Am. J. Bot. Subscribe to E-TOCs
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sauquet, H.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Sauquet, H.
Agricola
Right arrow Articles by Sauquet, H.
(American Journal of Botany. 2003;90:1293-1305.)
© 2003 Botanical Society of America, Inc.

Structure and Development

Androecium diversity and evolution in Myristicaceae (Magnoliales), with a description of a new Malagasy genus, Doyleanthus gen. nov1

Hervé Sauquet2

Laboratoire de Biologie et Évolution des Plantes vasculaires EPHE, Muséum national d'Histoire naturelle, 16 rue Buffon, 75005 Paris, France

Received for publication October 1, 2002. Accepted for publication April 24, 2003.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
Myristicaceae (Magnoliales) consist of 20 genera and nearly 500 species of lowland rainforest trees with a pantropical distribution. They are distinctive in having small, unisexual flowers with stamens fused into a synandrium, which consists of a single whorl of sessile anthers borne around a sterile central column. With its short filaments and more complex anther phyllotaxy, the Malagasy genus Mauloutchia represents a notable exception to this pattern. New scanning electron microscope (SEM) examinaitons of Brochoneura, Cephalosphaera, Knema, Mauloutchia, and Staudtia are incorporated into a broader review of androecium diversity across the family. These new results are discussed in the context of a phylogenetic study of the family, based on combined molecular and morphological data. The unusual synandrium of Mauloutchia, nested among genera with strictly sessile anthers fused to the column, appears to be secondarily derived. Furthermore, the diversity of patterns observed within the genus may be interpreted as the result of a stepwise transformation involving reappearance and elongation of filaments, increase of anther number, and modification of anther phyllotaxy. However, the question of the origin of stamen fusion in Myristicaceae remains unanswered and requires more developmental studies. Finally, a new Malagasy genus of Myristicaceae (Doyleanthus) is described, which is similar to Mauloutchia in most characters but fundamentally different in androecial traits.

Key Words: androecium evolution • Brochoneura • Cephalosphaera • Doyleanthus • Magnoliales • Mauloutchia • Myristicaceae • stamen evolution • Staudtia


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
With 20 genera and nearly 500 species, Myristicaceae are a medium-sized family of angiosperm trees with a wide pantropical distribution, mostly confined to lowland rainforest habitats. Along with Annonaceae, Magnoliaceae, and three monogeneric families (Degeneriaceae, Eupomatiaceae, and Himantandraceae), they belong to Magnoliales (sensu APG, 1998 ). Phylogenetic relationships within Magnoliales and Myristicaceae were recently investigated by Sauquet et al. (2003) , based on a combined analysis of a morphological matrix and multiple molecular data sets, the results of which are summarized in Fig. 1. Whereas unambiguously supporting Myristicaceae as the sister group of all remaining Magnoliales (referred to as suborder Magnoliineae) and resolving relationships within this clade, this study was less successful in reconstructing relationships among genera of Myristicaceae, mainly because of a very low level of molecular variation within the family. It did, however, support a monophyletic assemblage of four genera, two from mainland Africa (Cephalosphaera and Staudtia) and two from Madagascar (Brochoneura and Mauloutchia). Whereas morphological data suggested that this clade (referred to as mauloutchioids) is sister to all remaining Myristicaceae, molecular analyses placed it in a less basal position, with a clade of Otoba (a Neotropical genus) and [Coelocaryon + Pycnanthus] (two other mainland African genera) as its putative sister group (Fig. 1).



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 1. Phylogenetic relationships among Magnoliales and Myristicaceae, based on combined morphological and molecular data (Sauquet et al., 2003 ). Relationships within Myristicaceae are based on the majority-rule consensus tree obtained from the total combined analysis, rerooted according to maximum likelihood analyses of molecular data with reduced taxon sampling. Thicker branches were supported with a minimum decay index of 2 and bootstrap value of 65%. Several taxa were sampled for Laurales, Annonaceae, Brochoneura, and Mauloutchia, but these regions of the tree are not developed here (see Fig. 54 for internal topologies of Brochoneura and Mauloutchia).

