The classical example of dichotomy in a narrow sense, caused by an even longitudinal division of the AC, is a thallus bifurcation in an alga Dictyota dichotoma (Oltmanns, 1889; Goebel, 1928; van den Hoek et al., 1995). However, in most plant species with a single AC, this tetrahedral, lens- or wedge-shaped AC rarely undergoes a perfect longitudinal division, probably due to the reduction of the energy costs related to the new cell wall insertion during cell division (Schüepp, 1966; Cooke and Paolillo, 1980; Barlow, 1992). As a consequence, direct division of the AC is uncommon. It was reported to occur only in some ferns (Bierhorst, 1977). But even this interpretation was later criticized, mostly because the ACs for the new branches originated by formative divisions and not by a direct dissection of the AC (see Lyndon, 1998 for discussion).

In this pattern, the initial for a new branch originates not by the split of the original AC, but due to formative division in the adjacent segment. At the same time, the original AC maintains the growth of the main axis. Thus, it should be considered the pseudodichotomy (Schuster, 1984a), meaning that even if it looks like dichotomy, it is not formed by the meristem division. Such a branch development was described in detail in thalloid liverworts (Schuster, 1984a); in Psilotum nudum it was interpreted as dichotomy (Roth, 1963).

Dichotomy in a broad sense is applied here, understood as the division of the whole initial zone. The AC, which is usually distinct in non-dividing shoots, ceases its divisional activity prior to dichotomy and becomes indistinguishable at the apex. Then, two new initials are simultaneously initiated next to the inactivated original AC (Hagemann and Schulz, 1978; Hagemann, 1980; Mueller, 1982; Steeves and Sussex, 1989; Imaichi and Kato, 1991; Jernstedt et al., 1994; White and Turner, 1995; Lu and Jernstedt, 1996; Kato and Imaichi, 1997; Imaichi, 2008). This type of dichotomy occurs in shoots, rhizomes and roots, in some ferns and lycophytes (Table 1). In details, the changes of the meristem structure were analyzed in the rhizophores - the unique axial organs of Selaginella, bearing the root primordia at the tip (Imaichi and Kato, 1989, 1991; Jernstedt et al., 1994; Lu and Jernstedt, 1996; Kato and Imaichi, 1997). Such a dichotomy occurs also in the aerial shoots of Psilotum nudum, whereas in irregularly branched subterranean rhizomes, numerous ACs present at the apex can at random be selected for new branches or be inactivated. The mechanism differentiating the fate of ACs in the shoot and rhizome is not known (Takiguchi et al., 1997).

In shoots of Selaginella kraussiana two transient ACs are responsible for the apex growth. Here, before dichotomous branching the initial cells divide several times, producing a group of meristematic cells. Then, new ACs are selected for the two resulting axes (Harrison et al., 2007). Possibly, a similar branching pattern is also present in the notch meristems of Anthoceros, where intensive divisions of prismatic initial(s) result in a group of meristematic cells, which next split to form the furcated lobes (Schuster, 1984b).

This mechanism is typical of Lycopodiaceae s.l., where the entire meristem can divide into two even (isotomy) or uneven (anisotomy) parts (Figure 1C; Troll, 1937, 1948; Schoute, 1938; Øllgaard, 1979, 1990). In seed plants, dichotomy rarely occurs in gymnosperm shoots (Riding, 1976; Zagórska-Marek, 1985). However, the potential to branch dichotomously is preserved in conifers, as after colonization by mycorrhizal fungi, the lateral roots start to bifurcate (Figure 1E; Robertson, 1954; Wilcox, 1968; Faye et al., 1981; Piche et al., 1982). In angiosperms, shoot dichotomy is reported as a typical branching pattern only in several species, mostly in monocotyledons (Figure 1D).

