reached the same conclusion—that the major clades of Neornithes diverged deep in the
Cretaceous, largely at odds with the known fossil record.
The fossil record of Cretaceous bird remains that have been referred to extant lineages
is sparse—only a handful of records are represented by more than isolated elements
(Padian and Chiappe 1998; Chiappe and Dyke 2002). Hence the incompleteness of most
of these specimens referred to as Cretaceous neornithines renders few characters useful for
phylogenetic analyses (Clarke 1999; Dyke and Mayr 1999; Clarke and Chiappe 2001;
Dyke 2001a), and our currently poor understanding of the higher-level phylogenetic
relationships of extant lineages (Figure 12.2) further complicates their systematic
consideration (Cracraft 1988; Cracraft and Clarke 2001; Livezey and Zusi 2001). In spite
of problems, however, this scanty fossil record has been used to hypothesize the existence
of a number of extant lineages of birds prior to the end of the Mesozoic (e.g.
Pelecaniformes, Charadriiformes, Anseriformes, Galliformes, and Psittaciiformes), either
by direct interpretation or by using them for the temporal calibration of molecular
phylogenies. None the less, it is imperative that these Cretaceous reports of neornithines
are treated with extreme caution (Clarke 1999; Dyke and Mayr 1999; Clarke and
Chiappe 2001)—the earliest records of birds that are complete enough to be informative
for cladistic analyses (i.e. that present more than a few codeable characters for analysis and
as a result are useful for estimating the temporal divergences of the extant lineages), come
from rocks that are approximately 55 million years old, about 10 million years after the K-
T boundary. Although a few specimens represented by more than single bones do fill this
temporal gap (e.g. Vedding-Kristoffersen 2000), these have yet to be considered within
phylogenetic analyses. It would not be surprising if future studies of these, or even older
specimens, do support their placements within extant lineages, but such work has yet to be
undertaken.
How deep?In spite of inherent problems with the Mesozoic fossil record of the extant clades, the
presence of several immediate neornithine outgroups in the latest Cretaceous (e.g.
Ichthyornis Marsh 1880; Dyke et al. 2002; Limenavis Clarke and Chiappe 2001; Apsaravis
Norell and Clarke 2001) implies that the lineage leading to the Neornithes must have
Figure 12.2 Consensus cladogram to illustrate clade relationships at the base of Neornithes (based on
various sources, see text for details).
BASAL RADIATIONS WITHIN THE NEORNITHES 271