EXTANT SEED PLANTS

Plant woody, evergreen; nicotinic acid metabolised to trigonelline; primary cell walls rich in xyloglucans and/or glucomannans, 25-30% pectin [Type I walls]; lignins rich in guaiacyl units; true roots present, xylem exarch; shoot apical meristem complex; arbuscular mycorrhizae +; stem with ectophloic eustele, endodermis 0, xylem endarch; vascular tissue in t.s. discontinuous by interfascicular regions; vascular cambium + [xylem ("wood") differentiating internally, phloem externally]; wood homoxylous, tracheids +; tracheid/tracheid pits circular, bordered; sieve tube/cell plastids with starch grains; phloem fibers +; stem cork cambium superficial, root cork cambium deep seated; nodes ?; leaf vascular bundles collateral; leaves spiral, simple, axillary buds?, prophylls [including bracteoles] two, lateral; plant heterosporous, sporangia eusporangiate, on sporophylls, sporophylls aggregated in indeterminate cones/strobili; true pollen [microspores] +, mono[ana]sulcate, pollen exine and intine homogeneous, ovules unitegmic, crassinucellate, megaspore tetrad tetrahedral, only one megaspore develops, megasporangium indehiscent; male gametophyte development endo/exosporic, gametes two, with cell walls; female gametophyte endosporic, initially syncytial, walls then surrounding individual nuclei; seeds "large", first cell wall of zygote transverse, embryo straight, endoscopic [suspensor +], short-minute, with morphological dormancy, white, cotyledons 2; plastid transmission maternal; two copies of LEAFY gene, PHY gene duplication, mitochondrial nad1 intron 2 and coxIIi3 intron present.

MAGNOLIOPHYTA

Plant woody, evergreen; lignans, O-methyl flavonols, dihydroflavonols, triterpenoid oleanane, non-hydrolysable tannins, quercetin and/or kaempferol +, apigenin and/or luteolin scattered, cyanogenesis via tyrosine pathway, lignins derived from both coniferyl and sinapyl alcohols, containing syringaldehyde [in positive Maüle reaction, syringyl:guaiacyl ratio less than 2-2.5:1], and hemicelluloses as xyloglucans; root apical meristem intermediate-open; root vascular tissue oligarch [di- to pentarch], lateral roots arise opposite or immediately to the side of [when diarch] xylem poles; origin of epidermis with no clear pattern [probably from inner layer of root cap], trichoblasts [differentiated root hair-forming cells] 0; stem with 2-layered tunica-corpus construction; wood fibers and wood parenchyma +; reaction wood ?, with gelatinous fibres; starch grains simple; primary cell wall mostly with pectic polysaccharides; tracheids +; sieve tubes eunucleate, with sieve plate, companion cells from same mother cell that gave rise to the tube, the sieve tube with P-proteins; nodes unilacunar; stomata with ends of guard cells level with aperture, paracytic; leaves with petiole and lamina [the latter formed from the primordial leaf apex], development of venation acropetal, 2ndary veins pinnate, fine venation reticulate, vein endings free; flowers perfect, polysymmetric, parts spiral [esp. the A], free, numbers unstable, P not differentiated, outer members not enclosing the rest of the bud, A many, development centripetal, with a single trace, introrse, filaments stout, anther ± embedded in the filament, tetrasporangiate, dithecal, with at least outer secondary parietal cells dividing, each theca dehiscing longitudinally by action of hypodermal endothecium, endothecial cells elongated at right angles to long axis of anther, tapetum glandular, binucleate, microspore mother cells in a block, microsporogenesis successive, pollen subspherical, binucleate at dispersal, trinucleate eventually, tectum continuous, endexine compact, lamellate only in the apertural regions, pollen tube elongated, with callose plugs, penetrating between cells, growth rate moderate, siphonogamy occuring, nectary 0, G free, several, ascidiate, with postgenital occlusion by secretion, few [?1] ovules/carpel, ovules marginal, anatropous, bitegmic, micropyle endostomal, integuments 2-3 cells thick, megasporocyte single, megaspore lacking sporopollenin and cuticle, chalazal, female gametophyte ?type, stylulus short, stigma ± decurrent, wet [secretory]; P deciduous in fruit; seed exotestal; double fertilisation +, endosperm ?diploid, cellular [first division oblique, micropylar end initially with a single large cell, chalazal end more actively dividing], copious, oily and/or proteinaceous, embryo cellular ab initio; germination hypogeal, seedlings/young plants sympodial; Arabidopsis-type telomeres [(TTTAGGG)_n]; whole genome duplication, single copy of LEAFY and RPB2 gene, knox genes extensively duplicated [A1-A4], AP1/FUL gene, paleo AP3 and PI genes [paralogous B-class genes] +, with "DEAER" motif, SEP3/LOFSEP and PHYA/PHYC gene pairs.

Possible apomorphies are in bold. Note that the actual level to which many of these features, particularly the more cryptic ones, should be assigned is unclear, because some taxa basal to the [magnoliid + monocot + eudicot] group have been surprisingly little studied. Furthermore, details of relationships among gymnosperms will affect the level at which some of these characters are pegged.

NYMPHAEALES [AUSTROBAILEYALES [[CHLORANTHALES + MAGNOLIIDS] [MONOCOTS [CERATOPHYLLALES + EUDICOTS]]]]: vessels +, elements with scalariform perforation plates; pollen tectate-columellate, tectum reticulate [perforated]; nucleus of egg cell sister to one of the polar nuclei; ?genome duplication; "DEAER" motif in AP3 and PI genes lost, gaps in these genes.

PROTEALES [TROCHODENDRALES [BUXALES [GUNNERALES + CORE EUDICOTS]]]: ?

TROCHODENDRALES [BUXALES [GUNNERALES + CORE EUDICOTS]]: benzylisoquinoline alkaloids 0; euAP3 + TM6 genes [duplication of paleoAP3 gene: B class], mitochondrial rps2 gene lost.

BUXALES [GUNNERALES + CORE EUDICOTS]: ?

GUNNERALES + CORE EUDICOTS: Ellagic and gallic acids common, cyanogenesis via phenylalanine, isoleucine or valine pathways; micropyle?; PI-dB motif +, small deletion in the 18S ribosomal DNA common.

CORE EUDICOTS: Root apical meristem closed; flowers rather stereotyped: 5-merous, parts whorled, K and C distinct, K with 3 traces, A = 2x K, internal to the C whorl, (numerous, but then often fasciculate and/or centrifugal), pollen tricolporate, (nectary disc +), [G 5], [3] also common, compitum +, placentation axile, stigma not decurrent; endosperm nuclear; fruit dry, dehiscent, loculicidal [when a capsule]; euAP1 + euFUL + AGL79 genes [duplication of AP1/FUL or FUL-like gene], PLE + euAG [duplication of AG-like gene: C class], SEP1 + FBP6 genes [duplication of AGL2/3/4 gene].

SAXIFRAGALES [VITALES + ROSIDS]: Stipules +.

VITALES + ROSIDS: Anthers articulated [± dorsifixed, transition to filament narrow, connective thin].

ROSIDS: embryo long; genome duplication; chloroplast infA gene defunct, mitochondrial coxII.i3 intron 0.