Regardless of the structural differences related to the type of organ and organization of its meristem, the morphogenetic processes that accompany the branching are similar in all these plant groups. Dichotomy affects the apex geometry, forces the reorganization of the meristem structure, changes its divisional activity and cell differentiation, and has an impact on organogenesis (e.g., leaf initiation). Its first symptom is broadening of the apex in the plane perpendicular to the future division. In this early stage, the meristem zonation pattern is not disrupted, but the number of cells and volumes of particular meristematic zones increase by intensive cell proliferation. The distinctiveness of superficial layers is maintained during dichotomy progression, as well as the continuity of procambium and vascular tissues in the parental and dichotomous axes. Differentiation of the meristematic cells located between dichotomizing apices ceases the growth of the central part of the original meristem starting the separation of both branches (Härtel, 1937; Nolan, 1969; Tomlinson, 1970, 1971; Boke, 1976; Tomlinson and Posluszny, 1977; Faye et al., 1981; Piche et al., 1982; Laajanen et al., 2007; Raudaskoski and Salo, 2008; Gola and Jernstedt, 2011).

Broadening of the shoot apex can affect the organogenesis due to increasing size of the region of organ initiation. If the leaf initiation precedes the apex dichotomy, the leaf can encircle the enlarged stem (as in Flagellaria and Strelitzia; Tomlinson, 1970; Fisher, 1976) or be inserted in the increased meristem circumference. In the latter case, the pattern of leaf arrangement (phyllotaxis) can change (Schoute, 1938; Zagórska-Marek, 1985; Gola, 1996). With dichotomy progression, the initiation of new organs is usually maintained at the shoot circumference, but at the inner surface of furcation, organogenesis is repressed until the division is completed. It can be presumed, based on the analogy to the concave apices (Lintilhac, 2014), that this inner surface remains under the tension, which prevents the bulging of primordia. The restoration of organ initiation becomes possible after changes of the physical constraints and the local surface relaxation, similarly to the proposed buckling mechanism of leaf primordia formation (Green, 1996).

The ability to form dichotomous roots seems to be an inherent feature in Pinaceae (Robertson, 1954; Wilson and Field, 1984; Kaska et al., 1999). The intensification of the process, with repeated dichotomies resulting in so-called coralloid structures, is related to the colonization of roots by ectomycorrhizal fungi (Peterson and Bonfante, 1994). The fungal symbionts are probably the source of plant growth regulators, which stimulate morphogenetic changes, including root swelling and dichotomous branching (Barker and Tagu, 2000; Brundrett, 2002). Similar changes in the root architecture can be induced in the absence of fungi by exogenously supplied phytohormones or their inhibitors (Wilson and Field, 1984; Rupp et al., 1989; Kaska et al., 1999; Laajanen et al., 2007; Raudaskoski and Salo, 2008). Application of auxin transport inhibitors [N-(1-naphtyl)phthalamic acid (NPA); 2,3,5-triiodobenzoic acid (TIBA)], ethylene precursor [(1-aminocyclopropane-1-carboxylic acid (ACC)], or ethylene releasing compounds [2-chloroethylphosphonic acid (CEPA)] stimulates extensive dichotomous branching of Pinus roots, increasing the percentage of coralloid structures up to 25–30% (Kaska et al., 1999). It is suggested that the balance between auxin [indole-3-acetic acid (IAA)] and cytokinin, possibly mediated by the ethylene level, has a regulatory role in this process (Rupp et al., 1989; Kaska et al., 1999; Barker and Tagu, 2000; Laajanen et al., 2007; Raudaskoski and Salo, 2008). Possibly, the NPA-treatment increases the auxin concentration at the root tip, whereas the moderate level of the hormone at the meristem flanks induces dichotomous root formation. At the same time, high IAA concentration at the root tip stimulates the biosynthesis of ethylene, leading to cell differentiation in the central part of the original meristem and consequently, separation of dichotomous roots (Laajanen et al., 2007; Raudaskoski and Salo, 2008).

