| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
2Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs, Connecticut 06269-3043, USA; 3Instituto Multidisciplinario de Biología Vegetal (CONICET-Universidad Nacional de Córdoba), Casilla de Correo 495, 5000 Córdoba, Argentina; 4Jardín de Aclimatación de La Orotava (ICIA), Retama 2, E-38400 Puerto de La Cruz, Tenerife, Canary Islands, Spain
Received for publication February 6, 2006. Accepted for publication June 15, 2006.
ABSTRACT
We confirmed functional dioecy of Withania aristata via field and greenhouse studies. Male flowers are significantly larger. Female flowers bear stamens with no pollen; males bear 220 000 grains. Stigmata of male flowers senesce in buds. Anatomical observations confirm more ovules in females and an ovarian nectary in both sexes. We detected nectar in female flowers in the greenhouse but found no nectar in males. Thus, males offer pollen and females nectar. Females bear large numbers of fruits and, infrequently, male plants bear few significantly smaller fruits with few seeds. Outcrosses of females (self crosses impossible without pollen) yielded fruits in young buds, older buds, and open flowers. Self crosses of male flowers succeeded only with very young buds. Although functionally dioecious, this species manifests self-compatibility; however, no fruits are produced autonomously. Bee species (Lassioglossum, Amegilla, Apis) visit flowers and mature buds. Bud visits in which bees force petal tips apart, coupled with self-compatibility, may explain infrequent fruit on males. Thus, dioecy in W. aristata seems to have evolved from self-compatible ancestors, that leaky dioecy may have been favored during colonization, and, that despite autogamy and a low floral visition rate, this endemic enjoys a high rate of reproductive success.
Key Words: bees • Canary Islands • dioecy • flower morphology • leaky dioecy • reproductive biology • self compatibility
Studies of reproductive biology inform the nature of species, adaptation, speciation, hybridization, and systematics (Ornduff, 1969
; Anderson et al., 2002
; Neal and Anderson, 2005
). The Solanaceae shows a wide adaptive radiation that includes all forms of zoophily, i.e., its species can be pollinated by birds (e.g., Cocucci, 1999
; Kaczorowski et al., 2005
), moths (e.g., Arroyo and Squeo, 1990
; Vesprini and Galetto, 2000
; Raguso et al., 2003
), butterflies (e.g., Cocucci, 1995
, 1999
), bats (e.g., Voss et al., 1980
; Helversen, 1993
; Sazima et al., 2003
), bees (e.g., Anderson and Symon, 1988
; Sazima et al., 1993
; Bohs, 2000
), and flies (e.g., Galetto et al., 1998
; Cocucci, 1999
). Similarly diverse are the floral rewards, which include nectar (e.g., Galetto and Bernardello, 1993
, 2003
), pollen (e.g., Symon, 1979
; Lester et al., 1999
; Connolly and Anderson, 2003
), scents (e.g., Sazima et al., 1993
; Passarelli and Bruzzone, 2004
), and oil (e.g., Simpson and Neff, 1981
; Cocucci, 1991
). Within Solaneae, the solanaceous tribe with the most genera and species (Hunziker, 2001
), about 75% of the genera are bee-pollinated (Cocucci, 1999
), and, as in the Solanaceae generally, most flowers are hermaphroditic. However, the majority of the reproductively unusual dioecious taxa occur in this tribe (with the exception of Symonanthus from another subfamily: Anthocercidoideae; Haegi, 1981
; Hunziker, 2001
). The few reported cases of dioecious species are in the genera Solanum (e.g., Levine and Anderson, 1986
; Anderson and Symon, 1989
; Knapp et al., 1998
), Deprea (e.g., Sawyer and Anderson, 2000
), Dunalia (e.g., Hunziker, 2001
), Lycium (e.g., Minne et al., 1994
; Miller and Venable, 2002
), and Withania (Hepper, 1991
; Hunziker, 2001
).
The Solanaceae are not common on islands (Wagner et al., 1990
; Marticorena et al., 1998
; McMullen, 1999
); thus the biology, systematics, and natural history of the insular species are of particular interest. Herein, we address the reproductive biology for the single endemic species of Withania on the Canary Islands, Withania aristata (Aiton) Pauq., for which the reproductive biology turns out to be much more interesting than it initially appears (e.g., Bramwell and Bramwell, 2001
).
Withania, a small genus of 10–18 species—depending on species and generic boundaries—ranges from the Canary Islands, the Mediterranean region and northern Africa to India, China, and Japan (Hepper, 1991
; Hunziker, 2001
). Morphological and molecular data have generally indicated a systematic position among the physaloid genera in subfamily Solanoideae (Axelius, 1996
; Olmstead et al., 1999
).
The flora of the Macaronesian Canary Islands includes three Withania species (Bramwell and Bramwell, 2001
). Two of these—W. frutescens Pauquy and W. somnifera (L.) Dunal—are introduced continental species. In fact, W. somnifera is extremely wide-ranging, from the Canary Islands and Europe to India and Australia (Hepper, 1991
). The lone native species (locally known as "orobal") is the endemic W. aristata (Fig. 1), a frequent soft-wooded shrub found at low elevations on all the Canary Islands (Bramwell and Bramwell, 2001
). Withania species, especially W. somnifera used in ethnobotanical practices (e.g., Chevallier, 1996
), has been studied intensively because of its medicinal properties. Thus, although the reproductive biology of W. aristata has not previously attracted attention, the fruits, leaves, and bark are reported to have multiple uses in traditional medicine on the Canary Islands (Darias et al., 2001
).
|
, the Asian W. coagulans was the first member of the Solanaceae documented to be dioecious. Although the dioecy was not described in detail, it seems clear from the precise drawings of this species by Wight and Arnott (1850)
and Hooker's (1852
, p. 801) indication that it has "flores abortu dioeci," that some flowers bear functional stamens, and other plants have flowers with only shriveled, presumably nonfunctional stamens. Based on examination of herbarium collections, Hepper (1991)
proposed W. coagulans as dioecious as well, with plants bearing either male flowers with short styles or female flowers with short, sterile anthers and long styles. Withania adpressa, from northwestern Africa, shows a similar dichotomy between male and female flowers, but Hepper (1991)
was uncertain whether this species should be treated as monoecious or dioecious. The only article focusing on the reproductive biology of a Withania species (W. somnifera, Kaul et al., 2005
) indicates that it has hermaphroditic flowers and is autogamous.
