| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Anatomy and Morphology |
Laboratory of Plant Systematics, Institute of Botany and Microbiology, Katholieke Universiteit Leuven, Kasteelpark Arenberg 31, B-3001 Leuven, Belgium
Received for publication April 27, 2005. Accepted for publication September 26, 2005.
ABSTRACT
The floral morphological and developmental patterns in three species of Impatiens (Balsaminaceae), namely I. columbaria, I. hawkeri, and I. niamniamensis, were studied to contribute to a better understanding of floral evolution in the genus. Strangely enough, the highly diverse floral morphology and ontogeny of this horticulturally important genus have never been studied thoroughly (e.g., using scanning electron microscopic techniques). We discuss the position and the developmental sequence of the different perianth members. We hypothesized that in the course of evolution, the anterolateral sepals become reduced and that a morphocline can be recognized going from species with five sepals, over species with rudimentary sepals that fuse postgenitally with the anterior petal, to species where congenital fusion between these sepals and the anterior petal has taken place. Ovules generally are in one or two vertical rows per locule, but there are several vertical rows per locule in I. columbaria. The outer parts of the septa disintegrate to enable the explosive dehiscence of the capsules.
Key Words: Balsaminaceae • Ericales • floral development • floral evolution • floral morphology • Impatiens • scanning electron microscopy
The Balsaminaceae are a family of about 1000 species and two genera, the monotypic Hydrocera Blume and the large genus Impatiens L. (Fischer, 2004
; Stevens, 2004
). Impatiens is mainly distributed in the tropics and subtropics of the Old World, but several species occur in temperate Eurasia and North America. Native species are absent from South America and Australia. Hydrocera is a semiaquatic genus of the Indo-Malaysian region. Together with Marcgraviaceae and Tetrameristaceae s.l. (including Pelliciera Planch. & Triana), Balsaminaceae constitute the balsaminoid clade at the base of the Ericales (Anderberg et al., 2002
; Bremer et al., 2002
; Geuten et al., 2004
). The monophyly of the Balsaminaceae and of the genus Impatiens itself are well supported (Yuan et al., 2004
). Hydrocera can be characterized by its free petals and the berry-like capsular fruit, while Impatiens has four lateral petals connate in pairs and a five-valved, loculicid capsule. The sister-group relationship between Hydrocera and Impatiens is also confirmed by recent, molecular analyses (e.g., Yuan et al., 2004
). Based on the overall morphology and distribution, several groups can be distinguished within Impatiens, but the relationships among these groups remain unresolved. The taxonomic difficulties are probably due to the existence of a large number of intermediate groups and taxa (Grey-Wilson, 1980a
).
The often striking and beautifully colored flowers are hermaphroditic and are arranged in racemes, fascicles, or solitary in the axils of the leaves, or rarely pseudoterminally (Warburg and Reiche, 1895
). Bracteoles are missing. In the African Impatiens species, all inflorescences can be considered to be variations of racemes, while species from the Himalaya are sometimes characterized by inflorescences without clear racemose organization (Akiyama and Ohba, 2000
). For a detailed study of the inflorescence types of Impatiens, we refer to Akiyama and Ohba (2000)
.
The flowers are resupinate, and consequently, the floral parts are often named according to the position they acquire after resupination. However, in our ontogenetic descriptions, we will refer to the original positions of the organs and use the terms abaxial and adaxial. In the discussion, we will speak about the posterior sepal, i.e., the lower, spurred sepal, the anterior petal on the opposite side, and according to the situation, the anterolateral or posterolateral sepals or petals (Fig. 1).
|
, the stamens rupture at the base and are lifted by the gynoecium. The ovules are anatropous, bitegmic, and tenuinucellate, and they are arranged in one or two vertical rows on the axile placenta. The micropyle is directed upward and inward. One of the best-known characteristics of Impatiens is the explosive dehiscence of the five-valved, loculicid capsule, resulting in the English names busy-lizzy and touch-me-not. The valves of the fruit roll up inwardly and acropetally, which causes the seeds to be dispersed in all directions. What remains is the central axis of the fruit on which the seeds were attached, and at the top, the spirally winded valves. It may be clear that to effect such a dehiscence mechanism, a lot of tension has to be involved. Therefore, we paid special attention to the structure of the septa during our observations.
