Noncorrelated evolution between herbivore- and pollinator-linked features in Aristolochia chilensis (Aristolochiaceae)

The effect of mutualists (i.e. pollinators) and antagonists (i.e. herbivores) can have non-additive effects on plant fitness. This is often interpreted as evidence for correlated evolution on a suite of traits leading to an increase and decrease of the interaction of plants with mutualists and antagonists, respectively. This situation has been found to prevail in plants that have large floral and fruit displays but are not limited by pollinators for seed set. We suggest the alternative hypothesis, where plants limited by pollinators for seed set (e.g. deceit-pollinated plants) exhibit additive effects of pollinators and herbivores on fitness (i.e. noncorrelated evolution). Using a 2 × 2 factorial design, we tested this hypothesis by solely and simultaneously evaluating the effects of pollinators and the single herbivore, Battus polydamas archidamas , on female reproductive success of Aristolochia chilensis . Plants exposed to herbivores presented 2.6-fold greater herbivory than plants that excluded them. In addition, plants exposed to pollinators showed strong limitation by pollinators for seed set compared with other plants of the genus Aristolochia. However, only pollinators had a significant effect on fruit and seed set because plants that excluded them did not set fruits or seeds. Furthermore, herbivores and pollinators exerted additive effects on fruit and seed production. Collectively, these results indicate that herbivore- and pollinator-linked traits in A. chilensis exhibit noncorrelated evolution.  © 2007 The Linnean Society of London, Biological Journal of the Linnean Society , 2007, 91 , 239–245.     

Florivory indirectly decreases the plant reproductive output through changes in pollinator attraction

Species often interact indirectly with each other via their traits. There is increasing appreciation of trait‐mediated indirect effects linking multiple interactions. Flowers interact with both pollinators and floral herbivores, and the flower‐pollinator interaction may be modified by indirect effects of floral herbivores (i.e., florivores) on flower traits such as flower size attracting pollinators. To explore whether flower size affects the flower‐pollinator interaction, we used Eurya japonica flowers. We examined whether artificial florivory decreased fruit and seed production, and also whether flower size affected florivory and the number of floral visitors. The petal removal treatment (i.e., artificial florivory) showed approximately 50% reduction in both fruit and seed set in natural pollination but not in artificial pollination. Furthermore, flower size increased the number of floral visitors, although it did not affect the frequency of florivory. Our results demonstrate that petal removal indirectly decreased 75% of female reproductive output via decreased flower visits by pollinators and that flower size mediated indirect interactions between florivory and floral visitors.     

Noncorrelated effects of seed predation and pollination on the perennial herb Ruellia nudiflora remain spatially consistent

By simultaneously manipulating both seed predator and pollinator effects on the perennial herb Ruellia nudiflora at two sites in Yucatan (Mexico), the present study evaluated (1) whether a correlation (interaction) existed between seed predator and pollinator effects on R. nudiflora seed production and (2) whether such an interaction varied geographically. We used three populations per site, and a total of 20 plants per population ( N  = 120). Groups of five plants were randomly chosen at each population to simultaneously receive one of two seed predator and pollinator exclosure levels (present or excluded in each case). These two factors were fully crossed, resulting in each group being subjected to one of four possible combinations: pollinators excluded/herbivores present; herbivores excluded/pollinators present; herbivores excluded/pollinators excluded; or control (neither excluded). Response variables were the number of seeds produced per plant and the proportion of attacked fruits by seed predators per plant. Seed predators had a large impact on R. nudiflora seed production but did not show any preference for fruits from plants not excluded from pollinators. In addition, the pollination treatment was not significant, indicating no effect of pollinators on reproductive success. These findings resulted in a nonsignificant herbivory × pollination interaction, which was consistent across sites, indicating lack of correlated selection of these two guilds on R. nudiflora seed production.  © 2009 The Linnean Society of London, Biological Journal of the Linnean Society , 2009, 96 , 800–807.     