This mechanism of hormonal regulation corresponds well with the morphogenetic changes in dichotomizing roots of Pinus, but because it was proposed based on the auxin distribution in a model plant Arabidopsis, it should be validated. Nevertheless, it seems likely that specific hormone distribution and/or concentration can be a universal aspect of dichotomy regulation, as e.g., auxin shapes different developmental processes in all vascular plants and bryophytes (Cooke et al., 2002). In addition, it has recently been shown that the ratio between IAA and cytokinin regulates the dichotomous root branching in S. kraussiana, although the shoot dichotomy regulation by auxin is questionable in this species (Sanders and Langdale, 2013).

The knowledge concerning genetic background for dichotomy regulation and molecular signaling during this process is lacking. It can only be speculated that the regulation is based on genes involved in the cell division and differentiation homeostasis, affecting the size of the apex and the identity of meristematic cells.

The genetic machinery for the self-maintenance of the apical meristems is mostly deciphered in a model plant Arabidopsis. Here, based on the mutant phenotypes, it is possible to infer which genes could play a role in dichotomous branching. One of such Arabidopsis mutants is tonsoku (tsk), with a forked root reminiscing of dichotomy, and fasciated stems. The suggested role of tsk is to maintain the ordered structure of the meristem through the regulation of the cell cycle (Suzuki et al., 2004, 2005). However, the disorganization in tsk root meristems affects the expression of the other regulatory genes, e.g., SCARECROW (SCR; Suzuki et al., 2004). Contradictory, it was shown that the tissue-specific expression pattern of the SCR homolog is preserved in dichotomizing roots of Pinus. During dichotomy progression, the specificity of endodermis and root radial patterning are maintained, manifested by localization of this gene (Laajanen et al., 2007; Raudaskoski and Salo, 2008). These results are also in agreement with the cytohistological observation of dichotomizing apices.

The maintenance of the shoot apical meristem integrity in Arabidopsis requires the activity of WUSCHEL (WUS), which is antagonized by the CLAVATA genes (Laux et al., 1996; Fletcher et al., 1999; Reddy, 2008). As these genes are involved in the production and maintenance of initial cells and their mutation can result also in bifurcated stem phenotype (e.g., in clv3 or WUS overexpression mutants), their plausible role in dichotomy regulation can be therefore hypothesized. Interestingly, the homologous CLAVATA1-LIKE gene was found to play a potential role in the initiation of Pinus ectomycorrhizal roots (Heller et al., 2012). Furthermore, as recent research revealed the presence of WUS homologs in all plant groups (Lian et al., 2014) and the CLAVATA3 homolog was earlier described e.g., in Selaginella (Floyd and Bowman, 2007), the universality of the regulatory mechanism in plants can be suggested. However, it remains unknown, whether these homologous genes have a similar expression pattern and function in the apical meristem maintenance and if they are involved in dichotomy regulation.

Recently, the WUS-CLV3 interactions were simulated in the reaction–diffusion model to show different patterns of shoot apical meristem (SAM) development (Fujita et al., 2011). The assumption was that WUS promotes the growth of the apex (activator) whereas CLV suppresses the process (inhibitor). The pattern of dichotomous branching was generated in this model, when the increased level of the activator stimulated the cell proliferation leading to the meristem bifurcation due to spatial restrictions (SAM size limitation; Fujita et al., 2011). Thus, it can imply that the proper balance of these two factors can play a role in the meristem integration during dichotomy.

Finally, the class I KNOX (KNOTED-like homeobox) genes are hypothesized to suppress the cell differentiation within the SAM. They are found in all land plants and, in addition, they are supposed to be regulated by hormones (Veit, 2009). For example, a low level of auxin stimulates the class I KNOX genes and in turn promotes the meristematic activity in the SAM. Likely, the formation of the new apices at the flanks of dichotomizing meristem and simultaneously triggered differentiation in its center can result from the localized distribution of auxin and related expression of genes. Furthermore, in Selaginella KNOX/ARP (ASYMMETRIC LEAVES1, ROUGH SHEATH2, PHANTASTICA) interaction regulates the maintenance of the indeterminate growth of the apex vs. leaf formation. It was suggested that this expression pattern within the SAM can be related to the meristem dichotomy (Harrison et al., 2005).