| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Institut des Sciences de l'Evolution, UMR CNRS 5554, Université Montpellier II, Place Eugène Bataillon, 34095 Montpellier cedex 5, France; 2 Centre d'Ecologie Fonctionnelle et Evolutive, UPR CNRS 9056, 1919 route de Mende, 34293 Montpellier cedex 5, France; and 3 LACITO UPR CNRS 3121, 7, rue Guy Moquet, bât. 23, 94800 Villejuif, France; and Laboratoire de Botanique, Phytochimie et Mycologie, Faculté de Pharmacie, Université Montpellier I, 34060, Montpellier cedex 2, France
Received for publication May 27, 1998. Accepted for publication June 17, 1999.
ABSTRACT
Cup- or sometimes slit-shaped nectary glands on the rachis are a widespread trait in the legume subfamily Mimosoideae, especially in derived tribes. Their spotty occurrence in genera that appear to be basal has led to uncertainty about when in the mimosoid radiation this character evolved. Until now, specialized rachis glands were unknown in caesalpinioids thought to be related to ancestral mimosoids. We report here the occurrence of rachis glands in seven of the ten species of the Paleotropical genus Erythrophleum, a member of the Dimorphandra group of caesalpinioids thought to include the sister group(s) of mimosoids. The histological structure and location of Erythrophleum glands suggest homology with those of mimosoids; these glands are simpler structurally than rachis glands of any known mimosoid. The Erythrophleum glands differ from those of most mimosoids in the following respects: (1) they are smaller than glands of mimosoids; (2) the secretory surface is sunken in a pit capped by a small round pore rather than exposed on a broad concave or flat surface; (3) a smaller number of cells are involved in production and secretion of nectar; (4) vascular supply to the nectary is less extensive; and (5) mechanical support tissue (sclerenchyma) is less extensive and less organized. Rachis glands appear to be absent in the nine other genera included in the Dimorphandra group. We also report the occurrence of other secretory structures (patches of glandular trichomes) on the rachis of some Caesalpinieae and Mimoseae that lack specialized nectary glands and suggest that these patches of trichomes are primitive homologues of more organized glands. We discuss the significance of these glands and of the patches of trichomes for understanding relationships among primitive mimosoids and related caesalpinioids, and for understanding the origin of ant-guard defenses typical of many mimosoids.
Extrafloral nectaries are commonly present in Leguminosae. These glands show great diversity in their morphology and location on the plant (Zimmermann, 1932
; Elias, 1983
). Foliar nectary glands appear to have evolved numerous times independently in Leguminosae (McKey, 1989
). While many nectary types represented in the family have a fairly limited taxonomic distribution and are restricted to a single genus, subgenus, or even to one or a few species, the rachis glands found in the majority of mimosoids are very similar in structure. At least 44 of the ~60 genera of subfamily Mimosoideae possess flat-topped or concave nectary glands on the leaf rachis. Rachis glands of mimosoids, here termed interjugal glands, are usually located near the base of the rachis and/or at the intersection of pairs of leaflets or pinnae on the rachis of the pinnately or bipinnately compound leaf. They are most widespread in Acacieae (they are present in most sections of the very large genus Acacia but absent in the monotypic Faidherbia) and Ingeae (present in at least 17 of ~20 genera), both of which are considered derived lineages and are less frequent in tribe Mimoseae (present in 26 of 37 genera), considered a basal lineage in the subfamily (Elias, 1981a
; Lewis and Elias, 1981). They are absent in tribe Parkieae. One of the two genera currently included in this tribe, Pentaclethra, possesses extrafloral nectary glands (Bennett and Breed, 1985
), but they are located on stem axes and not on the rachis (G. Hartshorn, Organization for Tropical Studies, personal communication; D. McKey, personal observation).
Mimosoid rachis nectaries attract ants, and ant-guard defense against herbivores appears widespread in the subfamily (Elias, 1981a
; Polhill and Vidal, 1981
). Ant-plant interactions involving mimosoids exhibit great variation in specificity and in degree of protection afforded by ants. The role of ants opportunistically attracted by rachis nectaries in protection against herbivores has been demonstrated experimentally in Inga (Koptur, 1984
), and a similar role of facultative ant-plant interactions is suggested by observations in many other genera. In addition, rachis glands are an important trait of obligate and specific symbioses involving Acacia myrmecophytes (e.g., Janzen, 1974
) and in myrmecophytic species of a few other mimosoid genera (Davidson and McKey, 1993
). Elias (1981a)
briefly discussed evolutionary trends in mimosoid nectaries, viewing cup-shaped nectaries with branched vascular bundles and supporting tissues as derived relative to flattened or slightly concave nectaries. We have as yet no clear picture of the origin and subsequent evolutionary changes in anatomy of rachis glands or of how histological structure of the glands is related to ant attraction and functioning of ant-guard defenses.