The floral morphology is highly diverse, and information on the developmental and evolutionary patterns within the genus is sparse (Yuan et al., 2004
). The flowers are delicate structures, of which little remains in dried specimens. Therefore, herbarium material cannot be used to gain more insight on the morphology of the flower (Akiyama et al., 1991
). The floral ontogeny has also been scarcely investigated, but studies like Payer (1857)
help to illustrate the rich variety in the structure of the flower and may provide taxonomically useful characters.
The present study is part of a large study on floral development within Ericales, as defined by the Angiosperm Phylogeny Group (APG, 2003
). We studied flowers and floral buds of three Impatiens species. We will first describe the overall ontogenetic pattern, using our observations on I. niamniamensis Gilg, a species of tropical West and Central Africa (Grey-Wilson, 1983
). This frequently cultivated species is one of the most widespread, best-known, and most attractive species from Africa (Grey-Wilson, 1980a
). Subsequently, we present our results from the poorly known I. columbaria J.J. Bos, a species with an initially pentamerous calyx and some interesting gynoecium characteristics. It was discovered in 1985, and as far as we know, it can only be found in western Gabon (Grimshaw, 1998
). Finally, we have studied the highly diverse I. hawkeri W. Bull, which was added here to illustrate the initiation and inner structure of the gynoecium. It is distributed from New Guinea east to the Solomon Islands (Grey-Wilson, 1983
). According to Grey-Wilson (1980b)
, the group around I. hawkeri forms a complicated and highly variable aggregate. Moreover, it has many cultivars that have become popular pot plants (Grey-Wilson, 1983
).
These three species were selected because they encompass the basic variation in the flower morphology of the genus. Within Impatiens, two flower types can be distinguished on the basis of the spurred sepal (Grey-Wilson, 1980a
). In the first type (represented here by I. niamniamensis; Fig. 1A–C), it has a funnel-shaped or saccate appearance and gradually continues into the spur. In the second type (represented by I. hawkeri; Fig. 1D) the spurred sepal is much smaller and it bears a filiform spur, which commonly is much longer than the sepal itself. Impatiens columbaria was added because this species possesses a pentamerous calyx, while the other two species are characterized by a trimerous calyx.
MATERIALS AND METHODS
The material of I. niamniamensis (voucher n°FB/S2590 and n°FB/S2642) and I. columbaria (voucher n°FB/S2966) was obtained from the greenhouse collection of the National Botanic Garden in Meise, Belgium. Impatiens hawkeri (voucher n°PCV06) was grown by the second author at the Laboratory of Plant Systematics, K.U.Leuven. Voucher specimens are kept at the National Botanic Garden in Meise and the Institute of Botany and Microbiology, K.U.Leuven.
The material was fixed in FAA (40% formalin, acetic acid, 70% alcohol, 5 : 5 : 90) and the floral buds were dissected in 70% ethanol under a stereomicroscope (Wild M3; Leica Microsystems AG, Wetzlar, Germany) equipped with a cold light source (Schott KL 1500; Schott-Fostec LLC, Auburn, New York, USA). To dry the material, the buds were washed twice for 5 min with 70% ethanol, for a further 5 min with a mixture (1 : 1) of 70% ethanol and DMM (dimethoxymethane), then the material was placed in pure DMM for 20 min. The samples were critical point dried using liquid CO2 in a BAL-TEC CPD030 (BAL-TEC AG, Balzers, Liechtenstein). The material was mounted onto stubs using Leit-C and then gold-coated with a sputter coater (SPI Supplies, West Chester, Pennsylvania, USA). Observations were made using a JEOL JSM-5800 LV scanning electron microscope (JEOL Ltd., Tokyo, Japan) at the National Botanic Garden in Meise and a JEOL JSM-6360 microscope at the Laboratory of Plant Systematics, K.U.Leuven.
For the light microscopic observations, dehydrated floral buds of Impatiens niamniamensis were embedded in Kulzer's Technovit 7100 (Kulzer Histo-Technik, Wehrheim, Germany). Serial sections, 5 µm thick, were stained with toluidine blue and mounted with Entellan (Merck, Darmstadt, Germany). Photographs were made using a Leitz Dialux 20 (Leica Microsystems AG, Wetzlar, Germany) equipped with an Olympus DP-50 digital camera (Olympus, Tokyo, Japan).
