Embryology and relationships of Lauraceae (Laurales)

Embryology of Lauraceae, hitherto poorly known, was investigated on the basis of 35 species from 23 genera to contribute to a better understanding of familial and generic relationships. Data from the genera investigated and from the literature show that the genera of Lauraceae are very similar embryologically, but that differences do exist in a few characters among the genera. Based on comparisons with other families of Laurales, Lauraceae consistendy had a pachychalazal ovule or seed with a ramified raphal vascular bundle at chalaza (an apomorphy) in common with Hernandiaceae. However, since several core lauralean families such as Amborellaceae, Monimiaceae, and Gomortegaceae are little known embryologically, these must be studied for critical comparison. Wimin Lauraceae, Cassytha is clearly distinct from the rest of the family in having an ab initio cellular type endosperm (a plesiomorphy, also reported in Umbellularia ) instead of a nuclear type endosperm (an apomorphy) as in the rest of the family, in lacking the nucellar cap and in having the micropyle formed by bom the inner and outer integument, facts supporting the traditional taxonomic placement of Cassytha in its own subfamily and the remainder of the family in the other subfamily. Widiin the rest of the family, the amoeboid tapetum (an apomorphy) distinguishes 15 genera ( Actinodaphne, Cinnamomum, Laurus , etc.) from the six genera with the glandular tapetum. In addition, a mature embryo sac protruding from the nucellus (an apomorphy) distinguishes five genera ( Beibchmiedia, Caryodaphnopsis, Cryptocarya, Endiandra, Potamria) and one species of Ocotea (O. rubra ) from the rest of the family. These results can properly be incorporated in a future suprageneric classification. The distinctness of Ocotea rubra wimin the genus is also discussed.     

Developmental morphology of the ovules of Amborella trichopoda (Amborellaceae) and Chloranthus serratus (Chloranthaceae)

The developmental morphology of the outer integument in the pendent orthotropous ovules of Amborella trichopoda (Amborellaceae) and Chloranthus serratus (Chloranthaceae) was studied. In both species the outer integument is semiannular at an early stage and becomes cup-shaped but dorsiventrally somewhat asymmetric at later stages. The outer integument, which is initiated first on the concave and lateral sides of the ovule, differs from that of the anatropous ovules of other basal families with the outer integument semiannular at an early stage or throughout development. The bilateral symmetry of the outer integument is shared by these orthotropous and anatropous ovules. The developmental pattern of the outer integument and ovule incurving characterize the ovule of the Amborellaceae and Chloranthaceae, which is not equivalent to typical orthotropous ovules of eudicots. A phylogenetic analysis of ovule characters in basal angiosperms suggests that anatropous ovules with cup-shaped outer integuments and orthotropous ovules were derived independently in several clades and that the ovules of Amborella and Chloranthus might also be derivative.     

Embryology of Gomortegaceae (Laurales): characteristics and character evolution

The embryological characteristics of Gomortegaceae, which are poorly understood, were investigated on the basis of Gomortega nitida, the only species of the family, to understand better the evolution of this group within Laurales. Comparisons with other Laurales and Magnoliales (a sister group of Laurales) show that Gomortega has many embryological features in common with the other lauralean families. Notably, Gomortega shares a testa without or with at best only a poorly developed mesotesta as a synapomorphy with all other Laurales. The genus further shares anthers dehisced by valves as a synapomorphy with the other Laurales (except for Calycanthaceae and Monimiaceae), and a non-multiplicative testa and bisporangiate anther as synapomorphies with Atherospermataceae and Siparunaceae (although the non-multiplicative testa occurs as a homoplasy in Monimiaceae, and the bisporangiate anther in Monimiaceae pro parte, Lauraceae pro parte and Hernandiaceae, respectively). Gomortega shows simultaneous cytokinesis to form pollen grains, a one-celled ovule archesporium and non-specialized chalaza, all or part of which may be synapomorphies shared with Atherospermataceae. Gomortega appears to have no embryological autapomorphies, but further comparison with Atherospermataceae is required.Kweon Heo and Yukitoshi Kimoto contributed equally to this work.     