Although careful comparative studies including a large number of genera are lacking, the rachis glands of mimosoids have all been assumed to be homologous structures based on overall similarities in structure and position on the plant (McKey, 1989
). This hypothesis implies a single origin with divergent specialization. However, neither morphological nor phylogenetic evidence is sufficiently strong to favor the hypothesis of homology over the plausible alternative of parallel evolution from similar, but unknown precursor structures. The possibility of parallel evolution is further suggested by the existence of rachis glands in two related genera of the caesalpinioid tribe Cassieae, Chamaecrista and Senna (Irwin and Barneby, 1981
), and in the genus Desmodium of the papilionoid tribe Desmodieae (Pascal, Motte-Florac, Lafisca, and McKey, unpublished data). These two groups are only very distantly related to each other and to mimosoids (Doyle, 1994
; Doyle et al., 1997
; see Fig. 1).
|
) and in flowers (Rudall, Myers, and Lewis, 1994
). Secretory structures of leaf rachises have not been examined. Rachis glands have been thought to be absent from caesalpinioid genera closely related to mimosoids (e.g., Polhill and Vidal, 1981
), and these taxa have so far offered no clues to the origin of mimosoid glands. We report here the occurrence of "hollow" nectary glands ("Hohlnektarien": Zimmermann, 1932
) on the rachis of the bipinnately compound leaves of species of the genus Erythrophleum, a member of the Dimorphandra group of tribe Caesalpinieae. McKey (1989)
and Keller (1994)
signalled the presence of these glands, but presented no detailed description of their structure. The Dimorphandra group is a set of ten morphologically isolated genera placed in tribe Caesalpinieae and considered to include the closest living relatives of ancestral Mimosoideae (Polhill and Vidal, 1981
; Doyle, 1994
; Doyle et al., 1997
; Fig. 1). The glands of Erythrophleum are similar in structure to those of mimosoids (and very different from those of Senna, Chamaecrista, and Desmodium), but are smaller and morphologically simpler than those of most mimosoids. We describe the structure of the Erythrophleum glands and discuss their significance for understanding the evolution of structures that could be related to biotic defenses against herbivores in mimosoids. We also show that in a number of other Caesalpinieae and Mimoseae that lack organized rachis nectaries there exist patches of glandular trichomes densely concentrated at the same interjugal positions. We discuss possible evolutionary relationships between these patches of trichomes and organized rachis glands. MATERIALS AND METHODS
Anatomical studies of rachis glands
Histological sections of 17 species were examined (Table 1). All species studied were trees, shrubs, or lianas with pinnately or bipinnately compound leaves. Glandular structures were located on the rachis at the intersection of the leaflets or the pinnae, respectively (sometimes near the base of the rachis and not at interjugal position). The sources of materials and authorities for species names are detailed in the Appendix.
|
Survey of occurrence of rachis glands
In addition to the 17 species of which histological sections were studied, we also examined numerous Caesalpinieae and Mimosoideae for the presence of secretory tissue (glands, glandular trichomes) on the rachis using a binocular microscope. Species studied, sources of material, and authorities of species names are detailed in the Appendix. Table 2 summarizes the results of this survey.
|
(Fig. 1). Our aims were (1) to examine representatives of all groups positioned between the two groups in which rachis nectaries have been reported, i.e., Chamaecrista and Senna on the one hand, and the Mimosoideae on the other; and (2) to study the rachis of mimosoids with and without such glands. After our sampling was completed, Doyle et al. (1997)
published a revised phylogeny of Leguminosae based on rbcl sequences. The revised estimation of relationships does not affect our conclusions. RESULTS
Morphology of rachis glands
Following Zimmermann's (1932)
suggested classification of nectaries, three types are present on the rachis of Leguminosae: Hochnektarien, Grubennektarien, and Hohlnektarien. Elias (1981a)
adopted the same categories, terming them elevated, pit, and hollow nectaries, respectively. The rachis glands of Senna, Chamaecrista, and Desmodium are elevated nectaries with glandular tissue distributed over the surface (Pascal, Motte-Florac, Lafisca, and McKey, unpublished data), and this secretory tissue is supported by a stalk. These structures are very different from those of mimosoids and Erythrophleum and will be considered in a separate paper. In Mimosoideae and Erythrophleum, we found only pit and hollow nectaries.