The reproductive biology of W. aristata presents a number of puzzling features. Flowers have been described as being either unisexual or hermaphrodite, with unclear differences in the fruit set of different individuals (Webb and Berthelot, 1845
; Hepper, 1991
; Bramwell and Bramwell, 2001
; Hunziker, 2001
). As part of a broader survey of the reproductive biology of many Canarian species (Anderson et al., 2005
), our first studies of this endemic were confusing. Anthers seemed to dehisce and shrivel in many flowers, fruit set varied dramatically among plants, and the infrequent insect visits were by bees that only visited buds.
Thus, we were intrigued to turn our full attention to elements of the natural history of this notable and unusual island endemic. Here, via field and glasshouse analyses, we report detailed studies of the reproductive biology and mating system of this interesting taxon. We show that this solanaceous genus too, manifests some of the notable reproductive variation that characterizes an increasing number of carefully studied species in this family. Not only are the Solanaceae important economically (Heiser, 1987
), but, surprisingly in many ways, they turn out to be remarkable in terms of reproductive biology as well, providing a number of reproductive variations that serve as examples of the benefit of detailed studies.
MATERIALS AND METHODS
Seven natural populations of W. aristata (Table 1) were studied in detail on Tenerife (Canary Islands, Spain) in January 2003, January and June 2004, and May 2005; vouchers have been deposited in the University of Connecticut G. S. Torrey Herbarium (CONN). For anatomical studies (accessions 5031, 5039, and 5061, Table 1), flowers were fixed in 70% ethanol, dehydrated in an ethanol-xylol series, and embedded in Paraplast (Oxford Labware, St. Louis, Missouri, USA). Serial cross and longitudinal sections were cut at 10 µm, mounted serially, and stained with safranin-fast green-hematoxylin and observed with a compound microscope.
|
|
|
Counts of pollen grains and ovules per flower, as well as estimates of pollen viability, were obtained from flower buds preserved in 70% ethanol for three flowers per plant. The accessions examined (see Table 1) were 5031 (10 male plants), 5036 (one male plant), 5037 (10 female plants), 5038 (one male plant), 5039 (one male plant), and 5061 (one male plant). Pollen grains and ovules were counted using buds prior to anther dehiscence. Ovaries were gently squashed on microscope slides and their ovules counted. Pollen grain numbers per flower were estimated with a haemocytometer, using the methods of Lloyd (1965)
as modified by Anderson and Symon (1989)
. Viability of ca. 400 pollen grains per flower was tested by staining with 1% aniline blue in lactophenol. Nectar sugar concentration was measured with a Bausch and Lomb hand refractometer (Rochester, New York, USA).
Controlled pollinations, as well as additional counts of pollen grains, ovules, and estimates of pollen viability were carried out using plants grown in the greenhouse. Crosses were attempted utilizing all six plants in all possible combinations (given that males have pistils, but females produce no pollen): male x self, male x nonself male, and female x male. Each crossing combination was treated as a separate event, rather than pooling the data by plant. This approach is supported by the fact that the combinations generally succeed or fail; i.e., there are not gradations of crossing success (except for the female x male comparison among the three floral stages).
The fate of unmanipulated flowers in the pollinator-free greenhouse environment was also tracked. Because insects were observed forcing their way into unopened flower buds in nature, we considered the possibility of bud pollination in this species, and crosses were attempted at all three stages of floral development outlined above.
The diameter of fruits was measured at maturity, and seeds were extracted, counted, and then sown in soil-less potting mix. Seed pots were kept evenly moist in warm greenhouse conditions, and germination was assessed after 2 months. Additional germination trials were carried out with open-pollinated, field-collected seeds from the following accessions: 5036 (
), 5037 (), 5038 (
), 5039 (), 5060 (), and 5061 ().
Insects visiting W. aristata flowers or buds were observed, photographed, or collected. Observations were made in all wild populations in all field expeditions for a total of c. 100 plants for more than 100 h, at the sites given in Table 1. Periods of observation ranged from 10 min to 1 h during daylight hours (from 0900 to 1600 hours). In addition, from July to September 2005, two female and two male plants from the greenhouse-cultivated plants were placed in the garden of the Torrey Life Sciences Building (Storrs) and exposed to non-native North American pollinators for observations of bee behavior and fruit set. Insects were deposited in the Biological Collections at the Ecology and Evolutionary Department (University of Connecticut) and were identified by Francisco La Roche-Brier.
RESULTS
Floral features
Wild plants flower abundantly and simultaneously (both sexes) year round, with a peak in the northern hemisphere spring. Flowers are pentamerous, actinomorphic, and pendant (Figs. 1, 2) producing a weak odor that is similar to lilacs (genus Siringa, Oleaceae). Both buds and open flowers are green to chartreuse. The calyx is gamosepalous and campanulate with five long linear lobes (Figs. 1B, 2). The corolla is campanulate with five lobes as long as or slightly longer than the tube (Figs. 1B, 2).
A superficial examination of the flowers suggests that there are no sexual differences, thus perhaps explaining the treatment in floras as regular hermaphrodite flowers. However, both male and female plants can be recognized with striking differences in fruit production among individuals. Effectively, female plants bear many fruits at a time, whereas male plants have a large number of flowers, but bear no fruits (with unusual exceptions, see Fruit features). Although the corolla of male flowers is significantly larger, it is not an obvious feature, and the variation of corolla lengths masks any sexual difference. Thus, we conclude that no secondary sexual differences in architectural or other vegetative characters can be documented.