Embryology of Hortonioideae and Monimioideae (Monimiaceae,Laurales): characteristics of the ‘lower’ monimioids

We investigated the embryology of the ‘lower’ monimioids, i.e. Monimioideae (Monimia, Palmeria and Peumus) and Hortonioideae (Hortonia), which are poorly described embryologically. Our results show that, contrary to what has been reported in the literature, ‘lower’ monimioids show very little variation in their embryological characters. Comparisons with Mollinedioideae (a large derived subfamily in Monimiaceae) and other families in Laurales show that the ‘lower’ monimioids are relatively consistent in sharing predominantly isobilateral tetrads of microspores and megaspores, a non‐specialized chalaza, and a mesotestal–endotestal seed coat (with tracheoidal cells of the meso‐ and endotesta). It is likely that, while the shared successive cytokinesis during meiosis of microspore mother cells supports the Monimiaceae–Hernandiaceae–Lauraceae clade obtained by molecular evidence, no synapomorphies exist to support a sister‐group relationship of Monimiaceae with Hernandiaceae or Lauraceae. Instead, the lack of hypostase in ovules and/or young seeds, the lack of endosperm in mature seeds and the amoeboid tapetum in the anther are likely synapomorphies of Hernandiaceae and Lauraceae. © 2008 The Linnean Society of London, Botanical Journal of the Linnean Society, 2008, 158 , 228–241.     

Gynoecium diversity and systematics of the Magnoliales and winteroids

Carpel and ovule structure was compared in representatives of all 11 families of the Magnoliales (Annonaceae, Canellaceae, Degeneriaceae, Eupomatiaccae, Himantandraceae, Magnoliaceae, Myristicaceae) and winteroids (Austrobaileyaceae, Illiciaceae, Sehisandraceae, Winteraceae). Special attention was paid to features that are constant at family level. Bisexual flowers are always protogynous. In all representatives studied the carpels are closed at anthesis. Caipel closure is attained in three different ways: (1) postgenital fusion of inner surfaces (Degeneriaceae, Eupomatiaccae. Winteraceae), or (2) occlusion by secretion (Austrobaileyaceae, Sehisandraceae), or (3) a combination of (1) and (2): in Annonaceae, Canellaceae, Myristicaceae there is a conspicuous secretory canal in the innermost part of the ventral slit; in Illiciaceae and Magnoliaceae there is a narrow canal in the innermost part of the ventral slit; and in Himantandraceae the ventral slit is postgenilally fused in the style but completely open in the ovary. In most families the carpels have a double stigmalic crest or they have two tips in the transversal symmetry plane (i.e. at right angles to the median plane). Stigmas are unicellular papillate in most families but the papillae are bi-to multicellular (uniseriate) in Degeneriaceae and Eupomatiaceae. An unusual cryptic exlracarpellary compitum was found in Himantandraceae and Sehisandraceae. Intrusive oil cells were found in the carpel epidermis of Illiciaceae and Sehisandraceae. Mature ovules vary in length between 0.15 and 1.1 mm. The outer integument is fully annular (not semiannular) in Degeneriaceae, Himantandraceae, Canellaceae, Myristicaceae, and Illiciaceae. A rudimentary aril occurs in Canellaceae, and originates at the same site as in arillate Annonaceae and Myristicaceae. The results most strongly support an Annonaceae-Myristicaceae-Canellaceae alliance, to some degree also an Eupomatiaccac-Degeneriaceae-Himantandraceae-Magnoliaceae alliance, and an Illiciaceae-Schisandraceae-Winteraccae-Austrobaileyaceae alliance.     

Chromosome evolution in the Laurales based on analyses of original and published data

We present a summary of currently available chromosome information for all seven families in the order Laurales on the basis of original and previously published data and discuss the evolution of chromosomes in this order. Based on a total of 53 genera for which chromosome data were available, basic chromosome numbers appear consistent within families: x = 11 (Calycanthaceae); x = 22 (Atherospermataceae and Siparunaceae); x = 19 (Monimiaceae); and x = 12 and 15 (Lauraceae). The Hernandiaceae have diverse numbers: x = 15 (Gyrocarpoideae) and x = 18 and 20 (Hernandioideae). Karyotype analyses showed that Hennecartia, Kibaropsis, and Matthaea (all Monimiaceae) contained two or three sets of four distinct chromosomes in 38 somatic chromosomes, suggesting that 2n = 38 was derived by aneuploid reduction from 2n = 40, a tetraploid of x = 10. In light of the overall framework of phylogenetic relationships in the Laurales, we show that x = 11 is an archaic base number in the order and is retained in the Calycanthaceae, which are sister to the remainder of the order. Polyploidization appears to have occurred from x = 11 to x = 22 in a common clade of the Siparunaceae, Atherospermataceae, and Gomortegaceae (although 2n = 42 in the Gomortegaceae), and aneuploid reduction from x = 11 to x = 10 occurred in a common clade of the Hernandiaceae, Lauraceae, and Monimiaceae. To understand chromosome evolution in the Lauraceae, however, more studies are needed of genera and species of Cryptocaryeae.     

