Supplementary Discussion

The genus Acacia comprises approximately 1350 species1 and belongs together with the monotypic African genus Faidherbia Chev.2,3 to the tribe Acacieae, which forms part of the subfamily Mimosoideae (Fabaceae). Acacia currently is divided into three subgenera: subg. Acacia (pantropical, ca. 160 species), subg. Aculeiferum Vassal (pantropical, ca. 235 species) and subg. Phyllodineae (DC) Seringe (syn. subg. Heterophyllum Vassal; ca. 960 species, predominantly confined to Australia). Pedley4 proposed generic rank for these subgenera, namely Acacia, Senegalia Raf. and Racosperma C. Martius, but this concept of narrowly circumscribed genera was not widely accepted.

A number of studies have analyzed morphological, anatomical, palynological and phytochemical characters as well as molecular markers to ascertain the taxonomic status of Acacia and to elucidate the phylogenetic relationships within the genus and the tribe Acacieae1-13. Restriction fragment analyses of chloroplast DNA13-15 and sequence information from different chloroplast loci and nuclear rDNA (ITS, 5S rDNA) have contributed much to a better understanding of the phylogenetic relationships within the genus Acacia and the tribes Acacieae, Mimoseae and Ingeae, respectively9,12,16-18. Molecular studies in general indicate that the tribes Acacieae and Ingeae as well as the genus Acacia are not monophyletic, whereas Acacia subg. Acacia and subg. Phyllodineae each appear monophyletic12,13,15-18. However, the topologies of the cladograms differ remarkably in the relationships of the three subgenera of Acacia, Faidherbia, and the tribe Ingeae. A recent sequence analysis17 included a broad taxon sampling of subfamily Mimosoideae and reported a close relationship among Acacia subg. Acacia and certain members of tribe Mimoseae, while subg. Aculeiferum, the tribe Ingeae, and Acacia subg. Phyllodineae form a derived clade. Additionally, cladistic analysis of morphological, anatomical, phytochemical and palynological data support Acacia as a polyphyletic taxon and correspond to the molecular evidence10. Furthermore the analyses show Acacia subg. Acacia as monophyletic and distinct from the other two subgenera, which is also supported by a study of seed protein profiles11.

In the New World, Acacia subg. Acacia is represented by 57 taxa, which are characterized by their stipular thorns, leaves with well-developed extrafloral nectaries, pollen arranged in 16-grain polyads, and a single pollen colporate with columellae. This subgenus turned out difficult to subdivide on the basis of morphological characters19,20. The same remains true even for molecular analyses: A phylogeny based on trnK/matK sequences1,16,18 shows a week resolution, although the subg. Acaciaper se is well supported and all New World species form a monophyletic group.

All species of Acacia included in our study are neotropical and assigned to subg. Acacia. In detail, A. farnesiana, A. macracantha, A. pennatula, and A. cochliacantha belong to the non-myrmecophytic group of this subgenus, while A. chiapensis, A. cornigera, A. collinsii,A. globulifera and A. hindsii are myrmecophytes. Among these species, only A. farnesiana, A. pennatula and A. cochliacantha had already been included in molecular phylogenetic investigations, while none of the obligate myrmecophytes has so far been used for such studies. We combined trnK intron data presented by Miller and Bayer18 with data achieved in our own analysis and found that all Acacia species investigated are, as expected, members of the monophyletic subg. Acacia and are located within the neotropical group (Supplementary Figure 1). The myrmecophytic Acacia species (constitutive EFN flow) are the sister group to a monophyly comprising the non-myrmecophytes A. macracantha, A. pennatula and A. cochliacantha (bootstrap value = 62). A. farnesiana is part of the neighbour group of these species, together with A. schaffneri and A. caven (bootstrap value = 54). Although the bootstrap support for these relationships is low, a paraphyly of the induced EFN flow with regard to constitutive EFN flow is indicated.

The sister group to Acacia subg. Acacia is formed by Prosopis julifora and Leucaena leucocephala (inducible EFN flow, both Mimoseae), while Piptadenia flava (also inducible) is forming an own clade. The monophyly of Acacia subg. Acacia has a high bootstrap support (100), it is thus highly unlikely that the Mimoseae investigated can be considered as a part of this monophyly.

The genus Mimosa, which is used in most studies as outgroup for Acacia, forms part of tribe Mimoseae and is therefore of limited value as an outgroup when phylogenetic relationships among the different groups of this tribe and the various subgenera of Acacia are considered17. The placement of the taxa investigated in our study in the phylogeny presented by Luckow et al.17 is shown in Supplementary Figure 2. In this analysis, which has used a member of the Caesalpinioideae rather than of the Mimosoideae as the outroup, Piptadenia flava appears more closely related to Mimosa and Acacia subg. Acacia than Prosopis juliflora and Leucaena leucocephala. According to this analysis, inducible EFN flow occurs both in different lineages within the Mimoseae and in Acacia subg. Acacia. Inducible EFN flow via the same biochemical pathway as reported here has recently been reported forPhaseolus lunatus21, a member of subfamiliy Faboideae rather than Mimosoideae (Supplementary Figure 2). The inducible EFN secretion in Phaseolus therefore appears homologous to EFN induction in the Mimosoideae, indicating that this pattern has a wide distribution in the whole family of Fabaceae.