ROSID II: Flavonols +; (cambium storied); petiole bundle(s) annular; ca 2 ovules/carpel, style +; endosperm scanty.

Based on a study of the genome of Arabidopsis, De Bodt et al. (2005, see also Maere et al. 2005) suggest there was a duplication of the whole genome some 109-66 million years before present, although given the uncertainty over the dating of this duplication and relationships within rosids, exactly where this should go on the tree is unclear.

Huerteales are strongly supported as being monophyletic (Peng et al 2003: three gene analysis, sampling sketchy). However, relationships within the rosid II group are somewhat uncertain, and it remains unclear just what group is sister to Brassicales. The clade [Malvales + Sapindales] is a possible candidate (Soltis et al. 2000; Peng et al 2003: both weak support), Endress and Matthews (2006) note some features perhaps more common in these two families than elsewhere, and there is some support for this clade in Soltis et al. (2007a). However, Bausher et al. (2006) in an analysis of whole chloroplast genomes found strong support for the clade [Brassicales + Malvales], but only one species from the three orders, and no Huerteales, were included (see also Jansen et al. 2007; Moore et al. 2007). There was also some support for this topology in analyses by Savolainen et al. (2000) and Hilu et al. (2003); Alford (2006) suggested that [Huerteales (Perrottetia not included) + Brassicales + Malvales] might be sister to Sapindales. In line with these latter studies, Worberg et al. (2007b) recovered the relationships [Sapindales [Huerteales [Brassicales + Malvales]]], with strong support, and also for the monophyly of each of the four orders. However, in studies including the mitochondrial matR gene, although the rosid II clade was recovered, relationships within it were unclear (Zhu et al. 2007).

SAPINDALES Dumortier  Main Tree, Synapomorphies.

Interesting secondary compounds, ethereal oils, myricetin +; (secretory cells/tissue +); vessel elements with simple perforations; silicified wood or wood with SiO₂ grains in all major families [esp. Anacardiaceae and Burseraceae]; tension wood +; petiole bundle(s) annular; branching from previous innovation, petioles leaving a prominent scar; leaves spiral, odd-pinnately compound, conduplicate; A 2x K, filaments articulated, (exine distinctly striate), disc well developed, G opposite C or odd member adaxial, few ovules/carpel; seed coat?; (embryo green). - 9 families, 460 genera, 5670 species.

Sapindales contain ca 3% eudicot diversity. Associated with the accumulation of noxious secondary metabolites in this group, specialised herbivores are found on many of them. Thus the hemipteran Calophya eats largely Anacardiaceae, Burseraceae, Simaroubaceae and Rutaceae (Burckhardt & Basset 2000) - plus a couple of records from entirely unrelated families. Galls are quite common, perhaps especially on Sapindaceae and Anacardiaceae (Mani 1964).

Gums and resins occur in both the Rutaceae-Meliaceae-Simaroubaceae and Burseraceae-Anacardiaceae groups (Nair 1995). Stratified phloem may be quite widespread (in some Meliaceae, Burseraceae and Simaroubaceae, at least: M. Ogburn, pers. comm.), also Sapindaceae. Teeth, when present, have a clear glandular apex broadening distally and with a foramen and two accessory veins (or one, the other going above the tooth: Hickey & Wolfe 1975). The stipules may clearly be modified leaflets and have been described as pseudostipules or metastipules. The latter have been defined as having the morphology of true stipules, yet there is good reason to believe that they are derived from pseudostipules (Weberling & Leenhouts 1965). The flowers are often imperfect, and staminate and carpellate flowers have well-developed pistillodes and staminodes respectively, so they can be difficult to distinguish. Septal cavities have been noticed in Cneorum (Rutaceae) and Koelreuteria (Sapindaceae), but they do not secrete nectar (Caris et al. 2006, cf. septal nectaries in monocots). The style in at least some Rutaceae and Sapindaceae is hollow (Lersten 2004).

For general relationships, see Gadek et al. (1996), while Pell (2004) notes some deletions and insertions that may characterise groupings within the clade. Muellner et al. (2007) present a two-gene tree with quite good sampling; their results suggest the poorly supported basal relationships in the tree here, a fair bit of resolution elsewhere, excepting only a moderately-supported sister group relationship between Meliaceae and Simaroubaceae (cf. Gadek et al. 1996), so a trichotomy including Rutaceae is shown in the tree here. Biebersteiniaceae, Kirkiaceae and Nitrariaceae are poorly known morphologically, etc. Note that in the past Bretschneidera and Akania have been associated with Sapindales, Bretschneidera in particular looking very like a member of Sapindaceae, and morphological phylogenetic analyses may continue to suggest this position; they are to be included in Brassicales, q.v., (Ronse de Craene & Haston 2006).

Includes Anacardiaceae, Biebersteiniaceae, Burseraceae, Kirkiaceae, Meliaceae, Nitrariaceae, Rutaceae, Sapindaceae, Simaroubaceae.

Synonymy: Acerales Lindley, Aesculales Bromhead, Burserales Baskerville, Citrales Dumortier, Julianales Engler, Leitneriales Engler, Meliales Lindley, Nitrariales Doweld, Rutales Perleb, Terebinthales Dumortier - Burseranae Doweld, Rutanae Takhtajan, Sapindanae Doweld - Rutidae Doweld - Aceropsida Endlicher, Aesculopsida Brongniart, Rutopsida Meisner

BIEBERSTEINIACEAE Endlicher   Back to Sapindales

Perennial herbs; vessels?; nodes?; leaves (2-3 compound), leaflets lobed, toothed, stipules petiolar, lobed or not; inflorescence racemose; C clawed, nectary glands at base of antisepalous stamens, pollen 3-celled, 1 apical pendulous epitropous ovule/carpel, micropyle endostomal, embryo sac tetrasporic, 16-celled [Penaea type], styles separate, impressed, apically connate, stigma capitate; fruit a schizocarp, columella persisting, K ± accrescent; testa 0, exotegmen thick-walled, not lignified, endotegmen tracheidal; endosperm development?, embryo somewhat curved, cotyledons foliaceous; n = 5.

1/5. Greece to Central Asia (Map: from Heywood 2007; Muellner et al. 2007). [Photos - Collection]

• Biebersteiniaceae are perennial, glandular-hairy herbs that have odd-pinnate leaves with lobed to compound leaflets and petiolar stipules. Ihe inflorescence is erect and racemose and the flowers are large. The fruit has a single seed per carpel and a persistent columella.

At least some species are foul-smelling. The petalline stamens are longest; Takhtajan (1997) described the ovules as being unitegmic. The anatomy of Biebersteinia is largely unknown.

Although previously associated with Geraniaceae (Geraniales), molecular data place it in Sapindales, albeit with a long branch (Bakker et al. 1998). The herbaceous habit is unusual in the order, but ethereal oils (no oxygenated sequiterpenes, high proportion of aliphatic hydrocarbons), single ovule/carpel, etc., are all congruent with a position here. Its exotegmen is rather like that of Ledocarpaceae and Vivianaceae (both Geraniales), especially when young, since both exotesta and endotegmen are tanniniferous (Boesewinkel 1988), but it has distinctive flavones and methyl ethers quite like Rutaceae in part, but unlike Geraniaceae (Bate Smith 1973; Greenham et al. 2001).