Embryology of Siparunaceae (Laurales): characteristics and character evolution

Embryological characters of Siparunaceae, which are poorly understood, were studied on the basis of two constituent genera, an African Glossocalyx and a South American Siparuna, to better understand their evolution within Laurales. These two genera have many embryological characteristics in common with the other lauralean families. Noticeably, they share the multi-celled ovule archesporium (uncertain in Glossocalyx) as a synapomorphy with all the other lauralean families except Lauraceae, the anthers dehisced by valves as a synspomorphy with all the other lauralean families except Calycanthaceae and Monimiaceae, and the bisporangiate anther as a synapomorphy with Gomortegaceae and Atherospermataceae. Siparunaceae are, however, distinct from all other laularean families in having unitegmic ovules that were derived from bitegmic ovules, probably due to an elimination of the outer integument. Likewise, the lack of the testa (i.e., developed outer integument), the "endotegmic" seed coat, and the perichalazal seed at maturity are also characteristics of Siparunaceae. Within the family, Siparuna differs from Glossocalyx in having plural tetrads of megaspores and plural, starchy-rich, one-nucleate, tubular embryo sacs (autapomorphies). On the other hand, Glossocalyx is characterized by having bilaterally flattened seeds (autapomorphy). Although functional aspects of those autapomorphies are uncertain, both Glossocalyx and Siparuna show evolution in different embryological characters.     

Gynoecium diversity and systematics in basal monocots

Gynoecium and ovule structure was comparatively studied in representatives of the basal monocots, including Acorales (Acoraceae), Alismatales (Araceae, Alismataceae, Aponogetonaceae, Butomaceae, Hydrocharitaceae, Junc‐aginaceae, Limnocharitaceae, Potamogetonaceae, Scheuchzeriaceae, Tofieldiaceae), Dioscoreales (Dioscoreaceae, Taccaceae), and Triuridaceae as a family of uncertain position in monocots. In all taxa studied the carpels or gynoecia are closed at anthesis. This closure is attained in different ways: (1) by secretion without postgenital fusion (Araceae, Hydrocharitaceae); (2) by partly postgenitally fused periphery but with a completely unfused canal (Alismataceae, Aponogetonaceae, Butomaceae, Limnocharitaceae, Scheuchzeriaceae, Dioscoreaceae, Taccaceae); (3) by completely postgenitally fused periphery but with an unfused canal in the centre (Acoraceae, Tofieldiaceae); (4) by complete postgenital fusion and without an (unfused) canal (Juncaginaceae, Potamogetonaceae). In many Alismatales (but without Araceae) carpels have two lateral lobes. The stigmatic surface is restricted to the uppermost part of the ventral slit (if the carpel is plicate); it is never distinctly double‐crested (Butomaceae?). Stigmas are commonly unicellular‐papillate and secretory in most taxa. The locules are filled with a (often) mucilaginous secretion in a number of taxa. Superficial (probably intrusive) ethereal oil cells were found in the carpel wall of Acorus gramineus (as in Piperales!). Idioblasts in carpels are otherwise rare. A number of basal monocots has orthotropous ovules, which is perhaps the plesiomorphic condition in the group. The presence of almost tenuinucellar (pseudocrassinucellar) ovules is relatively common (Acoraceae, many Araceae, some Alismatales s.s.), whereas completely tenuinucellar ovules are rare and do not characterize larger groups. However, crassinucellar ovules occur in the largest number of families among the study group (basal Araceae, many Alismatales s.s.) The outer integument is always annular in orthotropous ovules. The inner integument is often lobed and it mostly forms the micropyle, whereas the outer integument is always unlobed. Gynoecium structure supports the isolated position of Acoraceae as sister to all other monocots. However, in an overall view, if compared with all other families, Acoraceae clearly shows the greatest similarities with Araceae.     

