Intra-plant floral variation in Cleome viscosa L. and its possible significance in breeding system

Transcription

1 Indian Journal of Experimental Biology Vol. 53, July 2015, pp Intra-plant floral variation in Cleome viscosa L. and its possible significance in breeding system Shveta Saroop* & Veenu Kaul Department of Botany, University of Jammu, Jammu , Jammu & Kashmir, India Received 03 March 2014; revised 06 May 2014 Cleome viscosa L., an annual rainy season weed, is cosmopolitan in distribution. Two naturally growing populations of C. viscosa from Jammu, J & K, India have been studied for floral variation at an intra-plant level and its possible role in its life cycle. Plants of both the populations bear flowers which exhibit tremendous intra-plant variation in size (large and small) and sex (hermaphrodite, staminate and pistillate). The average number of flowers per plant varied significantly and so did their structural and functional details. Greater propensity, however, was towards hermaphroditism at both plant and flower levels. The large and small sized flowers differed in their morphology and reproductive features; the former were significantly larger than the latter. Anthesis, anther dehiscence and stigma receptivity were coupled in all flower types. This functional aspect along with the structural proximity between stamens at two lengths and pistil further facilitated selfpollination. However, conspicuous floral display attracted diverse pollinator fauna (Apis dorsata, Halictus albescens, Nomia curvipes and N. elliotii) which in turn mediated cross pollination. Nevertheless, each floral type contributed towards plant s fitness in its own way. Hermaphrodite flowers exhibited both self and cross pollination and assured survival by setting fruits and seeds with the large sized counterparts more productive. All these floral variations seemed to impart flexibility to the pollination system and provide fitness over the short flowering season. Keywords: Asian spiderflower, Hermaphrodite, Insect pollinators, Pistillate, Pollination, Staminate Angiosperms exhibit enormous variation in many traits viz. size, structure, symmetry, display and sex expression of the flowers. Modifications of these over time eventually led to rapid diversification and formation of new species 1,2. Driven by various selective agents such as environment, resource patterning between sexual (male and female) functions, and pollinator behavior; these traits are stabilized by natural selection according to their functional and adaptive significance 3. These innovations have been analyzed not only in various species but also within the individuals of a species inhabiting different geographical ranges 4-8. Species of genus Cleome (Cleomaceae) are known for such variations in a single plant. They, provide a suitable system to analyse this variation critically and establish its possible adaptive significance 9,11. Cleome viscosa L, an annual weed, growing luxuriantly in and around Jammu (Jammu & Kashmir, India) is one such species. It propagates sexually through seed only. Barring an initial but brief vegetative phase, the plants, largely, have simultaneous vegetative and reproductive phases. Inflorescence, being racemose *Correspondence: shvetasaroop@gmail.com (SS), veenukaul@yahoo.co.in (VK) raceme, is characterized by continuous development of floral buds, flowers and fruits in the axils of young leaves 11. An individual plant bears flowers of 2-3 sex expressions; staminate, pistillate and hermaphrodite. Hermaphroditism is the primary and common sex expression. In some flowers, however, sterility of either male or female functions but not both at a time, leads to production of pistillate and staminate flowers, accordingly 11. Further, intergrading flower types, i.e., those bearing variable number of staminodes are also formed. In any case, female sterility is frequent than male sterility. The inter-sexualism is thus, one-sided always and doesn t lead to complete sterility of a flower or the whole plant. Additionally, the gender differentiation is further accompanied with variation in the sizes of these flower types. Floral diversity of this magnitude has a direct bearing on the mating system of the species and hence, the genetic structure of its populations Although work on the reproductive biology of C. viscosa has been worked out to some extent by Chauhan et al. 15, the floral variation at intra-plant level has not been studied. Recently, Kumari et al. 16 have reported extensive foliar variation among plants of C. viscosa. They differentiated between various plants in a population on the basis of leaf size and