Pit or cup-shaped (Keller, 1994
) nectaries on the leaf rachis are well known in mimosoids. We prefer the term "nectary with an apical depression" to describe these glands, because the degree to which the nectary is elevated above the surface of the rachis, as well as the extent of depression at the apex, is more or less pronounced depending on the species. The nectaries always lack a stalk and hence are not elevated nectaries in the sense of Elias (1981a)
. We studied two species chosen to be representative of this variation. In Inga feuillei (Ingeae), a large nectary is located on the rachis at the insertion of each pair of leaflets of the once-pinnate leaf. The cup-like gland is similar in basic structure to that of many mimosoids. In this species, the cup cavity is relatively large (diameter 2 mm) and bears a large secretory surface (Figs. 2–4). The other species studied was Acacia sphaerocephala (Acacieae). In contrast to the usual form of rachis nectaries in Acacia (a cup-like structure similar to the Inga gland), the nectary (Figs. 5–7) is a prominent structure that is elongate ("boat-shaped gland": Janzen, 1974
) rather than cylindrical. As in Inga and many other mimosoids, the tip of the gland is concave (Fig. 5).
|
|
) nectaries are less well-known among mimosoid rachis glands. We prefer to use the phrase "dome-like with a pore" to describe these nectaries. The presence of a pore is important because it distinguishes species with a secretory structure from species with superficially similar dome-like protuberances on the rachis (usually in interjugal position), but with no evident secretory tissues (such as Stachyothyrsus, see below). We studied two species, Mimosa guilandinae and Erythrophleum suaveolens, with this nectary type. In M. guilandinae (Figs. 8–10), the gland (in interjugal position) is a prominent dome (mean height 1.5 mm) with a basal diameter of 2 mm, and bearing a central pore 0.1 mm in diameter. The prominent dome is clearly delimited from adjacent tissues by its high stomatal density and a glabrous surface. The area surrounding the base of the nectary is covered with glandular and eglandular trichomes (Fig. 8). In E. suaveolens, in which the anatomy of rachis glands has previously not been described, a small nectary (Figs. 11–14) is located on the adaxial surface of the rachis, at the point of insertion of each of the 3–4 pairs of pinnae. The nectary has a mean height of 1 mm above the surface of the rachis, and a mean outside diameter of 2.0 mm (measured at the based of the dome). Externally, the Erythrophleum nectary is cryptic, a small domed structure with a central pore (mean inside diameter of the pore 0.1–0.2 mm) being the only indication of a glandular structure. The base of the gland is surrounded by a large number of eglandular trichomes. Similar trichomes are sparsely distributed on the surface of the nectary itself and are also present inside the aperture of the pore (Fig. 11).
|
|
|
Inga feuillei (Ingeae)
The raised body of the nectary (Figs. 3, 4) is supported by a cylinder of sclerenchyma fibers. Within this cylinder, the base of the nectary is composed of a large area of vascular tissue (xylem and phloem) and the apical one-third is occupied by cells inferred to be secretory cells (Pascal, 1994
). Between the vascular tissue and the secretory cells lies a thin, horizontally oriented layer of sclerenchyma cells. These cells have slightly thickened walls and are canaliculate, presumably allowing passage of materials from vascular tissue to the secretory cells. In this gland, a large area of secretory tissue is exposed to the surface (~3·10-3 cm2). This surface is lined with epidermis, and sclerenchyma fibers scattered throughout the secretory tissue presumably add rigidity. This is the only gland in which we observed such sclerenchyma fibers embedded in secretory tissue.
Acacia sphaerocephala (Acacieae)
The base of the nectary is surrounded by glandular hairs similar to those of Delonix and Dimorphandra. In this species (Figs. 6, 7) vertical ridges of sclerenchyma fibers support the raised nectary, one external (forming a single oblong ring around the glandular tissue) and one central. Vascular supply to the nectary is composed of a ring of xylem and phloem lying just interior to the external ring of sclerenchyma fibers. Secretory cells are disposed on either side of the central ridge of sclerenchyma fibers. The secretory tissue is more deeply embedded in the body of the nectary than in Inga, and a larger area of secretory tissue is exposed to the surface (~15·10-3 cm2), because of the greater length of the gland. As in Inga, a thin layer of canaliculate sclerenchyma cells with slightly thickened walls (cells oriented vertically in this species rather than horizontally) lies between the vascular tissue and the secretory cells.
