The Radiation Of Modern Birds

Modern birds, the Neornithes (see Box 9.2), are hugely diverse today. The traditional assumption has been that they radiated entirely within the Tertiary, following the extinction of the dinosaurs, pterosaurs and other Mesozoic groups. Indeed, Feduccia (2003) has argued that there was a bottleneck in avian diversity at the KT boundary. The previously diverse Enantiornithes, Hesperornithiformes, Ichthyornithiformes and other groups died out at the KT boundary, and the Neornithes radiated dramatically only after that.

On the other hand, there have been persistent reports by palaeornithologists of Cretaceous representatives of modern bird groups, some even from the earliest Cretaceous. This 'early origins' view gained strong support from molecular clock studies of birds (Hedges et al., 1996; Cooper and Penny, 1997), which suggested that the modern bird orders originated in the Early Cretaceous, some 100 Myr ago or more. This leaves a long gap between the molecular estimate to the oldest fossils confidently assigned to modern orders (65-50Myr ago). Does this indicate that the molecular dates are overestimates, perhaps not taking account of a speeding up of molecular evolution immediately after the KT event (Benton, 1999b),or that there is a major gap in the fossil record through the Late Cretaceous?

What of the 100 or more records of Neornithes that have been reported from the Cretaceous? It turns out that most of these supposed earliest representatives of flightless birds, ducks, flamingos, pelicans, loons, woodpeckers and the like have been rejected mainly because the specimens are too incomplete to show diagnostic characters or, in some cases, were wrongly dated. So far, all Cretaceous records of palaeognaths have proved to be unacceptable (Hope, 2002).Among neog-naths, Hope (2002) and Dyke and Van Tuinen (2004) recognize valid specimens of Anseriformes (ducks and geese), Gaviiformes (loons) and Pelecaniformes (pelicans), and are less certain of possible representatives of Galliformes (game birds), Charadriiformes (shore birds), Procellariiformes (albatrosses and relatives) and Psittaciformes (parrots). Virtually all of these records are based on single bones, quite different from the quality of the record of non-neornithine birds in the Cretaceous.

The most complete Cretaceous neornithine is Po-larornis from the latest Cretaceous, just below the KT boundary, in Antarctica (Chatterjee, 2002), represented by a partial skeleton (partial skull, four cervical vertebrae, fragment of the sternum, complete left femur and partial right femur, half of the left tibiotarsus). The skull (Figure 9.9(a)) looks superfically like that of a hesperor-

Fig. 9.9 One ofthe more complete Cretaceous neognath specimens, Polarornis,from the latest Cretaceous of Antarctica: (a) skull in lateral view; (b) reconsructed skeleton in dorsal view; parts known from fossils are shaded. (Courtesy ofSankar Chatterjee.)

Fig. 9.9 One ofthe more complete Cretaceous neognath specimens, Polarornis,from the latest Cretaceous of Antarctica: (a) skull in lateral view; (b) reconsructed skeleton in dorsal view; parts known from fossils are shaded. (Courtesy ofSankar Chatterjee.)

20 cm

20 cm nithiform (Figure 9.8(c)), but it lacks teeth and has a neognath palate. It has been identified as a loon (Order Gaviiformes, Family Gaviidae) on the basis of features of the skull, femur and tibiotarsus. The hindlimbs (Figure 9.9(b)) suggest that Polarornis swam underwater using expanded feet, like hesperornithiforms and modern loons; its wings are unknown.

So what can we understand from the apparently divergent molecular and palaeontological evidence? Both sides of the debate have perhaps exaggerated their positions. It is wrong to deny the existence of any neor-nithines of modern affinity in the Cretaceous (Benton, 1999b; Feduccia, 2003). Clearly some modern orders were represented (Dyke, 2001; Hope, 2002; Kurochkin et al., 2002), but it is important to note that the fossils are generally far scrappier than the other Cretaceous bird fossils, and that many, such as Polarornis, are found in sediments that lie not far below the KT boundary. Some critics had suggested that the molecular dates must be right and the fossil record wrong; in particular, the diverse radiation of neornithines through the Late Cretaceous implied by some molecular evidence cannot be checked by fossils because the record is so poor. Fossil birds have been found in Upper Cretaceous sediments from many continents, but it is only in the last 10Myr of the Cretaceous that neornithines appear (Dyke, 2001; Hope, 2002).

Reappraisals of the molecular evidence give diverging results. For example, the split between Anseriformes and Galliformes has been dated at 68 Myr ago by Waddell et al. (1999),85Myr ago by Paton et al. (2002) and 90Myr ago by Van Tuinen and Hedges (2001). These all lead, however, to estimates of the timing of the palaeognath-neognath split at 110-120 Myr ago, considerably older than the oldest fossils. Will more fossils be found to plug the gap between 115 and 70 Myr ago, or will the molecular ages be recalibrated?

0 0

Post a comment