Acacias belong to the family Mimosaceae, one of the three families belonging to the large and important legume order known as the Fabales. The other two families are the Caesalpiniaceae and the Fabaceae. Some authorities treat these three families as subfamilies of the family Leguminosae. They are all related by their leguminous fruits. The Mimosaceae are essentially tropical or subtropical trees or shrubs, and are separated from the other two families by having regular and radially symmetric flowers with 10 or more stamens. In Australia, Mimosaceae has probably more than 1000 species in 15 genera.

Fig. 3. The genus Acacia in the legume families

Before 1754, plants which are now recognised as Acacia were placed into one or other of several genera - in Australia, often Mimosa. In that year, Philip Miller formally described Acacia as a genus, basing it on the African species A. nilotica. His circumscription of the genus was, however, very loose and included 24 species. In 1783 Lamarck divided the genus into two sections (based on the nature of the stipules), and in 1825 de Candolle recognised 258 species which he divided into four sections (using leaf characters), while Sprengel, in 1826, divided his 188 species into three sections.

In 1842 Bentham undertook a comprehensive review of the genus. His circumscription was more exacting, and he distinguished Acacia from the other genera in the family as comprising those plants having more than 10 stamens which are free or not united at their bases. He divided the genus into six series as follows:-

Gummiferae - mostly African (some American and about 5 Australian) species having bipinnate leaves and spinescent stipules;

Vulgares - mostly American (some African and Asian and 2 Australian) species having bipinnate leaves and spines but not spinescent stipules;

Filicinae - American species, very similar to the Vulgares;

and the three Australasian series:

Phyllodineae - all species with phyllodes;

Botrycephalae - leaves bipinnate with gland(s) and flower heads in racemes or panicles, stipules (if present) not spinescent;

Pulchellae - leaves bipinnate with gland(s) and flowers in heads or spikes on simple peduncles, stipules not spinescent.

In 1864, Bentham, in his Flora Australiensis expanded the above classification to eleven series and a number of subseries for the 293 Australian species which he then recognised, and again in 1875 he made some modifications to his original classification for the now 432 species which he included.

Subsequently, many authors have published other classifications, the most notable and recent being those of Vassal⁶² and Pedley.^46,65 In all of these there have been attempts to rationalise and incorporate the basic morphological and chemical affinity and evolutionary relationships of species. Whilst these recent accounts have clarified and shed much light on the inter-relationships of the species of Acacia most workers continue to use, as a useful framework, the classification, at least at the higher levels, as laid down by Bentham in 1842.

BENTHAM 1842	VASSAL 1972	PEDLEY 1978	PEDLEY 1986
ser. Gummiferae	sgen. Acacia	sgen. Acacia	gen. Acacia
ser. Vulgares ser. Filicinae	sgen. Aculeiferum	sgen. Aculeiferum	gen. Senegalia
ser. Phyllodineae	sgen. Phyllodineae	sgen. Phyllodineae	gen. Racosperma
	sect. Uninervae	sect. Phyllodineae sect. Alatae	sect. Racosperma
ser. Botrycephalae		sect. Botrycephalae
		sect. Lycopodiifolia	sect. Lycopodiifoliae
	sect. Heterophyllum	sect. Plurinervia	sect. Plurinerves
		sect. Juliflorae
ser. Pulchellae	sect. Pulchelloidea	sect. Pulchellae	sect. Pulchellae

A simplified table comparing the classifications of Bentham, Vassal and Pedley

As can be seen from the above table, Vassal raised the rank of the series Gummiferae to that of subgenus, to become subgenus Acacia , and combined the series Vulgares and Filicinae into a single subgenus Aculeiferum. The three Australasian series he placed in the subgenus Phyllodineae (as Heterophyllum in error). This he subdivided into three sections. Section Uninervae contains those species whose phyllodes have one mid-nerve or are without phyllodes and also includes the bipinnate series Botrycephalae as well. Section Heterophyllum contains those species having phyllodes with two or more main longitudinal nerves. The third section, Pulchelloidea, contains some species from Bentham's series Pulchellae together with species from other phyllodinous series.

Pedley's in his 1978 treatment accepted Vassel’s subgenera Acacia and Aculeiferum. Subgenus Phyllodineae (as Heterophyllum in error) he divided into seven sections. Section Phyllodineae contains the phyllodinous species except the section Alatae which comprises a group of Western Australian species whose phyllodes are concurrent with the branchlet and form two wings along it; section Botrycephalae comprises the bipinnate species mostly from eastern Australia; and section Lycopodiifoliae is mostly a tropical group having small, whorled phyllodes. Sections Plurinerves and Juliflorae together comprise Vassal's section Heterophyllum. The Plurinerves have globular flower heads whilst the flowers are in spikes in the Juliflorae. The section Pulchellae is similar to Vassel’s section Pulchellae.

In 1986 Pedley revised his 1978 classification by raising the rank of the subgenera to that of genera and reducing the number of sections from seven to four in the predominately Australian genus Racosperma by combining the Botrycephalae, Phyllodineae and Alatae into one section, section Racosperma, and recombining the Plurinerves and Juliflorae into section Plurinervia. Whilst there may be good grounds for splitting Acacia into two or three genera, Pedley's 1986 classification has not, at present, been generally adopted. However, his 1978 classification has gained wider acceptance and is being adopted for the Flora of Australia⁶⁴ treatment of Acacia and is followed in this work.

In recent years there has been an increasing tendency to use numerical and cladistic procedures to establish inter- and infra-generic relationships. A numerical analysis⁷⁰ of available data on 177 Australian acacias reinforced Vassels separation of the Australian species into the sections Phyllodineae and Heterophyllum. From the results of a more recent cladistic analysis⁵⁰ the authors suggest that subgenus Acacia is derived from a different ancestor to the ancestor of the other two subgenera (namely Aculeiferum and Phyllodineae) and also supports Pedley's 1986 classification of dividing subgenus Phyllodineae into two main groups, one comprising the Juliflorae and Plurinerves, and the other containing the remainder of the Phyllodineae (sections Phyllodineae, Botrycephalae, Alatae, Lycopodiifoliae, Pulchellae). The analysis was, however, inconclusive in many respects. To complicate matters further, a recent serological investigation into the subgenus Phyllodineae reinforces some previously held views, such as the close relationship between the ‘racemose’ group of the uninerves and the Botrycephalae but in other respects produced some surprising results, such as suggesting that the Juliflorae is an unnatural grouping and that there is no close relationship between the Juliflorae and the Plurinerves. More work is required before definitive phylogenetic groupings can be resolved which will show the natural and evolutionary relationships within the genus Acacia.