2 SAROOP & KAUL: INTRA-PLANT FLORAL VARIATION IN CLEOME VISCOSA 469 accordingly named them as large and small leaved morphs. They also found contrasting reproductive features with respect to the size of flowers, fruits and other related attributes of the two foliar morphs. None of them revealed any significant variation in leaf size; neither could any distinct link be formed between foliar and floral variation. In this study, two populations growing naturally in two areas of Jammu district (J & K) were surveyed for quantification of size and gender polymorphism in flowers borne by an individual plant of C. viscosa. An attempt has also been made to determine its possible role in the breeding system of the species. Material and Methods Cleome viscosa is commonly found in Jammu during May-September when the day temperature fluctuates from 26.9 to C; relative humidity from to 100 % and total rainfall between and mm. For the present study, seeds of C. viscosa were collected from plants growing naturally along roadsides of Bahu Plaza, a commercial complex, and Directorate of Distance Education (Jammu University Campus), Jammu, J & K, India in The two sites are separated by a distance of approximately 3 km; the latter being more secluded. Ripe and mature seeds were harvested, shade dried for 2 days (to ensure total dryness), and stored in plastic bags under room temperature (25-30 C). The stored seeds were sown in the experimental beds ( m 2 ) of Botanical Garden (University of Jammu, Jammu) in May 2011, germinated within 3-4 days. Many seedlings survived and reached reproductive age. Of these, 30 randomly selected plants (n=15 per population) were used for the present study. Since the two populations investigated did not differ significantly in their morphological (t (34) =1.9; P >0.05) and reproductive features (t (38) =0.92; P >0.05), the data were pooled 11. Plants of Cleome viscosa are andromonoecious i.e., they bear staminate and hermaphrodite flowers 11. Externally, these flowers cannot be distinguished from each other. Their categorization into male and hermaphrodite is possible only after they open up and expose their essential organs completely. However, here, all flowers irrespective of their sex was divided into two major size groups; large (floral length >10 mm) and small (floral length 10 mm). Morphological data were, therefore, collected from mature and fully opened flowers of all types (n=20 per type) selected randomly from each plant. Measurements of various floral organs were undertaken with the help of a scale and vernier callipers. Occasionally, some flowers of both the sizes and sexes also bear variable number of staminodes (1-7). Rarely, all stamens are sterile making the flowers pistillate. However, detailed studies on these flowers couldn t be carried out due to their limited availability. The flowers of different sizes and sex in an individual plant (n=30) were determined on a day-today basis during July and August. The data obtained per plant were then pooled to calculate the proportion of different flower types on a single plant. Pollen and ovule productions of large and small hermaphrodite flowers (n=4 flowers of each type from 5 plants) were determined as described by Kaul et al. 13,17. The pollen:ovule ratios were estimated for all hermaphrodites by dividing the average pollen count per flower with ovule count of the same flower 13,18. The amounts of pollen produced by large and small staminate flowers were also quantified keeping the sample size uniform. For pistillate flowers, the pollen produced by anthers of staminodes was determined 18, while the ovule number was obtained from 2 small and 3 large flowers borne by one and three plants, respectively. Pollen count at the level of plant was calculated by first multiplying the average number of pollen grains per flower type with the average number of that type per plant and adding the figures so obtained. Same method was repeated for ovule count. The pollen and ovule counts per plant were divided to yield pollen:ovule ratio per plant. Events of floral biology like anthesis and anther dehiscence were recorded from 20 fully mature yet-to-open floral buds. Close and regular monitoring of these marked flowers revealed that flower opening is coupled with maturation of male and female sexual phases. Anthesis and anther dehiscence is simultaneous and soon thereafter stigma attains receptivity 11. Flowers of C. viscosa L. are bright yellow and high pollen producers. Nectar, though secreted, was in traces making it difficult to quantify. Together, these features attract many insects. To monitor their number, foraging behaviour and frequency of visits, visual observations were made at regular intervals of time throughout the day as done earlier and data recorded. The duration of visits, number of flowers visited per minute and the visiting period of each pollinator were also noted. The specimens of each