Mimosa guilandinae (Mimoseae)
The rachis nectary of this species (Figs. 9, 10) is a prominent cone-shaped dome with a small pore at the tip. The secretory tissue is embedded within this dome, with only a very small area being exposed directly to the surface at the apical pore (~27·10-6 cm2). Support and mechanical protection of the gland are assured by a cone-shaped layer of sclerenchyma fibers located just beneath the surface. Just interior to this layer of sclerenchyma fibers is a cone-shaped layer of vascular tissue (xylem and phloem). There is also a central cone of vascular tissue protruding upward from the base of the nectary. Just interior to these vascular tissues is a thin layer of canaliculate sclerenchyma cells with slightly thickened walls, which, as in the other mimosoid glands, separates the vascular tissue from the mass of secretory cells that occupies most of the interior of the cone-shaped dome.
Erythrophleum suaveolens (Caesalpinieae)
In this species (Figs. 12–14) the dome of the nectary is less prominent than in M. guilandinae, and the secretory tissue is deeply embedded in the rachis itself. Sclerenchyma fibers are present in the tissue covering the secretory cells but, in contrast to the other species described above, these fibers do not form a distinct and continuous layer surrounding the nectary. Vascular supply to the gland (xylem and phloem) is less extensive in this species. As in the other species, the secretory tissue is separated from the vascular supply by a layer of cells. In Erythrophleum, however, this separation is effected by a thicker layer of parenchyma rather than by a thin layer of canaliculate sclerenchyma cells. The roughly spherical mass of secretory tissue is not directly exposed to the surface, but is instead separated from the tip of the apical pore by a short canal lined with eglandular trichomes.
Stachyothyrsus stapfiana (Caesalpinieae)
The rachis of plants of this species bears a small dome-like protuberance (Figs. 17, 18) at the points of insertion of each pair of pinnae. However, this structure lacks a pore and transverse sections confirmed the absence of secretory tissues. Beneath the epidermis the dome is composed of parenchyma and scattered sclerenchyma cells.
|
Are Erythrophleum interjugal glands homologous with those of mimosoids?
The rachis glands of Erythrophleum share several anatomical traits with those of the mimosoids examined. First, the secretory tissue of the gland is more or less invaginated (and in this respect very different from Chamaecrista, for example); the gland is ringed or covered by supporting tissue (sclerenchyma fibers); and vascular supply to the gland is separated from secretory tissues by layers of cells (parenchyma or canaliculate sclerenchyma cells). These anatomical similarities support the hypothesis that the glands of Erythrophleum are homologous to those of mimosoids.
The Erythrophleum glands appear in many respects to be simpler morphologically than those of the mimosoids examined: (1) the mass of secretory cells is relatively small; (2) this secretory tissue is deeply embedded in the rachis, connecting to the exterior only via a small apical pore; (3) vascular supply to the nectary is less extensive; (4) separation of vascular supply from the secretory tissue is by unspecialized parenchyma cells, rather than by canaliculate sclerenchyma cells; and (5) supporting tissues (sclerenchyma fibers) covering or surrounding the nectary are less highly organized. Based on the phylogenetic position of Erythrophleum compared to mimosoids (Fig. 1), we postulate that these are relatively primitive character states.
In many respects, the glands of Mimosa (in the "basal" mimosoid tribe Mimoseae) appear to be intermediate in morphological complexity between those of Erythrophleum and those of Inga and Acacia, representatives of "derived" mimosoid tribes. Secretory tissue of the Mimosa glands is connected to the surface by a small pore, as in Erythrophleum, but a small area of secretory tissue is directly exposed, whereas in Erythrophleum a canal separates the secretory tissue from the mouth of the pore. In the nectaries of the "derived" mimosoids, the secretory tissue is much larger, with a greater area exposed to the surface (e.g., Inga, ~3·10-3 cm2 compared to Erythrophleum, ~27·10-6 cm2), the glands are not embedded in the rachis and have more extensive supporting tissue. The anatomical comparison of nectaries of these four genera, coupled with existing phylogenetic hypotheses, thus suggests trends in nectary evolution that can be tested with further phylogenetic data.