The sex ratio in wild populations is 1 : 1 (total of 126 male and 115 female plants, i.e., 52.3% vs. 47.7%; 
2-tests, P > 0.05, no significant differences from the expected ratio). We observed no signs of monoecy, e.g., female plants bearing some flowers with functional anthers, in either wild or greenhouse plants.
Even though both flower types have all floral whorls, male and female flowers can be differentiated. The outer whorls of male flowers are significantly larger than those of female flowers at the same developmental stage, with the calyx and corolla can be up to 40% longer in males (Table 2). But, as noted, the variation among flowers and plants is great enough that this is not a reliable character. In both flower types, stamens are included, equal in size, and inserted at the base of the corolla tube, forming a staminal column surrounding the ovary (Fig. 2B, C). In addition, anthers are tetrasporangiate and bithecal (Fig. 3A, B) having longitudinal dehiscence that occurs in stage 2 buds. The differences are related to the presence of pollen—anthers in female flowers are sterile and have no pollen (Fig. 3A, C), whereas stamens in male flowers have pollen (Figs. 3B, D)—and to statistical differences in anther length (longer in males in all developmental stages [Table 2]; supported as well by the analyses, not shown, with the data pooled by plant). In both sexes, anthers reach their maximum size early, notably in stage 1 buds, shrinking by the time the flowers open; this phenomenon is more marked in male flowers. In female flowers, anthers have no pollen (Fig. 3A, C) and, as soon as flowers open, their thecae become dry and brown. On the other hand, male flowers yield large numbers (more than 200 000) of pollen grains that are highly stainable (= viable) (Fig. 3B, D; Table 3). Some pollen is still present in open flowers with shriveled anthers.
|
|
All plants bear flowers with normal-looking ovules (Fig. 3E, F). Male and female flowers have a significantly different number of ovules, with females bearing more ovules (Table 3). The pollen/ovule ratio in males is about 17 200 (Table 3).
There is a circular nectary at the base of the ovary (Fig. 3C) in both sexes. However, nectar is very hard to observe in flowers from wild populations. In fact, in most flowers examined, we found no evidence of nectar. In contrast, nectar is easily detected in female flowers from greenhouse grown plants, but rarely in male flowers. Greenhouse-grown female flowers produce small amounts of nectar (<3 µL) with a mean sugar concentration of 36% (N = 25 flowers, four individuals).
Fruit features
The fruit is a globose berry, dark green when immature, turning orange-red when ripe, and surrounded throughout development by a conspicuous accrescent calyx (Fig. 1C). In greenhouse conditions, berries mature about 9 weeks after hand pollination. Fruits are often shed with the dried calyx and the pedicel still attached. Sporadically, the calyx expands and persists for weeks, or even months, around a presumably unpollinated ovary. Such false fruit set seems to be especially likely on vigorously-growing shoots on female plants.
By late spring (i.e., May) on Tenerife, females bear large numbers of fruits (Table 4). Infrequently, some males produce fruits (Table 4). When male and female fruit sizes and seed numbers are statistically compared, female fruits are larger and bear more seeds (Table 4).
|
|
Unmanipulated flowers in pollinator-free greenhouse conditions never yielded fruits, in either females (N = 750 flowers) or males (N = 600 flowers) in the two successive years of cultivation. Cultivated plants exposed to pollinators in a Connecticut garden for three summer months (July to August) were visited by wasps and small bees common in the area. In this environment, the exposed female plants set copious fruit, but the male plants set none.
Seeds produced by all plant types on Tenerife are viable and germinate, as do seeds from all greenhouse crosses (Table 6). In general, germination rates are moderate, reaching a maximum of 70%. However, seeds field-collected from male plants showed significantly lower germination rates (27%; Table 6). There are no statistical differences among the rates of seed germination from field-collected female plants and artificial crosses of female or selfed males (Table 6). Preliminary observations of young plants derived from these experimental sowings reveal no discernable effects of cross type on plant size or apparent health.
|
|
Floral morphology and anatomy and results from experimental crosses support the conclusion that W. aristata is dioecious. David Symon called Hepper's attention to the possibility that this species was dioecious, based on informal observation of cultivated plants in Australia (see Hepper, 1991
). In spite of these hints at the unusual sexual system in Withania, detailed studies have not followed.
Both male and female flowers have all whorls, but there are striking differences between them. Male flowers bear a much larger calyx, corolla, and anthers, and female flowers bear anthers, but they are devoid of pollen. Although the ovaries show no differences in the first stage of bud growth, female flowers subsequently develop longer styles. Most important, female stigmata are functional during anthesis, including from the early bud stage we identified, but in male flowers stigmata become nonfunctional well before flowers open—hence, their treatment as males. Plants with male flowers do not produce fruits, but exceptionally, fruits are found on male plants in the field in very low quantities. These observations are in accordance with the results of experimentally selfed or outcrossed males, where hand pollinations of flowers at an early bud stage (before the stigma and style wither) yielded around 50% fruit set. Thus, W. aristata is self compatible, in accordance at least with most other dioecious and andromonoecious solanums known (Anderson and Symon, 1989
). The high pollen/ovule ratio for male flowers supports Cruden's (1977)
category for obligate xenogamy, and is analogous to the pollen/ovule ratio for male flowers of another dioecious species, Solanum appendiculatum (Mione and Anderson, 1992
). Thus, the presumably residual self-compatibility reflects phylogenetic ancestry, but not current reproductive biology.