RESULTS
Impatiens niamniamensis Gilg
The flowers have long, slender pedicels and occur in clusters of two to six (to eight) in the axils of the leaves. The color of the flowers varies, but our material had red-yellow flowers: the spur is orange-red and the petals pale yellowish-green to whitish-green. The flower primordia arise spirally along the axis. They are initiated in the axil of a bract (Fig. 2A). Meanwhile, the first two sepal primordia are initiated on both sides of the flower primordium (Fig. 2B). Subsequently, the primordium of the large, spurred sepal originates at the adaxial side of the flower (Fig. 2C). Almost immediately afterward, another primordium becomes visible on the opposite side of the floral apex (Fig. 2D). This abaxially developing primordium will differentiate into the anterior petal. In the zone in between the large sepal and the anterior petal, the four remaining petal primordia arise in two successive pairs, the upper pair slightly before the lower pair (Fig. 2E). The petal primordia on the adaxial side, underneath the developing sepal, are somewhat smaller than the posterolateral petals (Fig. 2E). It is striking that in the development of the perianth, a clear distinction between the development of calyx and corolla is missing. The calyx arises in two stages, and the initiation of the anterior petal is intermediate with respect to the development of calyx and corolla (Fig. 2D–E). The four lateral petal primordia grow out independently, while the anterior petal always grows first (Fig. 2F). By common zonal growth at the base of the corolla, the petals fuse postgenitally (Fig. 2G). Next, five stamen primordia are initiated on an inner whorl, alternating with the corolla (Fig. 2H). They arise simultaneously, but due to differences in growth rate among the stamens, the androecium develops a zygomorphic appearance (Fig. 2I). The stamens on the abaxial side develop somewhat faster (Fig. 2I). While the developing stamens curve inward, the anthers have started to differentiate (Fig. 3A). The nearly triangular anthers are dithecal and tetrasporangiate; the connective has a papillose surface. During their development, the anthers become closely associated with each other and eventually are connivent (Figs. 3B, 4A). The sporangia of adjacent anther lobes will fuse before the pollen is released through a slit-like opening at the top of the anthers (Fig. 4A). On the floral apex, below the androecium, the gynoecium is initiated (Fig. 3C). Five locules are defined by the inwardly growing septa, which fuse at the center only at the base (Fig. 3C). The ovary closes at the top, and five stigma lobes become apparent (Figs. 3D, 4B). The young gynoecium as a whole has a barrel-like appearance, because a style seems to be lacking (Fig. 3D–E). The stigmas show little differentiation and can hardly be distinguished from the rest of the gynoecium (Fig. 3E). In each of the five locules of the spindle-shaped, superior ovary, 10–15 unitegmic ovules develop; they are anatropous and possess a long funicle (Figs. 3F–K, 4C–E). The ovules within a locule are arranged in one vertical row (Fig. 4C) and develop in a basipetal order (Fig. 3F–H). Although they are all attached above each other (Figs. 3I, 4C), they will develop alternately to the left and to the right occupying the available space in the locule in the most optimal way (Fig. 3F–H, J–K). In the young gynoecium we can see that the normal looking septa form massive separations between the locules (Fig. 3F– G). However, when looking at older stages, we can observe that the septa disintegrate in their outer parts: both walls of the locules become separated from each other (Figs. 3F, J–K, 4D–F). The fused inner margins of the carpels fall apart as the tissue between them disintegrates (Figs. 3J–K, 4E–F). In sections through the ovary, we can see that this disintegration is restricted to the peripheral parts of the septa, toward the ovary wall; more to the center, the massive structure of the septa is retained (Figs. 3I, 4C–D). It is striking that the massive parts of the septa are characterized by a high abundance of raphide bundles (Fig. 4D; arrowed). Figure 3I shows that some of the septa have already ruptured, breaking the connection between the central column and the ovary wall. When compared to the ovary wall, the septa appear to be rather thin and relatively delicate structures (Figs. 3I, 4E–F). We note that I. niamniamensis is characterized by the presence of a sepal spur that is slightly bilobed at the tip (Fig. 3L).
|
|
|
|
|
|
|
In the past, Balsaminaceae were often placed in Geraniales (e.g., Cronquist, 1981
) or in Sapindales (e.g., Scholz, 1964
). Now, its position in the Ericales s.l. is well supported (Savolainen et al., 2000
; Soltis et al., 2000
; Anderberg et al., 2002
; Bremer et al., 2002
; APG, 2003
; Geuten et al., 2004
). Although Impatiens is a popular pot plant and garden ornamental, few species are cultivated, and generally speaking, the genus is little studied. In particular, I. balsamina L., I. glandulifera Royle, I. hawkeri, and I. walleriana Hook. f. are widely grown (Wood, 1975
; Grey-Wilson, 1983
).
