The plant life cycle is often affected by animal–plant interactions. In insect-pollinated plants, interaction with pollinators is very important. When pollen transfer due to a lower abundance of pollinators limits seed production, selection pressures on plant traits related to plant attraction to pollinators might occur, e.g. on flowering phenology, height or number of flowerheads. Landscape changes (e.g. habitat fragmentation or changed habitat conditions) may cause plant–pollinator systems to lose balance and consequently affect population dynamics of many plant species. We studied the relationship between measured plant traits, environmental variables and pollinator preferences in Scorzonera hispanica (Asteraceae), a rare perennial, allogamous herb of open grasslands. We estimated the pollen limitation by comparing seed set of supplemental-pollinated plants with that of open-pollinated ones. Pollinators selected plants based on position within the locality (isolated plants close to trees) rather than on their traits. In spite of a high proportion of undeveloped seeds on the plants, we demonstrated that they are not pollen limited. Instead, seed set and weight of seeds was correlated with plant size traits (height and flowerhead number), with larger plants producing more and larger seeds. This suggests that the studied plants are likely resource limited. Overall, the results suggest that pollinators are not a selection factor in this system, in contrast to studies on various plant species, including self-compatible species of the Asteraceae. The lack of any effect of pollinators in the system may be caused by a strong negative effect of ungulate herbivores, which could play a decisive role in functioning of the system.

Introduction

Biotic and abiotic interactions during the flowering period may influence plant flowering strategy due to selection pressures on plant reproductive traits (Elzinga et al. 2007; Ehrlen & Munzbergova 2009). Floral visitors, for example, exert important selection pressure on flowering traits in insect-pollinated plants (Ashman et al. 2004). One necessary condition for the evolution of a plant trait due to pollinators is that the trait is heritable and at the same time relevant for the plant–pollinator relationship. The occurrence of pollen limitation in a species is a second necessary condition for the existence of selection pressure (Ramsey 1995; Knight et al. 2005). Pollen limitation arises when seed production of the maternal plant is limited by pollen receipt (reviewed in Ashman et al. 2004; Knight et al. 2005). Pollen limitation is expected to be more common in obligate outcrossers than in self-compatible plants (Knight et al. 2005). Pollen limitation is also expected in species with specialised pollinators when the species occur in small populations (Agren 1996; Milberg & Bertilsson 1997; Cheptou & Avendano 2006). However, in some cases, pollen limitation may arise even in large populations, e.g. as a consequence of a reduction in abundance of flowerheads by herbivores and in species with generalist pollinators (Pilson 2000 in Liliaceae; Knight 2003 in Asteraceae).

Pollen receipt is not the only factor limiting seed production. Resource availability within a microsite and resource acquisition by a single plant or single inflorescence may also affect seed set (Herrera 1991; Mustajarvi et al. 2001; reviewed in Diggle 1997, 2003).

To understand pollinator selection pressures on plant flowering, we need to separate the effect of limited resource availability from the effect of pollen limitation and determine the role of the pollen limitation. However, studies seldom separate these factors (Agren 1996; Pflugshaupt et al. 2002; Sandring & Agren 2009). The methodical principle is based on comparing open-pollinated plants with those that receive supplemental hand-pollination. Plants not suffering from pollen limitation do not have extra resources available for maturation of ovules fertilised after supplemental pollen receipt, so their seed set does not differ from that of open-pollinated plants. When plants suffer from pollen limitation, supplemental-pollinated plants have a larger seed set than open-pollinated plants (reviewed in Knight et al. 2005).

Pollen limitation combined with pollinator preference exerts selection pressures on flowering traits. Many studies have found significant relationships between attraction of pollinators and floral display, both in field studies (Willson 1979; Ohashi & Yahara 1998) and in manipulation experiments (Andersson 1996; Abraham 2005) for a wide range of plant families (e.g. Asteraceae, Malvaceae). Pollinator abundance often varies over the flowering season, and several studies found pollen limitation especially in the later part of the flowering period (Ramsey 1995 in Liliaceae; Santandreu & Lloret 1999 in Ericaceae; Elzinga et al. 2007 in Polemoniaceae).