Although we did not calculate the flower/fruits ratio for plants in the field, our extensive observations support a high percentage of fruit set on female plants. Given the impossibility of selfing in female flowers (no pollen), the lack of support for autogamous (or apomictic) fruit set in experimental greenhouse settings of either sex, and, in spite of few floral visitors being observed at any given time, we conclude that these flowers enjoy a high rate of cross pollination. All visitors observed were bees, endemic, native, or introduced (Izquierdo et al., 2004
). The behavior of native bees forcing their way into un-opened (but mature) buds, coupled with the self-compatibility of W. aristata, may be responsible for the infrequent fruit set in male plants which, otherwise, does not occur spontaneously. For instance, in the two male, cultivated specimens that we grew for 2 years in greenhouses and in a gardens, we never obtained fruits without manipulation.
Little information is available for pollinators of other Withania species. The only report is for W. somnifera, a species with smaller, nonpendant, generalist flowers, visited in India by bees, butterflies, and flies (Kaul et al., 2005
).
Flowers of each gender may be considered as offering different rewards. Females offer nectar and males pollen. In greenhouse-grown plants, females always produced nectar and males did not. In other dioecious species, Eckhart (1999)
found that the quantity of nectar may differ between male and female flowers, but Eckhart's survey did not include Solanaceae. Interestingly, anatomical observations in both flower types suggest that the nectary should be functional. This species produces tiny amounts of nectar in comparison to other solanaceous, nectariferous flowers of the same size (e.g., Galetto and Bernardello, 1993
, 2003
). And, as noted, in the field, we essentially could not detect nectar production. Presumably, this lack of nectar is due to visitor activity or environmental conditions (it is dry and virtually always windy along the coast where the plants grow). The floral nectary is similar to those observed in other Solanaceae (Hunziker, 2001
).
Recent accounts suggest that about 10% of flowering plant species are dioecious, but these studies also show that dioecy has evolved repeatedly, being represented in nearly half of the angiosperm families (Renner and Ricklefs, 1995
; Gerber et al., 1999
; Webb, 1999
). Within the Solanaceae, cosexuality is common and dioecy is comparatively rare (Sawyer and Anderson, 2000
; Hunziker, 2001
). Gender dimorphism has evolved independently in six lineages, primarily in the tribe Solaneae (Sawyer and Anderson, 2000
) and has been reported for as few as six of the more than 90 genera that comprise the family and only about 20 species (i.e., less than 1% of the species). Interestingly, virtually all dioecious cases in the Solanaceae are referred to as functional dioecy (Anderson, 1979
; Anderson and Symon, 1989
) or cryptic dioecy (Mayer and Charlesworth, 1991
), as we report here for Withania. Clearly, functional or cryptic dioecy is a regular phenomenon apparently not just in Solanum, but in other members of the Solanaceae. It is important to understand this syndrome, and its extent, because the reproductive system has a clear impact on the systematics, diversity, and evolution of a lineage (Anderson et al., 2002
). The existence and nature of such cryptic systems stimulates the study of other taxa that might otherwise continue to be considered typically hermaphroditic. And, the anatomy, morphology, development, and distribution of characters may help us to understand the evolution of fundamental features of flowers such as pollen (e.g., Zavada and Anderson, 1997
; Zavada et al., 2000
) and pistils.
Our data suggest that dioecy in W. aristata has evolved from self compatible (SC) ancestors. In Withania, only the hermaphroditic W. somnifera was analyzed previously. It is also SC, but, in contrast, is highly selfing and has a low pollen/ovule ratio (817; Kaul et al., 2005
). Self compatibility is likely ancestral in Withania, as is true for much of Solanum (Whalen and Anderson, 1981
). In several plant groups, gender dimorphism is purported to have evolved from SC hermaphroditic ancestors, a fact that has been interpreted as a mechanism to promote outcrossing and to avoid self-fertilization and the resulting negative consequences of inbreeding depression (Lloyd, 1976
; Charlesworth and Charlesworth, 1978a
, b
; Anderson and Symon, 1989
; Sakai and Weller, 1999
). However, in some other Solanaceae, such as Lycium (Miller and Venable, 2002
) and possibly Deprea (Sawyer and Anderson, 2000
), gender dimorphism has evolved on a phylogenetic background of self-incompatibility (Richman and Kohn, 2000
). Unfortunately, no data are available on the mating systems of the other dioecious genera of the family (Dunalia and Symonanthus), so broader conclusions are hard to draw at this point. But, it is just such studies that are needed in order to provide the foundation for a generalized understanding of the evolution of mating and breeding systems.
Most dioecious Solanum, with the exception of S. appendiculatum, are derived from SC progenitors (Whalen and Anderson, 1981
; Levine and Anderson, 1986
; Anderson and Symon, 1989
; Knapp et al., 1998
). There are other similarities between Solanum and Withania sexual systems. The dioecy is cryptic in both, in the sense that the sexes are morphologically hermaphroditic, but functionally unisexual. The scent of both is weak or absent. Both are pollinated by bees, though Solanum flowers require "buzzing" to extract the pollen from poricidal anthers, whereas the pollen in Withania falls free of the longitudinally dehiscent anthers. Pollen is a significant reward in both, though it is the only reward in Solanum, as is the case for the male flowers of W. aristata. And the sex ratio is the same for both (1 : 1). There are a number of differences as well. The flowers in Withania are comparatively dull in coloration (green vs. white or yellow), Solanum flowers completely lack nectar in both sexes, and though most species of the dioecious solanums seem to have evolved on a platform of self-compatibility, males never bear seed-bearing fruits (G. J. A. observed apparent "fruits" on males of S. appendiculatum in the field in Mexico; however, these were all seedless, bearing only larvae of an ovary-parasitic insect).
