MATERIALS AND METHODS

Historically, vegetative shoots ascribed to the Araucariaceae have sometimes been described using the generic name Araucarites; however, Zijlstra and Konijnenburg-van Cittert (2000) proposed that this name be restricted to megasporangiate cones and isolated bract-scale complexes. This practice is acceptable when applied to early representatives of the Araucariaceae with uncertain sectional affinities or to more poorly preserved material. However, the name Araucarites should not be applied to material assigned to a particular section, as such a determination must indicate affinity with the genus Araucaria.

The generic name Araucaria is used here with reference to the new Holyoke specimen, but it is not formally named as a new species as it is represented by a single, isolated specimen that cannot be definitively diagnosed from several Araucaria section Eutacta bract-scale complexes. Some other comparative taxa are referred to here as Araucarites based on the original descriptions until a thorough revision of the fossil record of the family is available.

The new Araucaria bract-scale complex was collected from the South Hadley Falls Member of the Portland Formation (Olsen et al. 2003, 2005), from strata that crop out just below the dam in the middle of the Connecticut River at Holyoke, Massachusetts (Figure 1). The Portland Formation contains the youngest rocks of the Hartford Group, Newark Supergroup and has been dated by palynostratigraphy (Cornet and Waanders 2006) and vertebrate biochronology (Lucas and Huber 2002) to span the Hettangian, Sinemurian, and possibly Pliensbachian stages of the Early Jurassic. The South Hadley Falls Member is of Hettangian age, and consists of gray and red lacustrine shale and siltstone beds arranged in ~20 m thick cycles, each of which has been interpreted as the depositional product of the 20 Ky precessional cycle (Olsen 1986; Olsen et al. 2003, 2005). At Holyoke Dam and nearby exposures, the South Hadley Falls Member contains an abundant, allochthonous, low-diversity flora strongly dominated by conifer branches with common equisetalian stem fragments and some cycadeoid leaves. Other fossils from the South Hadley Falls Member include fossil insect larvae (Huber et al. 2003), fragmentary fishes, and occasional theropod dinosaur and crocodilian footprints (Olsen et al. 2003, 2005). Although this locality has been known for some time (see discussion in McDonald 1992) no detailed systematic work on the plant material has been published despite the use of the conifer shoots in paleoecological studies (e.g., Cornet and Waanders 2006). Based on the abundance of Pagiophyllum and Brachyphyllum morphotype leafy shoots, ovulate cone scales putatively similar to Hirmeriella, and the dominance of dispersed Classopollis pollen (Cornet et al 1973) at this and coeval localities in the Hartford and Deerfield Basins, most of this material has been assumed to represent the important Mesozoic conifer family Cheirolepidiaceae. However, it has been previously suggested that some of the shoots may be araucarian (Cornet and Waanders 2006). In fact, the cuticular preservation of the leafy shoots is poor (contra Cornet and Waanders 2006) and the familial affinities of any of the specific leafy shoots remain uncertain. An ongoing restudy of ovulate cone scales and pollen cones from Holyoke Dam does suggest dominance of the Cheirolepidiaceae; however, the bract-scale complex described here provides unequivocal macrofossil evidence of the Araucariaceae in this flora.

No special preparation of the fossil was performed; however, it was immersed in ethanol to increase contrast for photography (Figure 2.1). The specimen exhibits considerable relief, which makes it possible to delimit several important characters under incident light, such as the position of the ovule and the presence and orientation of the wings. A thin layer of carbonaceous material is present, but the application of the transfer technique was not attempted because it is the only specimen, and attempts at transferring other fossils from this locality were unsuccessful. The fossil bract-scale complex specimen (# J 1430) will be deposited in the collections of the Paleobotany Division of the Natural History Museum and Biodiversity Research Center at the University of Kansas. Comparative material of extant Araurcaria species was examined and photographed in the L.H. Bailey Hortorium of Cornell University.

Although the value of fossils in phylogenetic reconstruction is controversial, the time of first appearance of groups in the geological record is widely used in calibrating phylogenies based on molecular evidence. Furthermore, fossils are sometimes used to describe the level of agreement between the temporal sequence of taxa in the fossil record and the order of branching on phylogenic trees. Such congruence studies normally employ one of two approaches; 1) methods that evaluate the number of inconsistencies between the phylogeny and temporal data (Norell and Novaceck 1992; Huelsenbeck 1994), and 2) methods that measure calibrated "ghost lineages" or phylogenetically implied gaps (Sidall 1998; Wills 1999; Pol and Norell 2001) by which the absolute temporal disparity is evaluated (Brochu and Norell 2000). In this paper, the second approach is used to evaluate the congruence between the current phylogenetic hypotheses for the sections of Araucaria and the temporal information from the fossil record of the genus. Specifically, we used the analysis of Setoguchi et al. (1998), which produced one hypothesis of relationships between the sections of Araucaria and that of Gilmore and Hill (1997), which produced two topologies–one perfectly congruent with that of Setoguchi et al. (1998) in terms of the sectional relationships and one different. The Gilmore and Hill (1997) study used fewer Araucaria species, but it included at least one species of each monophyletic section of the genus (sensu Setoguchi et al. 1998).

The first appearance datum (FAD) used for each taxon is based on the literature for sections Araucaria, Bunya, and Intermedia (e.g., Stockey 1994; Setoguchi et al. 1998). The FAD for section Eutacta is based on the Holyoke bract-scale complex described here. The Manhattan Stratigraphic Measure (MSM) method originally proposed by Sidall (1998) and later modified by Pol and Norell (2001) designated as MSM* was utilized. In addition, the age uncertainty was considered with the randomization approach for age ranges (Pol and Norell 2006) using the MSM* to calculate the stratigraphic fit for each of 1000 random replicates. This analysis was performed using the phylogenetic analysis software program TNT (Goloboff et al. in press).