There has been a hearty debate regarding the proper phylogenetics and taxonomy of the order Apiales. The debate is fueled by the ambiguous nature of the trends in morphological evolution of the Apiales. Repeated parallel evolution combined with instances of convergence has led to difficulty in determining the ancestral character states of the Apiales, and therefore has left the door open for differences of opinion. There are two major opposing hypotheses about the proper placement of the Apiales within the evolutionary tree. There is agreement about the higher level classification of Apiales – Kingdom Plantae, Phylum Magnoliophyta, Class Magnoliopsida – but the debate centers on the ordinal, interfamilial and intrafamilial relationships (Plunkett, 1996).
The old hypothesis, based on non-molecular examinations of plant morphology, placed the Apiales in the subclass Rosidae, the largest subclass of the angiosperms. The strongest proponents of this hypothesis are Leon Takhtajan and Arthur Cronquist. Plants belonging to the Apiales order were characterized as dicots with compound leaves, separate petals, an inferior ovary with a single ovule in each locule, and flowers in umbels (Cronquist, 1988). Regarding the Apiales' relationship to other angiosperm groups, some believed that Apiales was closely related to Cornales, but Arthur Cronquist (1988) believed there was a closer relationship between Apiales and Sapindales, based on the presence of similar chemicals in the two orders.
With regards to subdivision of the Apiales, the old hypothesis divided the Apiales into two families, the Apiaceae (=Umbelliferae) with 3000 species and the Araliaceae with 700. The old hypothesis further divided the Apiaceae family into three subfamilies, Apioideae, Saniculoideae, and Hydrocotyloideae. The Apioideae was the largest of the three and contained all the plants that bear flowers in a compound umbel, while the Saniculoideae had simple umbels. It was viewed that the Hydrocotyloideae represented an intermediate form between the Araliaceae and the Apiaceae, and this led many to believe that the Apiaceae originated directly from the Araliaceae, i.e. the Araliaceae might be paraphyletic (Cronquist, 1988).
However, in light of molecular evidence, especially analysis of rbcL DNA sequences, a new hypothesis has surfaced that radically differs from the old hypothesis. Regarding the Apiales' relationship to other angiosperm groups, phylogenetic trees based on the DNA sequences show that the Apiales are not as closely related to Cornales, but should be placed in the subclass Asteridae near the Aquifoliaceae (hollies), Gunneraceae, and Eucommiales (Nandi, 1998). Moreover, with regards to subdivision of the Apiales, the DNA evidence supports a sister-group relationship between Apiales and Pittosporaceae, which was previously classified in subclass Rosidae, order Rosales (Plunkett, 1996).
Support for the alliance between Apiales and Pittosporaceae also comes from similarities between the two groups in five classes of biochemicals, in ovule morphology, and in base chromosome number. Nevertheless, Cronquist refused to acknowledge the inclusion of Pittosporaceae in Apiales, stating that Pittosporaceae originated from lower rosids, while Apiales was more advanced. He based his observations on leaf shape and ovary position, but the DNA evidence suggests that these characteristics are misleading in determining phylogenetic relationships.
The DNA data supports the Apiales as a monophyletic group. Within the Apiales, Araliaceae is paraphyletic, whereas the Apiaceae is a monophyletic clade. The new hypothesis further divides the Apiaceae family into similar subfamilies, including the Apioideae (monophyletic), Saniculoideae (monophyletic), and Hydrocotyloideae (polyphyletic). The Hydrocotyloideae forms four separate clades and is included in both the Apiaceae and the Araliaceae. This may explain why some had thought of Hydrocotyloideae as an intermediate form between the two families, but the rbcL data does not support the hypothesis that Apiaceae was derived from Araliaceae (Plunkett, 1996).
In summary, the two hypotheses differ in three major areas: 1) Apiales relationship to other angiosperm groups -Old hypothesis: The order Apiales belonged to subclass Rosidae and was related to orders Cornales and Sapindales. -New hypothesis: The Apiales belongs to subclass Asteridae and is not as closely related to Cornales. 2) Subdivision of the Apiales -Old hypothesis: The Apiales was divided into two families, the Apiaceae and the Araliaceae. -New hypothesis: The Apiales is divided into three families, the Apiaceae, the Araliaceae, and the sister group Pittosporaceae. 3) Subfamilies -Old hypothesis: The Apiaceae subfamily Hydrocotyloideae represented an intermediate from between Araliaceae and Apiaceae, so the Apiaceae originated from the Araliaceae. -New hypothesis: Hydrocotyloideae was confused as an intermediate form because it forms four clades that appear in both the Apiaceae and the Araliaceae, but the Apiaceae is not derived from the Araliaceae.
There are still questions that need answering, but the combination of molecular tests and morphological studies will shed light on the Apiales' place in the evolutionary tree.