There are associations—as documented by Renner and Ricklefs (1995)
for other dioecious species in general—between dioecy and various ecological and morphological traits in W. aristata. Withania is biotically dispersed, and has a shrubby growth form and a tropical distribution. Effectively, seed dispersal is endozoic; W. aristata fruits are reddish, fleshy, and edible (often bird-associated dispersal features). Two endemic lizards in the genus Galottia are reported to be natural dispersal agents in the Canaries (Valido and Nogales, 1994
, 2003
; Valido et al., 2003
). Valido and Nogales (1994)
also reported that seeds passed through a lizard gut had a significantly increased germination rate. Birds may be dispersers as well, considering the rich bird fauna in the Canary Islands (Delgado, 2001
) and the fruit features of W. aristata (e.g., orange fruits at maturity).
Some oceanic islands have a large proportion of endemic dioecious species (Bawa, 1980
; Givnish, 1982
; Sakai and Weller, 1999
), among which New Zealand and Hawaii are notable (Carlquist, 1974
; Sakai et al., 1995a
, b
; Webb, 1999
). In contrast with these two archipelagoes, the Canary Islands bear only 3% (Helfgott et al., 2000
), a percentage close to that of other islands such as the Azores, Galapagos, Reunion, Aldabra, and Bermuda (Baker and Cox, 1984
). The general argument for dioecy on islands is that selection for outcrossing in small, colonizing, hermaphroditic populations favors separation of the sexual functions (Carlquist, 1974
; Baker, 1967
; Bawa, 1980
; Thompson and Barrett, 1981
). Alternatively, the incidence of dioecy on islands may simply reflect its incidence in the source flora of the nearest continents (Baker and Cox, 1984
).
Withania aristata shows weak gender plasticity, in that a few male plants sometimes produce a very few fruits. This kind of dioecy—perhaps akin to subdioecy (see Delph and Wolf, 2005
) or leaky dioecy (Baker and Cox, 1984
)—has been reported in different plant families and island systems for several species (e.g., Baker and Cox, 1984
; Cox, 1990
; Weller et al., 1990
; Sakai and Weller, 1991
; Ladley et al., 1997
; Percy and Cronk, 1997
; Humeau et al., 1999
, 2000
; Litrico et al., 2005
). On one hand, "leaks" in the dioecious system, in an evolutionary sense, are probably not adaptive, but instead reflect the origins of dioecy, the strength (weakness) of the selection for full sexual separation, and/or the relative recency of the dioecious system. But on the other hand, leaky dioecy may facilitate the establishment after long-distance dispersal of colonizers with sexual variability (Baker and Cox, 1984
). The ability of isolated males to undergo occasional sexual reproduction may be important and heavily selected for, preventing the extinction of subpopulations that have become unisexual through drift (Percy and Cronk, 1997
). Apropos to W. aristata, we can only speculate, but we did observe one exemplar population where there was only one female plant (out of seven). Thus, the evolution of dioecy in W. aristata may be relatively recent, and gender plasticity may have been favored during colonization of the archipelago. Its founders may have expressed sexual variation when they first arrived on the islands. The existence of a related species from Morocco and Algeria (W. adpressa) with unisexual flowers suggests this possibility, although its reproductive system has not been studied (Hepper, 1991
). There are other explanations possible for the leakiness. For instance, Percy and Cronk (1997)
also suggest that when full dioecy is approached, the selection pressure for further loss of female function might be negligible.
As is well known, there are many more recorded vascular plant extinctions from islands than from continental areas (Reid and Miller, 1989
; Frankham, 1997
). Island species, generally with few populations and few individuals, are especially vulnerable to human-induced disturbance (e.g., direct predation, habitat degradation or loss, introduction of plant and animal exotic species, loss of pollinators; Groombridge, 1992
; Vitousek et al., 1995
; Whittaker, 1998
; Vamosi et al., 2006
). The Canary Islands have been inhabited by humans for ±3 millennia, but serious human disturbance began in the 15th century when the islands were annexed by Spain (Fernández-Palacios and Martín-Esquivel, 2001
). Human impact has grown. The main threats today are the continuous loss of natural habitat to tourist and residential developments, agriculture, overgrazing, invasive plants, off-road vehicles, and fires (Bramwell, 1994
), together with natural erosion. In contrast to the 100+ endemic species in the archipelago that are endangered (Bañares et al., 2003
), W. aristata is fairly widespread (Bramwell and Bramwell, 2001
). In addition, unlike dioecious species from other archipelagoes (e.g., Farwig et al., 2004
), W. aristata naturally produces good fruit set. Thus, this distinctive island endemic may not be in immediate danger of extinction. Nonetheless, like other endemics care must be taken to conserve this species, whose persistence also means preserving the plant–pollinator relationship (Nabhan et al., 1998
; Vamosi et al., 2006
) and the natural areas where its populations grow. The unusual reproductive system we document herein increases the importance of such preservation. Protection of W. aristata presents some challenges because populations are restricted to lower elevations (from sea level to 600 m) where human disturbance is higher (from Canary banana and wine-grape production and seaside tourism), and inexorably, there are fewer undeveloped, "wild" areas.
Our studies suggest that the reproductive biology of other Withania species should be studied in detail, particularly those known to have unisexual flowers (the Asian W. coagulans is known to be dioecious), and some sort of dicliny is suggested (from morphological studies of specimens) for W. adpressa from North Africa and W. qaraitica from Oman. Phylogenetic analyses of the group would provide a foundation critical to tracing the evolution of the different mating systems. Finally, reproductive studies need to include living plants and experimental studies to accompany field and specimen study. Our results imply that it is imperative to perform careful reproductive studies even at the mature bud stages.
FOOTNOTES
1 The authors thank C. Martine and C. Morse for help with horticultural and experimental aspects of this work, P. Neal for extensive field and greenhouse assistance and discussion of conceptual issues, F. La Roche-Brier for identification of insects, K. E. Theiss, T. Mione, and an anonymous reviewer for comments on the manuscript, and V. Kask for help with the illustrations. The American Philosophical Society, the University of Connecticut Office of the Provost and the College of Liberal Arts and Sciences, the University of Connecticut Research Foundation, CONICET, and Universidad Nacional de Cordoba (Argentina) provided financial support. 