There is an assumption that during the evolution of flowering, plants evolved towards an optimal strategy, in which the costs of attraction balance the benefits for seed maturation. It might thus be expected that no strong selection pressures on flowering traits act at any given time. However, when the long-term balance between plants and their pollinators is disrupted, for example due to habitat fragmentation or species diversity loss, pollen limitation may occur and plant–pollinator interactions may represent strong selection pressures (Ghazoul 2005 in Asteraceae; Steffan-Dewenter & Tscharntke 1999 in Brassicaceae).

The aim of this study was to identify selection on traits related to flowering in a rare perennial, allogamous plant species Scorzonera hispanica L. (Asteraceae). Narrow specialist plant–pollinator relationships are rare in the Asteraceae, and pollen transfer by generalist pollinators is more common (Ghazoul 2005; Ellis & Johnson 2009). Nevertheless, several studies have found pollen limitation or pollinator-mediated selection pressures in the Asteraceae, Pilson (2000) in Helianthus annuus and Cheptou & Avendano (2006) in Crepis sancta, but see the experimental studies on Heterotheca subaxillaris (Olsen 1997) and Centaurea scabiosa (Ehlers 1999).

Scorzonera hispanica in our study area of fragmented dry grasslands typically has a high proportion of aborted seeds. Previous studies in the system demonstrated that the populations are highly genetically variable and that the high proportion of aborted seeds is thus probably not a consequence of inbreeding (Munzbergova & Plackova 2010). These facts raise the question of whether undeveloped seeds within flowerheads are caused through pollen limitation or a lack of resources. Pollen limitation in the system is likely to arise due to strong turnover of land use in the area in the last century, leading to high levels of fragmentation and thus isolation of the single habitats (Chylova & Munzbergova 2008). Many species, including S. hispanica, that are restricted to fragments of dry grasslands are currently rare in the landscape, with only a few populations in the region (Knappova et al. 2012). In addition, the selection on flowering traits in the system is likely affected by high and variable levels of mammalian herbivory, the abundance of which has recently increased in the area (Hemrova et al. 2012). Simultaneously, there is a high variability in plant traits (mainly height and number of flowerheads), which might affect attraction of plants to pollinators. To identify the potential selection pressures in this system, we asked the following questions: (i) is seed set of S. hispanica limited by pollen receipt or resource availability; and (ii) what are the criteria for pollinator choice and, consequently, potential selection pressures on species reproductive traits, including floral display and flowering phenology? To answer these questions, we studied the relationship between measured plant traits and pollinator preferences on tagged plants in the field. To estimate the rate of pollen limitation, we compared supplemental-pollinated plants with the open-pollinated ones.

Material and Methods

Study system

The study system is situated in an area of dry grasslands in northern Bohemia, Czech Republic. In the past, the area was covered with a fine-scale mosaic of pastures and fields. At present, large areas of arable fields surround the remaining grasslands. Most of the localities are now abandoned, forming a mosaic of grasslands with expanding shrubs and trees (Chylova & Munzbergova 2008). The flower stalks of S. hispanica are often grazed by ungulates (Hemrova et al. 2012).

The study population is situated at Holy vrch, which is a mild, south-facing slope with a mosaic of open grasslands and shrubs, and represents one of the largest populations in the area with ca. 1600 flowering individuals. The dry grasslands can be classified as belonging to the Bromion erecti Koch 1926 community (Ellenberg 1988), and are undergoing a slow succesional process towards oak or hornbeam forest (Chylova & Munzbergova 2008). The locality is seldom visited by people and provides enough space and plant individuals for manipulative experiments.

Scorzonera hispanica L. (Asteraceae) is a perennial herb, the centre of its distribution range being the Iberian Peninsula, with scattered occurrence in dry grasslands in Central and Southern Europe. Occasionally it is cultivated for its edible root and is locally naturalised (Chater 1976). In the study area it is, however, considered native. Rare occurrences of distinct populations in private gardens are possible, but we do not have any information on this.

The plant has a single rosette and one flowering stalk with one to seven yellow flowerheads, opening successively from the uppermost to the lowest. The flowers within flowerheads open from the outer towards the centre over 3–5 days, depending on the weather (personal observation). The flowering period in the study region, northern Bohemia, Czech Republic, is from late May to the beginning of July, with a peak of flowering in June. In the study area, the most common flower visitors are beetles (Mordellidae, Buprestidae, Dasytidae, Oedemeridae, Cerambycidae and Chrysomelidae) and bees (Apidae, Megachilidae and Halictidae; Červenkova & Münzbergová personal observations). While many species of the Asteraceae may be autogamous or even apomictic, previous experiments with the studied species indicate no developed seeds arise without pollen transfer. The plant can thus be classified as self-compatible but not capable of spontaneous selfing, so a pollinator is needed in all cases (Banga 1961; Munzbergova & Plackova 2010).