As to our knowledge, the species investigated in this study are the only mimosoid species for which the inducibility of EFN flow has been studied. Species with constitutive EFN flow are closely related to species with inducible EFN flow, and the latter trait also characterizes members of other phylogenetic groups. This supports the hypothesis that the constitutive EFN flow has evolved from the inducible state.

1.Maslin, B. R., Miller, J. T. & Seiger, D. S. Overview of the generic status of Acacia (Leguminosae: Mimosoideae). Austr. Syst. Bot.16, 1-18 (2003).

2.Vassal, J. Apport des recherches ontogeniques et seminologiques a l´ etude morphologique, taxonomique et phylogenique du genre Acacia. Bull. soc. d´hist. nat. Toulouse108, 105-247 (1972).

3.Vassal, J. in Advances in legume systematics, part 1 (eds. Polhill, R. M. & H., R. P.) 169-171 (Rocal Botanical Gardens Kew, London, UK, 1981).

4.Pedley, L. Derivation and dispersal of Acacia (Leguminosae), with particular reference to Australia, and the recognition of Senegalia and Racosperma. Bot. J. Linn. Soc.92, 219-245 (1986).

5.Evans, C. S., Qureshi, M. Y. & Bell, E. A. Free amino acids in the seeds of Acacia species. Phytochemistry16, 565-570 (1977).

6.Guinet, P. & Vassal, J. Hypotheses on the differentiation of the major groups in the genus Acacia (Leguminosae). Kew Bull.32, 509-527 (1978).

7.Guinet, P. in Advances in legume systematics, part 2 (eds. Polhill, R. M. & H., R. P.) 835-55 (Royal Botanical Gardens Kew, London, UK, 1981).

8.Guinet, P. The genus Acacia (Leguminosae, Mimosoideae): its affinities as borne out by its pollen characters. Plant Syst. Evol.suppl. 5, 81-90 (1990).

9.Playford, J., Appels, R. & Baum, B. R. The 5S DNA units of Acacia species (Mimosaceae). Plant Syst. Evol.183, 235-247 (1992).

10.Chappill, J. A. & Maslin, B. R. in Advances in legume systematics, part 7 (eds. Crisp, M. D. & Doyle, J. J.) 77-99 (Royal Botanical Gardens Kew, London, UK, 1995).

11.Lamarque, A. L. & Fortunato, R. H. Taxonomic significance of seed protein profiles in Acacia, with special reference to Acacia emiloana. Austr. Syst. Bot.16, 35-39 (2003).

12.Miller, J. T. & Bayer, R. J. in Advances in legume systematics, part 9 (eds. Herendeen, P. S. & Bruneau, A.) 181-200 (Royal Botanical Gardens Kew, London, UK, 2000).

13.Robinson, J. & Harris, S. A. A chloroplast DNA phylogeny of the genus Acacia Miller (Acacieae, Leguminosae). Bot. J. Linn. Soc.132, 195-222 (2000).

14.Bukhari, Y. M., Koivu, K. & Tigerstedt, P. M. Phylogenetic analysis of Acacia (Mimosaceae) as revealed from chloroplast RFLP data. Theor. Appl. Genetics98, 291-298 (1999).

15.Clarke, H. D., Downie, S. R. & Seigler, D. S. Implications of chloroplast DNA restriction site variation for systematics of Acacia (Fabaceae: Mimosoideae). Syst. Bot.25, 618-632 (2000).

16.Miller, J. T. & Bayer, R. J. Molecular phylogenetics of Acacia (Fabaceae: Mimosoideae) based on the chloroplast matK coding sequence and flanking trnK intron spacer region. Am. J. Bot.81, 1205-1224 (2001).

17.Luckow, L., T., M. J., Murphy, D. J. & Little, D. P. in Advances in legume systematics, part 10 (eds. Klitgaard, B. B. & Bruneau, A.) 197-222 (Royal Botanical Gardens Kew, London, UK, 2003).

18.Miller, J. T. & Bayer, R. J. Molecular phylogenetics of Acacia subgenera Acacia and Aculeiferum (Fabaceae: Mimosoideae), based on the chloroplast matK coding sequence and flanking trnK intron spacer. Austr. Syst. Bot.16, 27-33 (2003).

19.Ross, J. H. Towards a classification of the African Acacias. Bothalia11, 107-113 (1973).

20.Ross, J. H. A conspectus of the African Acacia species. Memorial Bot. Soc. South Africa44, 1-155 (1979).

21.Heil, M. Induction of two indirect defences benefits Lima bean (Phaseolus lunatus, Fabaceae) in nature. J. Ecol.92, 527-536 (2004).