A further trend in the evolution of secretory structures in mimosoids is suggested by examination of Table 2. Based on our limited sample, glandular trichomes on the rachis (usually concentrated at the interjugal position) occur in the caesalpinioid tribe Caesalpinieae (Delonix, Dimorphandra) and in the basal mimosoid tribe Mimoseae (Adenanthera, Leucaena, Mimosa), where they sometimes co-occur with nectaries. Glandular trichomes in this position are absent from our sample in the "derived" tribe Ingeae. A more extensive sample is necessary to evaluate the suggested trend.
The occurrence of rachis glands in widely separated taxa in all three legume subfamilies, along with the occurrence of concentrations of glandular hairs in interjugal position in some nectary-less members of both Caesalpinieae (Delonix, Dimorphandra) and Mimoseae (Adenanthera), suggest there may be a widespread predisposition to evolve secretory structures in this location. As in the case of ant-attractive elaiosomes in Papilionoideae (Berg, 1979
), it may be difficult to distinguish homology from "homoiology" (Riedl, 1979
), parallel evolution in a series of closely related taxa. The significance of dome-like protuberances at interjugal position lacking evident secretory structures, as observed in Stachyothyrsus, is unclear. It is perhaps noteworthy that such structures were in our sample (Table 2) restricted to the Dimorphandra group and the Sclerolobium group, which in the revised phylogeny of Doyle et al. (1997)
appears as sister group to (Dimorphandra group + Mimosoideae).
Distribution of rachis glands and the distinction between mimosoids and "caesalpinioids"
It is widely accepted that current classification only poorly reflects phylogenetic relationships among basal mimosoids and allied caesalpinioid taxa (e.g., Lewis and Elias, 1981; Zandee and Geesink, 1987
; Doyle et al., 1997
). New characters may be informative. If rachis glands of Erythrophleum are considered homologous to those of mimosoids, this may indicate that this genus is more closely related to nectary-bearing taxa of Mimoseae than to other genera of the Dimorphandra group, which in turn may be closely allied to basal mimosoids that lack rachis glands. Some evidence supports such a pattern. For example, Pentaclethra (placed in the mimosoid tribe Parkieae) and Dimorphandra are considered to be probably more closely related to each other than to other genera in their respective tribes (Elias, 1981b
). Both lack rachis nectaries. Among the five genera of Mimoseae considered most basal by Lewis and Elias (1981), Dinizia, Aubrevillea, and Fillaeopsis all lack rachis glands, while two others, Cylicodiscus and Stryphnodendron, both possess such glands. Are these latter genera more closely allied to Erythrophleum than to other Mimoseae? Lewis and Elias (1981) note the similar overall appearance of Cylicodiscus and Erythrophleum, but we are aware of no cladistic analysis that addresses this question. There is little reason to expect that rachis glands are any less subject to the homoplasy that has affected many other characters such as pollen grains (single or in polyads) or aestivation of perianth elements (valvate or imbricate), in which mosaic patterns of character distribution across taxa have persistently frustrated attempts to discern generic relationships. Molecular data from neutrally evolving loci may be necessary to resolve relationships in this group.
Functional significance of evolutionary trends in rachis glands
If, as we postulate based on their phylogenetic distribution, the simpler glands (e.g., those of Erythrophleum and Mimosa) are plesiomorphic and the more complex structures (e.g., in Acacia and Inga) are derived, several evolutionary trends are suggested: (1) increase in volume of secretory tissue and of vascular supply to it, leading to increased rates of nectar secretion (L. Pascal, unpublished data); (2) increase in the area of secretory tissue directly exposed to the surface of the gland, facilitating nectar collection by ants; and (3) increase in mechanical support of the larger and more exposed secretory tissue. In the extreme case of Inga, isolated sclerenchyma cells even occur directly in the nectariferous tissue, presumably providing horizontal support for the broadly flattened central surface of the nectary.
Evolution of ant-guard relationships in mimosoids
The study suggests the existence, in several genera of caesalpinioids and in mimosoid lineages considered to be basal, of secretory structures that are similar to the rachis glands of most mimosoids, but simpler in structure and characterized by very low rates of secretion (L. Pascal, unpublished data). These include not only the well-organized, sunken glands of Erythrophleum, but also glandular trichomes, which are usually concentrated in, or even restricted to, dense patches in interjugal position. Detailed anatomical and functional comparisons of these structures may shed light on the origins of ant-guard relationships in mimosoids and on the selective pressures that have led to their diversification.
