New supraordinal classification of Foraminifera: Molecules meet morphology

The classification of Foraminifera has a long history going back to the beginning of the 19th century and the work of d'Orbigny (1826) who established the order Foraminifera and proposed the first taxonomic system based on the growth plan of foraminiferal tests. d'Orbigny's successors have developed diverse systems based on the morphology of fossil and recent tests, differing principally by the importance given to form and chamber arrangement versus wall composition and structure (reviewed in Cifelli, 1990). The primary division of Foraminifera into single-chambered Monothalamia and multi-chambered Polythalamia (Schultze, 1854) was progressively replaced by a classification based on the presence or absence of pores (Reuss, 1861, Carpenter et al., 1862). In the 20th century, wall characteristics gained more and more importance and became the main criterion to distinguish higher-level groups in Foraminifera (Pokorny, 1963, Loeblich and Tappan, 1964, Hohenegger and Piller, 1975). In the seminal work of Loeblich and Tappan, 1988, Loeblich and Tappan, 1989, Loeblich and Tappan, 1992, Foraminifera were divided into 12 suborders that mainly differ by mineralogical and ultrastructural features of the test wall. In the most recent modifications of this classification (Sen Gupta, 1999, Mikhalevich, 2004, Kaminski, 2005), the number of orders (or classes/subclasses) increased to 16, but the foundations of this system remained unchanged (Table 1).

A few important attempts were carried out to group suborders into higher level taxa (Hohenegger and Baal, 2004, Hohenegger, 2011). The noticeable classification proposed by Mikhalevich, 1998, Mikhalevich, 2000, Mikhalevich, 2004 and Mikhalevich and Debenay (2001) was based on a "macrosystem" dividing Foraminifera into seven classes and resting upon "the whole organization of the test" rather than composition and ultrastructure of the test wall. This system revealed some interesting tendencies in the evolution of Foraminifera mainly based on morphological test patterns. Many taxonomic studies of Foraminifera concerned revisions of lower-level taxa (e.g., Hottinger, 1980, Gudmundson, 1994, Revets, 1996) but only a few tried to build up a more general system, such as the classification of agglutinated foraminiferans, whose updated versions are published on a regular basis (Kaminski, 2004).

The lack of progress in higher-level classifications of Foraminifera was mainly due to difficulties in inferring evolutionary relationships between major groups defined exclusively by morphological features and the sheer number of taxa involved. Despite the excellent fossil record, phylogenetic schemes of foraminiferal evolution are limited to textural and morphologic characters of tests (e.g., Cushman, 1948, Grigelis, 1978, Tappan and Loeblich, 1988, Vachard et al., 2010). This situation has changed with the advent of molecular studies that shed new light on the evolution of Foraminifera (Pawlowski, 2000, Bowser et al., 2006). The majority of molecular phylogenies were based on analyses of partial SSU and LSU rDNA sequences (Holzmann and Pawlowski, 2000, Pawlowski, 2000, Pawlowski et al., 2002a, Pawlowski et al., 2002b, Pawlowski et al., 2003). Because of their unusual length (> 3000 nucleotides) complete SSU sequences were only obtained for a few species, mainly representatives of the order Rotaliida (Pawlowski et al., 1996, Schweizer et al., 2008). Molecular phylogenies were also inferred from actin (Flakowski et al., 2005), tubulin (Habura et al., 2006) and RNA polymerase (Longet and Pawlowski, 2007), but the number of species analyzed in these studies was very small. Recent analysis of combined sequence data confirmed major trends in the evolution of Foraminifera inferred from single gene phylogenies (Groussin et al., 2011). However, up to now no formal attempt has been made to modify the higher-level classification of Foraminifera by including molecular data.

Here, we propose a new higher-level system of Foraminifera, based on molecular data. We present an updated version of a SSU rDNA phylogeny based on complete sequences obtained from representatives of almost all foraminiferal orders, including 23 new sequences. We discuss this phylogeny with reference to other multigene studies and we describe the basic morphological features for new molecular groupings.