Field experiment

In the peak of the flowering period, from 10 to 18 June 2010, we chose and tagged 204 fertile plants in the same phase of opening of the upper (first opening) flowerhead. A total of 99 plants were monitored in the first run (10–13 June), and a second group of 105 plants was monitored in the second run (14–18 June). We started the experiment on the first day of opening of the upper flowerheads. Because of the possibility of different fecundity of flowerheads within the plant (personal observation; reviewed in Herrera 1991; Diggle 1997, 2003), we only worked with the upper flowerheads (one flowerhead per plant). For each plant, we recorded height, number of flowerheads, length of the upper flowerhead (measured as length of the whole closed flowerhead including the ovary), number of open flowerheads of S. hispanica within 1 m, height (cm) and cover (%) of surrounding herbs within a 0.5-m radius and occurrence of trees or shrubs within a 1.0-m radius. Different radii were chosen for herbs or shrubs and trees because of their different heights and thus different assumed impacts on microclimate, plant performance and pollinator preference. The very rare co-occurrences of a shrub and a tree within the radius around a plant were recorded as ‘tree’. For the number of surrounding open flowerheads, we used the mean number for the whole 3-day period of the observation.

The flowerheads opened every day from about 06:30 to 11:00 h, depending on the weather, which was sunny or partly cloudy throughout the whole experiment. During that time, each flowerhead was observed for pollinator visits every 1.5 h for a total of 30 s (three times per day per flowerhead on average). The number and species (or higher taxonomic group) of visitors were recorded. Flowerhead visitors were identified in the field or collected for later identification. The visitation rate was estimated as the number of observed visitors per 30-s period.

To estimate the impact of pollen limitation, half of the tagged plants were randomly selected for supplemental hand-pollination. The plants received hand-pollination every day of the observation in the period in relation to the highest length and wetness of the stigma, which indicates its receptivity. The pollen was transferred (after the 30 s observation period) to the stigmas using a paintbrush, from at least three pollen donor plants randomly chosen among plants 2–10 m from the focal plant.

The pollinator monitoring and supplemental hand-pollination was repeated every day until all flowers in the flowerheads had withered (3–4 days). At the end of the flowering, the seeds were left to mature and collected 3 weeks later. Thereafter, we recorded the number of developed (visually full) and aborted seeds and the weight of developed seeds per flowerhead. Mean seed weight was estimated on the basis of the number and weight of developed seeds within each flowerhead. When measuring seed weight, we used whole achenes, i.e. seeds including carpels (pappus). We were unable to analyse 32 out of the 204 studied plants because there were browsed by ungulate herbivores.

Statistical analyses

The role of single variables in the study system was tested with path analyses (structural equation modelling) using the amos 5 software (Small Waters Corp., Chicago, IL, USA). We designed three models for the number of developed seeds, proportion of developed seeds and mean seed weight as dependent variables. The causal relationship among plant height, number of flowerheads, length of upper flowerhead, number of open flowerheads of S. hispanica within 1.0 m, height and cover of the surrounding vegetation within a 0.5-m radius, occurrence of trees or shrubs within a 1.0-m radius, insect visitation rate and the supplemental pollination, and plant performance (number of developed seeds, proportion of developed seeds and mean seed weight) were estimated. Simultaneously, vegetation cover, individual plant height and flowerhead number were each affected by a latent variable representing the residual variation. For the structure of models, see Fig. 1. The significance of the relationships was estimated using the generalised least squares method because the residuals were over-dispersed. Length of the flowerhead was not significant in any model, so it was removed from the final models. The surrounding vegetation height and vegetation cover were closely correlated. As vegetation cover performed better in the model, we did not incorporate vegetation height into the final models.

