia were a suc-
cessful and geographically widespread group in the early
Mesozoic, but were already in decline by the mid-Cretaceous.
They are represented today by a single genus, the New Zealand
Sphenodon. The Squamata (lizards, snakes and amphisbaeni-
ans) radiated in the Jurassic, continued successfully into the
Cretaceous, and are the largest and most diverse group of
living reptiles, with more than 6000 extant species.
Allowing for the metataxon status of Iguanidae and
Agamidae (see Discussion for an explanation of metataxa), all
living families of lizards either had known representatives in
the Late Cretaceous or their presence may be assumed by
extrapolation from the phylogenetic tree (e.g. by the existence
of sister taxa). All living families therefore represent deﬁnite or
inferred survivors from the Cretaceous. Of the groups occur-
ring in the upper Maastrichtian, only two are unknown from
the lower Tertiary—the mosasaurs and the polyglyphanodont
Mosasaurs were large (up to 15 m) marine predators of
the Late Cretaceous (Turonian–Maastrichtian). In North
America, they are known to have survived into the late
Maastrichtian (Lower Hornerstown Formation, New Jersey;
Moreno Formation, California; Gordon Bell pers. comm.
1995, contra Sullivan 1987), with representatives of both
mosasaurine and plioplatycarpine lineages. Their decline dur-
ing the late Maastrichtian would not be surprising. Mosasaurs
were amongst the top predators of the Cretaceous marine
ecosystem and anything that perturbed that ecosystem would
have aﬀected them.
A smaller Late Cretaceous group (North America, China,
Mongolia; Estes 1983), the polyglyphanodont teiids, also failed
to survive the K–T event, but these durophagous lizards were
probably already in decline. They reached their peak diversity
in the Campanian, and continued into the early Maastrichtian
in North America (North Horn Formation, Utah; Estes 1983;
Sullivan 1987). Only Haptosphenus, tentatively assigned to
the polyglyphanodonts (Estes 1983), has been recorded from
the upper Maastrichtian Hell Creek Formation of Montana.
For most of the latest Cretaceous, snake records are re-
stricted to a number of primitive families (aniliids, boids,
madtsoiids and nigerophiids), all of which crossed the K–T
boundary. Higher snakes (colubroids, Werner & Rage 1994)
have recently been recorded from the middle Cretaceous
(Albian–Cenomanian) of the Sudan, but apparently went
through a major radiation in the Tertiary. Like snakes, the
burrowing amphisbaenians are predominantly a Tertiary
group although they are now known from the Campanian of
China and Mongolia (Wu et al. 1993) and clearly crossed the
Only one study (Archibald & Bryant 1990) has followed a
complete lepidosaur assemblage through the K–T boundary
interval (Lancian–Puercan) in one restricted area (northeastern
Montana). Of 11 squamate species present in the Lancian
study area, four (three lizards, one snake) survived into the
Tertiary, although two more have been recorded from Tertiary
localities outside the study area and two were already rare in
the late Cretaceous. The only common Lancian lizards to be
lost completely were three teiid species. Bearing in mind the
Late Cretaceous decline of the polyglyphanodonts, this may
tell us more about teiid history than about the fate of
squamates in general at the K–T boundary.
No known family of terrestrial squamates became extinct
at the K–T boundary and, with the exception of teiids and
the marine mosasaurs, the limited evidence available suggests
that squamates suﬀered no major decline at this time. If there
was a Late Cretaceous biotic crisis, squamates survived it,
possibly because of their generally small size, variable metab-
olism, and ability to retire to favourable microhabitats. This
said, the late Maastrichtian and early Palaeocene squamate
record is still very limited. Squamates are preserved (and then
recovered) only under a limited number of conditions, and the
picture may change radically as the squamate fossil record
The clade Archosauria includes crocodilians, pterosaurs, dino-
saurs and birds. Birds are now almost universally accepted as
the sister group of maniraptorans, small advanced carnivorous
dinosaurs from which they are descended.
Ten crocodile families are represented in the Maastrichtian.
The last records of the Goniopholidae, Peirosauridae, Doli-
chochampsidae and Paralligatoridae are in the Maastrichtian
(Benton 1993). The Baurusuchidae may span the K–T bound-
ary, based on a specimen of Cynodontosuchus from the El
Molino Formation, Vila-Vila, Bolivia. The deposits in the
Vila-Vila area have been correlated with the mammal-rich
Tiupampa Beds, originally supposedly Maastrichtian in age,
but dated in several more recent papers as early Paleocene
(Gasparini et al. 1991).
Five families have both Maastrichtian and Tertiary repre-
sentatives. The Thoracosauridae ranged from the Maastrich-
tian to the late Paleocene, and Trematochampsidae from the
Maastrichtian to the mid Eocene. The last records of the
Dyrosauridae are in the upper Eocene in Asia; one genus,
Hyposaurus, spans the K–T boundary, occurring in the Upper
Cretaceous in North and South America and the Palaeocene to
Eocene in north and west Africa. The two remaining families,
Alligatoridae and Crocodylidae, include living represen-
tatives; two alligatorines and the crocodylid (or possibly stem
crocodylid) Leidyosuchus sternbergi, have been recorded from
both the Lancian and Puercan in Montana (Archibald &
Bryant 1990) and so crossed the K–T boundary at the species
level. All the families discussed here are terrestrial–freshwater,
except for the Dyrosauridae which are freshwater–marine. The
group as a whole shows a 50% family survival rate across the
K–T boundary. The only discernible trend apparent from this
pattern is that no large crocodiles, such as the giant North
American crocodylid Deinosuchus, survived.
The Azhdarchidae, which originated in the Berriasian, is
the only pterosaur family recorded from the Maastrichtian.
The azhdarchids include the last representatives of the ptero-
dactyloid pterosaurs which attained their highest diversity in
the mid-Cretaceous with ten families recorded (Wellnhofer
1991). All except the azhdarchids became extinct prior to
the Maastrichtian. Two genera are known from upper
Maastrichtian deposits; Quetzalcoatlus from the Javelina For-
mation in Texas and Arambourgiania from marine deposits in
Jordan (Benton 1993). No Tertiary pterosaur remains are
known and their extinction pattern was apparently a gradual
decline through the Late Cretaceous.
282 N. MLEOD ET AL.
More has probably been written about the extinction of
dinosaurs than of any other group of organisms. Families
recorded from the Maastrichtian are listed in Table 1. In
addition, there are indeterminate theropod footprints from
Morocco (Weishampel 1990) and an euornithopod from
Antarctica (Hooker et al. 1991); Maastrichtian dinosaurs are
thus known from all continents except Australia. Late
Maastrichtian records (Table 1) show little decline at family
level except in Asia and South America, though this probably
reﬂects the absence of relevant strata in those areas rather than
positive evidence for a fall in diversity.
The western interior of North America is the only area
where a continuous terrestrial sequence encompasses the K–T
boundary, permitting documentation of changes in faunal
diversity. Consequently, it has been the focus of several
detailed studies. In the upper Campanian (Judithian), the
dinosaur fauna is diverse, with 45 taxa determined to generic
level in 13 families (Weishampel 1990) from the Judith River
Formation of Alberta. In the upper Maastrichtian (Lancian)
Hell Creek Formation in Montana, the corresponding ﬁgures
are 24 genera in 12 families (Weishampel 1990). This decline in
dinosaur diversity mirrors that shown by Sloan et al. (1986)
based on older data, recording a decline from 30 genera in the
Judithian (=middle Campanian) down to seven at the K–T
Archibald & Bryant’s (1990) high resolution study of the
non-marine fauna across the K–T boundary section in north-
east Montana showed variable extinction rates for diﬀerent
vertebrate groups. Analysis of their raw data to allow for
factors such as palaeobiogeographical variation, disappear-
ances that represent speciation events and rarity of taxa,
produced an estimated 64% species survival rate (conversely, a
36% extinction rate) for the non-marine vertebrate fauna. The
survival rate for non-avian dinosaurs was 0%. Twelve dinosaur
genera (13 species) from the Bug Creek interval, overlying the
Lancian at the top of the Hell Creek Formation, are the last
recorded in the sequence (Table 1).
There is no consensus over the age of the Bug Creek interval.
Its fauna contains taxa otherwise unknown after the Lancian
(particularly dinosaurs) and others unknown before the
Puercan (especially mammal taxa). Archibald & Bryant (1990)
included the Bug Creek interval in the Puercan on the basis of
the mammal faunas, thus dating it as earliest Palaeocene.
Accepting this age, the presence of dinosaur remains in the Bug
Creek interval can be explained in one of two ways; either they
survived into the Paleocene, or they were reworked from the
Hell Creek Formation. Archibald & Bryant (1990), in common
with other studies, adopted the latter explanation on preser-
vational and taphonomic grounds since there are no articu-
lated dinosaur remains in the Bug Creek channel sediments.
They interpreted their results in terms of geologically rapid
change, but not a sudden catastrophic mass extinction at the
A detailed ﬁeld study of dinosaur diversity by Sheehan et al.
(1991) claimed that no statistically meaningful decline was
apparent through the Hell Creek Formation. These ﬁndings
were claimed to support a sudden extinction event at the K–T
boundary. Williams (1994) observed that the nature of the
Table 1. Dinosaur distribution in the Maastrichtian
Stratigraphic occurrence Last recorded genus
Maastrichtian Europe* North America†
Tyrannosauridae NA Present Present Tyrannosaurus
Ornithomimidae As, NA Present Present ?Ornithomimus
Elmsauridae As, NA Present Present Chirostenotes
Oviraptoridae As, NA Present —
Dromaeosauridae As, NA Present Present cf. Dromaeosaurus
Troodonidae As, NA Present Present Troodon indet.(=Paronychodon)
Abelisauridae SA Present —
Deionocheiridae As Present —
Noasauridae SA Present —
Therizinosauridae As Present —
Aublysodontidae NA Present Present
Camarasauridae As Present —
Diplocidae As Present —
Titanosauridae As, Eu, In, NA, SA Present Present Magyosaurus
Nodosauridae Eu, NA Present Present Struthiosaurus
Ankylosauridae As, NA Present Present Ankylosaurus
Hypsilophodontidae Eu, NA Present Present Thescelosaurus
Iguanodontia incertae sedis Present Present Rhabdodon
Hadrosauridae As, Eu, NA Present Present Telmatosaurus Edmontosaurus
Pachycepalosauridae As, NA Present Present Stygimoloch
Protoceratopsidae As, NA Present Present Pachycephalosaurus
Ceratopsidae NA Present Present Triceratops
*Sînpetru Formation, Hateg basin, Romania.