The floral structure, and in particular the perianth parts, vary, not only in color, but also in shape: characters such as the anterior petal, the lateral petals, and especially the spurred sepal are extremely variable, even within the same species (Hooker and Thomson, 1859
). The rich variation in the floral structure may be linked to coadaptation with pollinators. Several authors (e.g., Grey-Wilson, 1980a
; Travers et al., 2003
), for instance, find evidence of a relationship between nectar spur curvature and different sets of pollinators. On the other hand, Wilson (1995)
suggests that selection for more successful visitation based on pollinator behavior might be much more important with respect to evolution of floral characters than adaptations to improve the mechanical fit between pollinator and flower.
Organization of the perianth
As long as the flowers of Impatiens have been studied, the precise relationships among the different floral parts have remained unclear (Grey-Wilson, 1980c
). Relationships are blurred by the resupination of the flowers and the use of different descriptive terms regarding the position of the floral parts. In the past, several hypotheses have been postulated (for an overview of the old literature, see Payer, 1857
). Generally, it is assumed that in most Impatiens species the number of sepals is reduced to three. However, from our data from I. columbaria, a species not particularly known as having a pentamerous calyx (Grimshaw, 1998
), five sepals appear to be initiated early in the ontogeny. In later stages, the anterolateral sepals are no longer observable.
Röper (1830)
was the first author who stated that the perianth of Impatiens consists of a calyx and a corolla that both are pentamerous. Payer (1857)
studied the floral development of I. glandulifera and confirmed, like most of his colleagues, the findings of Röper (1830)
. According to Payer (1857)
, the two anterolateral sepals often stay rudimentary or disappear in later stages. Although we have no doubts about the conclusions of Payer (1857)
, we would like to stress that the rudimentary sepals are not initiated in all Impatiens species, as is clearly shown by our results from I. niamniamensis and I. hawkeri. Here, we can only observe an enlargement of the floral apex on the abaxial side, but anterolateral sepals are not initiated.
Warburg and Reiche (1895)
postulate that the initiation of the calyx follows a 2/5-spiral. According to these authors, only three sepals characterize most Impatiens species; the development of the third and the fifth are suppressed. If present, they are visible as small structures that have shifted towards the median axis (Warburg and Reiche, 1895
). As opposed to what one would expect, the anterolateral sepals (sepals three and five) are initiated last, after the appearance of the posterior sepal (sepal four). Warburg and Reiche (1895)
remark that in species with five sepals, the fourth one develops before the third one, indeed, which—according to them—may be the result of the rudimentary nature of the third and the fifth sepal. Our results support their descriptions.
Grey-Wilson (1980c)
studied the floral anatomy of Impatiens and found that some species contain rudiments of the vascular traces of the often-missing sepal pair. From their position, he concludes that the anterolateral pair has been reduced. This is supported by the fact that in species with five sepals, the anterolateral pair is always smaller and thinner than the posterolateral one; moreover, it is positioned more to the inside (Grey-Wilson, 1980c
). Our ontogenetic results from I. columbaria confirm the conclusions of Grey-Wilson (1980c)
: the rudimentary sepals develop in anterolateral position, and they are inserted somewhat higher up on the floral apex (more to the inside) when compared to the posterolateral sepals. They are also closely connected with the adjacent anterior petal. On top of that, the moment of initiation of the anterior petal differs from what one would expect: the anterior petal develops simultaneously with or immediately after the initiation of the posterior sepal, and commonly, well before the lateral petals appear.
The "disappearance" of the rudimentary sepals during ontogeny in some species can be explained through postgenital fusion with the adjacent anterior petal (which is often positioned on the same whorl) or through resorption by the further developing receptacular tissue. In species lacking the anterolateral sepal primordia, we could speak of a similar kind of fusion, but then it would be congenital. We believe that in the course of evolution the anterolateral sepals do not actually disappear, but gradually fuse with the anterior petal. As a result, in some species this is not a petal in the strict sense, because it is composed of parts from the calyx and the corolla that have become united. The often partly sepaloid appearance of this organ (which for convenience we will further describe as the anterior petal) supports this hypothesis (already formulated by Ramadevi and Narayana, 1989
), as does the fact that the anterolateral petals, when present, are positioned more to the inside of the flower.