Additional information is taken from Baillon (1874), Kunth (1912), Hegnauer (1989, as Geraniaceae: chemistry), Boesewinkel (1997) and Tzakou et al. (2001: fatty acids).

NITRARIACEAE Berchtold & J. Presl   Back to Sapindales

Shrubs; ß-carbalin alkaloids +, ethereal oils?; cork in inner cortex; wood storied; nodes?; mucilage cells +, throughout plant or not; cuticle waxes 0 (platelets, rodlets); stomata anomocytic; leaves two per node, adjacent (spiral), simple or multifid, ?ptyxis, stipules minute or foliaceous, ± cauline; A (10-)15, (antepetalous A in pairs; flower 3-4-merous, A 4, opposite K [Tetradiclis]), pollen zonocolporate, nectaries antepetalous [Nitraria] or extra/intrastaminal disc, G [2-4], 1 subapical apotropous or 6-many ?epitropous ovules/carpel, micropyle zig-zag or bistomal, style long to rather short, basal [Nitraria, Tetradiclis], stigma as commissural[?] lines down part of its length, dry; fruit a capsule, drupe, or berry; exo- and endotesta short palisade, enlarged [cells inflated - Tetradiclis] or not, endotegmen ± fibrous [Peganum] or not [Tetradiclis]; (endosperm copious); n = 7, 12, 30.

3[list]/16. Usu. ± arid regions in N. hemisphere, but little in America, also Australia (Map: Pan et al. 1999; Brummitt 2007; FloraBase 2007). [Photo - Flowers]

Nitraria often has two or three leaves at a node, although not all these have scarious stipules; the leaves are occasionally toothed or lobed. It also has apotropous ovules (Takhtajan 1997) and aliform-confluent xylem parenchyma. Peganum also may have two much-divided leaves at a node, in this case also with ± foliaceous and divided stipules; it has foliar raphides. The androecium is described as being obdiplostemonous by Eckert (1966); the 15 stamens may be in groups of three opposite the sepals, or there may be paired stamens opposite the petals (Ronse Decraene & Smets 1991a, 1992, 1996a; Ronse Decraene 1992; Ronse Decraene et al., 1996). Only the seeds in the central locelli of Tetradiclis are released when the capsule opens. Takhtajan (1997) says that stipules there are absent; they are present, if small. There is no endothelium (Kapil & Tiwari 1978), cf. Zygophyllaceae s. str. Basically, just about all features need careful examination throughout the group. I don't really understand the variation...

Molecular data suggest the relationships [[[Peganum + Malacocarpus] Tetradiclis] Nitraria] (Sheahan & Chase 1996), and assign the group to Sapindales. Phenetically, and also in features like wood anatomy, and perhaps also chemistry (Nag et al. 1995), there are similarities to Zygophyllaceae, although Nitrariaceae and Zygophyllaceae both grow in dry and warm habitats, which may account for some of these similarities. Indeed, they all used to be together; Takhtajan (1997) included this family (as three) in his Zygophyllales. Zygophyllales here include Zygophyllaceae s. str. and the monogeneric Krameriaceae and are not remotely close to Nitrariaceae.

Batygina et al. (1985) provide details of endosperm development, etc., Danilova (1996), of seed anatomy, and Sheahan and Cutler (1993), of anatomy; for chemistry, see Hegnauer (1973, 1990: as Zygophyllaceae).

Synonymy: Peganaceae Doweld (raphides; limited mucilage cells; stipules +; many ovules/carpel; fruit a loculicidal capsule or berry - 1-2 genera. Peganum, Malacocarpus), Tetradiclidiaceae Doweld (flowers (3-)4-merous; A = and opposite K, disc 0, G [(3) 4], each divided into three parts, 6 ovules/carpel borne on end of long placenta arising at base of G, 4 ovules in central locellus, 1 each in lateral locelli, style ± basal, stigmatic ridges down expanded apical portion; fruit a loculicidal capsule; n = 7)

Kirkiaceae + Anacardiaceae + Burseraceae: 2 ovules/carpel; fruits with 1 seed/carpel.

KIRKIACEAE Takhtajan   Back to Sapindales

Tree or shrub; ellagic acid +; nodes?; petiole bundle annular, with medullary bundles; cuticle waxes 0; stomata ?anomocytic; leaves ± opposite to spiral, leaflets serrate; plants monoecious, inflorescence dichasial, flowers small, 4-merous; K basally connate?, staminate flowers: A = and opposite K, filament articulation?, pollen syncolpate, disc broad, well developed, pistillode +; carpellate flowers: staminodes +, G [4(-8)], ?orientation, ovules pendulous, (1 ovule/carpel), styles closely adpressed, erect, finally spreading, stigmas punctiform; fruit a schizocarp, mericarps pendulous from columella; testa?; endosperm ?type, slight, embryo curved; n = ?

2/6. Tropical and S. Africa, Madagascar (Map: from Brummitt & Stannard 2007).

• Kirkiaceae are woody plants that have often rather closely-set odd-pinnate leaves with toothed leaflets. Ihe inflorescence is dichasial and the flowers are small and 4-merous. The fruit is a ridged or angled schizocarp with persistent if slender columella strands.

Kirkiaceae were previously placed in Simaroubaceae, but they lack quassinoids and limonoids. The wood of Pleiokirkia is reported to smell like honey (Schatz 2001).

For some information on anatomy, see Jadin (1901) and on chemistry, see Nooteboom (1967).

Anacardiaceae + Burseraceae: biflavonoids [alone in order]; (vessel elements with scalariform or reticulate perforations); vertical intercellular secretory canals in phloem, this surrounded by a light-coloured, sinuous, sclerenchymatous band [not easy to see]; cuticle waxes usu. 0; leaves compound, leaflets opposite; (plants dioecious); flowers rather small, K often connate, C little longer than K, A obdiplostemonous when flowers isomerous, palynologically indistinguishable, (disc extrastaminal); fruit an (operculate) drupe, endocarp cells in a mass, lignified, not oriented.

The basic endocarp condition for this clade seems to be that of an unoriented mass of sclerified and often crystalliferous cells (Wannan & Quinn 1990). Such an endocarp characterises Spondiadoideae, and it is also found in Burseraceae as well as in Buchanania, Campnosperma and Pentaspadon, included in Anarcardioideae, as by Pell (2004: unfortunately, not sequenced). Indeed, Buchanania is possibly sister to remaining Anacardioideae (Aguilar-Ortigosa & Sosa 2004), appropriate, given its endocarp anatomy. It has been suggested that an operculum may be derived twice in Anaacardiaceae (Pell & Urbatsch 2001), but it is also found in fruits of Burseraceae and perhaps is plesiomorphic within the whole clade.

For chemistry, see Hegnauer (1964, 1989), for general developmental information, see Bachelier and Endress (2007a).