 
In contrast to Magnoliineae, flowers of Myristicaceae are inconspicuous, usually very small, and always unisexual, with a single perianth cycle of three tepals (on average) that are fused to various degrees. All female flowers are uni-carpellate and uni-ovulate, and their ovary exhibits very little diversity across the family. Conversely, androecial features have played an essential role in the definition of genera and represent key characters for their identification (Sinclair, 1958 ). The androecium architecture of Myristicaceae also represents one of their most distinctive and enigmatic features. In all male flowers of the family, stamens are fused into a synandrium, which consists of a sterile column and a collection of 2–60 anthers fused to various degrees to this column (e.g., Fig. 4). It should be noted that the term synandrium has been used in two different ways in the literature. Whereas some authors (e.g., Endress, 1994a ) have used it as a synonym of the whole androecium when stamens are fused, de Wilde (2000) restricted it to the upper part of the androecium with fused anthers and used the term androphore for the sterile common stalk of the fused stamens. It is the former definition that I will adopt in this paper. In most genera, the only distinguishable unit is a bisporangiate theca, which led to some confusion in the literature regarding anther number, as anthers were sometimes considered to consist of one theca and sometimes a pair of thecae (as in most angiosperms). However, various anatomical studies (see Armstrong and Tucker, 1986 , p. 1132) clarified this point by favoring the latter interpretation.



View larger version (196K):
[in this window]
[in a new window]
  		Figs. 2–11. SEM micrographs of Brochoneura and Cephalosphaera. 2–4. B. acuminata (2, H. Sauquet 19; 3–4, H. Sauquet 24). 2. Part of inflorescence, showing sessile flowers. 3. Open male flower (with one tepal removed). 4. Synandrium. 5–8. B. vouri (H. Sauquet 15). 5. Part of inflorescence, showing sessile flowers. 6. Synandrium in flower bud (with upper part of perianth removed). 7. Staminal column with one anther (two other anthers removed). 8. Sessile anther, removed from the column. 9–11. C. usambarensis (H.J. Schlieben 2861). 9. Top view of synandrium. 10. Two sessile anthers, removed from the column. 11. Synandrium. Figs. 2–4, 6–11 , scale bar = 100 µm; Fig. 5 , scale bar = 1 mm.

 
Furthermore, the anthers are usually strictly sessile and arranged in a single cycle (borne as a crown around the upper part of the column), except in the genus Mauloutchia (endemic to Madagascar), in which the anthers are attached to the column by short filaments and seem to be arranged in a more complex, nonwhorled phyllotaxy (e.g., Fig. 26). Based on these and other characters, it was hypothesized that Mauloutchia was the most primitive genus in the family and was basal to all the remaining genera (Warburg, 1897 ; Walker and Walker, 1981 ), which explained the presence of filaments in this genus and their absence in all other Myristicaceae. It also made sense to interpret the sterile column bearing the anthers as consisting of completely fused filaments. However, preliminary results from the phylogenetic study of Sauquet et al. (2003) led me to reconsider the basal status of Mauloutchia in Myristicaceae and to look more closely at androecium structures in this genus and the putatively related genera Brochoneura, Cephalosphaera, and Staudtia.



View larger version (197K):
[in this window]
[in a new window]
  		Figs. 22–37. SEM micrographs of Mauloutchia. 22–29. M. chapelieri (22–23, 26–29, H. Sauquet 30; 24–25, R. Capuron 24900 SF). 22. Male flower at anthesis. 23. Detail of the staminal column (see Fig. 26 ), showing the insertion of anthers. 24. Top view of synandrium. 25. Synandrium. 26. Synandrium in flower bud (with one tepal and several anthers removed). 27. Detail of the staminal column (see Fig. 26 ), showing insertion of the anthers. 28. Anther with long filament, removed from the column. 29. Two anthers sharing a common filament. 30–31. M. coriacea (R. Capuron 11789 SF). 30. Synandrium in flower bud (with one tepal removed). 31. Anther with short filament, removed from the column. 32–34. M. heckelii (H. Sauquet 6). 32. Part of inflorescence, showing sessile flowers. 33. Top view of synandrium. 34. Synandrium. 35–37. M. parvifolia (10159 SF). 35. Open male flower (with one tepal removed). 36. Synandrium. 37. Anther with short filament, removed from the column. Figs. 22, 26, 30, 32, 35 , scale bar = 1 mm; Figs. 23, 27–29, 31, 33–34, 36–37 , scale bar = 100 µm; Figs. 24–25 , scale bar = 500 µm