†Bug Creek interval, northeast Montana (may be earliest Palaeocene in age, see text).
NA, North America; As, Asia; SA, South America; Eu, Europe; In, India.
THE K–T BOUNDARY TRANSITION 283
analysis of Sheehan et al. (1991) might not detect a gradual
decline at the top of the Hell Creek sequence. There remains,
despite concentrated collecting, a barren zone of about 3 m
below the K–T boundary, in which dinosaur bones and teeth
are increasingly scarce. The highest unreworked bone comes
from about 60 cm below the boundary clay—well below the
iridium layer. Evidence from the Bug Creek channels and the
highest few metres of the Hell Creek Formation is consistent
with a decline in the dinosaur population, perhaps a steep and
accelerating decline but not a sudden catastrophe (Williams
1994). Further, Hulbert & Archibald (1995) analysed the
quantitative methodology used by Sheehan et al. (1991) to
argue for a rapid end-Cretaceous dinosaur extinction event,
and concluded that there was no statistical support for either a
sudden or gradual decline in the number of dinosaur taxa at
the close of the Cretaceous.
Diversity changes and extinction patterns in the western
interior of the USA provide a test case for the demise of the
dinosaurs. There are no means of determining if this was a
local or global phenomenon due to lack of K–T terrestrial
sequences elsewhere. An upper Maastrichtian non-marine
fauna, including dinosaurs (Table 1), is known from the
Sînpetru Formation (once regarded as Danian) in Romania
(Grigorescu et al. 1994), but is overlain by middle to upper
Miocene marine strata. Dinosaurs are certainly present in the
uppermost Cretaceous of Asia; the highest occurrence known
is a nest site from magnetochron 29R in south central China
(Hansen 1990). This establishes a latest Maastrichtian age
for the strata involved, but evidence of diversity and rate
of decline is lacking. Dinosaur egg shell and protungulate
mammal material from the Vilquechio Formation of Peru was
dated as Paleocene by comparison with North American
mammals (Kérourio & Sigé 1984), but more recent charophyte
data and K–Ar dating of overlying volcanic tuﬀs suggest an
early Maastrichtian age (Mourier et al. 1988).
In summary, non-avian dinosaurs apparently became extinct
in the late Maastrichtian following a decline in diversity over
several million years, and possibly an accelerated decline in the
latest Maastrichtian, at least in North America. The last
unreworked remains occur well below the iridium layer which
marks the K–T boundary in North America, and the small
amount of data from other continents indicate that no dino-
saurs survived into the Palaeocene. However, there is no
evidence for the timing and nature of non-avian dinosaur
extinction in the rest of the world.
Recent discoveries have revolutionized the understanding of
bird evolution in the Mesozoic and early Tertiary, although
much of their fossil history is still missing. Two major
groups of birds were coeval in the Late Cretaceous. The
Enantiornithes were the dominant land bird group, and they
coexisted with a separate lineage, the Ornithurae (including the
Hesperornithiformes) and the Neornithes from which modern
birds derive (Chiappe 1995). Enantiornithes did not survive
beyond the Maastrichtian. Neither did several groups of
ornithurines, but there is no positive evidence that these
extinctions occurred at the K–T boundary. The toothed
hesperornithiforms, in particular, have no unequivocal post-
Campanian record. According to Feduccia (1995), birds
appear to have suﬀered a bottleneck across the K–T boundary,
a massive extinction from which only transitional shorebird
morphotypes survived as the basis for a Tertiary bird radi-
ation. However, fragments of several genera of Late Cretaceous
birds can be assigned to four orders of living birds: Anseri-
formes, Gaviiformes, Procellariformes and Charadriformes.
The phylogenetic relations between these groups suggest that
most, if not all modern bird orders must have diﬀerentiated
before the end of the Cretaceous (Chiappe 1995, see also refer-
ences therein), and therefore passed through the K–T boundary.
These results are supported by modern bird 12s rDNA sequence
studies (Cooper 1994).
Archosaurs fared badly within the Maastrichtian with the
total extinction of pterosaurs, non-avian dinosaurs, and the
enantiornithine birds, together with several ornithurine
groups. The crocodiles are the only amphibious archosaur
group and the only one in which a signiﬁcant number of
families passed through the boundary.
The record of megaﬂoras and microﬂoras across the K–T
boundary is best known from the western USA with additional
microﬂoral data from Antarctica. Johnson & Hickey (1990)
based their analysis of megaﬂoras on 25 000 specimens from
more than 200 localities in Montana, Wyoming, North Dakota
and South Dakota. This summary shows that megaﬂoral
changes were taking place well before, at, and after the
regional K–T boundary, as identiﬁed by Ir anomalies. Johnson
& Hickey (1990) divided this turnover pattern into four
diﬀerent climatic zones that suggest progressive warming of
In contrast to the megaﬂoral record, the terrestrial micro-
ﬂoral record in the USA is typiﬁed by the abrupt disappear-
ance of most Upper Cretaceous palynomorphs coincidental
with the Ir anomaly, the existence of a very short-term fern
abundance peak (the so-called ‘fern spike’) just above the Ir
anomaly, and the progressive appearance of fully Danian
palynomorph taxa (Nichols & Fleming 1990). (Note: plant
megafossils are regarded as more accurate indicators of
taxonomy, phylogenetic relationship, and species richness,
whereas terrestrial palynomorphs are regarded as more accu-
rate chronostratigraphical indicators of the presence of gener-
alized higher taxonomic groups.) These palynomorph turnover
patterns suggest that certain angiosperm taxa endemic to
North America may have been severely eﬀected by the K–T
event, and that gymnosperms, ferns, other pteridophytes, and
mosses were aﬀected to a much lesser degree. In Canada,
however, Sweet et al. (1990) have shown that palynomorph
turnover patterns were more progressive, with disappearances
occurring before, at, and after the K–T boundary. Moreover,
these authors reported no fern spike in their sections, but
rather angiosperm dominance extending from the Upper
Cretaceous through the lowermost Danian. Sweet et al. (1990)
concluded that morphologically complex palynomorphs were
diﬀerentially susceptible to the K–T event, but oﬀered no
speculations as to what these patterns might reﬂect from
functional or adaptive points of view.
In the high southern latitudes of New Zealand and
Antarctica, the K–T megaﬂoral and palynomorph record is
very diﬀerent from that of western North America, with no
signiﬁcant turnovers being recorded from the Seymour Island
sections (Johnson & Greenwood 1993; Askin & Jacobson
New palaeoclimatological data from the entire Upper Creta-
ceous (e.g. Huber et al. 1995), and the upper Maastrichtian in
284 N. MLEOD ET AL.
particular (e.g. Keller et al. 1993; Barrera 1994; Barrera &
Keller 1990, 1994), have shown that the latest Cretaceous was
a time of sustained global temperature and marine productiv-
ity changes, of the type known to be correlated with major
extinction events in other parts of the geological column. These
data suggest that extinction mechanisms may have operated
throughout the entire Upper Cretaceous rather than being
concentrated within a very narrow time interval at its very end.
The data reviewed above indicate that, aside from a few
microfossil groups (e.g. calcareous nannoplankton, dinoﬂagel-
lates, planktonic foraminifera), most clades experienced either
long-term decline throughout the Upper Cretaceous (e.g.
ostracodes, scleractinian corals, bryozoa, ammonite cephalo-
pods, bivalves, echinoid echinoderms, archosaurs) or were
relatively unaﬀected in terms of extinction levels by the K–T
boundary event (e.g. diatoms, radiolaria, benthic foraminifera,
brachiopods, gastropods, holothurian, asteroid, ophiuroid,
and crinoid echinoderms, ﬁsh, amphibians, lepidosaurs, terres-
trial plants). A few groups (e.g. benthic foraminifera) also
exhibit changes in faunal composition across the highest
Maastrichtian–lowest Danian interval.
It is uncertain whether increased levels of faunal and ﬂoral
turnover in the uppermost Maastrichtian are sampling arte-
facts (e.g. a Signor–Lipps eﬀect), or a response to environ-
mental changes (e.g. sea-level ﬂuctuations, increased rates of
volcanism) that are known to have occurred during the latest
Maastrichtian. The Signor–Lipps eﬀect (Signor & Lipps 1982)
arises from a number of causes, including interactions between
sample size and relative abundance, diﬀerential susceptibility
to diagenesis, and/or facies or palaeoecologically controlled
distributions. These factors can result in the last observed
appearance of a fossil taxon in a particular section or core not
accurately reﬂecting its local or global last appearance datum
(Fig. 9). Thus, Signor & Lipps (1982, p. 291) concluded that
gradual extinction patterns prior to a mass extinction do
not necessarily eliminate catastrophic extinction hypotheses.
However, uncertainty is implicit in the Signor–Lipps eﬀect.
This was recognized by Signor & Lipps (1982, p. 295) who
cautioned ‘Our arguments should not be interpreted as support
for the impact hypothesis or any other theory invoking a
catastrophe as the extinction mechanism. The evidence at
hand for most environments is as compatible with a gradual
extinction event as with a catastrophic one’.
Signor & Lipps’ (1982) conclusion has been veriﬁed by
Marshall (1995) who used stratigraphical conﬁdence intervals
to examine whether the observed progressive extinction pat-
terns of Maastrichtian ammonites from Seymour Island were
compatible with a catastrophic scenario. Marshall (1995,
p. 731) concluded that while the ‘pattern of [ammonite] dis-
appearances is consistent with a sudden mass extinction at the
Cretaceous–Tertiary boundary . . . computer simulation of the
Seymour Island ammonite fossil record indicates a wide range
of other extinction scenarios, including gradual extinctions
ranging over as much as 20 m that are [also] consistent with the
ammonite fossil record’. Marshall (1995) went on to observe
that without saturation collecting it may be impossible to
separate catastrophic from progressive extinction patterns
based on the fossil record alone. Marshall’s (1995) conclusions
diﬀer strongly from Ward’s (1995a,b) claim that his study
provided unambiguous evidence for catastrophic K–T faunal
turnover among high-latitude invertebrate faunas.