ANACARDIACEAE R. Brown, nom. cons.   Back to Sapindales

Trees or shrubs (climbers); exudate black or becoming blackish; (cork cortical); pith loose, shining; vertical resin ducts +; wood often fluorescing; nodes usu. 3:3; petiole with annular wing bundles; leaflets not articulated, margins toothed or not, base of petiole often swollen; inflorescence cymose, flowers (3-)5(-7)-merous; (K, C 0), A (1-10<), when 5, opposite K, (stamens on disc; disc 0; andro/gynophore +), 1 apical to basal usu. apotropous uni- or bitegmic ovule/carpel, micropyle zig-zag (endostomal), funicle often long, placental obturator common, styles separate (single), terminal to gynobasic, stigma capitate (lobed), dry; drupe often asymmetrical, ± flattened (K much accrescent); seed often ± pachychalazal, (exotestal cells [and hypodermis] thickened), endotegmen usu. ± thickened, lignified; endosperm oily (and starchy), embryo often curved (cotyledons folded - Mangifera); n = 7-12, 14-16, 21.

70[list]/600. Tropical, also temperate (Map: from Heywood 1976; George 1985; Meusel et al. 1978). Two groups below. [Photo - Flower, Fleshy fruit, Dry fruits.]

1. Spondoideae Arnott

Biflavonoids; also alkylcathechols and alkylresorcinols 0; A obdiplostemonous, G [(1-[pseudomonomerous?])4-5(-12)] and as many locules, ovule pendulous; exocarp thick.

10/115. Tropical.

Synonymy: Spondiadaceae Martynov

2. Anacardioideae Link

(Biflavonoids, 5-deoxyflavonoids, also alkylcathechols and alkylresorcinols [phenols with unsaturated side chains - allergenic] +); leaves simple (compound); A variable, G 1 [pseudomonomerous?], [3(-6, partly syncarpous, Buchanania)], antesepalous carpel alone fertile, ovule variable in position, (unitegmic, funicle massive, with outgrowths - Pistacia, etc.; ponticulus +), (styles connate); exocarp thin, epidermis lignified, endocarp with up to three layers of palisade lignified sclereids, internal to these a crystalliferous layer [= stratified].

60/485: Semecarpus (60). Largely tropical, also temperate.

A gall-forming jumping plant louse, the hemipteran Calophya, is notably common on Schinus (Burckhardt & Basset 2000). In Pistacia, and perhaps other genera, chalazogamy occurs, the pollen tube moving from the funicle via the ponticulus, an outgrowth of the funicle that bridges the gap between it and the chalaza (Martínez-Pallé & Herrero 1995).

For the limits of Rhus, which seem best narrowly drawn (i.e., restricted to ca 35 species in the genus), see Yi et al. (2006, and references). Pistacia and Amphipterygium (see Julianaceae below) have distinctive ovules; both are wind pollinated, dioecious, with reduced flowers, unitegmic ovules with a massive funicle, etc. (Bachelier & Endress 2007b).

Synonymy: Blepharocaryaceae Airy Shaw (inflorescence compact, involucrate), Comocladiaceae Martynov, Julianaceae Hemsley (A extrorse; carpellate flowers lacking perianth, surrounded by involucre), Lentiscaceae Horaninow, Pistaciaceae Adanson, Podoaceae Franchet (leaves opposite, pinnate, carpellate flowers lacking perianth), Schinaceae Rafinesque, Vernicaceae Link

• Anacardiaceae may be fairly readily recognised because of their often black and/or rather resinous-smelling exudate; the leaves often have black discolorations and the blade may dry a distinctive grey colour. In herbarium specimens, the dry, black exudate may entirely cover the cut surface of the twig, or it may form a distinctive black ring around the periphery. The leaves are often odd-pinnate and the leaflets are opposite to alternate. The flowers are small, and the fruits are often flattened, operculate drupes with excentric style or styles. Note that Anacardiaceae like Pachycormus have thin, brown, flaking bark and vegetatively look quite like Burseraceae...

Anacardiaceae are noted for the sometimes extremely violent allergenic reactions their exudates cause; catechols, resorcinols and other types of phenolic compounds, often in a mixture, as in urushiol, are involved. About a quarter of the genera - all Anacardioideae - have such compounds.

Aphids (Fordinae) that form distinctive galls are closely associated with species of Pistacia (Inbar 2009). There is a variety of winged fruit types in Anacardioideae. These include fruits adnate to broad bracts (Dobinea), samaras (Loxopterygium), the flattened peduncle of the inflorescence (Amphipterygium), much enlarged sepals (Parishia), and much enlaged petals (Swintonia); the evolution of these fruit types seems to be correlated with the adoption of a drier habitat (Pell & Mitchell 2007).

Spondiadeae and some Rhoeeae - the clade includes Pegia, Tapirira and Cyrtocarpa, Spondoideae (see Aguilar-Ortigosa & Sosa 2004 and Pell 2004, the latter with a list of included genera) - are sister to the rest of the family (Anacardioideae: Pell & Urbatsch 2000, 2001; Pell 2004 and Mitchell et al. 2006 for a list of genera, but see below); wind-dispersed taxa in this latter subfamily do not form a single group (Pell & Mitchell 2007, c.f. Pell & Urbatsch 2001). Buchanania in some analyses is quite well supported as sister to Anacardioideae (Aguilar-Ortigosa & Sosa 2004; Wannan 2006), consistent both with its chemistry, endocarp anatomy (see above), carpel number of 4-6, and different position of the fertile carpel, but its position is not fixed in others (Pell & Mitchell 2007, cf. abstract). Note that Campnosperma, so far included in only one study (Chayamarit 1997, sampling limited, relationships suggested are unlike those in other studies, no support values), has an endocarp similar to that of Buchanania and the fruit is sometimes two-locular. The morphology of neither of these genera is accounted for in the Anacardioideae as characterized here; if they are, the subfamily will probably lack much in the way of apomorphies.

Branching in Anacardium may occur on the current flush. Wind-pollinated taxa often lack a disc, also petals. Mangifera has one or two stamens borne inside the disc. In Anacardium the single stamen is on an oblique plane of symmetry; more generally, the position of the carpel, when single, suggests tht the flower has oblique symmetry.

For general information, see Ding Hou (1978), for fruit anatomy, Wannan and Quinn (1990), for floral morphology, Wannan and Quinn (1991), for general chemistry, Young (1976), for chemistry of Julianaceae, see Hegnauer (1966, 1989), for seed anatomy, see von Teichman (1991, 1994, and references), and for wood anatomy, see Gupta and Agarwal (2008). Pell (2004) covers the morphology of the whole family in a phylogenetic context, which she provides; Mitchell et al. (2006) focus more on Spondiadoideae.

BURSERACEAE Kunth, nom. cons.   Back to Sapindales

Trees or shrubs; bark often flaky, light grey; colorless to white resinous exudate common; tannins?; snail glands [curled ± uniseriate glandular hairs] common; (cork deeper in Santiria); pith frequently with vascular strands; nodes usu. 5:5; secretory canals rare; sclereids in stem; (petiole bundle with medullary strands); leaflets (with pellucid dots), ?ptyxis, margins often toothed, petiolules and petioles often pulvinate, stipules petiolar or cauline, laciniate to entire, or 0; flowers 3-5(-7)-merous; (hypanthium +), K induplicate-valvate, C induplicate-valvate, (A = and opposite K; pollen psilate, striate), ovules epitropous, ventral carpel bundles fused bundles of adjacent placentae, nucellar cap +, micropyle endostomal, style usu. short; drupe often angled, often only one loculus developing, stone with valves, (pseudaril [pericarpial in origin] +; mesocarp quite frequently splits down loculicidal radius; fruit a schizocarp, septifragal with columella), (winged pyrenes), K deciduous; (exotesta with shortly radially elongate but unthickened cells), endotesta lignified, ± tracheidal; reserves hemicellulosic; n = (11-)13, 23.