5 Author for correspondence (gregory.anderson{at}uconn.edu
), phone 860-486-4555, fax 860-486-6474
LITERATURE CITED
Anderson G. J.. 1979. Dioecious Solanum of hermaphroditic origin is an example of a broad convergence. Nature 282: 836-838.[CrossRef]
Anderson G. J. Bohs L. Santos-Guerra A. Bernardello G. Opel M. R. Martine C.. 2005. The few and the odd: the Solanaceae from the Canary Islands and the African connections of the endemic Solanum vespertilio and S. lidii Botany 2005: Annual Meeting of the Botanical Society of America,Austin, Texas, USA. Website http://www.2005.botanyconference.org, abstract 161.
Anderson G. J. Johnson S. D. Neal P. R. Bernardello G.. 2002. Reproductive biology and plant systematics: the growth of a symbiotic association. Taxon 51: 637-653.[CrossRef][ISI]
Anderson G. J. Symon D. E.. 1988. Insect foragers on Solanum flowers in Australia. Annals of the Missouri Botanical Garden 75: 842-852.[CrossRef][ISI]
Anderson G. J. Symon D. E.. 1989. Functional dioecy and andromonoecy in Solanum. Evolution 43: 204-219.[CrossRef][ISI]
Arroyo M. T. K. Squeo F.. 1990. Genetic incompatibility in the endemic Patagonian genus Benthamiella (Solanaceae). Gayana Botanica 47: 51-55.
Axelius B.. 1996. The phylogenetic relationships of the physaloid genera (Solanaceae) based on morphological data. American Journal of Botany 83: 118-124.[CrossRef][ISI]
Baker H. G.. 1967. Support for Baker's law as a rule. Evolution 21: 853-856.[CrossRef][ISI]
Baker H. G. Cox P. A.. 1984. Further thoughts on dioecism and islands. Annals of the Missouri Botanical Garden 71: 244-253.[CrossRef][ISI]
Bañares A. Blanca G. Güemes J. Moreno J. C. Ortiz S.. 2003. Atlas y libro rojo de la flora vascular amenazada de España Dirección General de Conservación de la Naturaleza, Madrid, Spain.
Bawa K. S.. 1980. Evolution of dioecy in flowering plants. Annual Review of Ecology and Systematics 11: 15-39.
Bohs L.. 2000. Insights into the Witheringia solanacea (Solanaceae) complex in Costa Rica. II. Insect visitors and pollination biology of W. asterotricha and W. meiantha. Biotropica 32: 80-89.[ISI]
Bramwell D.. 1994. Canary Islands. In S. D. Davis, V. H. Heywood, and A. C. Hamilton [eds.] Centres of plant diversity 89-91 World Wide Fund for Nature and International Union for the Conservation of Nature, Oxford, UK.
Bramwell D. Bramwell Z.. 2001. Flores silvestres de las Islas Canarias Editorial Rueda, Madrid, Spain.
Carlquist S.. 1974. Island biology Columbia University Press, New York, New York, USA.
Charlesworth B. Charlesworth D.. 1978a. A model for the evolution of dioecy and gynodioecy. American Naturalist 112: 975-998.[CrossRef][ISI]
Charlesworth D. Charlesworth B.. 1978b. Population genetics of partial male sterility and the evolution of monoecy and dioecy. Heredity 41: 137-154.[ISI]
Chevallier A.. 1996. The encyclopedia of medicinal plants Dorling Kindersley, London, UK.
Cocucci A. A.. 1991. Pollination biology of Nierembergia (Solanaceae). Plant Systematics and Evolution 174: 17-36.[CrossRef][ISI]
Cocucci A. A.. 1995. Floral mechanisms in the tribe Salpiglossidae (Solanaceae). Plant Systematics and Evolution 194: 207-230.[CrossRef][ISI]
Cocucci A. A.. 1999. Evolutionary radiation in neotropical Solanaceae. In M. Nee, D. E. Symon, R. N. Lester, and J. P. Jessop [eds.] Solanaceae IV: advances in biology and utilization 9-22 Royal Botanic Gardens, Kew, UK.
Connolly B. A. Anderson G. J.. 2003. Functional significance of the androecium in staminate and hermaphroditic flowers of Solanum carolinense (Solanaceae). Plant Systematics and Evolution 240: 235-243.[CrossRef][ISI]
Cox P. A.. 1990. Pollination and the evolution of breeding systems in Pandanaceae. Annals of the Missouri Botanical Garden 77: 816-840.[CrossRef][ISI]
Cruden R. W.. 1977. Pollen–ovule ratios: a conservative indicator of breeding systems in flowering plants. Evolution 31: 32-46.
Darias V. Martín-Herrera D. Abdala S. de la Fuente D.. 2001. Plants used in urinary pathologies in the Canary Islands. Pharmaceutical Biology 39: 170-180.[ISI]
Delgado G.. 2001. Flora vertebrada terrestre. In J. M. Fernández-Palacios and J. L. M. Esquivel [eds.] Naturaleza de las Islas Canarias 213-217 Turquesa Ed., Tenerife, Canarias, Spain.
Delph L. F. Wolf D. E.. 2005. Evolutionary consequences of gender plasticity in genetically dimorphic breeding systems. New Phytologist 166: 119-128.[CrossRef][ISI][Medline]
Eckhart V. M.. 1999. Sexual dimorphism in flowers and inflorescences. In M. A. Gerber, T. E. Dawson, and L. F. Delph [eds.] Gender and sexual dimorphism in flowering plants 123-148 Springer-Verlag, New York, New York, USA.