Wood (1975)
mentions (but rejects) an old interpretation of the perianth, which was thought to have four sepals and four petals. The fourth sepal (in fact the anterior petal) was reported to have a petaloid appearance and had an incision in the center. According to this interpretation, the incision suggests that this organ is in fact composed of two fused sepals. The latter is interesting with respect to our hypothesis in which the anterior petal is composed of several perianth parts as well. Floral anatomy also supports our hypothesis because the vascular traces for the anterolateral sepal pair are found at both sides (and only slightly to the outside) of the trace for the anterior petal (Grey-Wilson, 1980c
). Ramadevi and Narayana (1989)
studied the floral anatomy of Impatiens and found that from the ring-like vascular tissue in the pedicel first the traces to the posterolateral sepals diverge. Next, again two traces diverge, one leading to the posterior sepal, the other one splitting up into three bundles for the anterior petal and the anterolateral sepals, all three of them entering the composed perianth part formed by the union of the anterior petal and the anterolateral sepals (Ramadevi and Narayana, 1989
). In I. elegans Bedd., they observed that the vascular traces of the lateral petals, and the common trace for the composed perianth part are arranged on the same whorl (Ramadevi and Narayana, 1989
).
Undoubtedly, the ontogeny of many species is insufficiently known. Hence, we expect that many other species will show rudiments of the anterolateral petals in their early ontogenetic stages. Furthermore, in I. hawkeri we occasionally observed rudimentary anterolateral sepals, even in mature flowers, whereas most flowers do not have any signs of these sepals. Likewise, Hooker and Thomson (1859)
already mentioned that anterolateral sepals may be absent or present within the same species. Some species appear to be very plastic regarding the development of anterolateral sepals, or in other words, regarding the degree of congenital fusion of these sepals with the anterior petal.
Fusion of the lateral petals
The different fruit type and the presence of five free petals morphologically separate Hydrocera from Impatiens. Within the family, free petals are considered to be plesiomorphic. Impatiens is generally characterized by the presence of lateral petals connate in pairs. Nevertheless, the degree of fusion often varies among species (and maybe within species as well). In the past, it was unclear if these were two bilobed petals rather than four petals connate in pairs (Grey-Wilson, 1980c
). According to Warburg and Reiche (1895)
, most species possess three petals, and they explain this feature by assuming that from the theoretically five petals, the four lateral ones are fused. They find evidence for this hypothesis in the presence of a central incision in the upper margin of the bilobed structures. However, Grey-Wilson (1980c)
proved on the basis of floral anatomy that each of the petals has its own independent vascular trace and that all petals develop separately. Our floral ontogenetic results support this conclusion: the lateral petals are formed from four separate petal primordia. We think that the fusion between the lateral petals on either side of the flower is insufficiently studied to use it as a morphological character to separate Hydrocera from Impatiens. It cannot be excluded that future studies will reveal Impatiens species with five free petals as well, when the enormous diversity in shape, size, and fusion of the lateral petals is taken into account. On top of that, petals of Impatiens always develop from separate primordia and as far as known, they are never fused congenitally. A lot depends, of course, on where the borderline between "free" and "fused" is drawn. According to Wood (1975)
, for example, the lateral petals of I. walleriana are mainly free and only slightly fused at the very base. The situation in this species might as well be stated as free (Warburg and Reiche, 1895
).
Androecium
The zygomorphy present in the corolla is found in the androecium as well; as we described, the anterior stamens grow larger than the posterior ones. The scale-like appendages on the inner and upper side of the filaments form a kind of cap that partly covers the gynoecium. The five stigmas are coherent and in many species, they only spread after the androecium has been dropped.
The anthers lie closely together and adjacent anther lobes fuse (Raghuveer and Narayana, 1994
). The four sporangia that are involved, merge to form a common space, which contains the pollen of both thecae (cf. Fig. 4A; arrow). The pollen is released through a slit at the top of this common space via the pressure created by the swollen cells in the connective region of the anther (Loew, 1892
; Warburg and Reiche, 1895
). The pollen is presented on the depression enclosed by the edges of the connivent anthers, the so-called Pollenstreufläche (Loew, 1892
). Due to the resupination of the flower, the Pollenstreufläche is positioned below the stigmatic region. As a result, self-pollination is avoided (Loew, 1892
). When the pollinators search for nectar in the sepal spur, they are loaded with pollen from the Pollenstreufläche. According to Loew (1892)
, in some species in which the stigmas do not open, pollen from another flower on the head of the pollinators is positioned in the pollination chamber (cf. Loew, 1892
), which is reached by a slit between the anterior stamens. The stamen appendages form a small crown or funnel, with five lobes catching the pollen. The so-called pseudostigmas (Loew, 1892
) bring the pollen into the neighborhood of the stigmas. In many other species, the androecium is shed, and the coherent stigmas spread and expose their receptive surface, as we mentioned before. According to Warburg and Reiche (1895)
, the stamen appendages might as well have a function in avoiding self-pollination.