18[list]/550. Tropical. Two groups below.[Photo - Leaf, Flower, Fruit.]

1. Beiselia

Vessel elements with scalariform perforations; G [9-12], ovules superposed.

1/1: Beiselia mexicana. Mexico.

The Rest

G [(2-)3-5], stigmatic head formed by postgenital fusion; cotyledons folded, palmately lobed.

17/550: Commiphora (190, 150 from Africa), Protium (85), Canarium (75), Bursera (50). Tropical, but esp. America and N.E. Africa (Map: from Rzedowski 1978; D. C. Daly, pers. comm.).

• Burseraceae are recognisable by their often scented resinous exudate and smooth, flaking bark; compound, odd-pinnate leaves with opposite, often long-petiolulate and more or less pulvinate leaflets that have prominent fine venation especially when dry; and petioles, petiolules and rachis that tend to be brownish and more or less scurfy. The swollen leaf base is often concave adaxially. The calyx is often connate and 3- or 4-lobed, and the angled drupes have a wrinkled surface when dry. The resinous exudate often smells of almonds (myrrh, frankincense), and in some genera the leaflets are strikingly symmetrical. Anacardiaceae like Pachycormus have thin, brown, flaking bark and vegetatively look quite like Burseraceae.

The split between Bursera and Commiphora has been dated to some 120 million years before present, with diversification within the former genus starting some 70 million years before present (Becarra 2005); these dates seem something of an overestimate. Weeks and Simpson (2007) suggest that divergence of Commiphora from the clade now represented by the E. Asian B. tonkinensis occured some 53-41.6 million years before present (Eocene), although Commiphora itself did not diversify until 32.3-23.2 million years before present, Neogene aridification of Africa occuring more or less at that time (see Weeks et al. 2005 and Weeks & Simpson 2007 for further details on the complex biogeography of the family). For coadaptive relationships between Burseraceae, especially Bursera itself, and herbivorous chrysomelid beetles (Blepharida), and how the latter deal with the toxic terpene-containing resins the plants contain, see Becerra (1997, 2003 and references) and Becerra et al. (2001: particularly interesting).

Beiselia (B. mexicana is the only species) has simple cotyledons; molecular studies suggest that it is sister to the rest of the family (e.g. Clarkson 2002). This has considerable implications for character evolution, including cotyledon morphology; Beiselia also has probably derived features, like its gynoecium with its 9-12 carpels. Commiphora may be embedded in Bursera, but the support is weak (Weeks et al. 2005).

There can be foliaceous stipules; these are usually interpreted as being reduced leaflets of the compound leaf. A few genera (e.g. Garuga) have a well-developed hypanthium; the disc is rarely extrastaminal (Triomma). Odd G abaxial in 4-merous Amyris (Schnizlein 1843-1870, fam. 244).

For general information, see Lam (1931, 1932), Leenhouts (1956), and Forman et al. (1989), and for pollen morphology, see Harley and Daly (1995: Protieae) and Harley et al. (2005: considerable variation).

Synonymy: Balsameaceae Dumortier, Neomangenotioideae Leroy

SAPINDACEAE Jussieu, nom. cons.   Back to Sapindales

Woody; quebrachitol [cyclitol], toxic saponins, cyclopropane [non-protein] amino acids +, ellagic acid 0 (+); cork also outer cortex (pericyclic - Dodonaea); latex of sorts not uncommon; (vessel elements with scalariform perforations; petiole bundle with cortical or adaxial bundles); cuticle waxes 0 (platelets, rodlets); leaves spiral or opposite, even-pinnate or ternately or palmately (bi)compound (simple), leaflets opposite or not, articulated, ptyxis also conduplicate-plicate, margins entire to toothed (rachis winged; petiolar stipules +), colleters common; inflorescence paniculate, the flowers often in clusters, imperfect; pedicels articulated, flowers 4-5-merous, [K5 C₄; K4 C₄, etc], C (0, 5+), often clawed, hairy, and with various ± complex appendages, disc outside A, A (4-[Glenniea])8(-16), often hairy, G [(2) 3(-6)], ovules variously curved, sessile, often apotropous, funicular obturator +, micropyle endo(bi)stomal, (style hollow; styles +), stigma strongly 3-lobed or not, dry or wet; fruit a capsule, or a schizocarp with 1-seeded samaras; seed often pachychalazal, chalazal/integumentary arils and sarcotesta common, testa vascularised, exotesta palisade (not), unlignified, (mesotestal cell walls thickened and lignified; endotesta crystaliferous, exotegmen fibrous, lignified or not); endosperm 0, starchy, embryo curved, the radicle in a pocket of the testa, cotyledons spiral or not; n = esp. 10-12 [climbers] and 14-16 [non-climbers]; germination hypogeal or epigeal.

135[list]/1580: - four subfamilies below. ± World-wide. (Map: from Herzog 1936; Meusel et al. 1978; George 1985). [Photo - Flower, Fruit, Fruit.]

1. Xanthoceroideae Thorne & Reveal

Phloem stratified; leaves deciduous; disc with golden, pollen spiny, horn-like glands, 6-8 ovules/carpel.

1/1: Xanthoceras sorbifolia. N. China.

Hippocastanoideae + Dodonaeoideae + Sapindoideae: flowers often strongly obliquely or vertically [Aesculus] monosymmetric, 2 ovules/carpel.

Leaves opposite, palmate, (pinnate, simple); (disc inside A - Acer), (nucellar cap + [Acer]).

5/130: Acer (110: for a phylogeny, etc., see Li et al. 2006). N. temperate, some tropical, usu. mountains.

Synonymy: Aceraceae Jussieu (cuticle wax crystalloids quite common; leaves opposite; nectary may be inside the A, stigma dry), Aesculaceae Berchtold & J. Presl, Hippocastanaceae A. Richard (stigma dry; n = 20), Paviaceae Horaninow

3. Dodonaeoideae Burnett

?

16/135: Dodonaea (70). Pantropical-warm temperate, esp. Australia/Southeast Asia.

Synonymy: Dodonaeaceae Link (stigma wet)

4. Sapindoideae Burnett

(Lianes with branch tendrils; secondary thickening anomalous;) (pollen oblate, triporate - Serjania, etc.), 1 (2) ovules/carpel.

111/1185: Serjania (215), Paullinia (195), Guioa (65), Cupaniopsis (60). Pantropical. Paullineae (8 genera) contain one third of the species in the family.