Details are in the caption following the image — **Figure 1**
Open in figure viewer PowerPoint

Path model depicting the hypothesised causal relationships between environmental variables, plant traits, pollen transfer and resulting seed number. Width of each arrow is proportional to the standardised path coefficients (see legend for scale), and dotted lines indicate negative paths. Asterisks indicate values significantly different from 0 (P < 0.05). Numbers in bold are estimates of the proportion of total variance explained by all relationships (squared multiple correlations) for each dependent (endogenous) variable. Variables and paths representing unmeasured residual variation are not included for simplicity. The model fit was significant (χ² = 35.5, df = 14, P = 0.01)

Because the model for the number of developed seeds and proportion of developed seeds gave very similar results, only results for the number of developed seeds is shown in the results. See the Supplementary information for other diagrams showing results for mean seed weight and proportion of developed seeds. In addition to the above-described tests, we performed several other tests. Their results, however, did not differ from the results of the above tests and are thus not presented. Specifically, we included insect visitation rate separately for different pollinator functional groups (Coleoptera + Heteroptera, Hymenoptera and Diptera) in the model. We also performed all the tests only for plants from open grasslands, excluding plants under the woody cover from the data set (64 out of 168 plants were excluded). Finally, we also performed all the tests separately for the single observation runs and single observation days.

Results

The flowerheads contained from 0 to 90% developed seeds (37 ± 2%, mean ± SE); the remaining the seeds was aborted. No seeds appeared damaged by insect seed predators. In total, 24 different insect species were observed visiting flowers of S. hispanica at the locality (Table S1). Because of flowerhead morphology, all the insect species were considered as potentially effective pollinators.

The number of developed seeds per flowerhead and proportion of developed seeds per flowerhead were significantly affected by plant height (Figs 1 and 2) and by the number of flowerheads. Neither the insect visitation rate nor the supplemental pollination had any significant impact on the seed set. However, both plant height and flowerhead number were affected by vegetation cover and the occurrence of trees and shrubs. This might be evidence for a limitation from resource availability rather than from pollen transfer (Fig. 1). The insect visitation rate was negatively affected by the number of surrounding flowerheads of S. hispanica (Figs 1 and 3). The visitation rate was higher in flowerheads close to a tree. In the proximity of shrubs or in open vegetation, the visitation rate was lower (Fig. 4).

Insect visitation rate was independent of plant height and slightly positively affected by flowerhead number. The overall model with number of developed seeds per flowerhead was significant (χ² = 35.5, df = 14, P = 0.01). The model explaining the proportion of developed seeds (Figure S1) had a very similar fit (χ² = 35.6, df = 14, P = 0.01), and causal relationships were the same, with very similar standardised path coefficients, except for the non-significant impact of the number of flowerheads on the proportion of developed seeds.

The mean seed weight per flowerhead was not significantly affected by any related variable (Figure S2). As in previous models, plant height and flowerhead number were affected by the vegetation cover and occurrence of trees and shrubs; moreover, the vegetation cover and occurrence of trees and shrubs were correlated. The causal relationships among independent variables and insect visitation rate were the same as in the previous models. The visitation rate differed between the first and second experimental run, with a 2.2-fold higher visitation rate in the first experimental run (R² = 0.055, P < 0.001; Figure S3).

Discussion

This study demonstrated several significant relationships among the environmental variables, plant traits and seed set. The plant visitation by pollinators was affected by several environmental variables, but pollinators did not have any direct impact on seed set. On the basis of the positive relationship between resource availability (evaluated as surrounding vegetation cover) and plant height and flowerhead number per plant, we suggest that seed set in the system is resource-limited rather than pollen-limited. This conclusion is also supported by the fact that hand-pollination did not increase seed set in S. hispanica. Pollinators thus probably do not represent an important selection agent in this system.

The results also suggest that directional selection on the flowering traits does not occur here. S. hispanica does not suffer from pollen limitation, and the pollinator interactions with plant traits are rather weak. These results contradict results of studies using similar experimental methods to detect pollen limitation in Lythrum salicaria (Lythraceae), Prunus mahaleb (Rosaceae) and Arabidopsis lyrata (Brassicaceae), respectively (Agren 1996; Pflugshaupt et al. 2002; Sandring & Agren 2009). On the contrary, the results are in agreement with Ehlers (1999), who found no pollen limitation in C. scabiosa (Asteraceae), or Olsen (1997), who excluded pollen limitation in H. subaxillaris (Asteraceae) using the hand-pollination method. This may indicate that species from the Asteraceae are less likely to be pollen-limited than plants from other families due to their generalist flower morphology.

Pollen limitation is less common in self-compatible species than in species that are self-incompatible (Mustajarvi et al. 2001). Milberg & Bertilsson (1997), however, confirmed pollen limitation even in a self-compatible species. S. hispanica is self-compatible but it is not capable of spontaneous self-pollination, so pollinators are required for successful selfing. Pollen limitation could thus theoretically also occur in the self-compatible S. hispanica. Reasons for the lack of such limitation are discussed below.