 
Given the very small size of Myristicaceae flowers, scanning electron microscopy (SEM) represents an important tool for revealing clues essential to better understanding the origin and evolution of synandria in Myristicaceae. Until now, very few authors have looked at Myristicaceae synandria using SEM, and such studies focused on either Myristica (Armstrong and Tucker, 1986 ; Endress, 1994b , p. 37) or Compsoneura (J. Janovec, New York Botanical Garden, personal communication). In addition, a few anatomical studies of Knema, Horsfieldia, and Myristica provided insights into the synandrium structure of these genera (Wilson and Maculans, 1967 ; Siddiqi and Wilson, 1976 ; Armstrong and Wilson, 1978 ; Manilal, 1983 ; see also other anatomical work cited in Armstrong and Tucker [1986] ). Despite a great effort by de Wilde (2000) to provide an accurate description of the various types of synandria found in Asian Myristicaceae, only Warburg (1897) , Smith (1937) , and Sinclair (1958) have discussed the paths that the evolution of this structure may have taken in the family. Warburg hypothesized that anther fusion evolved several times in the family. Smith considered it impossible to polarize this character but stated that similar evolutionary trends occurred independently in several genera. Sinclair instead concluded that anther fusion was basic in Myristicaceae and that free anthers were secondarily derived.

In this paper, new SEM observations of the synandria of mauloutchioid genera are provided and their implications for the evolution of androecium characters in Myristicaceae are discussed in the context of the cladistic scoring of these characters published in Sauquet et al. (2003) and the phylogenetic relationships found in that study. In addition, a new Malagasy genus of Myristicaceae (Doyleanthus) is described, based on an unusual combination of morphological characters that unambiguously separate this new species, previously assigned to Mauloutchia, from the remaining species of this genus.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 LITERATURE CITED
 
Pickled flowers were prepared for scanning electron microscopy by dissection and dehydration with 70% ethanol to 100% acetone using successive acetone baths (5 min at 80%, then 5 min at 90%, then three times for 1 min at 100%). The specimens were then critical-point dried, mounted on stubs, and sputter-coated with gold. Dry herbarium flowers were first boiled in water for rehydration, then dissected and dehydrated to 100% acetone using successive acetone baths (5 min at 50%, then 5 min at 80%, then 5 min at 90%, then 1 min at 100%). All SEM micrographs were taken at the Centre Interuniversitaire de Microscopie Électronique (C.I.M.E.), Université Pierre et Marie Curie, Paris, France, except for a few early ones, which were taken by Peter Endress at the Institute of Plant Biology, University of Zurich, Zurich, Switzerland.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 LITERATURE CITED
 
Summary of new observations
Descriptions of androecial SEM observations for each genus included in this study (Figs. 2–52) and detailed information on the observed voucher materials are provided as Supplemental Data accompanying the online version of this article (Appendix A). To summarize briefly these new observations, the androecia of the mauloutchioid Malagasy genera Brochoneura and Doyleanthus and African genera Cephalosphaera and Staudtia appear to be remarkably similar and consist in a cylindrical column of sterile tissue and a collection of 3–4 strictly sessile anthers, borne around the upper part of this column (cf. Figs. 4, 11, 16, and 51). On the other hand, the androecium of the fifth mauloutchioid genus, Mauloutchia (Malagasy), fundamentally differs from the other genera studied in its nonmonocyclic anther phyllotaxy, which may be possibly irregular (e.g., Figs. 22 and 25) or spiral (e.g., Figs. 36 and 42), as well as in the presence of short filaments in most species (Figs. 28, 31, 37, and 40). In addition, the examination of all six described species and two new species has revealed an unexpected intrageneric variation on androecial characters within Mauloutchia. Thus, the species of Mauloutchia examined in this study may be grouped into four categories, based on their androecium characters, as summarized by the schematic drawings in Fig. 53. (1) M. chapelieri: numerous, small anthers inserted at various levels on an elongate column; all anthers borne on filament-like structures, either individually or in pairs (cf. Figs. 23, 26, and 28–29). (2) M. heckelii, M. parvifolia, and M. sp. nov. 2: similar to (1), but with fewer anthers and shorter filaments (cf. Figs. 34, 36–37, and 49). (3) M. coriacea and M. rarabe: relative to (1) and (2), anthers are longer; anthers inserted near the base of the column longer than those inserted above, so that all apices seem to reach the top of the androecium (cf. Figs. 30–31, 38, and 40). (4) M. humblotii and M. sp. nov. 1: similar to (3), but with a shorter column and almost or possibly entirely sessile anthers (cf. Figs. 41 and 43–47). Furthermore, the occasional pairing of anthers observed in M. chapelieri (Fig. 29) casts doubt on the very nature of filament-like structures in this genus. They may either be secondarily diverging expansions of the column or actual filaments, with some stamens fused in pairs. Finally, one species of the Asian genus Knema was also examined under SEM (Figs. 18–21), illustrating the unique architecture that characterizes the androecium of all species in this genus. Often described as a peltate disc, this synandrium typically has the shape of an inverted cone, more or less triangular-shaped at the apex and bearing a crown of 3–25 small anthers (Figs. 18–19).