Synonymy: Allophylaceae Martynov, Koelreuteriaceae J. Agardh, Ornithrophaceae Martynov

• Sapindaceae are recognisable by their often spiral, pinnately-compound leaves that often have subopposite leaflets and a terminal rhachis tip, although many taxa have ternately/palmately-compound leaves. The base of the petiole is quite strongly swollen and the stem is ridged. The leaflets may be coarsely serrate and the fine venation is often prominent when dry. The inflorescences are paniculate, and the flowers are often borne in congested groups along the axis. The flowers are often rather small and are conspicuously hairy inside; there are usually eight stamens and complex folds or scales on the petals. The fruits have only one or two seeds per carpel and are often deeply lobed, or a single carpel only may develop, the other one or two carpels persisting at the base; samaras of various types are common. The radicular pocket in the testa is distinctive.

Cupaniopsis-type pollen is distinctive and is widespread in the fossil record, including several sites in Africa, although Sapindaceae with such pollen are no longer found there (Coetzee & Muller 1984).

Aesculus has bud scales, Billia has naked buds, but both branch from previous flush. Radlkofer (1892-1900) shows Serjania as having strongly obliquely symmetrical flowers, with the odd gynoecial member abaxial on the plane of symmetry. The abaxial corolla member is absent, but A are abaxial, 2 adaxial(?)-lateral members being missing. In Acer, the samaras are oblique (Schnizlein 1843-1870). The petals of Sapindaceae are often rather complex, and have a similarly complex set of terms used to describe them. Brizicky (1963) reports that the ovules may be epitropous, while those of Koelreuteria and other taxa are both epitropous (the lower ovule) and apotropous (the upper ovule) in the same loculus (Mauritzon 1936; Danilova 1996). The fruit looks like a follicle when only one carpel develops; dehiscence is, however, down the abaxial side. In many Sapindaceae, and some Anacardiaceae, the pericarp grows much faster than the seed, so what seem to be almost mature fruits can contain seeds that are still very small.

Preliminary studies suggested that Xanthoceras, with simply 5-merous, polysymmetric flowers, ovules arrranged in parallel (see also Magonia), and complex, golden nectaries borne outside the eight stamens, might be sister to the rest of Sapindaceae, and the genera included in the erstwhile Aceraceae and Hippocastanaceae formed monophyletic sister taxa, the combined clade being sister to the rest of the family (see Klaassen 1999; Savolainen et al. 2000a; Soltis et al. 2007a). A recent two-gene study (Harrington et al. 2005) has largely confirmed these results. In single gene analyses Xanthoceras did not have this position, being somewhat embedded in the family (no strong support, however); it was only in the joint analysis that is was sister to all other Sapinaceae (70% bootstrap, >95% posterior probability). Early morphological analyses (Judd et al. 1994) suggested a rather different set of relationships.

Sapindaceae are chemically similar in some respects to Fabaceae (e.g. both have non-protein amino acids: for a summary of these, see Fowden et al. 1979), and both have compound leaves, but they are unlikely to be immediately related.

For an early but still useful account of the family, see Radlkofer (1933, 1934), for seeds, see van der Pijl (1955), for chemistry, see Hegnauer (1964, 1966, 1973, 1989, 1990, also under Aceraceae and Hippocastanaceae), for pollen, see Muller and Leenhouts (1976), for wood anatomy, see Klaassen (1999) and Agarwal et al. (2005), for embryology, Tobe and Peng (1990), for chromosome numbers, Lombello and Forni-Martens (1998), for floral development of Koelreuteria, see Ronse Decraene et al. (2000b), and for fruits of Paullineae, see Weckerle and Rutishauser (2005).

Rutaceae + Meliaceae + Simaroubaceae: alkaloids, pentanortriterpenes +; cuticle waxes 0; (leaves trifoliate, simple); inflorescence branches cymose.

The triterpenoid limonoids, meliacins, cneorids, and quassinoids are biosynthetically related (e.g. Connolly et al. 1970; Evans & Taylor 1983) and often have a bitter taste.

Rutaceae + Meliaceae: tetranortriterpenes, flavones +; stigma capitate.

RUTACEAE Jussieu, nom. cons.   Back to Sapindales

Herbs to trees; pyranochromones, furanocoumarins +; (vessel elements with scalariform perforations); wood often fluorescing; (nodes 1:1; cuticle waxes platelets, rodlets, etc.); stomata various; leaves spiral to opposite, (rhachis winged), leaflets opposite (subopposite), usu. articulated, ptyxis also flat, margins entire to crenate (serrate), punctate because of schizogenous cavities, (stipules intrapetiolar - Metrodora); breeding system various; flowers (vertically or obliquely monosymmetric), (3-)5-merous; K (2-4), connate or free, C (0-4), often valvate?, (connate), A (2-many in a single whorl), obdiplostemonous, filaments ± flattened (connate), (androgynophore, gynophore +), G (1 [2-)5(-many)], variously connate to almost free (opposite K - Zanthoxylum), 1-many (apical) ovules/carpel, micropyle zig-zag (endostomal), style (connate only at apex; separate (and marginal)), impressed to ± gynobasic, stigma (lobed), dry or wet; fruit a loculicidal and part septicidal capsule, often strongly lobed, or follicle, mesocarp and endocarp separating, or berry (drupe); exotesta often mucilaginous, irregularly palisade, or lignified and fibrous (mesotesta sclerotic), exo-(and meso- and endo-) tegmen tracheidal; (endosperm +), (embryo curved); n = (7-)9(-11+).

161[list]/1815: Zanthoxylum (250: prickles in the stipular position), Melicope (150). Largely tropical (Map: from Meusel et al. 1978; Brummitt 2007). [Photo - Flower, Flower, Fruit.]

• Rutaceae are recognisable because of their often opposite, compound and punctate leaves; their cymose inflorescences; their flowers, with the stamens often arranged in a ring, thick filaments, conspicuous disc at the base of the ovary, and expanded stigma, are quite distinctive. Although the fruits are variable, many have a glandular-punctate pericarp, while others are dry, deeply lobed, and dehisce to reveal the black and shiny seeds.

Papilionidae-Papilioninae caterpillars are notably common (ca 1/3 of the records, 80% of the ca 550 species of Papilio - Zakharov et al. 2004) on Rutaceae; like the magnoliids, on which Papilioninae are also found, plants of Rutaceae have a variety of alkaloids that attract the butterflies. It is possible that Rutaceae were the original food plants for Papilio, since even those caterpillars which now eat Magnoliales will eat Rutaceae if they have to. Ca 250 species of Diosmeae are restricted to South Africa, large to the Cape Floristic Region (Trinder-Smith et al. 2007).

Rutaceae as circumscribed here are a variable group. They are very diverse in their secondary metabolites, some of which (essential oils, coumarins, etc.) are similar to those in Apiaceae (Hegnauer 1971), while the alkaloids are like those found in some magnoliids and are produced via nine or more different biosynthetic pathways. Rutaceae and associated "families" once segregated from them all have pyranochromones. Da Silva et al. (1988) surveyed these secondary metabolites, suggesting that an overhaul of the infrafamilial classification was in order. Adsersen et al. (2007) note the value of prenylated acetophenones as a marker for Xanthoxyleae (inc. Melicope, etc.).