Farwig N. Randrianirina E. F. Voigt F. A.. 2004. Pollination ecology of the dioecious tree Commiphora guillauminii in Madagascar. Journal of Tropical Ecology 20: 307-316.[CrossRef][ISI]
Fernández-Palacios J. M. Martín-Esquivel J. L.. 2001. Naturaleza de las Islas Canarias Turquesa Ed., Tenerife, Canarias, Spain.
Frankham R.. 1997. Do island populations have less genetic variation than mainland populations?. Heredity 78: 311-327.
Galetto L. Bernardello G.. 2003. Nectar sugar composition in angiosperms from Chaco and Patagonia (Argentina): an animal visitor's matter?. Plant Systematics and Evolution 238: 69-86.[ISI]
Galetto L. Bernardello G. Sosa C. A.. 1998. The relationship between floral nectar composition and visitors in Lycium (Solanaceae) from Argentina and Chile: what does it reflect?. Flora 193: 303-314.[ISI]
Galetto L. Bernardello L.. 1993. Nectar secretion pattern and removal effects in three species of Solanaceae. Canadian Journal of Botany 71: 1394-1398.[ISI]
Gerber M. A. Dawson T. E. Delph L. F.. 1999. Gender and sexual dimorphism in flowering plants Springer-Verlag, New York, New York, USA.
Givnish T. J.. 1982. Outcrossing versus ecological constraints in the evolution of dioecy. American Naturalist 119: 848-865.
Groombridge B.. 1992. Global diversity—status of the Earth's living resources. Compiled by the World Conservation Monitoring Centre Chapman & Hall, London, UK.
Haegi L.. 1981. A conspectus of Solanaceae tribe Anthocercideae. Telopea 2: 173-180.
Heiser C. B.. 1987. The fascinating world of the nightshades Dover Publications, New York, New York, USA.
Helfgott D. M. Francisco-Ortega J. Santos-Guerra A. Jansen R. K. Simpson B. B.. 2000. Biogeography and breeding system evolution of the woody Bencomia alliance (Rosaceae) in Macaronesia based on ITS sequence data. Systematic Botany 25: 82-97.[CrossRef][ISI]
Helversen O. V.. 1993. Adaptations of flowers to the pollination by glossophagine bats. In W. Barthlott, C. M. Naumann, K. Schmidt-Loske, and K. L. Schuchmann [eds.] Plant–animal interactions in tropical environments 41-59 Museum Alexander König, Bonn, Germany.
Hepper F. N.. 1991. Old World Withania (Solanaceae): a taxonomic review and key to the species. In J. G. Hawkes, R. N. Lester, M. Nee, and E. Estrada [eds.] Solanaceae III: taxonomy, chemistry, evolution 211-227 Royal Botanic Gardens, Kew, UK.
Hooker W. J.. 1852. Puneeria coagulans Stocks. Icones plantarum Reprinted by Lehre, Cramer; Codicote, Herts., Wheldon & Wesley; Stechert-Hafner, New York, New York, USA. 1966..
Humeau L. Pailler T. Thompson J. D.. 1999. Cryptic dioecy and leaky dioecy in endemic species of Dombeya (Sterculiaceae) on La Réunion. American Journal of Botany 86: 1437-1447.
Humeau L. Pailler T. Thompson J. D.. 2000. Variation in gender and flower-size dimorphism in the dioecious tree Dombeya ciliata, an endemic to La Reunion Island. Biotropica 32: 463-472.[ISI]
Hunziker A. T.. 2001. Genera Solanacearum: the genera of the Solanaceae illustrated, arranged according to a new system A. R. G. Gantner Verlag K.-G., Ruggell, Liechtenstein.
Izquierdo I. Martín J. L. Zurita N. Arechavaleta M.. 2004. Lista de especies silvestres de Canarias. Hongos, plantas y animales terrestres Consejería de Política Territorial y Medio Ambiente, Gobierno de Canarias, Spain.
Kaczorowski R. L. Gardener M. C. Holtsford T. P.. 2005. Nectar traits in Nicotiana section Alatae (Solanaceae) in relation to floral traits, pollinators, and mating system. American Journal of Botany 92: 1270-1283.
Kaul M. K. Kumar A. Sharma A.. 2005. Reproductive biology of Withania somnifera (L.) Dunal. Current Science 88: 1375-1377.[ISI]
Knapp S. Persson V. Blackmore S.. 1998. Pollen morphology and functional dioecy in Solanum (Solanaceae). Plant Systematics and Evolution 210: 113-139.[CrossRef][ISI]
Ladley J. J. Kelly D. Robertson A. W.. 1997. Explosive flowering, nectar production, breeding systems, and pollinators of New Zealand mistletoes (Loranthaceae). New Zealand Journal of Botany 35: 345-360.[ISI]
Lester R. N. Francisco-Ortega J. Mazin A. A.. 1999. Convergent evolution of heterandry (unequal stamens) in Solanum, proved by sporoderm SEM. In M. Nee, D. E. Symon, R. N. Lester, and J. P. Jessop [eds.] Solanaceae IV: advances in biology and utilization 51-69 Royal Botanic Gardens, Kew, UK.
Levine D. A. Anderson G. J.. 1986. Evolution of dioecy in an American Solanum. In W. G. D'Arcy [ed.] Solanaceae: biology and systematics 264-273 Columbia University Press, New York, New York, USA.
Litrico I. Pailler T. Thompson J. D.. 2005. Gender variation and primary succession in a tropical woody plant, Antirhea borbonica (Rubiaceae). Journal of Ecology 93: 705-715.[CrossRef][ISI]
Lloyd D. G.. 1965. Evolution of self-incompatibility and racial differentiation in Leavenworthia (Cruciferae). Contributions of the Gray Herbarium of Harvard University 195: 3-133.
Lloyd D. G.. 1976. The transmission of genes via pollen and ovules in gynodioecious angiosperms. Theoretical Population Biology 120: 571-585.