View larger version (197K):
[in this window]
[in a new window]
  		Figs. 12–21. SEM micrographs of Doyleanthus and Knema. 12–17. D. arillata (12–14, R. Capuron 27198 SF; 15–17, R. Capuron 18904 SF). 12. Synandrium. 13. Top view of synandrium. 14. Synandrium in flower bud (with upper part of perianth removed). 15. Open male flower (with one tepal removed). 16. Synandrium. 17. Three sessile anthers, removed from the column. 18–21. Knema sp. (J. Munzinger & F. Engelmann 236). 18. Synandrium in flower bud (with one tepal removed). 19. Synandrium in flower bud (with upper part of perianth removed). 20. Fragment of synandrium, with two sessile anthers and the insertion area of a third anther. 21. Overview of male flower bud (with one tepal removed), showing the pedicel and the bracteole. Figs. 12–13 , scale bar = 500 µm; Figs. 14–20 , scale bar = 100 µm; Fig. 21 , scale bar = 1 mm.

 


View larger version (191K):
[in this window]
[in a new window]
  		Figs. 38–52. SEM micrographs of Mauloutchia and Staudtia. 38–40. M. rarabe (R. Capuron 28149 SF). 38. Synandrium. 39. Top view of synandrium. 40. Male flower at anthesis, with a few anthers removed. 41–44. M. humblotii (41–42, H. Sauquet 11; 43–44, H. Sauquet 5). 41. Open male flower (with all tepals and a few anthers removed). 42. Top-oblique view of synandrium. 43. Staminal column with two anthers (all tepals and all remaining anthers removed). 44. Almost sessile, basifixed anther, removed from the column. 45–48. M. sp. nov. 1 (R. Capuron 23961 SF). 45. Open male flower (with all tepals removed). 46. Synandrium in flower bud (with two tepals and a few anthers removed). 47. Almost sessile, basifixed anther, removed from the column. 48. Open male flower (with one tepal and a few anthers removed). 49–50. M. sp. nov. 2 (R. Capuron 28666 SF). 49. Open male flower (with two tepals removed). 50. Synandrium in flower bud (with one tepal removed). 51–52. S. kamerunensis (A. Hladik 1842). 51. Synandrium. 52. Upper part of synandrium, showing the sessile anthers. Figs. 38–40 , scale bar = 500 µm; Figs. 41–47, 49–52 , scale bar = 100 µm; Fig. 48 , scale bar = 1 mm

 


View larger version (39K):
[in this window]
[in a new window]
  		Fig. 53. Schematic drawings illustrating the diversity of synandrium structures found in Myristicaceae. Anther numbers are given in brackets for each genus or species. General distribution information and species numbers are also provided

 
Summary of Myristicaceae
Based on these SEM observations, additional personal observations of many herbarium specimens sampled from all genera in the family, and literature descriptions and drawings (including Warburg, 1897 ; Engler and Prantl, 1959 ; Capuron, 1972 , 1973 ; and de Wilde, 2000 ), all potentially informative androecial variations found across the family and within Mauloutchia were scored as cladistic characters included in the morphological data matrix of Sauquet et al. (2003 ; characters 49–65; see Appendix B online). In addition, schematic drawings were made for all genera of Myristicaceae to summarize the diversity of androecia in the family (Fig. 53). However, particular caution is needed in consulting such drawings for the largest genera, such as Knema, Horsfieldia, and Myristica, as they do not account for intrageneric variation, especially regarding the shape of the column (see, for instance, the comparative semi-schematic drawings for over 90 species of Horsfieldia in de Wilde [2000] p. 59–61).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
LITERATURE CITED
 