Gynoecium
Traditionally, the gynoecium is always considered to be five-carpellate, as is the case in the species studied here. However, Shimizu and Takao (1982)
have shown that some species of Impatiens possess tetramerous gynoecia with four-locular ovaries.
In general, the style is reported to be very short or missing: a clear distinction between style and ovary cannot be observed. Nevertheless, the septa are not fused in the upper part of the ovary, i.e., the hemisymplicate zone sensu Leinfellner (1950)
. This zone has no ovules and might as well be considered to be a stylar zone. The situation could easily be compared with that of a lot of other Ericales, in which the septa continue in the internal lobes of the style, leaving a central, stylar canal with as many branches as locules. Shimizu and Takao (1982)
mention a stylar canal for all species they have investigated. Ramadevi and Narayana (1989)
describe the ovary as five-locular in the ovule-bearing part and unilocular at the top because of the presence of incomplete septa there. It can be argued that the lobes on the top of the gynoecium do not represent five individual stigmas, but rather one composed stigma, consisting of several lobes. Similar cases can be found in, for example, Ericaceae, where the stylar lobes protrude at the surface of the stigma.
Boesewinkel and Bouman (1991)
investigated ovule development in Impatiens. They conclude that ovules are bitegmic, unitegmic, or intermediate. The changeover from bitegmic (plesiomorphic) to unitegmic (apomorphic) is caused by the fusion of the dermal integument initials together with a shift and a growth restriction of the outer integument primordium (Boesewinkel and Bouman, 1991
). Due to the fusion, a common zone develops in intermediate species, which is only divided in two individual integuments at the top (Boesewinkel and Bouman, 1991
). Nonetheless, completely unitegmic species also occur, for instance, I. niamniamensis studied here. Unitegmic ovules are particularly typical of sympetalous groups (Boesewinkel and Bouman, 1991
). The combination of bitegmic and tenuinucellate ovules, as is commonly found within Impatiens, is less widespread. It is interesting that this situation is also typical within another group of Ericales s.l., namely, the primuloid clade.
Ovule arrangement
Shimizu and Takao (1982)
distinguish between uniseriate and biseriate arrangement of the ovules within a locule. Species with both types occur as well. Moreover, these species often possess intermediate types of arrangement, making the exact insertion of the ovules difficult to judge (Shimizu and Takao, 1985
). According to the same authors, a reductive trend can be followed in the number and the position of the ovules: from several to one ovule or none and from biseriate to uniseriate (Shimizu and Takao, 1982
). Because of the alternation in the arrangement of the ovules, uniseriate species may seem to have biseriately arranged ovules in transverse sections, and on the other hand, transverse sections of biseriate species might have only one row of ovules.
Shimizu et al. (1996)
studied six species from the subgenus Acaulimpatiens Warb. and found that the ovules in a locule are arranged in four to eight vertical rows. According to Shimizu et al. (1996)
, this character does not occur outside this subgenus, with the exception of I. siamensis T. Shimizu, in which the ovules are arranged in three to four rows. Nevertheless, it is evident from our results that in I. columbaria, ovules are arranged in several series as well. Up to the present, nothing is known about the fruit of I. columbaria (Grimshaw, 1998
). Most probably, future studies will reveal that also other species of Impatiens show this feature.
Fruit dehiscence
The particular dehiscence mechanism of the fruits is made possible by the disintegration of the internal tissue of the septa. It is unknown whether the raphide bundles, which are, in fact, needle-shaped crystals of calcium oxalate, play a role in the disintegration of the septal tissue. Possibly, oxalic acid is formed, which might be able to destroy the cellulose walls in the outer parts of the septa. A similar process occurs in the disjunctive tissue of the anthers of some Ericaceae (Matthews and Knox, 1926
). As a result of the particular development of the septa, the presence of a completely septate ovary with axile placentation is not an obstacle to explaining the explosive dehiscence of the fruit. Wood (1975)
describes the septa of Impatiens as delicate structures that are compressed by the developing ovules. The explosive dehiscence of the fruit is the result of the combination of an outer, highly turgescent epidermis tissue and swollen parenchyma cells beneath the epidermis, and the nonturgescent tissue further in (Warburg and Reiche, 1895
). Raghuveer et al. (1993)
studied the anatomy and dispersal of what they called the dehiscent fruit of Hydrocera triflora (L.) Wight & Arn. They maintained that the five-seeded fruit is a capsular berry that opens septicidally: the wall of the imbibed fruit (the fruits float on and are dispersed by water) normally splits from the base upwards along the radii of the septa (Raghuveer et al., 1993
).