While Marshall (1995) applied stratigraphical conﬁdence
limits to his study of potential Signor–Lipps eﬀect bias,
MacLeod (1996a,b) employed an array of diﬀerent strategies
to test for signiﬁcant Signor–Lipps eﬀect bias in planktonic
foraminiferal data from El Kef (Tunisia), Agost (Spain),
Brazos River (Texas), Nye Kløv (Denmark) and ODP Site 738
(Kerguelen Plateau). These investigations also attempted to
identify the source of the bias (e.g. sample size, faunal compo-
sition, and palaeoecological factors). Although patterns of
variation consistent with a Signor–Lipps eﬀect were found in
each succession, the magnitude of the eﬀect was shown to be
insuﬃcient to account for the magnitude of the pre-boundary
faunal turnover in three of the ﬁve sections/cores. It is also
worth noting that the uppermost Maastrichtian changes
in ecological dominance patterns observed in benthic
foraminifera and bivalves suggest the presence of a period of
latest Maastrichtian global environmental change. This
result is consistent with the recognition of a stratigraphical
interval, within which the K–T boundary is embedded, that is
characterized by increased extinction intensity.
Attribution of seemingly anomalous occurrences of Creta-
ceous species in lower Danian strata to reworking and/or
remobilization of sediments has also been used to justify
catastrophic interpretations of what appear to be progressive
faunal turnover patterns. Certainly reworking and remobiliz-
ation of fossils do occur. However, in most cases it is exceed-
ingly diﬃcult to determine whether a particular specimen has
been reworked and, if so, over what stratigraphical interval. The
data presented above suggest that when K–T boundary succes-
sions are examined in detail, many characteristic Cretaceous
species are found in lowermost Danian strata, often several
metres above the boundary level. In some instances such species
even comprise the bulk of the fauna (see Fig. 5; Pospichal 1995).
Moreover, the simulations of Berger & Heath (1968) indicate
that non-biotic vertical mixing of sediments in marine pelagic
sediments can rarely account for specimen displacements of
Fig. 9. Diagrammatic explanation of the Signor–Lipps eﬀect. The
apparent distribution of fossil last appearance datums (LADs) will
always underestimate the true LAD distribution because the relative
abundance of fossils varies within a section or core while the sample
size usually remains constant or varies independently. Underlying
factors responsible for intraspeciﬁc abundance ﬂuctuations include a
wide variety of biological, ecological, and diagenetic processes. As a
result, it may be impossible to infer the correct pattern of either
local or global last appearances from the observed biostratigraphical
record. Note that recognition of the Signor–Lipps eﬀect confers no
ability to distinguish between alternative faunal turnover patterns.
THE K–T BOUNDARY TRANSITION 285
more than a few centimetres. Unless palaeontologists are to
regard every unexpected occurrence of a species as a priori
evidence for reworking throughout the stratigraphical column,
the hypothesis of reworking must be veriﬁed by independent
criteria. MacLeod (1994) reviewed the criteria that may be used
to infer reworking in fossil faunas. MacLeod & Keller (1994)
and MacLeod (1995a,b,c, 1996a,d) applied these tests to the
K–T planktonic foraminiferal record.
Although the K–T records of most fossil groups have yet to
be examined in detail for sampling problems and tested for
reworking, by far the most serious factor complicating our
interpretation of K–T extinction patterns is the current lack of
adequate phylogenies at the species and higher taxonomic
levels. Fortey (1983), remarking on the sudden extinction of
many trilobite families at the end of the Cambrian and the
sudden origination of many families in the Ordovician, pointed
out that this phenomenon may be a taxonomic artefact. Many
of the Cambrian families contain members more closely related
to Ordovician trilobites than to other members of their own
family. In other words, many Cambrian families are para-
phyletic (see Fig. 3). These families did not become extinct
because they were never real in the ﬁrst place.
Use of phylogenetic relationships in extinction studies has
been exempliﬁed by Norell (1992), Archibald (1993) and Smith
(1994). Consider the theoretical example illustrated in Figure
10, concerning four taxa (A–D) with observed stratigraphical
ranges as given in solid heavy lines. The phylogenetic tree is
shown to the right. This tree is constructed by identifying
morphological characters uniquely shared by A and B, C and
D, and A, B, C and D. The distribution of morphological
characters implies that A and B shared a unique ancestor ‘x’, C
and D shared a common ancestor ‘y’, while A, B, C, D shared
a common ancestor ‘z’. Inclusion of stratigraphical data
imposes some time limits on the occurrence of these ancestors
as well as the longevity of diﬀerent lineages.
This phylogenetic tree implies that sister-groups must be
of equal age. Thus, in Fig. 10, taxon B must have existed at
least as long as taxon A; similarly, D must have existed as
long as C; and A+B must have existed as long as C+D.
Consequently the age of taxa B and D must, of necessity, be
extended back as range extensions to match that of A and C
respectively. Additionally there must have been an undetected
lineage of species which occupied the time interval between
A+B and C+D; this is referred to as a ghost lineage (Norell
1992). Of course, we have no idea of how many additional
taxa may have occupied the time interval occupied by
either the range extension or the ghost lineage, and much of
palaeontological research is directed towards exorcising the
In this example it is assumed that taxa A–D are each
monophyletic and can be recognized as such because each
shows a unique character that serves as its morphological
ﬁngerprint. However, in palaeontological studies this is not
always the case. Often there are taxa (species or genera) to
which we give names without recognizing a unique attribute
(e.g. stratigraphically or geographically recognized species). If
one of these kinds of taxa should precede another taxon
without overlapping in time, there remains the possibility that
it is not monophyletic (it may be paraphyletic or a direct
ancestor). For instance if taxon C had no recognizable ﬁnger-
print it may have been ancestral to or part of D (Fig. 10c).
These kinds of palaeontological taxa have been called
metaphena by Smith (1994) and metataxa by Archibald
(1993). When counting extinctions these need to be left out of
consideration because the reality of their extinction cannot be
Phylogenetic considerations can give only a minimum idea
of diversity. But the consideration of phylogeny can suggest
patterns of extinction and origination which diﬀer from those
based on a taxonomic approach only. Smith (1994, ﬁg. 7.1)
worked through theoretical examples showing that phylo-
genetic considerations can lead to alternative estimations of
diversity, times of origination, and times of extinction. For
example, ﬁshes of the Superorder Osteoglossomorpha, the
most primitive group of living teleosts, have a fossil record
extending down to the Upper Jurassic. All modern species are
freshwater although there are marine fossil representatives.
Relationships amongst osteoglossomorphs are quite well
understood at both the family and generic level. The most
recent generic phylogeny is that of Li (1994) and this is plotted
in Fig. 11 against time. Like most ﬁsh records, species and
genera are often only recognized from one time interval (and
often one locality). Taking standing diversity through time, the
taxonomic approach counts only fossil occurrences (allowing
for Lazarus taxa, e.g. the ﬁsh genus Hiodon is not known from
all intervening stages between the Eocene and the Recent).
This leads to a pattern shown in the left side of the ﬁrst column
to the right where there is a burst of diversity in the Eocene.
(Note: the Eocene burst in diversity for this group is not due to
the Lagerstätten deposits mentioned above.)
The phylogenetically inferred pattern of standing diversity,
extinctions, and originations is the more faithful record of
the evolutionary history of the group. Consideration of the
phylogeny includes the counting of ghost lineages and range
extensions, and leads to the pattern in the right-hand side
of the ﬁrst column which shows not only a peak of diversity in
the Eocene but a smaller one in the mid-Cretaceous. Even
more interestingly, whereas the taxonomic summary seems
to suggest an exponential increase in standing diversity in
the Palaeocene (after a long, stable period of low diversity
throughout the Cretaceous), the phylogenetically corrected
data exhibit a sustained radiation in this group starting in
the Santonian and continuing through the Eocene with no
substantive change across the K–T boundary.
With respect to originations and extinctions, the taxonomic
approach yields the patterns in the left-hand side of the second
Fig. 10. Theoretical example illustrating the phenomena of range
extensions (thick dashed lines) and ghost lineages (narrow dashed
lines). Phylogeny at top right assumes all taxa are monophyletic.
Phylogeny at bottom right assumes that taxon C is a metataxon and
may be ancestral to D.
286 N. MLEOD ET AL.
and third columns respectively. These patterns match one
another closely because many genera are restricted to a single
stage. Clearly this is a reﬂection of a sampling problem which,
for originations, vanishes if a phylogenetic approach is
adopted by counting the number of inferred branching events
leading to the origin of new clades against each time stage
(right-hand of second column). Until summaries such as these
have been prepared for a large number of Cretaceous–Tertiary
clades the real magnitude and pattern of the K–T extinction
event will remain obscure.
The record of biotic transitions at and around the K–T
boundary is based on a mix of observations, ranging from
detailed and authoritative biostratigraphical data synthesized
from global sections and cores that represent a large suite of
depositional environments (e.g. planktonic foraminifera), to
taxonomic studies of higher categories based upon limited data
from a relatively small number of geographical and ecological
settings (e.g. lepidosaurs). Despite this variation, certain
conclusions can be drawn from the current database.
First, global events at the K–T boundary occurred within a
longer period of sustained biotic change. This longer episode
aﬀected diﬀerent groups at diﬀerent times, but most often
manifested itself as a progressive reduction in biotic diversity
throughout the Maastrichtian.