The discussion of the evolution of each androecial and various general floral characters, based on their a posteriori parsimony optimization on the phylogenetic hypothesis of Sauquet et al. (2003) , is provided as Supplemental Material accompanying the online version of this article (Appendix B). Figure 54 illustrates the evolution of one of these characters, the number of anthers in a synandrium. Despite the uncertainty remaining in the relationships among genera of Myristicaceae, it may be concluded that the ancestral flower of the family only had 2–12 anthers and that several independent increases occurred within the family, on the one hand in each of the three large Asian genera, Horsfieldia, Knema, and Myristica, and on the other hand within the genus Mauloutchia, where a gradual increase up to 60 anthers in M. chapelieri is conspicuous.



View larger version (36K):
[in this window]
[in a new window]
  		Fig. 54. Parsimony optimization of anther number in Myristicaceae, based on relationships found in a combined morphological and molecular phylogenetic analysis of Myristicaceae (Sauquet et al., 2003 )

 
Alternative hypotheses for the origin of stamen fusion in Myristicaceae
The most common view about the origin of the peculiar synandrium of Myristicaceae is that the central column is simply the result of complete fusion of the filaments (and connectives, depending on the genera) (e.g., Kühn and Kubitzki, 1993 , p. 461; de Wilde, 2000 , p. 14). The conditions observed in Mauloutchia may then be interpreted as simply the partial fusion of filaments, with free portions of filaments emerging from the fused portions that make up the column. However, congruent evidence from molecular data and other morphological characters (Sauquet et al., 2003 ) placing Mauloutchia in a derived phylogenetic position, well nested among genera with completely fused stamens, calls into question this simple interpretation by implying that completely fused filaments secondarily became partially free in this genus.

Another less common interpretation of the column is that filaments were indeed completely lost and that the common receptacle bearing the stamens elongated to give the central column of sterile tissue observed today (e.g., Armstrong and Tucker, 1986 ). But this explanation is also challenged by the derived position of Mauloutchia, which implies that after being lost, filaments reappeared in this genus.

In either case, it is possible that stamen fusion in Myristicaceae is related to their apomorphic unisexuality. It may be speculated either that a shift to unisexuality resulted in stamen fusion or overgrowth of the area of the apex where the gynoecium would have developed, or alternatively that a trend in filament fusion began first and eventually led to separation of the sexes, after which sterile tissue developed without constraints until the apical area was filled.

It may be that it was a combination of such processes that led to the androecium observed in extant Myristicaceae. Once Mauloutchia has been ruled out as a relict taxon branching off the lineage that led to other extant Myristicaceae, there is no intermediate androecium that might help us understand better what happened. In this respect, it is particularly regrettable that the fossil record of the family is so scarce and recent. The only fossil flower reported so far was found in the Dominican Tertiary amber and assigned to modern Virola (Poinar and Poinar, 1999 ).

Armstrong and Tucker's (1986) developmental study of two species of Myristica provides a few insights into the origin of the synandrium in Myristicaceae. First, the authors found that the androecium of these species was trimerous in initiation, with the first three stamen primordia arranged in an acyclic or helical pattern. Next, they emphasized that the column develops and extends as the anthers elongate and that neither the lower portion (stalk) nor the apical portion of the column is derived from the stamen primordia. They concluded that the column cannot be composed of fused filaments and instead results from the elongation of the receptacle.

A third, unexplored interpretation of the central sterile column is that it was derived from the complete fusion of inner staminodes, assuming that inner staminodes were present in the ancestor of Magnoliales and represent a synapomorphy of Laurales plus Magnoliales, as discussed in Appendix B online. However, the column tissue would then be expected to develop from stamen primordia, contrary to Armstrong and Tucker's (1986) observations.

A scenario for androecium evolution in Mauloutchia
Although the phylogenetic tree obtained for the species of Mauloutchia requires much further testing, as its resolution is a function of a small number of morphological characters that show variation at this level, it supports four unambiguous androecial transformations (Fig. 55), suggesting a scenario to account for the patterns observed in the genus.