Are organized rachis glands homologous with patches of glandular trichomes?
McDade and Turner (1997)
present evidence that in Aphelandra (Acanthaceae) large bracteal nectary glands are derived from patches of numerous small glands, which in turn appear to be homologous with simpler glandular trichomes. The phylogenetic distribution of patches of glandular trichomes in caesalpinioids and basal mimosoids (compare Table 2 and Fig. 1) suggests that in legumes, as well, the highly organized rachis nectaries of more "advanced" mimosoids may be similarly derived from concentrations of large numbers of glandular trichomes. The interjugal position, at the insertion of pairs of pinnae on the rachis, is well supplied with vascular tissue leading to the pinnae and to the continuation of the rachis, and it is perhaps for this reason that it appears to be a favored location for such glandular structures.
The conjecture that highly organized rachis glands are homologous with patches of glandular trichomes might be objected to on the grounds that in some species both structures are found, violating Patterson's (1982)
criterion of non-conjunction. This seeming difficulty could be resolved by considering that these structures, like many other traits of plants and other modular organisms, are affected by homeosis and may be examined in the framework of serial homology (Patterson, 1982
; Sattler, 1984, 1988
). A character may be transformed in one way at one position on the plant and in another way (or not at all) at another position. In species where organized rachis glands and glandular trichomes co-occur, they differ in position, the gland occupying the center and glandular trichomes being placed around the margin of the gland. The two may be homologous in the same way that foliage leaves, sepals, petals, and stamens are homologous. Comparative studies of development, combined with further phylogenetic data, could shed light on this hypothesis.
|
|
1 The authors thank Jack Fisher for his assistance in histology and his kind reception in his laboratory at Fairchild Tropical Garden and the Montgomery Foundation; Mrs. J. Bayonove for her help in scanning electron microscopy; Mr. P. Lafisca for assistance in cutting sections; the directors of Fairchild Tropical Garden, of The Kampong of the National Tropical Botanical Garden (both in Miami, Florida, USA), and of the "Service des Cultures" in the Arboretum of the National Museum of Natural History at Versailles (France), for kindly permitting us to collect material used in this study; the directors of the herbaria of the National Museum of Natural History (Paris), the National Herbarium of Cameroon (Yaoundé), and the University of Montpellier for allowing us to study specimens lodged in these collections; Laurence Gaume for collecting material from two species of Erythrophleum; Prof. J. J. Doyle (Cornell University) for permission to use Fig. 1 .
This paper is contribution number ISEM 99-035 from the Institut des Sciences de l'Evolution, Université Montpellier II. 
2 Author for correspondence (Fax: 33 4 67 41 21 38; e-mail: Pascal{at}cefe.cnrs-mop.fr
).
LITERATURE CITED
Bennett, B., and M. D. Breed. 1985 On the association between Pentaclethra macroloba (Mimosaceae) and Paraponera clavata (Hymenoptera: Formicidae) colonies. Biotropica 17: 253–255. [CrossRef][ISI]
Berg, R. Y. 1979 Legume, seed, and myrmecochorous dispersal in Kennedia and Hardenbergia (Fabaceae), with a remark on the Durian theory. Norwegian Journal of Botany 26: 229–254. [ISI]
Davidson, D. W., and D. B. McKey. 1993 The evolutionary ecology of symbiotic ant-plant relationships. Journal of Hymenoptera Research 2: 13–83.
Doyle, J. J. 1994 Phylogeny of the legume family: an approach to understanding the origin of nodulation. Annual Review of Ecology and Systematics 25: 325–349. [CrossRef][ISI]
———, J. L. Doyle, J. A. Ballinger, E. E. Dickson, T. Kajita, and H. Ohashi. 1997 Aphylogeny of the chloroplast gene rbc L in the Leguminosae: taxonomic correlations and insights into the evolution of modulation. American Journal of Botany 84: 541–554. [Abstract]
Elias, T. S. 1972 Morphology and anatomy of foliar nectaries of Pithecellobium macradenium (Leguminosae). Botanical Gazette (Crawfordsville) 133: 38–42.
———. 1981a Mimosoideae. In R. M. Polhill and P. H. Raven [eds.], Advances in legume systematics, 1, 143–152. Royal Botanic Gardens, Kew, Richmond, Surrey, UK.