Second, a much shorter term global biotic event appears to
have taken place close to the K–T boundary. This event is
most prominent among a few groups of marine microfossils
(e.g. calcareous nannoplankton, planktonic foraminifera),
which seem to have remained relatively unaﬀected by the long-
term Maastrichtian decline. Perhaps one of the questions we
should be asking is why marine plankton were resistant to the
longer-term Maastrichtian biotic crisis. The extent to which
this short-term, near-boundary event was inﬂuenced (or pre-
cipitated) by a bolide impact is uncertain. In most microfossil
lineages, with the possible exception of calcareous nanno-
plankton, decline in species numbers begins prior to the
occurrence of impact debris in various K–T boundary succes-
sions. The uppermost Maastrichtian decline in some of these
groups may be a sampling artefact (e.g. Signor–Lipps eﬀect) or
it may reﬂect a biotic response to terrestrially forced global
events (e.g. general climatic cooling due to intense volcanic
activity, eustatic sea-level change). Given the patchy occur-
rence and low relative abundance of most microfossil species it
will take sampling programmes of unprecedented intensity
(even by K–T standards), coupled with sophisticated forms of
data analysis, to resolve this question. It is interesting to note,
however, that there does seem to be a marked geographical
heterogeneity to the observed extinction pattern within this
short-term K–T event, with high-latitude faunas and ﬂoras
exhibiting reduced rates of extinction intensity. Such patterns
are diﬃcult to explain as sampling artefacts.
Third, the intensive scrutiny of lowermost Danian strata has
resulted in the discovery of many more examples of the
occurrence of Cretaceous morphotypes than were previously
thought to exist (e.g. calcareous nannoplankton, planktonic
foraminifera, dinoﬂagellates). Debates as to whether these
taxa represent true survivors or reworked components of
Cretaceous faunas are ongoing. In several instances geographi-
cal, morphological and isotopic tests of these seemingly
anomalous Cretaceous faunas in Danian beds have supported
the survivorship, rather than reworking, interpretation.
Fourth, biotic diversity needed a substantial time interval to
recover from the K–T event. Except for the lowermost Danian
interval, in which very reduced low and middle latitude faunas
predominate, rates of new species accumulation were unprec-
edentedly high in many (but not all) groups, no doubt reﬂect-
ing sustained radiation into vacant ecospace. Nevertheless,
Fig. 11. Generic-level phylogeny of teleosts of the Superorder Osteoglossomorpha (after Li 1994) plotted against time. To the right are paired
columns reﬂecting standing diversity, originations and extinctions. In each pair the left-hand columns counts occurrences in the fossil record
ignoring any phylogenetic framework while the right-hand columns take the phylogeny into consideration.
THE K–T BOUNDARY TRANSITION 287
new data (Jablonski 1995) suggest that North American
marine bivalve faunas lagged signiﬁcantly behind those of
Europe and Asia in terms of their recovery rate. It remains to
be seen whether this pattern is also present in other groups.
Fifth, the most tractable and signiﬁcant area in which
improvements can be made in the current K–T database is via
the determination of accurate species- and genus-level phylog-
enies. So long as detailed phylogenetic relationships between
the Cretaceous and Danian taxa remain unknown or subject
to arbitrary taxonomic convention, the magnitude, pattern,
and ecological–geographical character of both long- and short-
term K–T biotic events will resist accurate determination.
Many colleagues helped us assemble portions of the data presented
herein, including: G. Bell, R. Fortey, A. Gale, J. Hooker, G. Keller,
D. Jablonski, J. Pospichal, D. Ward, R. C. Whatley, and J.
Williams. Comments and suggestions during the JGS review by
S. G. Molyneux, A. Crame, and an anonymous reviewer were also
very helpful. M. Kaminski also acknowledges support from NATO
Collaborative Research Grant No. 890149. This article was based on
presentations and discussions that took place during a University
College London symposium entitled ‘The Cretaceous–Tertiary
Faunal Transition’, sponsored by the Natural History Museum–
University College London (NHM-UCL) Global Change and the
Biosphere Programme during November 1994, and represents a
contribution from that programme.
A, S., S, N., N, M., G, G., G, T.,
K,E.&A, B. 1993. Geological events at the Cretaceous–
Palaeogene boundary in Georgia (Caucasus). Geologica Balcanica,23,
A, D.V. 1973. The Nature of the Stratigraphical Record. John Wiley & Sons,
—— 1993. The New Catastrophism: The Importance of the Rare Event in
Geological History. Cambridge University Press, Cambridge.
A, L.W., A, W., A,F.&M, H. 1980. Extraterrestrial
cause for the Cretaceous–Tertiary extinction. Science,208, 1095–1108.
A, J.D. 1993. The importance of phylogenetic analysis for the assess-
ment of species turnover: a case history of Paleocene mammals in North
America. Paleobiology,19, 1–27.
A,J.D.&B, L.J. 1990. Diﬀerential Cretaceous–Tertiary extinc-
tion of nonmarine vertebrates; evidence from northeastern Montana. In:
S,V.L.&W, P.D. (eds) Global catastrophes in Earth history:
an interdisciplinary conference on impacts, volcanism, and mass mortality.
Geological Society of America, Special Papers, 247, 549–562.
A, U. 1979. The irregular echinoids and the boundary in Denmark. In:
B,T.&B, R.G. (eds) Cretaceous/Tertiary boundary events
symposium. I. The Maastrichtian and Danian of Denmark. University of
A,R.A.&J, S.R. 1996. Palynological change across the
Cretaceous–Tertiary boundary on Seymour Island, Antarctica: environ-
mental and depositional factors. In:ML,N.&K, G. (eds) The
Cretaceous–Tertiary mass extinction: biotic and environmental events.W.W.
Norton & Co., New York, 7–26.
A, H., B, E., B, A.D., C, J., C, C., E,
R., G-G, F., J, J.J., J-F, E., LL,
J., M, J.M., O-E, X., P-S, J., P,
J.E., R, J.C., R-L, J., S,J.L.&T, H. 1990. The
fossil vertebrates from Lan˜o (Basque Country, Spain); new evidence on the
composition and aﬃnities of the Late Cretaceous continental faunas. Terra
B, E. 1994. Global environmental changes preceeding the Cretaceous–
Tertiary boundary: Early-late Maastrichtian transition. Geology,22, 877–
——&K, G. 1990. Foraminiferal stable isotope evidence for gradual
decrease of marine productivity and Cretaceous species survivorship in the
earliest Danian. Paleoceanography,5, 867–870.
—— & —— 1994. Productivity across the Cretaceous-Tertiary boundary in high
latitudes. Geological Society of America Bulletin,106, 1254–1266.
B, M.J. 1993. Reptilia. In:B, M.J. (ed.) The fossil record 2.
Chapman & Hall, London, 681–715.
B,W.&H, G.R. 1968. Vertical mixing in pelagic sediments. Journal
of Marine Research,26, 134–143.
B, W.A., K, D.V., S, C.C., III & A, M.-P. 1995. A
revised Cenozoic geochronology and chronostratigraphy. In:B,
W.A., K, D.V., A, M.-P. & H, J. (eds) Geochronology,
time scales, and global stratigraphic correlation. SEPM (Society for
Sedimentary Geology) Special Publications, 54, 129–212.
B,T.&H, E. 1982. The terminal Cretaceous in Boreal shelf
seas: a multicausal event. In:S,L.T.&S, P.H. (eds) Geological
implications of impact of large asteroids and comets on the Earth. Geological
Society of America Special Papers, 190, 373–384.
B, N.N. 1958. Paleocene Radiolaria of western Kubanj. Trudy Vsesoy-
uznyi Neftegazovyi Nauchno-Issledovalelskii Institut (VNII), Krasnodarskii
—— 1960. New radiolarians from the Paleocene deposits of the Kubanj. Trudy
Vsesoyuznyi Neftegazovyi Naucho-Issledovalelskii Institut (VNII),
Krasnodarskii Filial,4, 199–207.
B, M.N. 1965. Massive extinctions in biota at the end of Mesozoic
time. Science,148, 1696–1699.
B,H.&Z, W.J. 1988. Dinoﬂagellate cysts, sea level changes
and planktonic foraminifers across the Cretaceous-Tertiary boundary at El
Haria, northwest Tunisia. Marine Micropaleontology,13, 153–191.
B, K. 1972. Cyclostomatous Bryozoa from the Upper Cretaceous and Danian
in Scandinavia. Stockholm Contributions in Geology, 26.
—— 1976. Cyclostomatous Bryozoa from the Paleocene and Maastrichtian of
Majunga basin, Madagascar. Geobios,9, 393–423.
B,E.M.&DD, P. 1993. Late Maastrichtian and Danian
ostracode faunas from northern Alaska: reconstructions of environment
and paleogeography. Palaios,8, 140–154.
B, J.R. 1991. A Paleocene coral-algal-sponge reef from southwestern
Alabama and the ecology of Early Tertiary reefs. Lethaia,24, 423–438.
C, I.J., K,G.&M, E. 1991. Cretaceous/Tertiary boundary
extinction pattern and faunal turnover at Agost and Caravaca, S. E. Spain.
Marine Micropaleontology,17, 319–341.
C, H. 1987. Extinctions et renouvellements fauniques chez les sélachiens
postjurassiques. Mémoires de la société géologique de France, new series,
——, D,C.&Z, J. 1993. Chondrichthyes. In:B, M.J. (ed.) The
fossil record 2. Chapman & Hall, London, 593–609.
C,S.M.&W, M.H. 1977. Immunological comparisons of caecilian
albumins (Amphibia: Gymnophiona). Herpetologica,33, 94–98.
C, L.M. 1995. The ﬁrst 85 million years of avian evolution. Nature,378,
C,A.G.&J, J.B.C. 1987. Clonal growth, algal symbiosis, and reef
formation by corals. Paleobiology,13, 363–378.
C,R.&G, S. 1994. K–T boundary extinction: geologically
instantaneous or gradual event? Evidence from deep-sea benthic foramini-
fera. Geology,22, 779–782.
C, G. 1990. A comparison of the evolution, diversity, and composition of
the Cainozoic Ostracoda in the deep water North Atlantic and shallow
water environments of North America and Europe. In:W, R.C. &
M, C.A. (eds) Ostracoda and Global Events. Chapman & Hall,
C, A. 1994. Calibration of mitochondrial 12s rDNA sequences indicate
many modern avian orders survived the K–T boundary impact. Geological
Society of America, Abstracts with Programs,26, A-395.