Nocturnal Versus Diurnal Pollination Success in Isertia laevis (Rubiaceae): A Sphingophilous Plant Visited by Hummingbirds

Abstract: Isertia laevis (Rubiaceae) possesses flowers with traits typical for the pollination syndrome of sphingophily. Diurnal flower observation showed that nine different hummingbirds (Trochilidae) and one flower piercer (Coerebidae) were frequent visitors. Their activity on the flowers peaked in the morning hours. Very low nectar volumes were found in the morning (8.00 h) in unbagged flowers. Nectar volumes, however, reached their peaks (27 μl) at night (2.00 h) in bagged, as well as in unbagged flowers. At night few individuals of sphingids were observed. Pollination experiments showed that flowers presented to nocturnal pollinators from 18.00 h to 6.00 h had low fruit set (14 %) but high seed set (59 %). Flowers accessible from 6.00 h to 18.00 h for diurnal flower visitors showed high fruit set of 63 % but low seed set of 14 %. This suggests that pollination of individual flowers is less effective during daytime. Regarding relative reproductive success, i.e., efficiency of pollination defined as fruit set x seed set, both diurnal and nocturnal pollinators, however, are equally successful. We conclude that frequently occurring, but not very effective pollinators contribute substantially to seed production, when the expected pollinators are scarce.     

ECOLOGICAL COSTS OF PLANT RESISTANCE TO HERBIVORES IN THE CURRENCY OF POLLINATION

In this paper, we examine how ecological costs of resistance might be manifested through plant relationships with pollinators. If defensive compounds are incorporated into floral structures or if they are sufficiently costly that fewer rewards are offered to pollinators, pollinators may discriminate against more defended plants. Here we consider whether directional selection for increased resistance to herbivores could be constrained by opposing selection through pollinator discrimination against more defended plants. We used artificial selection to create two populations of Brassica rapa plants that had high and low myrosinase concentrations and, consequently, high and low resistance to flea beetle herbivores. We measured changes in floral characters of plants in both damaged and undamaged states from these populations with different resistances to flea beetle attack. We also measured pollinator visitation to plants, including numbers of pollinators and measures of visit quality (numbers of flowers visited and time spent per flower). Damage from herbivores resulted in reduced petal size, as did selection for high resistance to herbivores later in the plant lifetime. In addition, floral display (number of open flowers) was also altered by an interaction between these two effects. Changes in floral traits translated into overall greater use of low-resistance, undamaged plants based on total amount of time pollinators spent foraging on plants. Total numbers of pollinators attracted to plants did not differ among treatments; however, pollinators spent significantly more time per flower on plants from the low-resistance population and tended to visit more flowers on these plants as well. Previous work by other investigators on the same pollinator taxa has shown that longer visit times are associated with greater male and female plant fitness. Because initial numbers of pollinators did not differ between selection regimes, palatability and/or amount of rewards offered by high- and low-resistance populations are likely to be responsible for these patterns. During periods of pollinator limitation, less defended plants may have a selective advantage and pollinator preferences may mediate directional selection imposed by herbivores. In addition, if pollinator preferences limit seed set in highly defended plants, then lower seed set previously attributed to allocation costs of defense may also reflect greater pollinator limitation in these plants relative to less defended plants.     

Unpaved roads disrupt the effect of herbivores and pollinators on the reproduction of a dominant shrub

Anthropogenic linear developments, such as trails and firebreaks, also called soft linear developments (SLD), can influence animal behavior, altering the ecological interactions in which animals are involved. For example, SLD can affect the behavior of pollinators and herbivores, but little is known about the combined effect of these three elements on plant reproduction.We evaluated the combined effect of SLD, insect pollinators and herbivores (ungulates) on three reproductive output variables (fruit set, seed set, and seed mass) of a Mediterranean shrub (Halimium halimifolium). We considered two different habitats (SLD verges vs. adjacent scrublands), two scenarios of herbivory (with and without ungulates), and three scenarios of pollinator activity (without pollinators, with manual pollination and with natural pollination).SLD had contrasting effects on H. halimifolium reproduction. In the absence of herbivores, overall fruit set was lower in the verges of SLD than in adjacent scrublands, probably due to lower flower pollination rates. Where herbivores were present, overall fruit set was similar between habitats, because ungulate browsing was lower in SLD verges than in adjacent scrublands. The quantity and weight of seeds per fruit was similar in both habitats, probably because all fertilized flowers received similar amounts of pollen.SLD can alter the interaction among pollinators, herbivores and plants, leading to changes in the reproductive performance of the latter. These changes can have strong negative impacts on endangered plants that rely on fruit and seed production to persist. However, SLD verges could be safe places for plants particularly sensitive to herbivory by ungulates.     