———. 1981b Parkieae. In R. M. Polhill and P. H. Raven [eds.], Advances in legume systematics, 1, 153. Royal Botanic Gardens, Kew, Richmond, Surrey, UK.
———. 1983 Extrafloral nectaries: their structure and distribution. In B. Bentley and T. Elias [eds.], Biology of nectaries, 174–203, Colombia University Press, New York, New York, USA.
Irwin, H. S., and R. C. Barneby. 1981 Cassieae. In R. M. Polhill and P. H. Raven [eds.], Advances in legume systematics, 1, 97–106. Royal Botanic Gardens, Kew, Richmond, Surrey, UK.
Janzen, D. H. 1974 Swollen-thorn acacias of Central America. Smithsonian Contributions to Botany 13: 1–131.
Keller, R. 1994 Clef de terrain pour l'identification des familles des plantes ligneuses tropicales à l'aide des seuls caractères végétatifs, Naturalia Monspeliensia, numero Hors série.
Koptur, S. 1984 Experimental evidence for defense of Inga (Mimosoideae) saplings by ants. Ecology 65: 1787–1793. [CrossRef][ISI]
Lersten, N. R., and J. D. Curtis. 1996 Survey of leaf anatomy, especially secretory structures, of tribe Caesalpinieae (Leguminosae, Caesalpinioideae). Plant Systematics and Evolution 200: 21–39. [CrossRef][ISI]
Lewis, G. P., and T. S. Elias. 1981 Mimoseae. In R. M. Polhill and P. H. Raven [eds.], Advances in legume systematics, 1, 155–168. Royal Botanic Gardens, Kew, Richmond, Surrey, UK.
McDade, L. A., and M. D. Turner. 1997 Structure and development of bracteal nectary glands in Aphelandra (Acanthaceae). American Journal of Botany 84: 1–15. [Abstract]
McKey, D. 1989 Interactions between ants and leguminous plants. Monographs in Systematic Botany of the Missouri Botanical Garden 29: 673–718.
Pascal, L. 1994 La sécrétion nectarifère de lianes de Guyane française, origine, composition et devenir du nectar. Thèse de l'Université Pierre et Marie Curie, Paris, France.
Patterson, C. 1982 Morphological characters and homology. In K. A. Joysey and A. E. Friday [eds.], Problems of phylogenetic reconstruction, 21–74. Academic Press, London, UK.
Polhill, R. M., and J. E. Vidal. 1981 Caesalpinieae. In R. M. Polhill and P. H. Raven [eds.], Advances in legume systematics, 1, 81–95. Royal Botanic Gardens, Kew, Richmond, Surrey, UK.
Riedl, R. 1979 Order in living organisms: a systems analysis of evolution. Wiley, London, UK.
Rudall, P. J., G. Myers, and G. P. Lewis. 1994 Floral secretory structures in Caesalpinia sensu lato and related genera. In I. K. Ferguson and S. Tucker [eds.], Advances in legume systematics, 6, 41–52. Royal Botanic Gardens, Kew, Richmond, Surrey, UK.
Sattler, R. 1984 Homology—a continuing challenge. Systematic Botany 9: 382–394. [CrossRef][ISI]
———. 1988 Homeosis in plants. American Journal of Botany 75: 1606–1617. [CrossRef][ISI]
Zandee, M., and R. Geesink. 1987 Phylogenetics and legumes: a desire for the impossible? In C. H. Stirton [ed.], Advances in legume systematics, 3, 131–167, Royal Botanic Gardens, Kew, Richmond, Surrey, UK.
Zimmermann, J. G. 1932 Über die extrafloralen Nectarien der Angio-spermen. Beihefte zum botanischen Zentralblatt 49: 99–196.
This article has been cited by other articles:
|
|
 |
|
  B. Marazzi, P. K Endress, L. P. d. Queiroz, and E. Conti Phylogenetic relationships within Senna (Leguminosae, Cassiinae) based on three chloroplast DNA regions: patterns in the evolution of floral symmetry and extrafloral nectaries Am. J. Botany, February 1, 2006; 93(2): 288 - 303. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. DIAZ-CASTELAZO, V. RICO-GRAY, F. ORTEGA, and G. ANGELES Morphological and Secretory Characterization of Extrafloral Nectaries in Plants of Coastal Veracruz, Mexico Ann. Bot., December 1, 2005; 96(7): 1175 - 1189. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