The diversity of pollinators in our system was quite high. Franzen & Larsson (2009) found opposite impact of different groups of pollinators on Knautia arvensis (Dipsacaceae); therefore we also separately tested the effect of each group of pollinators on S. hispanica, but no difference between pollinator groups was detected. As stated above, this may be related to the generalist flower morphology within the Asteraceae.

In spite of the failure to detect pollinator limitation in the system, varying abundance of flower visitors was observed during the experiment. Several studies have shown that pollen limitation varies within a season, being highest either at the beginning (Ramsey 1995) or at the end of the flowering period (Widen 1991; O'Neil 1999; Santandreu & Lloret 1999; Elzinga et al. 2007; Weber & Kolb 2011). This variation is likely because of variations in pollinator abundance, abundance of co-flowering species or abundance of antagonists (Ehrlen & Munzbergova 2009). Although the number of pollinators in our study decreased during the season, we do not consider selection for earlier flowering because pollinators do not seem to be a limiting factor in the system.

In our system, pollinators responded to the surrounding environment (i.e. the surrounding flowerheads or surrounding vegetation cover) rather than to the traits of the plants. The pollinator preference of more isolated plants with less surrounding flowerheads contrasted with similar studies that found an opposite pattern (Caruso 2002; Torang et al. 2006). However, our conclusion concerning the importance of surrounding vegetation for the pollinator visitation rate is congruent with conclusions in the review of Ghazoul (2005): that in areas with a lower density of flowerheads, pollinators spend more time on a single plant, and therefore the probability of recording the pollinator on a flowerhead is higher. In addition, the longer time spent on an individual flowerhead may translate into the possibility of increased pollen removal (Harder 1990; Neff & Simpson 1990), more pollen deposition on stigmas and/or more florets per flowerhead being successfully pollinated (Neff & Simpson 1990).

The absence of pollen limitation and pollinator-mediated selection pressure could indicate a balance in the plant–pollinator system at the locality studied. This explanation is tenable because the locality is still rather undisturbed, large and species-rich. However, the changes in the surrounding landscape are striking: an expanding urbanised area, development of photovoltaic plants, changes in the amount of intensively managed fields and succession in abandoned places (see also Chylova & Munzbergova 2008; Knappova et al. 2012). These processes drive changes in the ambient environment and population dynamics at many localities of S. hispanica and also present a serious also for the system at the locality. On the other hand, the absence of pollen limitation and pollinator-mediated selection could also be caused by a coincidence of mutualistic and antagonistic animal–plant interactions in the system, as the plants are heavily grazed by ungulates (Hemrova et al. 2012).

Opposing selection from mutualists and antagonists has been found, by Ehrlen et al. (2012). According to their study, some plant traits, including inflorescence height, can influence the total seed production both positively and negatively via pollinator and seed predator preferences. In another study, on Erysimum mediohispanicum (Cruciferae) grazed by ibex, a significant selection on flowering traits (e.g. flower number, plant height, petal length) was observed when the grazing ungulates were absent. When the ungulates were present, selection on floral traits completely disappeared (Gomez 2003). In our study system, the rate of ungulate herbivory (mainly by roe deer) is high. The role of pollinators could thus theoretically change between localities with different herbivore pressure. Vanhoenacker et al. (2013) suggest a decreasing role of selection mediated by pollinators with an increasing intensity of interaction, whereas selection mediated by antagonists increases together with the intensity of the interaction. According to this study, corroborated in the results of Hemrova et al. (2012), who found a rate of herbivory between 40 and 100% among localities of S. hispanica, we suppose that pollen limitation or pollinator-mediated pressure occurs in less browsed localities. A follow-up study at the landscape level would be needed to explore this.

The high proportion of undeveloped seeds in our study can most likely be explained by limited resource acquisition of the plant. This finding is supported by the significant relationship between the number of developed seeds and maternal plant height, and also between mean seed weight and flowerhead length in this study (Herrera 1991 for similar findings). In addition, Munzbergova & Plackova (2010) demonstrated, in the same system, that seed weight in S. hispanica was significantly affected by habitat conditions, while Munzbergova (2006) demonstrated that seed number increased with site productivity. This expectation is also confirmed in the significant decrease in seed number and mean seed weight per flowerhead from the uppermost to the lower flowerheads (personal observation).