Marticorena C. Stuessy T. F. Baeza C. M.. 1998. Catalogue of the vascular flora of the Robinson Crusoe or Juan Fernandez Islands, Chile. Gayana Botanica 55: 187-211.
Mayer S. S. Charlesworth D.. 1991. Cryptic dioecy in flowering plants. Trends in Ecology and Evolution 6: 320-325.
McMullen C. K.. 1999. Flowering plants of the Galapagos Cornell University Press, Ithaca, New York, USA.
Miller J. S. Venable D. L.. 2002. The transition to gender dimorphism on an evolutionary background of self-incompatibility: An example from Lycium (Solanaceae). American Journal of Botany 89: 1907-1915.
Minne L. Spies J. J. Venter H. J. T. Venter A. M.. 1994. Breeding systems in some representatives of the genus Lycium (Solanaceae). Bothalia 24: 107-110.[ISI]
Mione T. Anderson G. J.. 1992. Pollen–ovule ratios and breeding system evolution in Solanum section Basarthrum (Solanaceae). American Journal of Botany 79: 279-287.[CrossRef][ISI]
Nabhan G. P. Allen-Wardell G. Bitner R. Burquez A. Buchmann S. Cane J. Cox P. A. Dalton V. Feinsinger P. Ingram M. Inouye D. Jones C. E. Kennedy K. Kevan P. Koopowitz H. Medellin R. Medellin-Morales S.. 1998. The potential consequences of pollinator declines on the conservation of biodiversity and stability of food crop yields. Conservation Biology 12: 8-17.[CrossRef][ISI]
Neal P. Anderson G. J.. 2005. Are ‘mating systems' ‘breeding systems' of inconsistent and confusing terminology in plant reproductive biology? Or is it the other way around?. Plant Systematics and Evolution 250: 173-185.[CrossRef][ISI]
Olmstead R. G. Sweere J. A. Spangler R. E. Bohs L. Palmer J. D.. 1999. Phylogeny and provisional classification of the Solanaceae based on chloroplast DNA. In M. Nee, D. E. Symon, R. N. Lester, and J. P. Jessop [eds.] Solanaceae IV: advances in biology and utilization 111-137 Royal Botanic Gardens, Kew, UK.
Ornduff R.. 1969. Reproductive biology in relation to systematics. Taxon 18: 121-133.[CrossRef]
Passarelli L. Bruzzone L.. 2004. Significance of floral colour and scent in three Solanum sect. Cyphomandropsis species (Solanaceae) with different floral rewards. Australian Journal of Botany 52: 659-667.[CrossRef]
Percy D. M. Cronk Q. C. B.. 1997. Conservation in relation to mating system in Nesohedyotis arborea (Rubiaceae), a rare endemic tree from St. Helena. Biological Conservation 80: 135-145.[CrossRef][ISI]
Proctor M. Yeo P. Lack A.. 1996. The natural history of pollination Timber Press, Portland, Oregon, USA.
Raguso R. A. Henzel C. Buchmann S. L. Nabhan G. P.. 2003. Trumpet flowers of the Sonoran Desert: floral biology of Peniocereus cacti and sacred Datura. International Journal of Plant Sciences 164: 877-892.[CrossRef]
Reid W. V. Miller K. R.. 1989. Extinction: how serious a threat?. In W. V. Reid and K. R. Miller [eds.] Keeping options alive: the scientific basis for conserving biodiversity 31-56 World Resources Institute, Washington, D.C., USA.
Renner S. S. Ricklefs R. E.. 1995. Dioecy and its correlates in the flowering plants. American Journal of Botany 82: 596-606.[CrossRef][ISI]
Richman A. D. Kohn J. R.. 2000. Evolutionary genetics of self-incompatibility in the Solanaceae. Plant Molecular Biology 42: 169-179.[CrossRef][ISI][Medline]
Sakai A. Wagner W. L. Ferguson D. M. Herbst D. R.. 1995a. Biogeographical and ecological correlates of dioecy in the Hawaiian flora. Ecology 76: 2530-2543.[CrossRef][ISI]
Sakai A. Wagner W. L. Ferguson D. M. Herbst D. R.. 1995b. Origins of dioecy in the Hawaiian flora. Ecology 76: 2517-2529.[CrossRef][ISI]
Sakai A. K. Weller S. G.. 1991. Ecological aspects of sex expression in subdioecious Schiedea globosa (Caryophyllaceae). American Journal of Botany 78: 1280-1288.[CrossRef][ISI]
Sakai A. K. Weller S. G.. 1999. Gender and sexual dimorphism in flowering plants: a review of terminology, biogeographic pattern, ecological correlates and phylogenetic approaches. In M. A. Gerber, T. E. Dawson, and L. F. Delph [eds.] Gender and sexual dimorphism in flowering plants 1-31 Springer-Verlag, New York, New York, USA.
Sawyer N. W. Anderson G. J.. 2000. Dioecy in South American Deprea (Solanaceae). Biotropica 32: 291-298.
Sazima M. Buzato S. Sazima I.. 2003. Dyssochroma viridiflorum (Solanaceae): A reproductively bat-dependent epiphyte from the Atlantic rainforest in Brazil. Annals of Botany 92: 725-730.
Sazima M. Vogel S. Cocucci A. Hausner G.. 1993. The perfume flowers of Cyphomandra (Solanaceae): pollination by euglossine bees, bellows mechanism, osmophores, and volatiles. Plant Systematics and Evolution 187: 51-88.[CrossRef][ISI]
Simpson B. Neff J.. 1981. Floral rewards: alternatives to pollen and nectar. Annals of the Missouri Botanical Garden 68: 301-322.[CrossRef][ISI]
Symon D.. 1979. Sex forms in Solanum (Solanaceae) and the role of pollen collecting insects. In J. G. Hawkes, R. N. Lester, and A. D. Skelding [eds.] The biology and taxonomy of Solanaceae 385-397 Academic Press, London, UK.