3 470 INDIAN J EXP BIOL, JULY 2015 visitor were trapped, anaesthetized with chloroform and got identified in the Entomology Laboratory, Department of Zoology, of University of Jammu. The fruit set was largely determined for hermaphrodite flowers of both the sizes and for a few pistillate flowers as well. For this purpose, flowers of both the sexes, which opened every day, were tagged. The numbers before and after fruit formation were scored and fruit set per plant (n=30) estimated. The data were pooled and average fruit set per plant in nature/non experimental conditions (open pollination) calculated. In addition, 20 flowers (n=2 per 10 plants) were emasculated, tagged and kept undisturbed for in vivo pollination. This was done to see if insects transfer pollen from one plant to another and ensure effective pollination. Another set of equal sample size were emasculated and bagged to check fruit set due to non pseudogamous apomixis. The size and weight of individual fruits were measured to see variation, if any. Two fruits each from the large and small hermaphrodite flowers of a single plant were randomly selected for the purpose (Number of plants = 10). This was done to ensure that the flowers of two sizes are proportionally represented at the plant level. Seeds differentiating therein were counted and weighed together to obtain total number of seeds and seed weight per fruit. Seed set was also determined as seed to ovule ratio (n=20 large and small each). Fruit and seed dimensions were determined with a scale or vernier caliper, and the weight using electronic balance (MR-Ohyo 200). Statistical analysis Mean, standard deviation and coefficient of variation were calculated on the data so generated. Differences in the number of flowers of both the sexes and sizes produced by plants were quantified and statistically analyzed using student s t-test. One way ANOVA was applied to test differences in various morphological and reproductive parameters like: (a) length of flower, stamens and pistil; (b) number of stamens per flower type; (c) sex allocation (in terms of pollen and ovule production and P/O ratio) patterning among flowers of different sexes and sizes; and (d) reproductive output (fruit and seed set attributes) under non-experimental conditions/open pollination. Results Floral variation and morphology The observations have revealed that the plants of C. viscosa had flowers of varied sex expressions. They possessed hermaphrodite, staminate and pistillate flowers in addition to those with staminodes. However, andromonoecy predominated with % plants in a population differentiating staminate and hermaphrodite flowers. Ten to 36.6 % of these plants bore flowers having staminodes in addition to the stamens. Only one percent flowers of a few plants (1-3) were found to be functionally female. Further, the C. viscosa flowers of all sex were grouped into two major size groups; large and small (Fig. 1a-e). Fig. 2a shows variation in the average number of different flower types formed during initial (July) and later (August) parts of the short flowering season. Accordingly, the percentage frequency of the plants bearing them also differed (Fig. 2b). Large flowers of both the sex significantly outnumbered the small ones [for hermaphrodite t (48) = 7.69 and staminate t (56) = 15.79; P <0.05; Fig. 2a]. The pattern was parallel with that at the plant level (Fig. 2b. n general, however, hermaphrodite flowers ( =. differentiated and bloomed in higher numbers than the staminate ones ( = 22.6) making the sex ratio per plant biased towards hermaphroditism i.e., 1.7:1 (Table 1). Fig. 1 Flowers of Cleome viscosa L. (a) An inflorescence in three phases of reproductive cycle, buds, flowers and fruits; Stereo-photomicrographs of large and small counterparts of hermaphrodite (b & c); and staminate (d & e) flowers. Note the flower with rudimentary pistil (d & e) and variable number of stamens (b & e).

4 SAROOP & KAUL: INTRA-PLANT FLORAL VARIATION IN CLEOME VISCOSA 471 Fig. 2 (a) Frequency of occurrence of flowers of different types during July-Aug., peak flowering in C. viscosa in Jammu. The first bar represents the average number of each flower type per plant and the second; proportion of those with staminodes; (b) Percentage frequency of plants bearing different flower types. Table 1 Frequency occurrence of different flower types that differentiate on a single plant Flower type. E. Coefficient of variation (c.v) % No. of hermaphrodite ± 15.5 (12-75) 40.3 flowers No. of staminate 22.6 ± 9.2 (8-50) flowers Sex ratio 1.7 No. of large flowers ± 13.8 (10-66) i. Hermaphrodite ii. Staminate ± 9.9 (3-50) iii. Sex ratio 1.8 No. of small flowers 5.58 ±4.4 (1-21) 78.7 i. Hermaphrodite ii. Staminate 4.43 ± 4 (1-17) 92.7 iii. Sex ratio 1.26 n=30 Irrespective of the sex and size, all flowers were asymmetric, hypogynous, tetramerous and sometimes pentamerous. Floral architecture was also unique. Instead of circumambulating the reproductive apex (i.e., of an individual flower), petals were oriented sideways approx. ¾ ths with respect to the pedicel leaving a gap which probably serves as a gateway or guide for pollinators. Sex organs were positioned towards the petal end away from the gap providing some sort of structural proximity between sex organs (Fig. 1a). Results of one-way ANOVA revealed significant differences in floral length between large and small types of hermaphrodite (F (1,28) = 91.62; P <0.05) and staminate (F (1,28) = 193.4; P <0.05) flowers. The large and small sized flowers also differed in their other morphology and reproductive features with the former significantly larger than the latter (Tables 2 and 3). The stamens in androecium varied in length and number. The number per large flower ranged from in hermaphrodite and 6-19 in staminate. The corresponding figures for the small sized counterparts were 5-12 and The number of long and short stamens and their respective length also varied in each flower type (Table 2). The longer stamens exceeded the pistil length while smaller ones were below the level of pistil. Gynoecium, a bicarpellary syncarpous with a superior ovary, was well developed in all hermaphrodites measuring 7.49 mm in large and 4.8 mm in small types. Contrarily, the staminate flowers had a rudimentary pistil bearing aborted ovules that failed to set fruit. Pollen:ovule ratio Pollen count varied between flowers of different sizes and sexes. Highest production was in large hermaphrodite ( ±4233.5) followed by large staminate ( ±4356.5) flowers and their respective small ( ± and ±4502.5) counterparts. One way ANOVA on the log transformed pollen counts revealed the significant differences between large and small counterparts of hermaphrodite (F (1,28) = 45.75; P <0.05) and staminate (F (1,18) = 35.33; P <0.05) flowers. A similar trend was observed in ovule number; large flowers (178.15±6.3) had almost twice more ovules than the small ones (88.8±6.14) [F (1,28) = 76.54; P <0.05]. Consequently, the pollenovule ratios also varied (585.5±38.8 for large and 454.1±63.8 for small) despite that their averages was comparable. In pistillate flowers, pollen production was the least ( ± ) and non-viable. However, the average ovule number varied between in large and 60.5 in small pistillate flowers; the difference was highly significant (F (1,3) = 94.75; P <0.05). At the plant level, pollen:ovule ratio worked out to be , indicating that for a single ovule, ~800 pollen grains were available for siring.

5 472 INDIAN J EXP BIOL, JULY 2015 Characters Table 2 Morphometry of large and small morphs of hermaphrodite and staminate flowers. Large Mean±S.E (range) Hermaphrodite Small Mean±S.E (range) Large Mean±S.E (range) Staminate Small Mean±S.E (range) Flower length 12.9 ± 0.56 (10-19) 7.7 ± 0.31 (5-10) 12.1±0.03 (10-15) 7±0.024 (5-9) Total stamens/flower 18.95±0.77 (13-23) 8 ± 0.39 (5-12) 12.85±0.87 (6-19) 6.1±0.29 (3-10) No. of 7.85 ± 0.7 (2-13) 4.15 ± 0.26 (2-6) 5.55±2.9 (2-12) 3.6±1.16 (2-7) i. Longer stamen ii. Shorter stamen 11.1 ± 0.8 (6-20) 3.73 ± 0.25 (2-6) 7..3±2.9 (2-13) 2.6±0.66 (2-4) Filament length 10.2 ± 0.26 (9-13) 6.08 ±0.4 (3.5-10) 8.8±1.10 (6-9) 4.79±0.08 (3-5.9) i. Longer stamen ii. Smaller stamen 7.1 ± 0.24 (5-9) 4.48 ± 0.24 (2-7) 6.075±0.07 (4-7) 3.6±0.06 ( ) Pistil length 7.49± 0.35 (4.7-12) 4.8±0.09 ( ) 3.3±0.08 ( ) 2.18±05 ( ) Table 3 F-value of different floral characters between flower types. Character Hermaphrodite (large and small) (P <0.05) Staminate (large and small) (P <0.05) Flower length Total stamen no Long stamen no Short stamen no Long stamen length Short stamen length Pistil length * *P>0.05 Pollination and reproductive output Unique floral architecture of C. viscosa with bright yellow colour and production of ample pollen attracted a variety of insects viz. Apis dorsata, Halictus albescens, Nomia curvipes, and N. elliotii. Though these insects visited both hermaphrodite and staminate flowers indiscriminately during early morning hours yet foraging was intense on large flowers. All visitors were duly rewarded with nectar and abundant pollen. Foraging started at 7:00 a.m. and continued till 11:30 a.m. A. dorsata and N. curvipes visits were brief while those of H. albescens and N. elliotii were of longer duration. The number of flowers foraged by them in a minute was also more comparatively. During the foraging activity, these insects shifted from one flower to other and ensured transfer of pollen. This was further confirmed by the formation of 61.8% fruits in flowers emasculated and kept undisturbed for in vivo pollen deposition and fertilization. Same was true of pistillate ones (Fig. 3). Whereas, the emasculated and bagged flowers did not transform into fruits indicating that non pseudogamous apomixes was not operating. Fig. 3 Comparative fruit and seed set in large and small flower types. Fig. 4 Comparative fruit and seed related attributes of large and small flower types. Under non-experimental conditions also, the flowers of each size and sex were equally efficient in setting fruits and seeds (Fig. 3). The pistillate flowers of both the sizes had set significantly more fruits than their hermaphrodite counterparts (Fig. 3; t (31) = for large and t (22) = 2.625, P <0.05 for small). The fruits formed by hermaphrodite flowers of two sizes exhibited significant differences in their weight [F (1,38) = 78.3; P <0.05] and length [F (1,38) = 30; P <0.05]. A similar trend was observed for those of pistillate flowers [weight F (1,3) = 28.9 and length F (1,3) = 23.77; P <0.05]. Between sexes, however, the differences in fruit length [F (1,21) = in large and F (1,10) = 0.66; P >0.05 in small] and weight [F (1,21) = and F (1,10) = 0.78; P >0.05, respectively] were not significant (Fig. 4).

6 SAROOP & KAUL: INTRA-PLANT FLORAL VARIATION IN CLEOME VISCOSA 473 This pattern was repeated at the level of seed set. Surprisingly, seed production by small pistillate flowers was highest followed by those of large pistillate ones with lowest seed set recorded in small hermaphrodite ones (Fig. 3). Differences between sexes were not significant [F (1,21) = in large and F (1,10) = 3.93; P >0.05 in small] while those on account of size were in hermaphrodites [F (1,28) = 13.43; P <0.05] only. For pistillate F (1,3) = at P >0.05. Discussion Cleome viscosa L. has several interesting floral traits of evolutionary and functional significance. This study demonstrated the prevalence of its variation in sex and size of flowers within the individual plants. This variation was further expressed differently over the two flowering months of July and August. Initially, i.e., in July, the frequency of these flowers was low; an increase occurred gradually and considerably in successive developing inflorescences in August. Small sized flowers of all sex differentiated in lesser number than their large counterparts (Fig. 2a). Staminodal tendency/stamen sterilization was almost equal among flowers of all sizes and sexes; and apparently increased with the age of the plant and the time of differentiation of the flower. Notwithstanding this, hermaphrodites of all types were produced in greater number. This indicates that plants have a strong propensity towards hermaphroditism (Fig. 2 a & b). That, however, doesn t preclude the plant s tendency to develop flowers of other sex expressions. Although the differentiation of staminate flowers commenced early along with the hermaphrodite ones, their frequency was 3-times greater towards the later part of the season. Seemingly, the subtending fruits signal the termination of pistil in the otherwise hermaphrodite flowers leading to a preponderance of staminate flowers at the distal end 6,7. This variability in sex and size also differs between plants at the population level. Majority of plants in a population are andromonoecious with % plants differentiating into staminate and hermaphrodite flowers throughout their life cycle. Only 40-50% percent of these plants produce small sized flowers of both the sexes in addition to the large sized normal ones. Proportion of plants bearing flowers with variable number of staminodes was the least ( %) suggesting a slight tilt towards femaleness at flower, plant and populational levels. However, the gradual increase of these plants bearing small and intergrading (staminode bearing) flower types as observed in August (Fig. 2b), the latter part of the season, indicates a shift probably occurring in sex expression from andromonoecy to gynomonoecy. But the advantages of doing so are intriguing because assurance of female reproductive success at this critical juncture when resources and potential mates are meagre is quite low. Variation in floral features at the structural level gets reflected at the functional level as well. Floral display, stamen number, their corresponding lengths and pollen ovule counts differ significantly between the various size types; with large sized ones, particularly hermaphrodites, being superior over others. Hermaphrodite and pistillate flower contribute toward survival by setting fruits and seeds efficiently. In this study, surprisingly, fruit and seed production by pistillate flowers was significantly higher than hermaphrodites. This result is possibly an overestimate generated by low sample size of the pistillate flowers. Further, the large and small fruits differed significantly in their weight and size. The number of seeds differentiating therein also varied considerably between sizes. All these features confer on the species a potential to modulate pollination success at individual flower and plant levels. These differences among sexes lead to further flexibility in reproductive traits. Variations of this magnitude are not too frequent in flowering plants. Presence of male and hermaphrodite flowers similar to those found in the present case have been reported in andromonoecious species of Leptospermum, Solanum and Zigadens 4-5. But size differences and those in sexual organs within a flower on a single plant are not on record in these taxa. These workers have, however, drawn conclusions similar to ours. In C. viscosa, flowers of all types are zygomorphic, bright yellow and most of them are abundant pollen producers. Zygomorphy (monosymmetry), an interesting floral feature, characteristic of family Cleomaceae 19, is well presented in this species. This floral architecture facilitates the landing and orientation of pollinators toward floral rewards (pollen and nectar), and also provides structural proximity between sex organs (Fig. 1a-e). Coupled with an overlap in the maturation of two sexual phases these features permit both self and cross pollination as observed here in formation of fruits and seeds in both self and cross pollination treatments 11.

7 474 INDIAN J EXP BIOL, JULY 2015 In pistillate flowers, brightly yellow colored staminodes mimic the stamens, appear nectariferous and serve as an additional source of attraction 20,21. The flowers of C. viscosa attracting many pollinators viz. bees (Apis dorsata, A. indica, A. florea), butterflies (Pieris brassicae, Danaus plexippus), flies (Musca domestica) and ants (Camponotus compressus) is already on record 15. In the present investigation as well, all flower types were equally effective in luring insect visitors (namely, Apis dorsata, Halictus albescens, Nomia curvipes and N. elliotii) which in turn facilitated cross pollination. During the foraging activity, these insects shifted from one flower to other and ensured transfer of pollen. Moreover, the pollinators also don t discriminate between floral types although logically their visitation should be directly related to the quantity of rewards offered 22. The large flowers with greater rewarding capacity should be more appealing than the small ones and receive more insect visitors per flower or more visits per insect or both. This, however, remains to be ascertained. Our observations, however, differ from those of Kumari et al. 16. Instead of intra-plant floral variation they found small and large leaved plants of C. viscosa which inhabit different regions of northern India. According to them, the small leaved accessions undergo early flowering, bear greater number of leaves, pods and large sized flowers per plant. The large leaved ones are late blooming with less number of leaves and small sized flowers. But the size of pods and seeds in them are larger than those in small leaved counterparts. In the populations under present investigation, plants could not be distinguished on the basis of their leaf size. Instead, each plant had large and small sized flowers of different sex. Large sized flowers differentiated pistils of greater size bearing large number of ovules. These flowers also got transformed into large sized fruits with greater number of seeds therein and vice versa. It is quite possible that the number of populations studied here was less while that by Kumari et al. 16 studied wider germplasm. Nevertheless, each flower type of C. viscosa is significant and contributes towards plant s fitness in its own way. Staminate flowers are produced to minimize the wastage on those flowers which lack the ability to set fruit. As efficient pollen donors and insect attractants, these flowers not only boost insect visitation but also help in overcoming the loss caused by pollinators 12,13,23. Therefore, their differentiation helps in retrieving the costs of sex by exercising male function through pollen donation 21, Contrarily, the hermaphrodite and pistillate flowers assure survival by setting fruits and seeds efficiently (Figs. 3 & 4). All these variations within flowers on a single plant and between plants impose variability in breeding system. Cleome viscosa, as an individual plant, practices autogamy through hermaphrodites and xenogamy through staminate and pistillate flowers. This ensures mixed mating strategy which helps in easy proliferation and invasion of the species to new areas. Selfing offers opportunity to ensure reproduction when the availability of pollinators or potential mates is scarce, catering to the plant s need of immediate fitness. While cross pollination generates variability, assures adaptability in diverse climatic conditions and imparts evolutionary plasticity 12,13,26. The common angiosperms way of mixed mating system is favoured in many other annual plants, Cleome lutea, C. serrulata 10, Commelina benghalensis and C. caroliniana etc. These species have evolved various alternatives to undergo both self and cross-pollination. The first two species are andromonoecious. Commelina bengalensis, another andromonoecious species self-pollinates through cleistogamy and cross pollinates via chasmogamous flowers 13. Similarily, C. caroliniana cross pollinates but undergoes delayed selfing as well 12 whereas, Cardiospermum canescens and C. halicacabum enjoy the benefits of mixed mating through functional monoecy by producing staminate (bearing rudimentary pistillode) and pistillate (with indehiscent stamens) flowers 27. The present species appear to have evolved these mixed mating traits through variation in floral traits (sex expression and size) coupled with unique floral architecture and biology. Acknowledgement The authors are thankful to Dr. Sanjay Bhatia, Department of Zoology of University of Jammu for identifying the insect pollinators. The first author acknowledges University Grants Commission for providing financial assistance in the form of BSR (Basic Scientific Research) fellowship. References 1 Grant V, Modes and origins of mechanical and ethological isolation in angiosperms. Proc Natl Acad Sci USA, 91 (1994) 3.