D’H, S. 1995. Carbon isotopic recovery from mass extinctions: no
Strangelove oceans on geologic timescales. Geological Society of America
Abstracts with Programs,27, A-164.
D, P., D,D.T.&N, M. 1994. Phylogeny and systematics of the
Coleoidea. University of Kansas Paleontological Contributions,5, 1–15.
D, P., C, J.P., D, R., O, H.J., P, J.-P. & S,
R. 1982. Les ostracodes du Campanien terminal à l’Eocene inférieur de la
coupe du Kef, Tunisie Nord-orientale. Bulletin des Centres de Recherches
Exploration-Production Elf-Aquitaine,6, 307–335.
D, W.S. 1967. Palynology of the Upper Moreno Formation (late
Cretaceous-Paleocene) Escarpado Canyon, California. Palaeontographica,
Abteilung B, 118, 1–71.
D, P. 1973. Paleocene Radiolaria, DSDP Leg 21. In:B, R.E.
(eds) Initial Reports of the Deep Sea Drilling Project,21 U.S. Government
Printing Oﬃce, Washington, D. C., 787–817.
D, P.M. 1880. Sind fossil corals and Alcyonaria. Memoirs of the Geologi-
cal Survey of India. Palaeontologica Indica (Series 7 & 14), 1, 1–110.
288 N. MLEOD ET AL.
E, T.W. 1993. Late Cretaceous (Maastrichtian) calcareous nanno-
plankton biogeography with emphasis on events immediately preceding the
Cretaceous/Palaeocene boundary. Woods Hole Oceanographic Institution,
E, D.H., A, R.A., K,F.T.&Z, W.J. 1994. Iridium and
dinocysts at the Cretaceous–Tertiary boundary on Seymour Island,
Antarctica: implications for the K–T event. Geology,22, 347–355.
E, R. 1983. Sauria terrestria, Amphisbaenia. Handbuch der paläoherpetologie.
Gustav Fischer Verlag, Stuttgart.
F, A. 1995. Explosive evolution in Tertiary birds and mammals. Science,
F, J. 1985. Late Cretaceous and Palaeogene diatom biostratigraphy. In:
B, H.M., S,J.S.&P-N, K. (eds) Plankton strati-
graphy. Cambridge University Press, Cambridge, 713–762.
F, H.P. 1968. Upper Maestrichtian Radiolaria of California. Special
Papers in Paleontology,3, 1–82.
F, R.A. 1983. Cambrian–Ordovician trilobites from the boundary beds in
western Newfoundland and their phylogenetic signiﬁcance. Special Papers
in Palaeontology,30, 179–211.
F,D.L.&M, E.S. 1951. Paleocene Radiolaria from southeastern
Missouri. University of Missouri School of Mines and Metallurgy, Technical
G, A.S. 1986. Goniasteridae (Asteroidea, Echinodermata) from the late
Cretaceous of north-west Europe. 1. Introduction. The genera Metopaster
and Recurvaster.Mesozoic Research,1, 1–69.
—— 1987. Goniasteridae (Asteroidea, Echinodermata) from the late Cretaceous
of north-west Europe. 2. The genera Calliderma,Crateraster,Nymphaster
and Chomataster.Mesozoic Research,1, 151–186.
G, W.B. 1991. Selective extinction and survival across the Cretaceous/
Tertiary boundary in the northern Atlantic Coastal Plain. Geology,19,
G, B.G. 1993. Osteichthyes: basal actinopterygians. In:B, M.J.
(ed.) The fossil record 2. Chapman & Hall, London, 612–619.
G, S. 1996. Calcareous nannofossils at the Cretaceous–Tertiary bound-
ary. In:ML,N.&K, G. (eds) The Cretaceous–Tertiary mass
extinction: biotic and environmental events. W. W. Norton & Co., New
G, Z., C,L.M.&F, M. 1991. A new Senonian
peirosaurid (Crocodylomorpha) from Argentina and a synopsis of the
South America Cretaceous crocodiles. Journal of Vertebrate Paleontology,
G, P.M. 1990. The skeletal morphology and systematics of Recent and
fossil holothurians with particular reference to the Triassic/Jurassic. PhD
thesis, University of Exeter.
—— 1992. The skeletal morphology, systematics and evolutionary history of
holothurians. Special Papers in Palaeontology, 47.
G, R.N., C, J.I., K, G., M, B.A., O,R.K.,
O´-E,X.&S, J. (in press). The Cretaceous–Tertiary
boundary: the El Kef blind test. Marine Micropaleontology.
G, M.F. 1937. Studien über Foraminiferen aus der Kreide und dem
Tertiär des Kaukasus. I—Die Foraminiferen der ältesten Tertiärschichten
des Nordwestkaukasus. Problems in Paleontology, Moscow University
Laboratory of Paleontology,2–3, 349–410.
G, E.V. 1988. The correlation of heterofacial Maastrichtian deposits
of the Tadzhik Depression by radiolarians. Doklady Akademii Nauk
Tadzhikskoy SSR,31, 202–206.
G, P. 1979. Remarks on the regular echinoids in the Upper Maastrich-
tian and Lower Danian of Denmark. In:B,T.&B, R.G.
(eds) Cretaceous/Tertiary boundary events symposium. I. The Maastrichtian
and Danian of Denmark. University of Copenhagen, 72–73.
G, D., W, D.B., N, D.B., S, M., R, M.,
B,A.&T, V. 1994. Late Maastrichtian dinosaur eggs
from the Haneg basin (Romania). In:C, K., H, K.F. &
H, J.R. (eds) Dinosaur eggs and babies. Cambridge University Press,
H˚, E., K,C.R.&T, E. 1996. Benthic recovery subsequent
to the Cretaceous-Tertiary boundary—the European example. In:
R, J.E. (ed.) Sixth North American Paleontological Convention
Abstracts of Papers. The Paleontological Society Special Publications, 8,
H˚,E.&T, E. 1979. Distribution and types of bryozoan
communities at the boundary in Denmark. In:B,T.&B,
R.G. (eds) Cretaceous–Tertiary boundary events. I. The Maastrichtian and
Danian of Denmark. University of Copenhagen, Copenhagen, 78–91.
H, H.J. 1990. Diachronous extinctions at the K/T boundary: a scenario. In:
S,V.L.&W, P.D. (eds) Global catastrophes in Earth history:
an interdisciplinary conference in impacts, volcanism, and mass mortality.
Geological Society of America, Special Papers, 247, 417–423.
——, R, K.L., G,R.&K, H. 1987. Iridium-bearing
carbon black at the Cretaceous–Tertiary boundary. Bulletin of the Geologi-
cal Society of Denmark,36, 305–314.
H, B. 1991. Sequence stratigraphy, sea-level change and signiﬁcance for the
deep sea. International Association of Sedimentologists, Special Publications,
——, H,J.&V, P.R. 1987. Chronology and ﬂuctuating sea levels
since the Triassic. Science,235, 1156–1166.
H, P.E. 1986. The relationship of some fossil diatom genera to resting
spores. Proceedings of the Eighth International Diatom Symposium, 33–46.
——&F, F.W. 1983. Diatom resting spores: signiﬁcance and strategies.
In:F, G.A. (ed.) Survival strategies of the algae. Cambridge
University Press, Cambridge, 49–68.
H, D.M. 1988. Upper Cretaceous and Lower Paleocene diatom and
silicoﬂagellate biostratigraphy of Seymour Island, Eastern Antarctic
Peninsula. In:F,R.M.&W, M.O. (eds) Geology and
Paleontology of Seymour Island, Antarctic Peninsula. Geological Society of
America Memoirs, 169, 55–129.
——&G, R. 1990. Lower Cretaceous diatoms from ODP Leg 113, Site
693 (Weddell Sea). Part 2: resting spores, chrysophycean cysts, an endo-
skeletal dinoﬂagellate, and notes on the origin of diatoms. Proceedings of
the Ocean Drilling Program, Scientiﬁc Results,113, 403–425.
H, S.E., &P, B. 1987. A reﬁned nannoplankton zonation for
the Danian of the central North Sea. Abhandlungen der Geologischen
H, H. 1976. International Stratigraphic Guide: A Guide to Stratigraphic
Classiﬁcation, Terminology, and Procedure. John Wiley & Sons, New York.
H, S.B., N,R.A.&M, L.R. 1993. Caecilian phylogeny and
biogeography inferred from mitochondrial DNA sequences of the 12S
rRNA and 16S rRNA genes (Amphibia: Gymnophiona). Herpetological
H, C.J. 1991. Latest Cretaceous to Late Paleocene Radiolaria from
Marlborough (New Zealand) and DSDP Site 208. PhD thesis, University of
—— 1993. Latest Cretaceous to Late Paleocene radiolarian biostratigraphy: a
new zonation from the New Zealand region. Marine Micropaleontology,21,
—— 1996. Radiolarian faunal change through the Cretaceous–Tertiary transi-
tion of eastern Marlborough, New Zealand. In:ML,N.&K,
G. (eds) The Cretaceous–Tertiary Mass extinction: biotic and environmental
events. W. W. Norton & Co., New York, 173–204.
——&H, J.A. 1991. Well-preserved late Paleocene Radiolaria from
Tangihua Complex, Camp Bay, eastern Northland. Tane,33, 65–76.
H, J.J., M,A.C.&S, S.E.K. 1991. An ornithopod dinosaur
from the Late Cretaceous of West Antarctica. Antarctic Science,3, 331–
H, M.R. 1993. Fluctuations in ammonoid evolution and possible environ-
mental controls. In:H, M.R. (ed.) The Ammonoidea: environment,
ecology, and evolutionary change. Systematics Association Special Volumes,
H¨ , K.J. & M, J.A. 1985. A ‘‘strangelove’’ ocean in the earliest
Tertiary. In:S,E.T.&B, W.S. (eds) Natural variations:
Archean to present. American Geophysical Union Monographs, 32, 487–
H, B.T., H,D.A.&H, C.P. 1995. Middle-Late Cretaceous
climate of the southern high latitudes: stable isotopic evidence for minimal
equator-to-pole thermal gradients. Geological Society of America Bulletin,
H,S.H.&A, J.D. 1995. No stastistical support for sudden (or
gradual) extinction of dinosaurs. Geology,23, 881–884.