Rutaceae are also rather variable in flower and fruit, less so vegetatively. Kallunki (1992) illustrates the flowers of Erythrochiton fallax as having the median sepal adaxial, but their exact orientation, and how they are actually held, is unclear since the inflorescence can be up to 1.5 m long. The flowers of Galipeinae, to which Erythrochiton belongs, may have only two stamens plus staminodes, a connate corolla, filaments connate and forming a tube, or a corolla that is connate only because of the adnation of the filaments (Pirani & Menezes 2007). Peltostigma has a floral formula K3 C3 A9 G [?5], and looks almost lauraceous; Pilocarpus has an erect raceme and the calyx is reduced to a rim. Ovules of Glycosmis are unitegmic, and both apotropous and epitropous ovules are recorded from the family. Nucellar polyembryony is widespread. Triphasia has G [3], with the odd member adaxial, and the same is true of Cneorum tricoccon, which has 3-merous flowers (see Caris et al. 2006 for floral development).  

General relationships in a two-gene analysis: most of Rutaceae sister to [[Spathelia + Dictyoloma] [[Cneorum + Ptaeroxylum] Harrisonia]]], both big clades with moderate support; Harrisonia with sequences from only a single gene (Chase et al. 1999). Spathelia (chromones) and Dictyoloma (C valvate, 4-5 epitropous ovules, embryo curved) are a strong pair. Cneorum: chromones, limononoids +; leaves simple, not punctate; micropyle endostomal; fruit drupaceous, schizocarpic. Ptaeroxylum: chromones only; oil cells (nectaries) +; petiole bundle more or less cylindrical, of two opposed plates (arcuate - Bottegoa); stomata anomocytic to cyclocytic; wood anatomy similar to that of some Sapindaceae and fewer Rutaceae; flowers 3-merous, short androgynophore +, A 4-5, pollen reticulate, more like that of Rutaceae than Meliaceae, 1(-3) apotropous, campylotropous ovules/carpel, hypostase +; fruit a loculicidal capsule, carpels opening adaxially and separating from columella; testa multiplicative. Finally, Harrisonia: (2-methylchromones [ptaeroxylins] + limonoids; solitary oil cells, nectaries on leaves, stipules at leaf base and stipular thorns, filaments with scales, carpels ± connate, 1 apotropous ovule/carpel; testa multiplicative, exotestal cells large, outer walls thickened, exotegmen tracheidal). Jadin (1901) noted that anatomically Harrisonia was rather different from other Simaroubaceae its its heterogenous pith and lack of medullary secretory canals. Although it does not seem to have pellucid foliar gland dots, Fernando and Quinn (1992) found secretory cavities in the fruits and Fernando et al. (1995) suggested that its removal from Simaroubaceae was justified on both moelcular and morphological grounds.

Some of the fruit characters used to distinguish the classic subfamilies in Rutaceae are proving unreliable. Within core Rutaceae, [Aurantioideae (distinctive fruit with fleshy hairs in the loculi; fibrous exotegmen; endosperm 0; x = 9) [Chloroxylon + Ruta]] form a well-supported group (most other Rutoideae are immediately unrelated), while subfamilies other than Aurantioideae are interspersed in another group, within which relationships are uncertain. Zanthoxylum may be sister to the rest (see also Muellner et al. 2007); it is certainly part of a well-supported clade with distinctive isoquinoline alkaloids that may be sister to Ptelea. Euodia and relatives form a moderately supported clade, but other than that, relationships are unclear (Poon et al. 2007). Although Aurantioideae seem to be monophyletic (e.g. see also Morton et al. 2003), even there genera need overhauling (Scott et al. 2000; Samuel et al. 2001), as also in Galipeinae (Kallunki & Groppo 2007) and the largely South African Diosmeae (Trinder-Smith et al. 2007).

Cronquist (1981) included Cneorum in Sapindaceae; Airy Shaw (1966) associated Kirkia with Ptaeroxylaceae, but with hesitation: see also Dahlgren and van Wyk (1988), van der Ham et al. (1995) and White and Styles (1966) for information. Hegnauer (1990) included Ptaeroxylum in Meliaceae, although he noted it was chemically more similar to Rutaceae. Harrisonia has also been included in Rutaceae, as by Thorne (1992: no reasons given).

Savolainen et al. (2000b) suggested that Lissocarpaceae should be included here, but a position in Ebenaceae-Ericales is now strongly supported (Berry et al. 2001).

For gynoecial morphology in particular, see Gut (1966), for general information, see van der Ham et al. (1995) and White and Styles (1966), for general chemistry, see Hegnauer (1973, 1990, also 1964, 1989 as Cneoraceae), Waterman and Grundon (1983), and Mulholland et al. (2000, "Ptaeroxylaceae" in particular), for alkaloid chemistry, see Waterman (1975, 1999), for floral development of Cneorum, Caris et al. (2006b), for a review of relationships, Araújo et al. (2003), and for floral orientation, see Eichler (1878).

Synonymy: Amyridaceae Kunth, Aurantiaceae Durande, Boroniaceae J. Agardh, Citraceae Roussel, Cneoraceae Vest, Dictamnaceae Vest, Diosmaceae R. Brown, Diplolaenaceae J. Agardh, Flindersiaceae Airy Shaw (secretory cells in stem only, capsule septifragal [Meliaceae], but furoquinoline alkaloids, schizogenous cavities, filaments not flattened, free [Rutaceae]), Fraxinellaceae Nees & Martius, Jamboliferaceae Martynov, Pilocarpaceae J. Agardh, Ptaeroxylaceae J.-F. Leroy, Pteleaceae Kunth, Spatheliaceae J. Agardh, Zanthoxylaceae Berchtold & J. Presl

MELIACEAE Jussieu, nom. cons.   Back to Sapindales

Trees; bark often rather bitter; secretory cells with resin, etc. +; nodes 5:5; (hairs stellate); (leaves even-pinnate), leaflets opposite, not articulated (alternate; articulated; toothed); plants often dioecious, but flowers apparently perfect, (3-)5(-8)-merous; K often connate, C (-14; connate), A 2x C, connate (5-30 in a single whorl), G (1) [2-6(-many)], archesporium often multicellular, micropyle (bi)endostomal, nucellar cap +, placental obturator common, stigma wet; seeds often pachychalazal, coat vascularised, testa undistinguished but thick, endotesta crystalliferous, exotegmen fibrous [Trichilia, Swietenia] (not); embryo white [green - Trichilia]; n = 10-14.

52[list]/621 - 2 groups below. Pantropical, but largely Old World; plants of the lowlands (Map: see Wickens 1976; Pennington 1981; FloraBase 2006: approximate).

1. Melioideae Arnott

Buds naked; (nodes 3:3); (leaves 2-ranked [Turraea]); 1-2 epitropous (-many) ovules/carpel, stigma capitate; fruit a loculicidal capsule (berry, drupe, nut); seeds unwinged (winged - Quivisianthe), usu. with aril [funicular in Naregamia] or sarcotesta.

37/571. Aglaia (110 - see Muellner et al. 2005), Trichilia (85), Dysoxylum (80), Chisocheton (50), Guarea (50). Pantropical, but largely Old World.