Embryology and relationships ofGyrocarpus andHernandia (Hernandiaceae)

Two representative genera of Hernandiaceae,Gyrocarpus andHernandia, were investigated embryologically to contribute to a better understanding of their respective evolutionary position. Comparisons with other lauralean families using Chloranthaceae or Annonaceae (as a representative of Magnoliales) as an outgroup of Laurales (if present, plus other related taxa) support a lauraceous affinity for the two genera because of the presence of ramified raphal vascular bundles at the chalaza (a synapomorphy), but do not provide evidence for the separation of Hernandiaceae from Lauraceae.Hernandia rather shares with Lauraceae two apomorphies (i.e., the seed pachychalazy and the ruminate seed) which may be homoplasies judged from results of cladistic and molecular studies published elsewhere.Hernandia is greatly divergent from an ancestral line common withGyrocarpus and is even diversified within the genus. Based on evidence from embryology as well as from other sources, it seems best to accept two separate subfamilies in Hernandiaceae as usually have been accepted: one is a derived subfamily Hernandioideae, and the other a less specialized Gyrocarpoideae.     

COMPARATIVE MORPHOLOGY AND PHYLOGENY OF THE RANALES

Comparative morphological studies of woody Ranales have established the primitive status of the group and hence their key place in angiosperm phylogeny. Significant advances in our knowledge of some ranalian families have been made in recent years. An attempt is made in the present review to bring together a range of morphological data (vegetative and floral anatomy, palynology and embryology) on the Ranales (sensu lato), with particular reference to research work published after the publication of Eames's (1961) book, and to discuss the relationships of the families. Recent ontogenetic studies have shown that the carpel of Drimys is ascidial and not conduplicate as earlier suggested. The inclusion of Degeneria in the Winteraceae is not supported by morphological data. Melville's gonophyll theory has been shown to be inapplicable to the magnoliaceous flower. The pollen of Schisandra is interpreted as derived and specialized rather than primitive as previously supposed. The removal of Schisandra from Magnoliaceae is upheld by morphological evidence. Recent morphological studies do not support a close relationship between Schisandraceae and Illiciaceae suggested by earlier authors. The Canellaceae shows similarities to Winteraceae, Magnoliaceae, Illiciaceae, Eupteleaceae and Myristicaceae. Transitional types of division of pollen mother cells found in Winteraceae, Schisandraceae and Annonaceae and their probable phylogenetic significance have been discussed. The Annonaceae, Winteraceae, Degeneriaceae, Magnoliaceae, Schisandraceae and Cercidiphyllaceae share several embryological features in addition to similarities in floral structure. Ruminate endosperm is regarded either as an archaic feature retained in some taxa or as a later and parallel development in others. Thus its value in assessing relationships seems to be doubtful. Myristicaceae has been shown to be closely related neither to the the Annonaceae nor to the Lauraceae. The suggested relationship of Eupomatiaceae to Annonaceae is not supported by palynology. Floral cortical vascular systems in Magnoliaceae, Annonaceae, Calycanthaceae and Myristicaceae have been compared and it is concluded that they may be vestigial structures. A great deal of similarity has been found between Lauraceae and Calycanthaceae in wood, node, flower structure and embryology. Further floral anatomical evidence has been adduced to support the removal of Scyphostegia from Monimiaceae. The Hernandiaceae show similarities to some members of Monimiaceae while the Gyrocarpaceae resemble the Lauraceae, Gomortegaceae and certain other genera of Monimiaceae. Available evidence from wood and floral anatomy and embryology indicates close relationships among Lauraceae, Monimiaceae and Hernandiaceae. Vegetative and floral anatomical and embryological data seem to indicate a place for the Chloranthaceae in the ranalian complex. Recent anatomical studies in the Nymphaeaceae show that the floral structure is of a primitive type with similarities to the woody Ranales. Available morphological evidence is considered inadequate to express an opinion on the splitting of the family. Ceratophyllaceae is regarded as a highly reduced ranalian family derived most probably from a nymphaeaceous stock. The gynoecium in Berberidaceae is interpreted as monocarpellate. No evidence has been found to support the tricarpellate view. Berberidaceae, Lardizabalaceae and Menispermaceae share several embryological features, while at the same time showing evidence of specialization, each in its own way. Thus they might have arisen from a common stock and early diverged along different lines. The occurrence of several types of embryo sac in Ranunculaceae may well be an indication of specialization, but their probable taxonomic value, if any, is not yet clear. The occurrence of numerous primitive features in Paeonia has been suggested as an argument for its retention in the Ranales. No evidence has been found to preclude the inclusion of Dilleniaceae in the Ranales. On the other hand, as opposed to similarities in wood and pollen characters between Dilleniaceae and Theaceae, floral anatomical and embryological features offer a sharp contrast between the two. The Ranales are believed to be polyphyletic. It has been tentatively suggested that two major phyletic lines may be recognized in each of the woody and herbaceous series: the magnolialian and lauralian lines in the former and the nymphaealian and berberidalian lines in the latter.     

Floral development of Cardiopteris,with emphasis on gynoecial structure and ovular morphology

Cardiopteris is unique in the expanded Cardiopteridaceae for several distinctive features, including its gynoecial structure and ovular morphology. We studied the floral development of Cardiopteris to clarify floral morphology and document floral development. Cardiopteris has three carpel primordia, which are separate at their tips but congenitally fused at their bases. The synascidiate zone (the fused proximal part) develops into the unilocular ovary; the three discrete carpel apices diverge in development: the apex of the adaxial carpel differentiates into a style and stigma, while the apices of the two lateral-abaxial carpels elongate and develop into a fleshy appendage only after fertilization. The ovules are attached to the lateral-abaxial carpels. At anthesis, the ovules are ategmic and orthotropous without funicles (morphologically undifferentiated). Functional differentiation occurs in the three carpels of Cardiopteris: the adaxial one is the site of pollination, while the lateral-abaxial two produce ovules. The ategmic orthotropous ovule is unusual in Cardiopteridaceae and is an apomorphy of Cardiopteris.     

CLADISTICS OF THE MAGNOLIIDAE

Abstract A cladistic resolution is presented for the origin of the angiosperms based on a parsimony analysis of 49 taxa of Magnoliidae. Hamamelidae and Alismatidae, with gymnospermous outgroup comparisons for the polarization of 104 characters. The Magnoliidae is recognized as a paraphyletic assemblage of nine orders: Calycanthales, Magnoliales, Laurales, Illiciales, Lactoridales. Ranunculales, Aristolochiales, Piperales and Nymphaeales. The Calycanthaceae and Idiospermaceae are segregated as the new order Calycanthales, which is hypothesized to be the archetype for angiosperms. Excluding Winteraceae and Lactoridaceae, the Magnoliales is monophyletic. The Austrobaileyaceae is a first branch of Magnoliales, rather than lauralean. Excluding Amborellaceae and Calycanthales, the Laurales is monophyletic. The Chloranthaceae is a first branch of Laurales, rather than piperalean. The Amborellaceae and Winteraceae are early branches of Illiciales. The Lactoridaceae is isolated as the Lactoridales. Including Papaveraceae, the Ranunculales is monophyletic, with Lardizabalaceae as a first branch. The Ranunculales is more closely related to the Hamamelidae, forming the clade Tricolpates. The Aristolochiales, Piperales and Nymphaeales are successively more closely related to the Alismatidae, forming the clade Paleoherbs. The Nelumbonaceae are nymphaealean Paleoherbs, rather than Tricolpates. The Lactoridaceae is not a Paleoherb. These results support many aspects of the strobilar-flower hypothesis for the origin of the angiosperms, as well as the plesiomorphic character states of woody shrubs with simple, pinnatelyveined leaves.     

Reconstructing the ancestral angiosperm flower and its initial specializations

Increasingly robust understanding of angiosperm phylogeny allows more secure reconstruction of the flower in the most recent common ancestor of extant angiosperms and its early evolution. The surprising emergence of several extant and fossil taxa with simple flowers near the base of the angiosperms-Chloranthaceae, Ceratophyllum, Hydatellaceae, and the Early Cretaceous fossil Archaefructus (the last three are water plants)-has brought a new twist to this problem. We evaluate early floral evolution in angiosperms by parsimony optimization of morphological characters on phylogenetic trees derived from morphological and molecular data. Our analyses imply that Ceratophyllum may be related to Chloranthaceae, and Archaefructus to either Hydatellaceae or Ceratophyllum. Inferred ancestral features include more than two whorls (or series) of tepals and stamens, stamens with protruding adaxial or lateral pollen sacs, several free, ascidiate carpels closed by secretion, extended stigma, extragynoecial compitum, and one or several ventral pendent ovule(s). The ancestral state in other characters is equivocal: e.g., bisexual vs. unisexual flowers, whorled vs. spiral floral phyllotaxis, presence vs. absence of tepal differentiation, anatropous vs. orthotropous ovules. Our results indicate that the simple flowers of the newly recognized basal groups are reduced rather than primitively simple.     

Gynoecium diversity and systematics of the basal eudicots

Gynoecium and ovule structure was compared in representatives of the basal eudicots, including Ranunculales (Berberidaceae, Circaeasteraceae, Eupteleaceae, Lardizabalaceae, Menispermaceae, Papaveraceae, Ranunculaceae), Proteales (Nelumbonaceae, Platanaceae, Proteaceae), some families of the former ‘lower’ hamamelids that have been moved to Saxifragales (Altingiaceae, Cercidiphyllaceae, Daphniphyllaceae, Hamamelidaceae), and some families of uncertain position (Gunneraceae, Myrothamnaceae, Buxaceae, Sabiaceae, Trochodendraceae). In all representatives studied, the carpels (or syncarpous gynoecia) are closed at anthesis. This closure is attained in different ways: (1) by secretion without postgenital fusion (Berberidaceae, Papaveraceae, Nelumbonaceae, probably Circaeaster); (2) by a combination of postgenital fusion and secretion; (2a) with a complete secretory canal and partly postgenitally fused periphery (Lardizabalaceae, Menispermaceae, some Ranunculaceae, Sabiaceae); (2b) with an incomplete secretory canal and completely fused periphery (Tro-chodendron); (3) by complete postgenital fusion without a secretory canal (most Ranunculaceae, Eupteleaceae, Platanaceae, Proteaceae, all families of Saxifragales and incertae sedis studied here). Stigmas are double-crested and decurrent in most of the non-ranunculalian taxa; unicellular-papillate in most taxa, but with multicellular protuberances in Daphniphyllaceae and Hamamelidaceae. Carpels predominantly have three vascular bundles, but five in Proteales (without Nelumbonaceae), Myrothamnaceae and Trochodendraceae. The latter two also share ‘oil’ cells in the carpels. Stomata on the outer carpel surface are present in the majority of Ranunculales and Proteales, but tend to be lacking in the saxifragalian families. In basal eudicots, especially in the non-ranunculalian families there is a trend to form more than one ovule per carpel but to develop only one seed, likewise there is a trend to have immature ovules at anthesis. Ovule number per carpel is predominantly one or two. Proteales (without Nelumbonales) mainly have orthotropous ovules, the other groups have anatropous (or hemitropous or campylotropous) ovules. The outer integument is annular in the groups with orthotropous or hemitropous ovules, and also in a number of saxifragalian families with anatropous ovules. In Proteales the integuments are predominantly lobed but there is no distinct pattern in this feature among the other groups. Among Ranunculales two pairs of families (Lardizabalaceae/Menispermaceae and Bcrberidaceae/Papaveraceae) due to similarities in gynoecium structure can be recognized, which are not apparent in molecular analyses. The close relationship of Platanaceae and Proteaceae is supported by gynoecium structure but gynoecial features do not support their affinity to Nelumbonaceae. The alliance of Daphniphyllaceae with Hamamelidaceae s.l. is also supported.     

A structural investigation of the ovule in sugar beet, Beta vulgaris: Integuments and micropyle

The micropyle and the integuments of sugar beet (Beta vulgaris) ovules have been investigated by light and electron microscopy during differentiation and maturation of the ovule. The micropyle itself is formed by the inner integument which is surrounded by the outer integument at its base. The micropyle containts a fibrillar PAS+ substance and is often covered by a thin sheet or hymen. Both integuments are cuticle-covered thin sheets, each 2-few cell layers in thickness. In the outer integument an increase in starch accumulation occurs during ovule maturation and probably functions as nutrient storage for embryo development. The inner epidermis of the inner integument differentiates as the most conspicuous cell layer of the beet ovule. During growth and maturation of the ovule a system of small perinuclear vacuoles containing dense material increases steadily in these cells. At maturity this system fills up more than half of each cell and very dense material has accumulated in each vacuole. This vacuole content is highly refractive and contains tannins and/or polyphenols.     

The Arabidopsis SUPERMAN Gene Mediates Asymmetric Growth of the Outer Integument of Ovules

Arabidopsis superman (sup, also referred to as floral mutant10) mutants have previously been shown to have flowers with supernumerary stamens and reduced carpels as a result of ectopic expression of the floral homeotic gene APETALA3 (AP3). Here, we report that sup mutations also cause specific alterations in ovule development. Growth of the outer integument of wild-type ovules occurs almost exclusively on the abaxial side of the ovule, resulting in a bilaterally symmetrical hoodlike structure. In contrast, the outer integument of sup mutant ovules grows equally on all sides of the ovule, resulting in a nearly radially symmetrical tubular shape. Thus, one role of SUP is to suppress growth of the outer integument on the adaxial side of the ovule. Genetic analyses showed that the effects of sup mutations on ovule development are independent of the presence or absence of AP3 activity. Thus, SUP acts through different mechanisms in its early role in ensuring proper determination of carpel identity and in its later role in asymmetric suppression of outer integument growth.     

Homeotic Transformation of Ovules into Carpel-like Structures in Arabidopsis

Ovules are specialized reproductive organs that develop within the carpels of higher plants. In Arabidopsis, mutations in two genes, BELL1 (BEL1) and APETALA2 (AP2), disrupt ovule development. In Bel1 ovules, the inner integument fails to form, the outer integument develops abnormally, and the embryo sac arrests at a late stage of megagametogenesis. During later stages of ovule development, cells of the outer integument of a Bel1 ovule sometimes develop into a carpel-like structure with stigmatic papillae and second-order ovules. The frequency of carpel-like structures was highest when plants were grown under conditions that normally induced flowering and was correlated with ectopic expression in the ovule of AGAMOUS (AG), an organ-identity gene required for carpel formation. Together, these results suggested that BEL1 negatively regulates AG late in ovule development. Likewise, mutants homozygous for the strong AP2 allele ap2-6 sometimes displayed structures with carpel-like features in place of ovules. However, such abnormal Ap2 ovules are much less ovulelike in morphology and form earlier than the Bel1 carpel-like structures. Because one role of the AP2 gene is to negatively regulate AG expression early in flower development, it is possible that AP2 works in a similar manner in the ovule. A novel ovule phenotype observed in Bel1/Ap2-6 double mutants suggested that BEL1 and AP2 genes function independently during ovule development.     

Lauraceous flowers from the Late Cretaceous of North Carolina, U.S.A.

Three new taxa with clear affinity to extant Lauraceae are described from the Santonian/ Campanian ( c .83 My, Late Cretaceous) Neuse River locality in North Carolina, U.S.A. A new lauraceous genus, Neusenia , is established to accommodate an excellently preserved flower with tetrasporangiate anthers and psilate pollen grains. Two additional lauraceous taxa are described but not named due to incomplete preservation. The fossil taxa described in this paper represent a variety of evolutionary lineages within Lauraceae with respect to inflorescence structure and anther morphology, including both distinctly pedicellate flowers and sessile, closely crowded flowers, as well as tetrasporangiate and disporangiate anthers. In light of the co-occurrence of both tetrasporangiate and disporangiate anthers in the Neuse River flora, the plesiomorphic state of lauraceous anthers is discussed. Mapped on a recent cladogram of Laurales, tetrasporangiate anthers appear to be primitive within Lauraceae. Thus, disporangiate 2-valvate anthers must have evolved independently at least three times in Laurales (in Lauraceae, Hernandiaceae, and Atherosperrnataceae/Gomortegaceae). In Hernandiaceae and Atherosperrnataceae/Gomortegaceae such anthers are interpreted to have originated from tetrasporangiate 2-valvate anthers through reduction of the septum in each theca, while in Lauraceae they may have originated in the same way and/or from reduction of two pollen sacs in a tetrasporangiate 4-valvate anther.