View larger version (24K):
[in this window]
[in a new window]
  		Fig. 55. Parsimony optimization of four androecial characters in the mauloutchioid clade, based on relationships found in a combined morphological and molecular phylogenetic analysis of Myristicaceae (Sauquet et al., 2003 ). Thick bars on the cladogram represent unambiguous transformations. Numbers refer to original character numbers in the morphological data set of Sauquet et al. (2003)

 
This scenario starts with a typical synandrium of Myristicaceae, most likely with 3–4 completely fused stamens, as found in the other mauloutchioid genera. In some common ancestor of all species of Mauloutchia, the connectives of the anthers became free from the central column and the anthers therefore became basifixed, but with almost no filament, as observed in the basal species M. humblotii and M. sp. nov. 1. Possibly correlated with this shift to free anthers (but not necessarily), the simple androecium phyllotaxy began to become more complex, perhaps starting with the addition of a second whorl inside the original one or possibly with the addition of new stamens in a helical pattern, both suggested by Figs. 41–42 of M. humblotii. The latter interpretation is supported by Armstrong and Tucker's (1986) examinations of the earliest developmental stages of Myristica, although they were not able to determine whether the addition of new stamens followed an irregular or a spiral order. Whichever is true, the developmental androecium phyllotaxy was obscured in later stages, leading to the single whorl of completely fused anthers observed in the mature flowers. But it could be speculated that in Mauloutchia, where anther separation reaches a maximum in the family, the observed patterns of stamen arrangement represent true reversals to the original stamen phyllotaxy in ancestral stem-lineage Myristicaceae, before stamen fusion evolved.

Short filaments then began to develop. Later evolution of the synandrium in Mauloutchia probably involved elongation of both the column and the incipient filaments, as well as reduction of anther length, so that the anthers were inserted at different levels and their apices reached different levels. Along with this trend, a gradual increase in anther number soon obscured the original phyllotaxy of the genus, making it seem more irregular.

Finally, this transformation of the synandrium in Mauloutchia reached its ultimate expression in the species M. chapelieri, which most authors had taken as representative of the genus but instead turns out to be the end of an evolutionary trend. This most parsimonious scenario (based on current data) therefore completely reverses previous assumptions on the polarity of androecium evolution in Myristicaceae. However, it does not answer the question of the origin of stamen fusion in the family, as discussed earlier.

Doyleanthus arillata gen. et sp. nov
During the 1950s and 1960s, René Capuron, who was responsible for outstanding work on the tree flora of Madagascar, set apart a few specimens of Myristicaceae collected in Madagascar and ascribed them to a new species of Mauloutchia, which he provisionally named M. arillata on herbarium sheets (in P and TEF), because it has a fully developed aril (an exception in Mauloutchia, which usually has a rudimentary aril). However, because of his premature death in 1971, he never formally described this new species. Although this species resembles Mauloutchia in most characters, the present SEM study revealed that it differs fundamentally from all remaining species of the genus in its androecial characters (Figs. 12–17): the synandrium consists of only a single whorl of 3–4 strictly sessile anthers, almost entirely fused to a central sterile column, as observed in Myristicaceae outside Mauloutchia. In the morphological analysis of Sauquet et al. (2003) , this new species appeared to be sister to a clade consisting of the remaining species of Mauloutchia. However, when molecular data were combined with morphology, its position shifted to a deeper branch because Brochoneura became the sister group of Mauloutchia (Fig. 55). In some trees, it was linked with Cephalosphaera (Tanzania), supported by a deeply laciniate aril and a very similar synandrium, as well as by its unusual distribution in northern Madagascar (the remaining Malagasy genera are mainly distributed along the east coast, with only a few species reaching northeast Madagascar), suggesting a disjunction of the two taxa across the Mozambique Channel. Although no molecular data are currently available for this new species, and its exact placement remains unresolved, these results suggest it may not be the basal branch in Mauloutchia. In order to preserve the monophyly of Mauloutchia, I am therefore assigning this new species to a new genus, Doyleanthus. This name is dedicated to Prof. James A. Doyle (University of California, Davis), who played an essential role in my study of the morphology and phylogenetics of Myristicaceae.

Doyleanthus Sauquet gen. nov
Genus novum, Mauloutchiae Warb. affine a quo antheris sessilibus in unicum verticillum dispositis, atque dorso columnae staminali connatis differt. A Brochoneura Warb. inflorescentiis laxis, floribus pedicellatis, semine arillo haud rudimentali munito et pollinis granorum tecto granuloso haud laevi distinctum. A Haematodendro Capuron foliorum apice acuto vel acuminato, floribus monoeciis, semine arillo munito et pollinis granorum exinio sine columella, tecto continuo granuloso differt.