Special adaptations of the flower related to protandry, the particular floral structure with its zygomorphy and deviations in the perianth organization, along with the exceptional opening mechanism of the fruit linked to the special morphology of the ovary, make the flower of Impatiens a highly modified structure. Moreover, floral diversity in the genus is extremely high. In the present study, we investigated three species of Impatiens in order to comment on general trends in flower morphology and evolution. Nevertheless, detailed morphological and anatomical studies of a broader range of species are needed to contribute to a better understanding of floral evolution and fruit morphology in this species-rich genus.
FOOTNOTES
The authors thank M. Verhaegen for technical assistance with the SEM observations at Meise and the director of the National Botanic Garden of Belgium for floral material. This research is supported by a grant from the Research Council of the K.U.Leuven (OT/01/25) and the Fund for Scientific Research–Flanders (Belgium) (G.0268.04; 1.5.003.05N). 
2 Author for correspondence (e-mail: pieter.caris{at}bio.kuleuven.be
)
LITERATURE CITED
Akiyama S. H. Ohba M. Wakabayashi 1991 Taxonomic notes of the east Himalayan species of Impatiens. Studies of Himalayan Impatiens (Balsaminaceae). I. In H. Ohba and S. B. Malla [eds.], The Himalayan plants, vol. 2, 67–94. University of Tokyo Press, Tokyo, Japan
Akiyama S. H. Ohba 2000 Inflorescences of the Himalayan species of Impatiens (Balsaminaceae). Journal of Japanese Botany 75: 226-240
Anderberg A. A. C. Rydin M. Källersjö 2002 Phylogenetic relationships in the order Ericales s.l.: analyses of molecular data from five genes from the plastid and mitochondrial genomes. American Journal of Botany 89: 677-687
APG (Angiosperm Phylogeny Group). 2003 An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG II. Botanical Journal of the Linnean Society 141: 399-436[CrossRef]
Boesewinkel F. D. F. Bouman 1991 The development of bi- and unitegmic ovules and seeds in Impatiens (Balsaminaceae). Botanische Jahrbücher für Systematik, Pflanzengeschichte und Pflanzengeographie 113: 87-104
Bremer B. K. Bremer N. Heidari P. Erixon R. G. Olmstead A. A. Anderberg M. Källersjö E. Barkhordarian 2002 Phylogenetics of asterids based on 3 coding and 3 non-coding chloroplast DNA markers and the utility of non-coding DNA at higher taxonomic levels. Molecular Phylogenetics and Evolution 24: 274-301[CrossRef][ISI][Medline]
Cronquist A. 1981 An integrated system of classification of flowering plants. Columbia University Press, New York, New York, USA
Fischer E. 2004 Balsaminaceae. In K. Kubitzki [ed.], The families and genera of vascular plants, vol. 6, Flowering plants: Dicotyledons. Celastrales, Oxalidales, Rosales, Cornales, Ericales, 20–25. Springer, Berlin, Germany
Geuten K. E. Smets P. Schols Y.-M. Yuan S. Janssens P. Küpfer N. Pyck 2004 Conflicting phylogenies of balsaminoid families and the polytomy in Ericales: combining data in a Bayesian framework. Molecular Phylogenetics and Evolution 31: 711-729[CrossRef][ISI][Medline]
Grey-Wilson C. 1980a Impatiens of Africa. A. A. Balkema, Rotterdam, Netherlands
Grey-Wilson C. 1980b Impatiens in Papuasia. Studies in Balsaminaceae: I. Kew Bulletin 34: 661-688[CrossRef]
Grey-Wilson C. 1980c Some observations on the floral vascular anatomy of Impatiens. Studies in Balsaminaceae: VI. Kew Bulletin 35: 221-227[CrossRef]
Grey-Wilson C. 1983 A survey on Impatiens in cultivation. Plantsman 5: 86-102
Grimshaw J. M. 1998 Impatiens columbaria. Curtis's Botanical Magazine 15: 37-41
Hooker J. D. T. Thomson 1859 Praecursores ad floram Indicam. Balsaminaceae. Botanical Journal of the Linnean Society 4: 106-157