J, D. 1995. The biogeography of rebounds: comparisons among K–T
bivalves. Geological Society of America Abstracts with Programs,27, A-164.
——&R, D.M. 1995. Selectivity of end-Cretaceous marine bivalve extinc-
tions. Science,268, 389–391.
J, J.W.M. 1995. Late Cretaceous and early Cainozoic crinoid assemblages
from northeast Belgium and the southeast Netherlands. In:E, R.H.,
S,A.B.&C, A.C. (eds) Echinoderm Research 1995.A.A.
Balkema, Rotterdam, 185–196.
J, C.H., in press, a. Carrying on regardless: changes in the Cretaceous
echinoid genus Cyclaster at the Cretaceous–Tertiary boundary. Proceedings
of the 9th International Echinoderms Conference, San Francisco.A.A.
THE K–T BOUNDARY TRANSITION 289
——, in press, b. All change at the K–T boundary: echinoids from the
Maastrichtian and Danian of the Mangyshlak Peninsula, Kazachstan.
——&S, A.B., in press. Estimating extinction levels and changes in
diversity and disparity of echinoids across the Cretaceous–Tertiary bound-
ary. Proceedings of the 9th International Echinoderms Conference, San
Francisco. A. A. Balkema, Rotterdam.
J, M.B. 1987. Brachiopods from the Maastrichtian-Danian boundary
sequence at Nye Kløv, Jylland, Denmark. Fossils and Strata,20, 1–99.
J,C.C.&K, E.G. 1996. Maastrichtian extinction patterns of
Caribbean province rudistids. In:M,N.&K, G. (eds) The
Cretaceous–Tertiary Mass Extinction: Biotic and Environmental Changes.
W. W. Norton & Co., New York, 231–274.
J,K.R.&G, D. 1993. High-latitude deciduous forests and the
Cretaceous–Tertiary boundary in New Zealand. Geological Society of
America, Abstracts with Programs,25, A-50.
——&H, L.J. 1990. Megaﬂoral change across the Cretaceous–Tertiary
boundary in the northern Great Plains and Rocky Mountains, U.S.A. In:
S,V.L.&W, P.D. (eds) Global catastrophes in Earth History:
an interdisciplinary conference on impact, volcanism, and mass mortality.
Geological Society of America, Special Papers, 247, 433–444.
J, A.P. 1978. Diatom biostratigraphy on the generic level. Micropaleontol-
K, M.A. 1985. Evidence for control of abyssal agglutinated community
structure by substrate disturbance: results from HEBBLE area. Marine
——, G,J.F.&W, R.B. 1988. Life history and recolonization
among agglutinated foraminifera in the Panama basin. In:G,
F.M. & R¨, F. (eds) Proceedings of the Second International Workshop
of Agglutinated Foraminifera. Abhandlungen der geologischen Bundesan-
stalt (Wien), 41, 229–244.
K, E.G. 1984. The fabric of Cretaceous extinctions. In:B,
W.A. & C, J.A. (eds) Catastrophes and Earth History: The
New Uniformitarianism. Princeton University Press, Princeton, 151–246.
——&J, C.C. 1988. The morphological and ecological evolution of
Middle and Upper Cretaceous reef-building rudistids. Palaios,3, 194–216.
K, G. 1988a. Biotic turnover in benthic foraminifera across the
Cretaceous–Tertiary boundary at El Kef, Tunisia. Palaeogeography,
Palaeoclimatology, Palaeoecology,66, 153–171.
—— 1988b. Extinction, survivorship and evolution of planktic foraminifera
across the Cretaceous/Tertiary boundary at El Kef, Tunisia. Marine
—— 1989. Extended Cretaceous/Tertiary boundary extinctions and delayed
population change in planktonic foraminiferal faunas from Brazos River,
Texas. Paleoceanography,4, 287–332.
—— 1993. The Cretaceous–Tertiary boundary transition in the Antarctic Ocean
and its global implications. Marine Micropaleontology,21, 1–46.
——, B, E., S,B.&M, E. 1993. Gradual mass extinction,
species survivorship, and long-term environmental changes across the
Cretaceous–Tertiary boundary in high latitudes. Geological Society of
America Bulletin,105, 979–997.
——&B, C. 1991. Paleoenvironment of the eastern Tethys in the Early
Paleocene. Palaios,6, 439–464.
——, L,L.&ML, N. 1995. The Cretaceous/Tertiary boundary stratotype
section at El Kef, Tunisia: how catastrophic was the mass extinction?
Palaeogeography, Palaeoclimatology, Palaeoecology,119, 255–273.
——&L, M. 1989. Stable isotope, TOC and CaCO
the Cretaceous/Tertiary boundary at El Kef, Tunisia. Palaeogeography,
Palaeoclimatology, Palaeoecology,73, 243–265.
K, W.J. 1993. Ammonite faunas of the European Maastrichtian; diver-
sity and extinction. In:H, M.R. (ed.) The Ammonoidea: environment,
ecology, and evolutionary change. Systematics Association Special Volume,
K´,P.&S´, B. 1984. L’aport des coquilles d’oeufs de dinosairs de
Laguna Umayo à l’âge de la Formation Vilquechico (Pérou) et à la
compréhension de Perutherium altiplanense.Newsletters on Stratigraphy,
K, U. 1994. Di Entwicklung der kalkigen Nannofossilien und der kalkigen
Dinoﬂagellaten-Zysten an der Kriede/Tertiär-Grenze in Westbrandenburg
im Vergleich mit Proﬁlen in Nordjütland und Seeland (Dänemark). Berliner
Geowissenschaftliche Abhandlungen, Reihe E,12, 1–87.
K, P.M. 1974. Evolutionary trends and their functional signiﬁcance in the
post-Paleozoic echinoids. Paleontology Society Memoirs 5(Journal of
Paleontology,48, supplement), 1–95.
K, J.A., C,D.L.&G, A.M. 1986. Biological selectivity of
extinction: A link between background and mass extinction. Palaios,1,
K, G.E. 1983. Radiolarian complexes of Boreal regions in the lower
Paleocene. In:L,P.S.&M, E.V. (eds) The use of
microfauna in the study of sediments from the continents and oceans
(miscellaneous scientiﬁc reports). Proceedings of the All Union Petroleum
Scientiﬁc Research Institute for Geological Survey (VNIGRI). 84–112.
K,W.&H, S. 1994. Benthic foraminifera (Sonne-95 cruise). Berichte-
Reports Geologisch-Paläontologisches Institut der Universität Kiel,68, 215–
——&K, M.A. 1993. Changes in the community structure of deep water
agglutinated foraminifers across the K/T boundary in the Basque basin
(northern Spain). Revista Espan˜ola de Micropaleontología,25, 57–92.
L,J.&T´, P. 1909–1925. Essai de nomenclature raisonée des
L,A.&D, W.W. 1993. Phylogenetic relationships of the sala-
mander families: an analysis of congruence among morphological and
molecular characters. Herpetological Monographs,7, 77–93.
L, G.-Q. 1994. New osteoglossomorphs (Teleostei) from the Upper Cretaceous
and Lower Tertiary of North America and their phylogenetic signiﬁcance.
PhD thesis, University of Alberta.
L, S., MK,F.K.&T, P.D. 1993. Competition, clade
replacement, and a history of cyclostome and cheilostome bryozoan
diversity. Paleobiology,19, 352–371.
L, H. Y. 1991. Cretaceous (Maestrichtian) radiolarians: Leg 114. In:
.(eds) Proceedings of the Ocean Drilling Program,
Scientiﬁc Results. The Ocean Drilling Program, College Station, Texas,
L,J.H.&H, C.S. 1982. Origin, age, and evolution of Antarctic and
deep-sea faunas. In:E,W.G.&M, J.G. (eds) The environment of
the deep sea. Prentice-Hall, Englewood Cliﬀs, N. J., 324–356.
L,C.&O, R.K. 1992. Evolutionary radiation of microperforate plank-
tonic foraminifera following the K/T mass extinction. Journal of Foramini-
feral Research,22, 328–346.
M, C.E. 1986. Late Campanian-Maastrichtian ammonite fauna from
Seymour Island (Antarctic Peninsula). Journal of Paleontology, Memoirs,
ML, N. 1994. An evaluation of criteria that may be used to identify species
surviving a mass extinction. New Developments Regarding the K/T Event
and Other Catastrophes in Earth Histroy, LPI Contribution,825, 75–77.
—— 1995a. Cretaceous/Tertiary (K/T) biogeography of planktic foraminifera.
Historical Biology,10, 49–101.
—— 1995b. Graphic correlation of high latitude Cretaceous–Tertiary boundary
sequences at Nye Kløv (Denmark), ODP Site 690 (Weddell Sea), and ODP
Site 738 (Kerguelen Plateau): comparison with the El Kef (Tunisia)
boundary stratotype. Modern Geology,19, 109–147.
—— 1995c. Graphic correlation of new Cretaceous/Tertiary (K/T) boundary
sections. In:M,K.O.&L, H.R. (eds) Graphic Correlation. Society
for Sedimentary Geology Special Publications, 53, Tulsa, 215–233.
—— 1996a. Nature of the Cretaceous–Tertiary (K–T) planktonic foraminiferal
record: stratigraphic conﬁdence intervals, Signor–Lipps eﬀect, and patterns
of survivorship. In:ML,N.&K, G. (eds) The Cretaceous–
Tertiary mass extinction: biotic and environmental changes. W. W. Norton &
Co., New York, 85–138.
—— 1996b. Stratigraphic completeness and planktic foraminiferal survivorship
across the Cretaceous–Tertiary (K/T) boundary. In:M,A.&
W, R. (eds) Microfossils and Oceanic Environments. University of
Wales, Aberystwyth, 327–353.
—— 1996c. K–T Redux. Paleobiology,22, 311–317.
—— 1996d. Testing patterns of Cretaceous–Tertiary planktonic foraminiferal
extinctions at El Kef (Tunisia). In:R, G., F,D.&G,
S. (eds) The Cretaceous–Tertiary Event and Other Catastrophes in Earth
History. Geological Society of America, Special Papers, 307, 287–302.