Melia azederach is the neem tree. Guarea (tropical America) and Chisocheton (Malesia) both have indefinitely growing leaves (see Fukuda et al. 2003 for a phylogeny of the latter). In Guarea the apical part of the leaf is shoot-like in its gene expression (Tsukaya 2005), while in species of Chisocheton such as C. pohlianus the inflorescence may be epiphyllous (Fisher & Rutishauser 1990), flowers appearing between the leaflets (specimens have been misidentified as Rubiaceae!). In the latter genus the leaves can be rooted, and then they continue to grow for a long time. However, I do not know of any case where a fertile tree has been produced from a leaf. Sieve tube plastids with protein crystalloids and starch occur in Melia and Azederach. The filaments of Vavaea may be largely free.

Synonymy: Aitoniaceae Reveal & Hoogland

2. Swietenioideae Kostel.

Buds perulate (naked - Capuronianthus); (C connate; A at least partly free; nectary 0), (2 [Capuronianthus]) 3-many ovules/carpel, style-head discoid (capitate); fruit a septifragal capsule, valves falling off, columella persisting and seeds winged, or slight columella and seeds with woody or corky sarcotesta.

15/50: Entandophragma (15). Pantropical, but largely Old World.[Photo - Flower, Fruit.]

The wood of Swietenia is mahogany. The filaments of Cedrela and Toona are free. Capuronianthus has opposite leaves. There may be an earlier subfamilial name for the clade, but Swietenioideae is the name in common use.

Synonymy: Cedrelaceae R. Brown, Swieteniaceae Berchtold & J. Presl

• Meliaceae are recognisable by their usually spiral and odd-compound leaves with well-developed leaf buttresses, i.e. the leaf bases are swollen and more or less elongated vertically. The leaflets are often opposite and dry thin and brittle, with little in the way of tertiary venation evident. The bark may smell - sweet, like garlic, etc. - and milky exudates occur sporadically. The flower buds are often longer than broad, with petals much longer than the sepals, and most members have a staminal tube and a large stigma. The fruits and seeds are usually quite large. [Photo - Flower, Fruit.]

Muellner et al. (2006) discuss the biogeography of the family, suggesting its origin in Africa and subsequent dispersal; the family is ca 85-76 million years old, diversification within it beginning ca 75-67.5 million years before present.

Swietenioideae (Cedreloideae may be an earlier names, but is currently not used) and Melioideae are clearly monophyletic clades (Oon et al. 2000: one gene, Cedreloideae not well supported; Muellner et al. 2003: three genes; Muellner et al. 2006: rbcL alone, sampling better). Although the two subfamilies can be separated chemically based on limonoid types, recent work on Quivisianthe suggests that the distinction may not be that simple (Mulholland et al. 2000). Two Malagasy genera previously segregated as separate subfamilies, Quivisianthe and in particular Capuronianthus, are embedded in Melioideae and Swietenioideae respectively (e.g. Muellner et al. 2003, 2006). Within Melioideae, Melieae (probably including Owenia) are sister to the rest, but with only moderate support; relationships along the backbone of the rest of the rather pectinate ITS tree are poorly supported, but raher better resolved by rbcL data (Muellner at al. 2008).

Walsura often has leaflets with ± pulvinate petiolules and prominent reticulate venation, the stamens are ± free, and the fruit is often 1-seeded. Munronia is ± herbaceous. The simple-leaved Vavaea and Turraea look rather unlike other Meliaceae, except when in flower. Cedreleae are rather different florally from other Meliaceae, but features found there such as more or less free stamens may be derived, not plseiomorphous (cf. Gouvêa et al. 2008).

For chemistry, see Hegnauer (1969, 1990) and Mulholland et al. (2000), for embryology, etc., see Nair (1970), for a generic monograph, see Pennington and Styles (1975), for a monograph of the neotropical members, see Pennington (1981). I am grateful to David Kenfack for useful information.

SIMAROUBACEAE Candolle, nom. cons.   Back to Sapindales

Trees or shrubs; bark very bitter by simaroubilide quassinoids, ellagic acid +; wood often fluorescing; (nodes multilacunar); medullary secretory canals common; sclereids common, oil cells uncommon; (stomata paracytic); leaf ptyxis also supervolute-curved, leaflets opposite, not articulated, (flat surface glands +), margins coarsely toothed to entire, (stipules +, cauline [Picrasma, some Soulamea] or petiolar); flowers rather small, bisexual or not, (3-)4-6(-8)-merous; K connate or free (0), (C 0), A (5, opposite K, 10<), with lateral or basal scales or 0, G 1-5(-8), ± free to connate, 1 epitropous ovule/carpel, micropyle endostomal (bistomal? - Samadera), zig-zag, nucellar cap +, style +/0 (styles +), often ± basal, stigmas ± recurved, ± pointed, with elongated receptive zone, dry; fruit 1-2 seeded drupelet (also samara); seed (pachychalazal), with undistinguished testa (mesotegmen with reticulate thickenings) or scattered lignified cells, endotesta often slightly lignified, tegmen crushed; (hemicellulose reserve); n = 8-13.

19[list]/95. Largely tropical; a few (e.g. Ailanthus) temperate (Map: from Nooteboom 1962; Heywood 1978; Thomas 1990). [Photo - Flower, Fruit] [Photo - Fruit]

• Simaroubaceae can be recognised because the twigs have conspicuous pith - herbarium specimens are often notably light - and the bark is bitter, even when dry; the leaflets of the odd-compound leaves often have rather coarse, blunt, teeth and flat glands on the lower surface, sometimes close to the margin, and the rhachis collapses at the nodes. The flowers are rather small, the petals usually barely exceeding the sepals. The fruits are often separate drupelets or samaras.

Although the carpels may be ± free, there is often only a single style. Even in taxa with unitegmic ovules, the axis of the embryo and the micropyle are offset at sharp angle, hence the latter is zig-zag. There are reports of other than porogamous fertilisation in the family (cf. Anacardioideae: Rao 1970).

Ailanthus is known fossil from ca 52 million years before present (Corbett & Manchester 2004).

Picrasma, Holacantha and Castela form a clade with strong support that is sister to the rest of the family, and Ailanthus is sister to the remainder; much of the phylogenetic structure along the backbone of the family is well supported (Clayton et al. 2007a, esp. b). Leitneria is well embedded in the family.

Simaroubaceae have been very difficult to delimit, and molecular data suggest the excision of Suriana and its relatives (Surianaceae - Fabales), Harrisonia (Rutaceae), and Picramnia and Alvaradoa (Picramniaceae - unplaced Rosid: Fernando et al. 1995). Gumillea (ex Cunoniaceae) should perhaps be included here.

For chemistry, see Hegnauer (1973, 1990, also 1966, 1989, as Leitneriaceae), Abbe (1974: floral morphology/anatomy of Leitneria), and for other information, see Fernando and Quinn (1992: pericarp anatomy), Jadin (1901: vegetative anatomy), Boas (1913: vegetative anatomy), and Webster (1936: wood anatomy).

Synonymy: Ailanthaceae (Arnott) J. Agardh, Castelaceae J. Agardh, Holacanthaceae Engler, Leitneriaceae Bentham and J. D. Hooker (bark bitter!; plant ± dioecious; flowers reduced, C 0?, pollen reticulate, G 1, stigma decurrent; mesotegmen with reticulate thickenings, endosperm starchy; n = 16), Quassiaceae Bertolini, Simabaceae Horaninow, Soulameaceae Endlicher