In general, our results suggest that S. hispanica does not experience any selection pressure on the part of pollinators. Resource limitation is likely stronger than pollen limitation in this system. The realized preferences can hardly cause any selection pressure because pollinator choice was not affected by plant traits, but simply by the ambient environment. This situation is not static, however; any shift in population density, the rate of herbivory or the slightest change in landscape dynamics can disrupt the present functioning of the system.

Acknowledgements

The project was supported by grants GAUK 64709, P504/10/0456 and partly by RVO 67985939. We thank the entomologists Jiri Skuhrovec and Jakub Straka, who kindly helped with pollinator determinations, and one anonymous reviewer who provided many helpful comments on the manuscript.

Supporting Information

References

Abraham J.N. (2005) Insect choice and floral size dimorphism: sexual selection or natural selection? Journal of Insect Behavior, 18, 743–756.
10.1007/s10905-005-8737-1
Web of Science® Google Scholar
Agren J. (1996) Population size, pollinator limitation, and seed set in the self-incompatible herb Lythrum salicaria. Ecology, 77, 1779–1790.
10.2307/2265783
Web of Science® Google Scholar
Andersson S. (1996) Floral display and pollination success in Senecio jacobaea (Asteraceae): interactive effects of head and corymb size. American Journal of Botany, 83, 71–75.
10.1002/j.1537-2197.1996.tb13876.x
Web of Science® Google Scholar
Ashman T.L., Knight T.M., Steets J.A., Amarasekare P., Burd M., Campbell D.R., Dudash M.R., Johnston M.O., Mazer S.J., Mitchell R.J., Morgan M.T., Wilson W.G. (2004) Pollen limitation of plant reproduction: ecological and evolutionary causes and consequences. Ecology, 85, 2408–2421.
10.1890/03-8024
Web of Science® Google Scholar
Banga O. (1961) Breeding Scorzonera hispanica L by Polycross Method. Euphytica, 10, 49–58.
10.1007/BF00037205
Web of Science® Google Scholar
Caruso C.M. (2002) Influence of plant abundance on pollination and selection on floral traits of Ipomopsis aggregata. Ecology, 83, 241–254.
10.1890/0012-9658(2002)083[0241:IOPAOP]2.0.CO;2
Web of Science® Google Scholar
Chater A.O. (1976) Scorzonera. In: T.G. Tutin, V.H. Heywood, N.A. Burges, D.H. Valentine, S.M. Walters, D.A. Webb (Eds), Flora Europaea 4. Cambridge University Press, Cambridge, UK, pp 466–467.
Google Scholar
Cheptou P.O., Avendano L.G. (2006) Pollination processes and the Allee effect in highly fragmented populations: consequences for the mating system in urban environments. New Phytologist, 172, 774–783.
10.1111/j.1469-8137.2006.01880.x
CAS PubMed Web of Science® Google Scholar
Chylova T., Munzbergova Z. (2008) Past land use co-determines the present distribution of dry grassland plant species. Preslia, 80, 183–198.
Web of Science® Google Scholar
Diggle P.K. (1997) Ontogenetic contingency and floral morphology: the effects of architecture and resource limitation. International Journal of Plant Sciences, 158, S99–S107.
10.1086/297510
Web of Science® Google Scholar
Diggle P.K. (2003) Architectural effects on floral form and function: A review. In: T. Stuessy, E. Horandl, V. Mayer (Eds), Deep Morphology: Toward a Renaissance of Morphology in Plant Systematics. Koeltz, Konigstein, Germany, pp 63–80.
Web of Science® Google Scholar
Ehlers B.K. (1999) Variation in fruit set within and among natural populations of the self-incompatible herb Centaurea scabiosa (Asteraceae). Nordic Journal of Botany, 19, 653–663.
10.1111/j.1756-1051.1999.tb00675.x
Web of Science® Google Scholar
Ehrlen J., Munzbergova Z. (2009) Timing of flowering: opposed selection on different fitness components and trait covariation. American Naturalist, 173, 819–830.
10.1086/598492
PubMed Web of Science® Google Scholar