——&K, G. 1991. How complete are Cretaceous/Tertiary boundary
sections? A chronostratigraphic estimate based on graphic correlation.
Geological Society of America Bulletin,103, 1439–1457.
—— & —— 1994. Comparative biogeographic analysis of planktic foraminiferal
survivorship across the Cretaceous/Tertiary (K/T) boundary. Paleobiology,
M, R.F. 1985. Ostracoda of the Cretaceous–Tertiary contact sections in
central Texas. Gulf Coast Association of Geological Societies Transactions,
M, N. 1990. Revision der Kreide-Austern (Bivalvia: Pteriomorpha)
Aegyptens (Biostratigraphie, Systematik). Berliner Geowissenschaftliche
Abhandlungen, Reihe A,125, 1–231.
290 N. MLEOD ET AL.
M,I.I.&S, K.H. 1994. Cretaceous–Tertiary boundary events in
the area of Bjala, eastern Bulgaria—biostratigraphical results. Geologica
M, L., B, E.M., H,D.M.&MK, M.C. 1990.
Late Mesozoic and Cenozoic paleogeographic and paleoclimatic history of
the Arctic Ocean basin, based on shallow-water marine faunas and
terrestrial vertebrates. In:The Arctic Ocean Region. The Geology of North
America, 1, Geological Society of America, Boulder, 403–426.
M, C.R. 1995. Distinguishing between sudden and gradual extinctions
in the fossil record: predicting the position of the Cretaceous–Tertiary
iridium anomaly using the ammonite fossil record on Seymour Island,
Antarctica. Geology,23, 731–734.
M, R., S, F., S,G.&R, A. 1982. Palaeoenviron-
mental signiﬁcance of Maastrichtian biological communites in the
Pachino area (Sicily) and preliminary data on their distribution in the
Mediterranean Upper Cretaceous. In:M G, E. (ed.)
Palaeontology, essential of historical geology. S.T.E.M. Mucchi, Modena,
MG,G.&E, S. 1995. Albanerpetontid amphibians from the
Cretaceous of Spain. Nature,373, 143–145.
M, K.G. 1982. Cenozoic benthic foraminifera: case histories of paleoceano-
graphic and sea-level changes. In:B, T.W. (ed.) Foraminifera,
notes for a short course. University of Tennessee, Studies in Geology No. 6,
M,D.H.&MK, C.P. 1982. Response of marine plankton communities
to a global atmospheric darkening. In:S,L.T.&S, P.H. (eds)
Geological implications of impacts of large asteroids and comets on the Earth.
Geological Society of America Special Papers, 190, 297–303.
M, A.R. 1993. Amphibian-grade Tetrapoda. In:B, M.J. (ed.) The
fossil record 2. Chapman & Hall, London, 663–677.
M, M.M., S,A.N.&E, L.G. 1980. [Class Echinoidea.]
In:M, V.V. (ed.) [Evolution and change of invertebrates at the
boundary of the Mesozoic and Cenozoic]. Izdatel’stvo ‘Nauka’, Moskva [in
M,E.&V, A. 1995. Paleocene reef sediments from the Maiella
carbonate platform, Italy. Facies,32, 213–222.
M, T., B, P., B, M., B, E., C, H., C,
J-Y., F, M., H, K.F., M, M., N, C., P, D., R,
J., S´, B., T,Y.&T, P. 1988. The Upper Cretaceous-
Lower Tertiary marine to continental transition in the Bagua basin,
nothern Peru. Newsletters on Stratigraphy,19, 143–177.
M, E.A. 1990. Non-marine Cretaceous ostracodes from Argentina and
the paleobiogeographical relationships. In:W,R.C.&M,
C.A. (eds) Ostracoda and Global Events. Chapman & Hall, London,
N, J.S. 1994. Fishes of the world. John Wiley & Sons Inc., New York.
N, N.D. 1962. Paleontological gaps and geochronology. Journal of
N,D.J.&F, R.F. 1990. Plant microfossil record of the terminal
Cretaceous event in the western United States and Canada. In:S,
V.L. & W, P.D. (eds) Global catastrophes in Earth history: an interdis-
ciplinary conference on impacts, volcanism, and mass mortality. Geological
Society of America Special Papers, 247, 445–455.
N, K.B. 1909. Brachiopoderne i Danmarks Kridtaﬂejringer Det Kogelige.
Danke Videnskabernes Selskabs Skrifter,7, 129–178.
—— 1914. Some remarks on the brachiopods in the Chalk of Denmark.
Meddelelser fra Dansk Geologisk Forening,4, 287–296.
N, A. 1986. Paleocene radiolarians at Site 384, DSDP. News of Osaka
Micropaleontologists, Special Volume,7, 87–93.
—— 1992. Paleocene radiolarian biostratigraphy in the northwest Atlantic at
Site 384, Leg 43, of the Deep Sea Drilling Project. Micropaleontology,38,
N, F. 1897. Fauna of the Upper Cretaceous (Maestrichtian) of the Mari
Hills. Memoirs of the Geological Survey of India, Palaeontologica Indica,
Series 16, 1–80.
N, M.A. 1992. Taxic origin and temporal diversity: the eﬀect of phylogeny.
In:N,M.J.&W, Q.D. (eds) Extinction and phylogeny.
Columbia University Press, New York, 89–118.
O,R.K.&L, C. 1993. Controversies on the placement of Cretaceous-
Paleogene boundary and the K/P mass extinction of planktonic foramini-
fera. Palaios,8, 127–139.
P,Q.&W, R.C. 1990. The biostratigraphical sequence of Mesozoic
non-marine ostracod assemblages in northern China. In:W, R.C. &
M, C.A. (eds) Ostracoda and Global Events. Chapman & Hall,
P, C. 1993. Osteichthyes: Teleostei. In:B, M.J. (ed.) The fossil
record 2. Chapman & Hall, London, 621–656.
P,C.&S, A.B. 1987. Is periodicity of mass extinctions a
taxonomic artefact? Nature,330, 248–251.
—— & —— 1989. Periodicity in extinction: the role of systematics. Ecology,70,
P-N, K. 1979. Calcareous nannofossils from the Cretaceous between
the North Sea and the Mediterranean. International Union of Geological
Sciences, Series A,6, 223–272.
P,S.F.&F, A.G. 1977. Changes in calcareous nannoplankton in
the Cretaceous–Tertiary biotic crisis at Zumaya, Spain. Evolutionary
P, J.-P. 1983. Krithe and Parakrithe in the Kef section (northeast
Tunisia) around the Cretaceous–Tertiary boundary: paleohydrological
implications. In:M, R.F. (ed.) Applications of Ostracoda.
University of Houston, Houston, 510–519.
P, D., L,R.&D, T. in press. Deep-water aggluti-
nated foraminiferal changes and stable isotope proﬁles across the
Cretaceous/Paleogene boundary in the Rotwandgraben section, Eastern
Alps (Austria). Palaeogeography, Palaeoclimatology, Palaeoecology.
P, J.J. 1994. Calcareous nannofossils and the K/T boundary, El Kef:
No evidence for stepwise, gradual, or sequential extinctions. Geology,22,
—— 1995. Cretaceous/Tertiary boundary calcareous nannofossils from Agost,
Spain. In:F,J.A.&S, F.J. (eds) Proceedings of the 5th
International Nannoplankton Association Conference. University of
Salamanca, Salamanca, Spain, 185–217.
——, in press. Calcareous nannoplankton mass extinction at the Cretaceous/
Tertiary boundary: an update. In:R, G., F,D.&G,
S. (eds) The Cretaceous–Tertiary Event and Other Catastrophes in Earth
History. Geological Society of America, Special Papers, 307.
P, H.J. 1894. Brachiopoderne i den dansk Kridtformation. Danmarks
Geologiske Undersogelse,2, 1–59.
R, H.W. 1950. Cretaceous Asteroidea and Ophiuroidea with special
reference to the species found in Denmark. Danmarks Geologiske Under-
søgelse,77, 1–134, pls 1–18.
—— 1961. A monograph on the Cretaceous Crinoidea. Biologiske Skrifter det
Kongelige Danske Videnskabernes Selskab, 12.
—— 1972. Lower Tertiary Crinoidea, Asteroidea and Ophiuroidea from north-
ern Europe and Greenland. Kongelige Danske Videnskabernes Selskab
Biologiske Skrifter,19, 1–83, pls 1–14.
R,D.M.&B, G.E. 1988. Patterns of generic extinction in the fossil
record. Paleobiology,14, 109–125.
——&S, J.J., J. 1984. Periodicity of extinctions in the geologic past.
Proceedings of the National Academy of Sciences,81, 801–805.
R, Z. 1957. The Bilamellidea, nov. superfamily, and remarks on Cretaceous
Globorotaliids. Contributions of the Cushman Foundation for Foraminiferal
R,Z.&N, L.A. 1993. Cretaceous anurans from central Asia.
Palaeontographica Abteilung A,226, 1–54.
R, J. 1984. Les échinidés et la crise Crétacé-Tertiaire. Bulletin de la Section
des Sciences 1984,6, 133–147.
R, A.J.T. 1979. Lineages in Early Palaeogene calcareous nannoplankton.
Utrecht Micropaleontological Bulletin,22, 1–231.
R, B.R. 1988. From fossils to earth history: applied historical biogeography.
In:M,A.A.&G, P.S. (eds) Analytical biogeography: an
integrated approach to the study of animal and plant distributions. Chapman
& Hall, London, 437–481.
——&T, D. 1989. Extinction patterns and biogeography of scleractinian
corals across the Cretaceous/Tertiary boundary. In:J,P.A.&P,
J.W. (eds). Fossil Cnidaria 5. Memoirs of the Association of Australasian
Palaeontologists, 8, 355–370.
S, O. 1962. Neokatastrophismus? Deutsch Geologische Gesellschaft
Zeitschrift Jahrgang,114, 430–445.
S, B., K,G.&S, O. 1992. Stable isotope and foraminiferal
changes across the Cretaceous-Tertiary boundary at Stevns Klint,
Denmark: arguments for long-term oceanic instability before and after
bolide-impact event. Palaeoceanography, Palaeoclimatology, Palaeo-
S, R.M. 1986. Phylogeny reconstruction in paleontology. Van Nostrand
Rheinhold Co., New York.