Leinfellner W. 1950 Der Bauplan des synkarpen Gynözeums. Österreichische Botanische Zeitschrift 97: 403-436[CrossRef]
Loew E. 1892 Der Blütenbau und die Bestäubungseinrichtung von Impatiens Roylei Walp. Botanische Jahrbücher für Systematik, Pflanzengeschichte und Pflanzengeographie 14: 166-182
Matthews J. R. E. M. Knox 1926 The comparative morphology of the stamen in the Ericaceae. Transactions of the Botanical Society of Edinburgh 29: 243-281
Payer J. B. 1857 Traité d'organogénie comparée de la fleur. Victor Masson, Paris, France
Raghuveer M. L. L. Narayana B. S. M. Dutt 1993 Dehiscent fruit of Hydrocera triflora (Linn.) Wt. & Arn. (Balsaminaceae): its anatomy and dispersal. Rheedea 3: 12-14
Raghuveer M. L. L. Narayana 1994 Embryology of Balsaminaceae: I. Feddes Repertorium 105: 23-29
Ramadevi D. L. L. Narayana 1989 Floral anatomy of Balsaminaceae. In M. L. Trivedi, B. S. Gill, and S. S. Saini [eds.], Plant science research in India, 707–713. Today & Tomorrow's Printers & Publishers, New Delhi, India
Röper J. A. C. 1830 De floribus et affinitatibus Balsaminearum. [Publisher unknown], Basel, Switzerland
Savolainen V. M. W. Chase S. B. Hoot C. M. Morton D. E. Soltis C. Bayer M. F. Fay A. Y. de Bruijn S. Sullivan Y. L. Qiu 2000 Phylogenetics of flowering plants based on combined analysis of plastid atpB and rbcL gene sequences. Systematic Biology 49: 306-362[CrossRef][ISI][Medline]
Scholz H. 1964 Sapindales. In H. Melchior [ed.], A. Engler's Syllabus der Pflanzenfamilien II, 277–288. Borntraeger, Berlin, Germany
Shimizu T. S. Takao 1982 Taxonomic significance of the inner structure of the ovary in the genus Impatiens (Balsaminaceae). Botanical Magazine Tokyo 95: 89-99[CrossRef][ISI]
Shimizu T. S. Takao 1985 Taxonomic discussions on the four-carpellate species of Impatiens (Balsaminaceae). Acta Phytotaxa Geobotanica 36: 97-106
Shimizu T. S. Takao N. Utami A. Takao 1996 Anatomy of floral organs and its taxonomic significance in the genus Impatiens (Balsaminaceae), with special reference to subgen. Acaulimpatiens. Phytomorphology 46: 253-266
Soltis D. E. P. S. Soltis M. W. Chase M. E. Mort D. C. Albach M. Zanis V. Savolainen W. H. Hahn S. B. Hoot M. F. Fay M. Axtell S. M. Swensen L. M. Prince W. J. Kress K. C. Nixon J. S. Farris 2000 Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnean Society 133: 381-461[CrossRef]
Stevens P. F. 2004 Angiosperm phylogeny website [online, version 5, May 2004], http://www.mobot.org/MOBOT/research/APweb/
Takhtajan A. 1997 Diversity and classification of flowering plants. Columbia University Press, New York, New York, USA
Travers S. E. E. J. Temeles I. Pan 2003 The relationship between nectar spur curvature in jewelweed (Impatiens capensis) and pollen removal by hummingbird pollinators. Canadian Journal of Botany 81: 164-170[ISI]
Warburg O. K. Reiche 1895 Balsaminaceae. In H. G. A. Engler and K. A. E. Prantl [eds.], Die natürlichen Pflanzenfamilien, Teil 3, Abteilung 5, 383–392. Engelmann, Leipzig, Germany
Wilson P. 1995 Selection for pollination success and the mechanical fit of Impatiens flowers around bumblebee bodies. Biological Journal of the Linnean Society 55: 355-383[CrossRef]
Wood C. E. Jr. 1975 The Balsaminaceae in the southeastern United States. Journal of the Arnold Arboretum 56: 413-426[ISI]
Yuan Y.-M. Y. Song K. Geuten E. Rahelivololona S. Wohlhauser E. Fischer E. Smets P. Küpfer 2004 Phylogeny and biogeography of Balsaminaceae inferred from ITS sequences. Taxon 53: 391-403[CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  Flowering Newsletter bibliography for 2006 J. Exp. Bot., April 20, 2007; (2007) erm028v2. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