Ehrlen J., Borg-Karlson A.K., Kolb A. (2012) Selection on plant optical traits and floral scent: effects via seed development and antagonistic interactions. Basic and Applied Ecology, 13, 509–515.
10.1016/j.baae.2012.08.001
Web of Science® Google Scholar
Ellenberg H. (1988) Vegetation ecology of Central Europe. Cambridge University Press, Cambridge, UK.
Google Scholar
Ellis A.G., Johnson S.D. (2009) The Evolution of Floral Variation without Pollinator Shifts in Gorteria diffusa (Asteraceae). American Journal of Botany, 96, 793–801.
10.3732/ajb.0800222
PubMed Web of Science® Google Scholar
Elzinga J.A., Atlan A., Biere A., Gigord L., Weis A.E., Bernasconi G. (2007) Time after time: flowering phenology and biotic interactions. Trends in Ecology & Evolution, 22, 432–439.
10.1016/j.tree.2007.05.006
PubMed Web of Science® Google Scholar
Franzen M., Larsson M. (2009) Seed set differs in relation to pollen and nectar foraging flower visitors in an insect-pollinated herb. Nordic Journal of Botany, 27, 274–283.
10.1111/j.1756-1051.2009.00348.x
Web of Science® Google Scholar
Ghazoul J. (2005) Pollen and seed dispersal among dispersed plants. Biological Reviews, 80, 413–443.
10.1017/S1464793105006731
CAS PubMed Web of Science® Google Scholar
Gomez J.M. (2003) Herbivory reduces the strength of pollinator-mediated selection in the Mediterranean herb Erysimum mediohispanicum: consequences for plant specialization. American Naturalist, 162, 242–256.
10.1086/376574
PubMed Web of Science® Google Scholar
Harder L.D. (1990) Pollen removal by bumble bees and its implications for pollen dispersal. Ecology, 71, 1110–1125.
10.2307/1937379
Web of Science® Google Scholar
Hemrova L., Cervenkova Z., Munzbergova Z. (2012) The effects of large herbivores on the landscape dynamics of a perennial herb. Annals of Botany, 110, 1411–1421.
10.1093/aob/mcs064
PubMed Web of Science® Google Scholar
Herrera J. (1991) Allocation of reproductive resources within and among inflorescences of Lavandula stoechas (Lamiaceae). American Journal of Botany, 78, 789–794.
10.1002/j.1537-2197.1991.tb14480.x
Web of Science® Google Scholar
Knappova J., Hemrova L., Munzbergova Z. (2012) Colonization of central European abandoned fields by dry grassland species depends on the species richness of the source habitats: a new approach for measuring habitat isolation. Landscape Ecology, 27, 97–108.
10.1007/s10980-011-9680-5
Web of Science® Google Scholar
Knight T.M. (2003) Floral density, pollen limitation, and reproductive success in Trillium grandiflorum. Oecologia, 137, 557–563.
10.1007/s00442-003-1371-8
PubMed Web of Science® Google Scholar
Knight T.M., Steets J.A., Vamosi J.C., Mazer S.J., Burd M., Campbell D.R., Dudash M.R., Johnston M.O., Mitchell R.J., Ashman T.L. (2005) Pollen limitation of plant reproduction: pattern and process. Annual Review of Ecology, Evolution and Systematics, 36, 467–497.
10.1146/annurev.ecolsys.36.102403.115320
Web of Science® Google Scholar
Milberg P., Bertilsson A. (1997) What determines seed set in Dracocephalum ryuschiana L. an endangered grassland plant. Flora, 192, 361–367.
10.1016/S0367-2530(17)30806-X
Web of Science® Google Scholar
Munzbergova Z. (2006) Effect of population size on the prospect of species survival. Folia Geobotanica, 41, 137–150.
10.1007/BF02806475
Web of Science® Google Scholar
Munzbergova Z., Plackova I. (2010) Seed mass and population characteristics interact to determine performance of Scorzonera hispanica under common garden conditions. Flora, 205, 552–559.
10.1016/j.flora.2010.04.001
Web of Science® Google Scholar
Mustajarvi K., Siikamaki P., Rytkonen S., Lammi A. (2001) Consequences of plant population size and density for plant–pollinator interactions and plant performance. Journal of Ecology, 89, 80–87.
10.1046/j.1365-2745.2001.00521.x
Web of Science® Google Scholar
Neff J.L., Simpson B.B. (1990) Pollination biology of wild sunflower (Helianthus annuus L., Asteraceae). Israel Journal of Botany, 39, 197–216.