8 SAROOP & KAUL: INTRA-PLANT FLORAL VARIATION IN CLEOME VISCOSA Levin DA, The Origin, Expansion and Demise of Plant Species. (Oxford University Press, Oxford) 2000, Strauss SY &Whittall JB, Non-pollinator agents of selection on floral traits. In: Ecology and Evolution of flowers (eds.lawrence D Harder & Spencer CH Barrett, Oxford University Press, Oxford), 2006, Primack RB & Lloyd DG, Andromonoecy in New Zealand montane shrub manuka, Leptospermum scoparium (Myrtaceae). Am J Bot, 67 (1980) Solomon BP, Environmentally influenced changes in sex expression in an andromonoecious plant. Ecology, 66 (1985) Meagher TR, Linking the evolution of gender variation to floral development. Ann Bot, 100 (2007) Narbona E, Ortiz PL & Arista M, Functional andromonoecy in Euphorbia (Euphorbiaceae). Ann Bot, 89 (2002) Lambrecht SC & Dawson TE, Correlated variation of floral and leaf traits along a moisture availability gradient. Oecology, 151 (2007) Stout AB, Alternation of sexes and intermitted production of fruit in spider flower (Cleome spinosa). Am J Bot, 10 (1923) Cane JH, Breeding biologies, seed production and speciesrich bee guilds of Cleome lutea and Cleome serrulata (Cleomaceae). Plant Species Biol, 23 (2008) Saroop S, A preliminary analysis on seed to seed cycle of Cleome viscosa L. (M.Phil. Dissertation, University of Jammu, Jammu), 2011 (Unpublished data) 12 Kaul V & Koul AK, Floral phenology in relation to pollination and reproductive output in Commelina caroliniana (Commelinaceae). Aust J Bot, 56 (2008) Kaul V & Koul A-K, Sex expression and breeding strategy in Commelina benghalensis L. J Biosci, 34 (2009) Saroop S & Kaul V, Phenological events of Cleome viscosa L. growing in Jammu district. Int J Plant Rep Biol, 3 (2011) Chauhan S, Rana A, Sharma SB & Chauhan SVS, Reproductive biology of Cleome viscosa L. (Capparaceae). Ann For, 15 (2007) Kumari R, Tyagi A, Sharma V, Jain VK & Kumar S, Variability in the accessions from Aravali range assessed for domestication of the Cleomaceae biodiesel plant Linn. Indian J Nat Prod Resour, 3 (2012) Kaul V, Sharma N & Koul AK, Reproductive effort and sex allocation strategy in Commelina benghalensis L., a common monsoon weed. Bot J Linn Soc, 140 (2002) Cruden RW, Pollen-ovule ratios: a conservative indicator of breeding systems in flowering plants. Evolution, 31 (1977) Hall JC, Systma KJ &Iltis HH, Phylogeny of Capparaceae and Brassicaceae based on chloroplast sequence data. Am J Bot, 89 (2002) Endress PK, Diversity and evolutionary biology of tropical flowers (Cambridge University Press) Kaul V & Koul AK, Staminal variation and its possible significance in Commelina benghalensis L. and Commelina caroliniana Walter. Curr Sci, 103 (2012) Saroop, S. & Kaul, V. Correlation patterns among floral traits in Cleome viscosa L. a sexually polymorphic species. Curr Sci, 2015 (in press). 23 Lovett-Doust J, Floral sex ratios in andromonoecious Umbelliferae. New Phytol, 85 (1980) Schoen DJ, Male reproductive effort and breeding systems in a hermaphrodite plant. Oecology, 53 (1982) Huang SQ, Flower dimorphism and the maintenance of andromonoecy in Sagittaria guyanensis ssp. Lapula (Alismataceae). New Phytol, XX (2003) Goodwillie C, Kalisz S & Eckert CG, The evolutionary enigma of mixed mating systems in plants: occurrence, theoretical explanations, and empirical evidence. Annu Rev Ecol Syst, 36 (2005) Raju AJS, Ramana KV, Rao NG & Varalakshmi P, Monoecy and entomophily in Cardiospermumcanescens Wall. (Sapindaceae), a medicinally valuable herbaceous vine. Curr Sci, 101 (2011) 617.