Non-additive effects of herbivores and pollinators on Erysimum mediohispanicum (Cruciferae) fitness

In this study, the non-additivity of effects of herbivores and pollinator on fitness of the plant Erysimum mediohispanicum (Cruciferae) has been experimentally tested. The abundance and diversity of the pollinator assemblage of plants excluded from and exposed to mammalian herbivores, and the combined effect of pollinators and herbivores on plant reproduction were determined over a period of 2 years. Pollinator abundance was higher and diversity was lower on plants excluded from herbivores. Furthermore, the experimental exclusions demonstrated that both pollinators and herbivores affected plant fitness, but their effects were not independent. Herbivores only had a detrimental effect on plant fitness when pollinators were present. Similarly, pollinators enhanced fitness only when herbivores were excluded. This outcome demonstrates that the importance of pollinators for plant fitness depends on the occurrence of herbivores, and suggests that herbivores may hamper pollinator-mediated adaptation in plants.     

Pollination in Verbascum thapsus (Scrophulariaceae): the advantage of being tall

According to the "effective pollination" hypothesis, tall stature resulting from strong apical dominance attracts greater pollinator visitation, thus allowing larger pollen loads and/or greater outcrossing rates, which in turn produces more vigorous offspring with greater genotypic variability and/or less inbreeding depression. Components of this hypothesis were tested in Verbascum thapsus, which commonly grows unbranched to over 2 m tall with strong apical dominance suppressing all axillary meristems. A natural population survey indicated that plants with visiting pollinators were significantly taller than their nearest neighboring individuals not possessing a visiting pollinator. Plants in natural populations with excluded pollinators produced seeds via a delayed selfing mechanism. However, delayed selfing under pollinator exclusion resulted in only 75% of the seed set obtained with natural pollinators. Under natural pollination, emasculated flowers experienced a 50% reduction in pollen deposition by the time of flower closure but only a 5% reduction in seed set relative to intact flowers. Hence, taller plants attracted more pollinators and maximum seed set could not be achieved without pollinators. Comparison of seed set and seed mass in plants that were artificially selfed and artificially crossed (in both the greenhouse and in natural populations) indicated that plants were fully self-compatible with no evidence of early-acting inbreeding depression. However, this does not exclude the possibility that inbreeding depression is manifested in later life stages. The results suggest that V. thapsus has a mixed mating system with potential for reproductive assurance and various levels of outcrossing depending on variables affecting pollinator availability (e.g., population size).     

Attracting pollinators and avoiding herbivores: insects influence plant traits within and across years

Perennial plants interact with herbivores and pollinators across multiple growing seasons, and thus may respond to herbivores and pollinators both within and across years. Joint effects of herbivores and pollinators influence plant traits, but while some of the potential interactions among herbivory, pollination, plant size, and plant reproductive traits have been well studied, others are poorly understood. This is particularly true for perennial plants where effects of herbivores and pollinators may manifest across years. Here, we describe two experiments addressing the reciprocal interactions of plant traits with herbivore damage and pollination across 2 years using the perennial plant Chamerion angustifolium. We measured (1) plant responses to manipulation of damage and pollination in the year of treatment and the subsequent season, (2) damage and pollination responses to manipulation of plant size and flowering traits in the year of treatment, and (3) plant-mediated indirect interactions between herbivores and pollinators. We found that plant traits had little effect on damage and pollination, but damage and pollination affected plant traits in both the treatment year and the subsequent year. We found evidence of indirect effects between leaf herbivores and pollinators in both directions; indirect effects of pollinators on leaf herbivores have not been previously demonstrated. Our results indicate that pollen receipt results in shorter plants with fewer stems but does not change flower number, while leaf herbivory results in taller plants with fewer flowers. Together, herbivory and pollination may contribute to intermediate plant height and plants with fewer stems and flowers in our system.     