Web of Science® Google Scholar
Ohashi K., Yahara T. (1998) Effects of variation in flower number on pollinator visits in Cirsium purpuratum (Asteraceae). American Journal of Botany, 85, 219–224.
10.2307/2446309
CAS PubMed Web of Science® Google Scholar
Olsen K.M. (1997) Pollination effectiveness and pollinator importance in a population of Heterotheca subaxillaris (Asteraceae). Oecologia, 109, 114–121.
Web of Science® Google Scholar
O'Neil P. (1999) Selection on flowering time: an adaptive fitness surface for nonexistent character combinations. Ecology, 80, 806–820.
10.1890/0012-9658(1999)080[0806:SOFTAA]2.0.CO;2
Web of Science® Google Scholar
Pflugshaupt K., Kollmann J., Fischer M., Roy B. (2002) Pollen quantity and quality affect fruit abortion in small populations of a rare fleshy-fruited shrub. Basic and Applied Ecology, 3, 319–327.
10.1078/1439-1791-00126
Web of Science® Google Scholar
Pilson D. (2000) Herbivory and natural selection on flowering phenology in wild sunflower, Helianthus annuus. Oecologia, 122, 72–82.
10.1007/PL00008838
CAS PubMed Web of Science® Google Scholar
Ramsey M. (1995) Causes and consequences of seasonal variation in pollen limitation of seed production in Blandfordia grandiflora (Liliaceae). Oikos, 73, 49–58.
10.2307/3545724
Web of Science® Google Scholar
Sandring S., Agren J. (2009) Pollinator-mediated selection on floral display and flowering time in the perennial herb Arabidopsis lyrata. Evolution, 63, 1292–1300.
10.1111/j.1558-5646.2009.00624.x
CAS PubMed Web of Science® Google Scholar
Santandreu M., Lloret F. (1999) Effect of flowering phenology and habitat on pollen limitation in Erica multiflora. Canadian Journal of Botany – Revue Canadienne De Botanique, 77, 734–743.
10.1139/cjb-77-5-734
Web of Science® Google Scholar
Steffan-Dewenter I., Tscharntke T. (1999) Effects of habitat isolation on pollinator communities and seed set. Oecologia, 121, 432–440.
10.1007/s004420050949
CAS PubMed Web of Science® Google Scholar
Torang P., Ehrlen J., Agren J. (2006) Facilitation in an insect-pollinated herb with a floral display dimorphism. Ecology, 87, 2113–2117.
10.1890/0012-9658(2006)87[2113:FIAIHW]2.0.CO;2
CAS PubMed Web of Science® Google Scholar
Vanhoenacker D., Agren J., Ehrlen J. (2013) Non-linear relationship between intensity of plant–animal interactions and selection strength. Ecology Letters, 16, 198–205.
10.1111/ele.12029
CAS PubMed Web of Science® Google Scholar
Weber A., Kolb A. (2011) Evolutionary consequences of habitat fragmentation: population size and density affect selection on inflorescence size in a perennial herb. Evolutionary Ecology, 25, 417–428.
10.1007/s10682-010-9430-1
Web of Science® Google Scholar
Widen B. (1991) Phenotypic selection on flowering phenology in Senecio integrifolius, a perennial herb. Oikos, 61, 205–215.
10.2307/3545338
Web of Science® Google Scholar
Willson M.F. (1979) Sexual selection in plants. American Naturalist, 113, 777–790.
10.1086/283437
Web of Science® Google Scholar

Citing Literature

Volume16, Issue5

September 2014

Pages 967-972

Filename	Description
plb12142-sup-0001-FigS1.jpgimage/jpg, 48.3 KB	Figure S1. Path model depicting the hypothesised causal relationships between environmental variables, plant traits, pollen transfer and resulting developed seed ratio.
plb12142-sup-0002-FigS2.jpgimage/jpg, 47.6 KB	Figure S2. Path model depicting the hypothesised causal relationships between environmental variables, plant traits, pollen transfer and resulting mean seed weight.
plb12142-sup-0003-FigS3.jpgimage/jpg, 12.3 KB	Figure S3. Comparison of insect visitation rate between the two observation runs.
plb12142-sup-0004-TableS1.docxWord document, 14.3 KB	Table S1. List of the natural floral visitors on S. hispanica, showing the number and capture date of the determined individuals.

Pollen limitation and pollinator preferences in Scorzonera hispanica

Abstract

Introduction