S, P.M. 1985. Reefs are not so diﬀerent—they follow the evolutionary
pattern of level-bottom communities. Geology,13, 46–49.
——, F, D.E., H, R.G., B,C.B.&G, D. 1991.
Sudden extinction of the dinosaurs: Latest Cretaceous, Upper Great Plains,
U.S.A. Science,254, 835–839.
THE K–T BOUNDARY TRANSITION 291
S,V.N.&S, A.N. 1982. [The Mesozoic-Cenozoic boundary in
the organic development of the world.] Izdatel’stvo ‘Nauka’, Moskva. [in
S, P.W., III & L, J.H. 1982. Sampling bias, gradual extinction patterns
and catastrophes in the fossil record. In:S,L.T.&S, P.H.
(eds) Geological Implications of Impacts of Large Asteroids and Comets on
the Earth. Geological Society of America Special Papers, 190, 291–296.
S, P.S., G, E., R,B.R.&V, F.X. in press. Corals and
rudists in the Late Cretaceous: a critique of the hypothesis of competitive
displacement. VII International Symposium on Fossil Cnidaria and Porifera,
Madrid, September 12–15.
S, R.E., R, J.K., V V,L.M.&G, D. 1986. Gradual
dinosaur extinction and simultaneous ungulate radiation in the Hell Creek
Formation. Science,232, 629–633.
S, A.B. 1994. Systematics and the Fossil Record: Documenting Evolutionary
Patterns. Blackwell, London.
——, M, N.J., G,A.S.&R, B.R. 1995. Late Cretaceous
(Maastrichtian) macrofossil assemblages and palaeoenvironments from a
Tethyan carbonate platform succession, Oman Mountains. In:P,J.&
S, P.W. (eds) Palaeoenvironmental models for the benthic associ-
ation of Cretaceous carbonate platforms in the Tethyan Realm. Palaeoge-
ography, Palaeoclimatology, Palaeoecology,119, 155–168.
——&P, C. 1988. The inﬂuence of taxonomic method on the
perception of patterns of evolution. Evolutionary Biology,23, 127–216.
S, N.F. 1987. Cretaceous gastropods, contrasts between Tethys and the
temperate provices. Journal of Paleontology,61, 1085–1111.
S,R.P.&V Z, G.T. 1996. Extinction and surviorship of
southern Tethyan benthic foraminifera across the Cretaceous/Palaeogene
boundary. In:H, M.B. (ed.) Biotic recovery from mass extinction events.
Geological Society Special Publication No. 102, 245–258.
S,G.D.&C, S.D. 1988. Constructional azooxanthellate coral
communities: an overview with implications for the fossil record. Palaios,3,
S, N. I. 1975. Diatoms of the Cretaceous Period. Third symposium on
Recent and fossil diatoms, Nova Hedwigia, Supplement,53, 311–321.
S, J.D., F,R.E.&R, P.J. 1994. Paleocene isocrinids
(Echinodermata: Crinoidea) from the Kauru Formation, South Island,
New Zealand. Journal of Paleontology,68, 135–141.
S, R.B. 1979. An analysis of the ranges of spatangoid echinoid genera and
their bearing on the Cretaceous/Tertiary Boundary. In:C, W.K.
&B T. (eds) Cretaceous/Tertiary boundary events symposium. I.
Proceedings. University of Copenhagen, 78–82.
S,L.D.&K, J.P. 1990. Antarctic Paleogene planktonic foramini-
feral biostratigraphy: ODP Leg 113, sites 689 and 690. Proceedings of the
Ocean Drilling Program, Scientiﬁc Results,113, 549–569.
S, C.P., H,C.J.&W, G.J. 1995. Foraminiferal, radiolarian and
dinoﬂagellate biostratigraphy of Late Cretaceous to middle Eocene pelagic
sediments (Muzzle Group) at Mead Stream, Marlborough, New Zealand.
New Zealand Journal of Geology and Geophysics,38, 165–206.
S, R.M. 1987. A reassessment of reptilian diversity across the
Cretaceous–Tertiary boundary. Contributions in Science,391, 1–26.
S,F.&J, M.B. 1984. End-Cretaceous brachiopod extinctions in
the chalk of Denmark. Science,223, 1174–1177.
S, A.R., B,D.R.&L, J.F. 1990. Palynoﬂoral response to
K/T boundary events; A transitory interruption within a dynamic system.
In:S,V.L.&W, P.D. (eds) Global Catastrophes in Earth
History. Geological Society of America, Special Papers, 247, 457–469.
S, N.H.M. 1990. The extinction of rudist bivalves. D Phil Thesis, Open
University, Milton Keynes.
T, J.D., M,N.J.&T, C.N. 1980. Food specialization and the
evolution of predatory prosobranch gastropods. Palaeontology,23, 375–
T, P.D. 1993. Bryozoa. In:B, M.J. (ed.) The fossil record 2.
Chapman & Hall, London, 465–489
——&L, G.P. 1988. Mass extinctions and the pattern of bryozoan
evolution. In:L, G.P. (ed.) Extinction and survival in the fossil
record. Clarendon Press, Oxford, 99–119.
T,J.&N, D. 1994. Variation in Jurassic echinoid biodiversity
at ammonite zone levels: stratigraphical and palaeoecological signiﬁcance.
In:D, B., G, A., F´, J.-P. & R, M. (eds) Echinoderms
through time. A.A. Balkema, Rotterdam, 901–909.
T, H. R. 1981. Late Cretaceous nannoplankton and the change at
the Cretaceous–Tertiary boundary. In:W, J.E., W,E.&
D, R.G. (eds) The Deep Sea Drilling Project: a decade of progress.
Society of Economic Paleontologists and Mineralogists, Special Publi-
cations, 32, 355–394.
T, E. 1990. Late Cretaceous–Early Eocene mass extinction in the deep sea.
In:S,V.L.&W, P.D. (eds) Global Catastrophes in Earth
History. Geological Society of America, Special Papers, 247, 481–495.
T, E. 1976. Depositional environment and development of Danian
bryozoan biomicrite mounds (Karlby Klint, Denmark). Sedimentology,23,
T, J.C. 1957. Some planktonic foraminifera of the Type Danian and
their stratigraphic importance. U.S. National Museum Bulletin,215, 125–
V H, R. 1988. Echinoids from the early Palaeocene (Danian) of
the Maastricht area (NE Belgium, SE Netherlands): preliminary results.
Mededelingen van de Werkgroep voor Tertiaire en Kwartaire Geologie,25,
——, W, R.W., Z,G.&V B, M. 1987. Zeeëgels uit het
Krijt en Tertiair van Maastricht, Luik en Aken. Een atlas van de zeeëgels
uit het Campanin, Maastrichtin en Danin van Zuid-Limburg en aangren-
zende delen van Belgie en Duitsland. Publicaties van het Natuurhistorisch
Genootschap in Limburg,36, 1–96
V, O. 1989. Palaeocene calcareous nannofossil biostratigraphy. In:C,
J.A. & H, S.E. (eds) Nannofossils and their applications. Ellis
Horwood, Chichester, 267–310.
V, L.A. 1994. The dynamics of diversity of Gymnolaemata around the
Cretaceous-Paleogene crisis. In:Fossil and living Bryozoa of the Globe.
All-Russian Paleontological Society, Perm, 61.
W, P.D. 1995a. The K/T Trial. Paleobiology,21, 245–247.
—— 1995b. After the fall: lessons and directions after the K/T debate. Palaios,
——&K, W.J. 1993. Maastrichtian ammonites from the Biscay region
(France, Spain). Paleontological Society Memoirs,34, 1–58.
——, ——, ML,K.G.&M, J.F. 1991. Ammonite and inoceramid
bivalve extinction patterns in Cretaceous/Tertiary boundary sections of the
Biscay region (southwestern France, northern Spain). Geology,19, 1181–
W, A.A., K, L., C, M., R,T.H.&R, P.V. 1991. Early
Cretaceous labyrinthodont. Alcheringa,15, 327–332.
W, D.B. 1990. Dinosaurian distribution. In:W, D.B.,
D,P.&Oó, H. (eds) The Dinosauria. University of California
Press, Berkeley, 63–139.
W, P. 1991. The illustrated encyclopedia of Pterosaurs. Salamander
W,C.&R, J.-C. 1994. Mid-Cretaceous snakes from Sudan. A
preliminary report on an unexpectedly diverse snake fauna. Comptes
Rendus de l’Académie des Sciences, Paris,319, 247–252.
W, R. 1990. Ostracoda and global events. In:W,R.&M,
C. (eds) Ostracoda and Global Events. Chapman Hall, London, 3–24.
W,J.G.&M, B.A. 1992. Benthic foraminiferal changes across
the Cretaceous-Tertiary boundary in the deep sea; DSDP Sites 525, 527,
and 465. Journal of Foraminiferal Research,22, 81–113.
W, M.E. 1994. Catastrophic versus noncatastrophic extinction of the
dinosaurs: testing, falsiﬁability, and the burden of proof. Journal of
W, W.W. 1972. Stratigraphic distribution of diatom genera in marine
sediments in western North America. Palaeogeography, Palaeoclimatology,
W, C.W. ( C,J.H.&H, M.K.). 1986. Treatise on
invertebrate paleontology. Part L. Mollusca (revised), 4. The Geological
Society of America and the University of Kansas, 362.
W, X., B, D.B., R, A.P., D, Z., C, P. J., H,L.&
C, G. 1993. Oldest known amphisbaenian from the Upper Cretaceous of
Chinese Inner Mongolia. Nature,366, 57–59.
Y, C. 1993. Ostracode communities and their paleoecological environments for
the Cretaceous-Tertiary from southern Tibet, PR China. In:MK,
K.G. & J, P.J. (eds) Ostracoda in the Earth and Life Sciences. Balkema,
Received 18 April 1996; revised typescript accepted 6 November 1996.
Scientiﬁc editing by Stewart Molyneux.
292 N. MLEOD ET AL.