Positive effects of forest edges on plant reproduction: literature review and a case study of bee visitation to flowers of

Positive effects of fragmentation on plant reproduction are uncommon; in a literature review we found significant negative effects on fruit or seed set for 50 plant species, compared to 26 species showing no effect, and only nine affected positively. One of these is the declining New Zealand mistletoe Peraxilla tetrapetala (Loranthaceae), and here we investigate the mechanism of this positive effect. P. tetrapetala requires visits from native bird or bee pollinators to produce fruit. Fruit set was consistently pollen limited at several South Island sites because of a shortage of pollinators, but within a site at Lake Ohau, plants on forest edges had higher fruit set than those in the forest interior. Previous work showed that this difference was not caused by a shortage of resources in interior plants, but was associated with higher bird visitation rates to flowers on edges. In this study, we tested whether native bees also show a preference for edge flowers. At two sites (Ohau and Craigieburn) edge mistletoes had higher visitation by native bees (Hylaeus agilis and Leioproctus sp.) and higher fruit set. Some, but not all, of the higher visitation to edge flowers was explained by a preference amongst bees for flowers in direct sunshine. Therefore, P. tetrapetala experiences higher fruit set on edges because both of its main groups of pollinators (endemic birds and bees) visit edge flowers more often. The other eight published cases of positive effects of fragmentation on fruit set also all reported increased visitation rates by pollinators.     

Low legitimate pollen flow in distylic Turnera hermannioides (Passifloraceae) and its consequences on fruit and seed set

Turnera hermannioides is a ruderal distylic subshrub, native to NE-Brazil. In the Catimbau National Park, situated within the semi-arid Caatinga region, we studied the pollination ecology of this species, emphasizing (1) effective pollinators; (2) characteristics of short- and long-styled flowers; (3) intra- and intermorph pollen flow; and (4) fruit and seed set. Short and long-styled morphs differ in pollen size and ornamentation, stigmatic surface, style and stamen length and nectar production. The flowers are obligate intermorphic outcrossers and depend on animals for pollination. The flowers of T. hermannioides attracted insect visitors of 25 species, among them butterflies, beetles, but mainly bees. Polylectic bees, such as Apis mellifera, stingless bees, and solitary Callonychium brasiliense were the most frequent visitors and the principal pollinators. The frequency of visits, however, was very low (on average 1.9 visits/flower/day), resulting in a low and unbalanced legitimate pollen flow, i.e. a much lower number of pollen grains from short-styled flowers reaching stigmas of long-styles than vice versa, which in turn compromised fruit and seed set. Surprisingly, Protomeliturga catimbaui and P. turnerae, both oligolectic on flowers of the Turneraceae-clade of the Passifloraceae s.l., were very rare flower visitors and did not contribute significantly to the pollination of Turnera hermannioides. We discuss how the low visitation rate and the inefficiency of pollinators (i.e. inefficiency in promoting legitimate pollination) compromise pollen flow and, consequently, fruit and seed set in this species.     

Indirect effects of domestic and wild herbivores on butterflies in an African savanna