Typus generis: D. arillata Capuron ex Sauquet

Doyleanthus arillata Capuron ex Sauquet sp. nov. (Figs. 12–17 and 56)
Typus: R. Capuron 18904 SF, Madagascar, Sambirano, South of Ambanja, between Ankaramy and Ankingameloka, 7 XI 1958 (holo-, P!, male flowers and young fruits; iso-, TEF!, flowers).



View larger version (72K):
[in this window]
[in a new window]
  		Fig. 56. Doyleanthus arillata Capuron ex Sauquet gen. et sp. nov. (A, D–G, R. Capuron 18904 SF, type specimen; B–C, 5671 SF; H, R. Capuron 27198 SF, pickled sample). A. Flowering branch. B. Dry fruit. C. Laciniate aril around deflated seed. D. Young fruit. E. Pollen grain. F. Young inflorescence (each small spheroidal part is an immature cluster of flowers). G. Part of inflorescence with female flowers. H. Part of inflorescence with male flowers (see also Figs. 12–17 ). All drawings are by Agathe Berthelot

 
Tree 10–12 m tall. Twigs 1–3 mm in diameter, grayish, glabrous, spotted with numerous, minute lenticels. Leaves distichous, almost entirely glabrous. Petiole 7–10 mm long, canaliculated on the adaxial side. Terminal leaf bud 10 mm long, conduplicate, glabrescent. Leaf blade elliptic-oblong, 6–11.5 cm x 2.5–3.5 cm, base attenuate (to almost rounded), apex acuminate (to acute). Midrib canaliculated to flat above, raised below. Secondary veins cladodromous to festooned brochidodromous, at an angle of 70–80° with the midrib, with first arches forming almost halfway to the leaf margin. Tertiary veins reticulate, faint but visible, at least on the adaxial face. Inflorescences axillary, paniculate, of de Wilde's (1991) plural type, with numerous unclustered pedicellate flowers, softly pubescent axes, and persistent bracts. Flowers unisexual monoecious, with different sexes on the same twig but not in the same inflorescence. Male flowers with a 1–2 mm long ebracteolate pedicel and a pubescent 1–1.5 mm long perianth of three tepals, entirely open and spread at anthesis. Stamens completely fused into a synandrium consisting of 3–4 strictly sessile, yellow-orange anthers with connectives fused to a sterile, cylindrical column with a short stalk. Length of anthers about two-thirds of the total synandrium length. Pollen grains in monads, subglobose, monosulcate, aperture ulceroid with a sculptured membrane, tectum continuous, granulate-echinulate. Female flowers pedicellate and similar to male flowers in size, shape, and perianth. Ovary ovoid-conical with a dark stigma. Dry young fruit glabrous, obovoid-oblong, mucronate at apex. Dry mature fruit 3 x 2.5 cm, indehiscent, glabrous, subglobose, attenuate at base, obtuse at apex. Dry pericarp 3–5 mm thick, strongly lignified. Aril fully developed, deeply laciniate, dark red, with 2–3 mm wide laciniations reaching the apex of the seed.

Other specimens examined: R. Capuron 27198 SF, Madagascar, Sambava, Antsirabe Ava., Andrangana, 2–7 XII 1966 (P!, pickled flowers only; TEF!, flowers and pickled flowers); 5671 SF, Madagascar, Antsiranana, Montagne des Français, 11 IX 1952 (P!, fruits; TEF!, fruits).

Conclusions
This study demonstrates the lability of androecium evolution in Myristicaceae, as previously emphasized by Warburg (1897) , Smith (1937) , and Sinclair (1958) . A number of androecial characters thus appear to be homoplastic, including anther number, connective fusion, and shape of the sterile column. Contrary to the views of Warburg (1897) and Walker and Walker (1981) , but in agreement with those expressed by Sinclair (1958) , the present results imply that the most recent common ancestor of Myristicaceae had a synandrium with anthers entirely fused by their connectives and that androecia with free anthers evolved independently in a few genera. In addition, this study reveals a diversity of synandria in the Malagasy genus Mauloutchia, allowing us to understand better how secondarily derived filaments and a spiral or irregular stamen phyllotaxy may have evolved in this taxon, which is nested among Myristicaceae that have strictly sessile anthers arranged into a single cycle. However, little can be said about the origin of stamen fusion in the family, because Mauloutchia, the only genus with potentially intermediate androecial features, is unambiguously apomorphic in these characters. Additional histological and developmental studies across Myristicaceae or new fossil data may eventually shed light on the basic transformations in floral morphology that preceded the diversification of the extant family.