Indirect interactions driven by livestock and wild herbivores are increasingly recognized as important aspects of community dynamics in savannas and rangelands. Large ungulate herbivores can both directly and indirectly impact the reproductive structures of plants, which in turn can affect the pollinators of those plants. We examined how wild herbivores and cattle each indirectly affect the abundance of a common pollinator butterfly taxon, Colotis spp., at a set of long‐term, large herbivore exclosure plots in a semiarid savanna in central Kenya. We also examined effects of herbivore exclusion on the main food plant of Colotis spp., which was also the most common flowering species in our plots: the shrub Cadaba farinosa. The study was conducted in four types of experimental plots: cattle‐only, wildlife‐only, cattle and wildlife (all large herbivores), and no large herbivores. Across all plots, Colotis spp. abundances were positively correlated with both Cadaba flower numbers (adult food resources) and total Cadaba canopy area (larval food resources). Structural equation modeling (SEM) revealed that floral resources drove the abundance of Colotis butterflies. Excluding browsing wildlife increased the abundances of both Cadaba flowers and Colotis butterflies. However, flower numbers and Colotis spp. abundances were greater in plots with cattle herbivory than in plots that excluded all large herbivores. Our results suggest that wild browsing herbivores can suppress pollinator species whereas well‐managed cattle use may benefit important pollinators and the plants that depend on them. This study documents a novel set of ecological interactions that demonstrate how both conservation and livelihood goals can be met in a working landscape with abundant wildlife and livestock.     

Melliferous potential of Brassica napus L. subsp. napus (Cruciferae)

Morphophysiological characteristics of oilseed rape flowers, such as features of the nectaries, nectar production, and observations on honey bee visits and honey and seed yield were studied with the aim to evaluate the melliferous potential of this crop as well as its attractiveness to pollinators. Calculation of the theoretical maximal honey yield revealed that the actual amount of extracted honey was much lower than the potential yield, indicating that this bee pasture is underutilized. We found that honey bee pollination increased oilseed rape yield, i.e., seed production, by 12 % compared with the treatment in which pollinators were excluded.     

Pollination and reproduction enhance the invasive potential of an early invader: the case of Lythrum salicaria (purple loosetrife) in South Africa

The potential of an alien plant to spread rapidly and colonize new habitat may be related to the mode of reproduction and the ability to attract pollinators. Most studies focus on widespread invasive plants, in which pollinators are rarely limiting. Here, we assess the ability of a recent invader in South Africa, the tristylous Lythrum salicaria to self-reproduce and whether this can explain the delay between introduction and spread. This study was conducted in one of the largest known populations (a total of 7 populations in South Africa) of L. salicaria in the Liesbeek river in the fynbos biome. We assessed the importance of pollinators and autonomous selfing in L. salicaria by comparing seed set between pollinator excluded and naturally pollinated flowers. Overall, 5 pollinators (4 native and 1 alien) were recorded with Cape honeybees and Africa Monarch butterflies the most prominent. Seed and fruit set were significantly higher in open pollinated flowers compared to pollinator excluded flowers. Also, seed and fruit set in pollinator excluded flowers were higher in long and medium morphs compared to short morphs. Germination was high for seeds from pollinator, but also from pollinator excluded treatments. This shows that L. salicaria in South Africa is self-compatible to some extent, but it is frequented by pollinators, significantly increasing seed production. Despite L. salicaria being tristylous, all 3 morphs are present in South Africa and with a huge seed production, this species has the potential to become a major invader of rivers and wetlands in South Africa.

     

盗蜜对角蒿传粉者行为和生殖成功的影响

一些研究显示盗蜜对自交植物的结实和结籽没有显著影响。然而, 对于既有传粉者为其传粉实现异交又能通过自交实现生殖保障的兼性自交植物来说, 盗蜜对其生殖的影响还知之甚少。由于兼性自交植物可以自交, 盗蜜对其总体结实可能不会有显著影响, 但可能会通过影响传粉者行为而影响传粉者介导的结实。为了验证这一假说, 本研究以兼性自交的一年生角蒿(Invarvillea sinensis var. sinensis)为研究材料, 通过野外调查和控制实验, 探讨了盗蜜对传粉者介导的结实(传粉者行为)和总体结实率的影响。结果表明: 角蒿的盗蜜者和主要传粉者相同, 均为密林熊蜂(Bombus patagiatus)。熊蜂盗蜜频率平均为20.24% (范围为0-51.43%)。盗蜜对角蒿总体结实率、每果结籽数和每果种子重量没有显著影响。然而, 被盗蜜花的柱头闭合比率显著高于未被盗蜜花, 说明盗蜜影响传粉者的访花行为和传粉者介导的结实率。另外, 被盗蜜花的高度显著高于未被盗蜜花, 说明盗蜜者倾向于从较大较高的花上盗蜜。这些结果为全面认识盗蜜对植物生殖的影响提供了新的信息。     

Do flowers wave to attract pollinators? A case study with Silene maritima

To answer the question whether flowers wave to attract pollinators, we determine: (1) the heritability of floral mobility; (2) whether wavy flowers attract more insects; (3) does the duration of pollination affect seed set; and (4) the relationship between seed set and floral mobility. The pollination ecology of Silene maritima was investigated. Flowers on stalks of different waviness were used to investigate the effect of floral movement on pollinator visits. There is heritable variation in both direct and indirect estimates of floral mobility. The highest insect total visitation times were associated with medium length thin stalks that were visited more frequently and by more insect species. Although mean individual visit durations were less than those of less mobile flowers, this was compensated for by increased visits. Observations of controlled pollinations show that when the visit times are low, so is seed set and therefore low and high mobility flowers might suffer from reduced fitness. Combining these observations provides a mechanism that could be driving stabilizing selection for flower stalk traits, with a trade-off applying between waving to attract pollinators and not being too mobile as to prevent effective pollination. Further evidence for stabilizing selection is provided by the relationship observed in the field between seed set and floral mobility where the highest levels of fitness was associated with intermediate levels of floral waving.     

The birds and the seeds: opportunistic avian nectarivores enhance reproduction in an endemic montane aloe

Opportunistic bird pollination has become more evident in studies that confirm distinct differences in floral adaptations that attract opportunistic, rather than specialist, bird pollinators. Pollination syndrome studies investigating the effectiveness of different pollinator guilds on reproduction seldom do so by measuring seed viability. We studied pollination in Aloe peglerae, a high altitude endemic succulent of the Magaliesberg mountain range, previously thought to be largely sunbird (specialist) pollinated. Using field observation and pollinator exclusion treatments, i.e. (i) open to all visitors, (ii) bird excluded, and (iii) all visitors excluded, we established that birds contributed significantly more to fruit (2.3–5.6 times) and seed (1.3–1.4 times) set than insect or self-pollination, respectively. Overall, pollination by opportunistic avian nectarivores amplified seed production per aloe ~7 and 10 times compared to insects and self-pollination, respectively. One of three opportunistic nectarivores, the Cape Rock-Thrush (Monticola rupestris), played a significant role in pollination, contributing ~60 % of all probes at inflorescences. The difference in reproductive output between insect visited and visitor excluded flowers was not significant and suggests possible self-pollination in A. peglerae which is particularly unusual in Aloe species. Breeding system experiments would help clarify this. In assessing the effectiveness of pollinator guild on seed viability, we found no differences in percentage seed viability, seed germination or seedling emergence between exclusion treatments. Seed viability and germination were low and variable; however, ~19 % seedling emergence was observed across the treatments. Practically, the net effect of bird pollination may result in 8–12 times more potential seedlings compared to insect and self-pollination respectively. These findings highlight the importance of pollination by opportunistic avian nectarivores in Aloe.     

Floral herbivory at different stages of flower development changes reproduction in <Emphasis Type="Italic">Iris gracilipes</Emphasis> (Iridaceae)

Floral herbivores and pollinators are major determinants of plant reproduction. Because interaction of floral herbivores and pollinators occurs when herbivores attack the flowers in the bud and flower stages and because the compensatory ability of plants is known to differ according to the timing of herbivory, the effects of herbivory may differ according to its timing. In this study, we investigated the effects of floral herbivory at different stages on fruit production and seed/ovule ratio at two sites of large populations of the perennial herb, Iris gracilipes for 2 years. Herbivory at the bud and fruit stages both tended to have negative effects on fruit production, but the former caused more severe damage. On the other hand, herbivory at the flower stage tended not to have negative effects on fruit production because the degree of flower loss was smaller in the flower stage. Although herbivory decreased fruit production, flowers did not compensate for the damage by increasing the seed/ovule ratio because reproduction of I. gracilipes was limited by pollen availability rather than resources. These results indicate that floral herbivory at different stages has different effects on plant reproduction.     

Indirect costs counteract the effects of pollinator-mediated phenotypic selection on corolla size in the Mediterranean shrub Halimium atriplicifolium

Aims Larger corollas receive more pollinator visits but involve higher production and maintenance costs, especially under hot and dry conditions. This can result in indirect costs on reproductive output, which may counteract the effects of pollinator-mediated phenotypic selection on corolla size. In this study, I explored the relationship between corolla size and indirect costs and whether these costs counteract the effects of pollinator-mediated phenotypic selection on this trait in the Mediterranean shrub Halimium atriplicifolium. I hypothesized that (i) corolla production entails direct costs in dry mass, N and P, (ii) corollas entail significant indirect costs in terms of fruit and seed production, (iii) indirect costs increase with corolla size, (iv) this species may suffer pollen limitation to a certain degree and (v) indirect costs counteract the effects of pollinator-mediated selection on corolla size.Methods I compared fruit set and seed production of petal-removed flowers (R flowers) and unmanipulated control flowers (C flowers) and evaluated the influence of individual mean corolla size on relative fruit and seed gain of R compared to C flowers. I also estimated phenotypic selection on corolla size mediated by indirect costs and the combined effect of costs and pollinators (i.e. total selection).Important findings Corollas allocated sizeable amount of resources in terms of dry mass relative to the other floral structures. Fruit set and seed per fruit were significantly higher in R flowers, while individual mean corolla size showed a positive relationship with relative fruit gain. Phenotypic selection analysis revealed cost-mediated negative directional selection and absence of positive directional total selection on corolla size through fruit set. This translated into stabilizing total selection. These results suggest that Mediterranean environments can impose constraints on corolla size, counteracting advantages of larger corollas from the pollination point of view with increased indirect costs of such flowers.     

Pollination by flies, bees, and beetles of Nuphar ozarkana and N. advena (Nymphaeaceae)

Nuphar comprises 13 species of aquatic perennials distributed in the temperate Northern Hemisphere. The European species N. lutea and N. pumila in Norway, the Netherlands, and Germany are pollinated by bees and flies, including apparent Nuphar specialists. This contrasts with reports of predominant beetle pollination in American N. advena and N. polysepala. We studied pollination in N. ozarkana in Missouri and N. advena in Texas to assess whether (1) there is evidence of pollinator shifts associated with floral-morphological differences between Old World and New World species as hypothesized by Padgett, Les, and Crow (American Journal of Botany 86: 1316-1324. 1999) and (2) whether beetle pollination characterizes American species of Nuphar. Ninety-seven and 67% of flower visits in the two species were by sweat bees, especially Lasioglossum (Evylaeus) nelumbonis. Syrphid fly species visiting both species were Paragus sp., Chalcosyrphus metallicus, and Toxomerus geminatus. The long-horned leaf beetle Donacia piscatrix was common on leaves and stems of N. ozarkana but rarely visited flowers. Fifteen percent of visits to N. advena flowers were by D. piscatrix and D. texana. The beetles' role as pollinators was investigated experimentally by placing floating mesh cages that excluded flies and bees over N. advena buds about to open and adding beetles. Beetles visited 40% of the flowers in cages, and flowers that received visits had 69% seed set, likely due to beetle-mediated geitonogamy of 1st-d flowers. Experimentally outcrossed 1st-d flowers had 62% seed set, and open-pollinated flowers 76%; 2nd-d selfed or outcrossed flowers had low seed sets (9 and 12%, respectively). Flowers are strongly protogynous and do not self spontaneously. Flowers shielded from pollinators set no seeds. A comparison of pollinator spectra in the two Old World and three New World Nuphar species studied so far suggests that the relative contribution of flies, bees, and beetles to pollen transfer in any one population depends more on these insects' relative abundances (and in the case of Donacia, presence) and alternative food sources than on stamen length differences between Old World and New World pond-lilies.