   FOOTNOTES
 
1 The author thanks Thierry Deroin, James A. Doyle, Peter K. Endress, Annick Le Thomas, and two anonymous reviewers for critical comments on earlier drafts of this manuscript, Annick Le Thomas for assistance with the Latin diagnosis, and Agathe Berthelot for art work. Back

2 hsauquet{at}mnhn.fr Back


   LITERATURE CITED
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 LITERATURE CITED
 
APG (The Angiosperm Phylogeny Group). 1998 An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85: 531-553[CrossRef][ISI]

Armstrong J. E. S. C. Tucker 1986 Floral development in Myristica (Myristicaceae). American Journal of Botany 73: 1131-1141[CrossRef][ISI]

Armstrong J. E. T. K. Wilson 1978 Floral morphology of Horsfieldia (Myristicaceae). American Journal of Botany 65: 441-449[CrossRef][ISI]

Capuron R. 1972 Contribution à l'étude de la flore forestière de Madagascar. A. Haematodendron, genre nouveau de Myristicaceae. Adansonia, série 2 12: 375-379

Capuron R. 1973 Observations sur les Myristicacées de Madagascar: les genres Brochoneura Warb. et Mauloutchia Warb. Adansonia, série 2 13: 203-221

de Wilde W. J. J. O. 1991 The genera of Myristicaceae as distinguished by their inflorescences, and the description of a new genus, Bicuiba. Beiträge zur Biologie der Pflanzen 66: 95-125

de Wilde W. J. J. O. 2000 Myristicaceae. Flora Malesiana, Series I 14: 1-632

Endress P. K. 1994a Shapes, sizes and evolutionary trends in stamens of Magnoliidae. Botanische Jahrbücher für Systematik 115: 429-460

Endress P. K. 1994b Diversity and evolutionary biology of tropical flowers. Cambridge University Press, Cambridge, UK

Engler A. K. Prantl 1959 Die Natürlichen Pflanzenfamilie, vol. 17(a)II. Duncker & Humblot, Berlin, Germany

Kühn U. K. Kubitzki 1993 Myristicaceae. In K. Kubitzki, J. G. Rohwer, and V. Bittrich [eds.], The families and genera of vascular plants, vol. II, 457–467, Springer-Verlag, Berlin, Germany

Manilal K. S. 1983 Morphology and anatomy of Myristica flowers. Bulletin of Pure and Applied Sciences 2B: 7-13

Poinar G. Jr. R. Poinar 1999 The amber forest: a reconstruction of a vanished world. Princeton University Press, Princeton, New Jersey, USA

Sauquet H. J. A. Doyle T. Scharaschkin T. Borsch K. W. Hilu L. W. Chatrou A. Le Thomas 2003 Phylogenetic analysis of Magnoliales and Myristicaceae based on multiple data sets: implications for character evolution. Botanical Journal of the Linnean Society 142: 125-186[CrossRef]

Siddiqi M. R. T. K. Wilson 1976 Floral anatomy of the genus Knema (Myristicaceae). Biologia, Lahore 22: 127-141

Sinclair J. 1958 A revision of the Malayan Myristicaceae. Gardens' Bulletin, Singapore 16: 205-472

Smith A. C. 1937 The American species of Myristicaceae. Brittonia 2: 393-510[CrossRef]

Walker J. W. A. G. Walker 1981 Comparative pollen morphology of the Madagascan genera of Myristicaceae (Mauloutchia, Brochoneura, and Haematodendron). Grana 20: 1-17

Warburg O. 1897 Monographie der Myristicaceen. Nova Acta Academia Caesareae Leopoldiana-Caroliniana 68: 1-680

Wilson T. K. L. M. Maculans 1967 The morphology of the Myristicaceae. I. Flowers of Myristica fragrans and M. malabarica. American Journal of Botany 54: 214-220[CrossRef][ISI]





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sauquet, H.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Sauquet, H.
Agricola
Right arrow Articles by Sauquet, H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS