| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ecology |
University of Illinois at Chicago, Department of Biological Sciences (M/C 066), 845 W. Taylor Street, Chicago, Illinois 60607 USA
Received for publication November 15, 2006. Accepted for publication July 29, 2007.
ABSTRACT
Seed production may be limited because flowers do not get enough suitable pollen or because plants lack the resources to make seeds. We used replicated plantings to test factors that influence effects of bumblebee behavior on pollen limitation, as measured by the difference in seed set between hand- and naturally pollinated flowers, of Penstemon digitalis in patches of four to 41 flowering individuals. Seed set per flower was 376% higher in the largest as compared with the smallest Penstemon patches. This positive density dependence reflects activity of long-tongued bees, which (1) have higher effective density as patch size increases, (2) visit greater proportions of plants as patch size increases, and (3) visit smaller proportions of flowers per visited plant as patch size increases. Our results suggest that economics of flight and maneuverability of large, long-tongued bumblebees lead them to transfer more pollen between than within Penstemon plants in large patches. Density of smaller, short-tongued bumblebees was not positively associated with Penstemon seed set, but these bees may be important pollinators at low plant densities. Our experimental system indicates a clear positive relationship between activity of effective pollinators and seed set in a species capable of pollinating itself.
Key Words: bumblebee behavior • long-tongued bees • Penstemon • pollen limitation • positive density-dependent pollination • protandry • self-compatibility • short-tongued bees
The number of seeds produced by a plant reflects inherent capacity and environment. Inherent capacity is the number of ovules available for fertilization, while environment includes the pollen available to fertilize those ovules, pollinator effectiveness, and resources available to develop seeds from ovules that are fertilized. A general question in pollination biology is why some individual plants in a population produce far more seeds than others (see Salisbury, 1942
; Silvertown et al., 1993
). A consensus is that pollen limitation, the difference between seeds produced when flowers are hand-pollinated and when they are naturally pollinated by insects or other agents, distinguishes effects on seed production due to pollinator effectiveness from resource limitation. A substantial literature shows that pollen limitation is common among plants that do not self-fertilize (Knight et al., 2005
). The literature is much less clear in demonstrating pollen limitation in plants that can fertilize themselves. Here we experimentally test the hypothesis that foraging responses of bees within a pollinator system can result in pollen limitation in a plant capable of self-fertilization.
Pollinator response to plant or flower density may determine visitation rate and consequently the presence or absence of pollen limitation. The key issue in this paper is how different groups of pollinators respond to plant density. Positive density effects were first noted by Allee and his colleagues, who observed that animal reproduction is often enhanced in small populations as the number of individuals increases (Allee et al., 1949
; Allee, 1951
). An "Allee effect" is also evident in plant reproduction, where positive density dependence increases pollinator abundance and visitation, and increases seed set in grouped as compared with isolated plants (e.g., Sih and Baltus, 1987
; Kunin, 1993
; Ågren, 1996
; Groom, 1998
; see Forsythe, 2003
). Allee effects clearly occur in self-incompatible species that require pollinators for sexual reproduction (see reviews by Larson and Barrett, 2000
; Wilcock and Neiland, 2002
; Knight et al., 2005
). Consequently, isolated plants, such as those growing in fragmented or degraded habitats, undergo pollen limitation that is sometimes severe. An open question is whether plants capable of self-fertilization also suffer diminished pollination success at low densities.
Pollinator behavior has pervasive effects on pollination success. In flower patches of a single species, pollinator visitation rates and pollen quality are expected to be disproportionately low for isolated plants and increase with floral density until a patch becomes saturated with pollinators (Rathcke, 1983
). The effect of low plant density on seed production is most pronounced for self-incompatible flowering plants that cannot attract enough pollinators carrying pollen from other individuals of the same species (Feinsinger et al., 1991
; Steffan-Dewenter and Tscharntke, 1999
; Bosch and Waser, 2001
; see Wagenius, 2006
) or that have so few potential mates that insufficient pollen is available even if pollinators are common (Molano-Flores and Hendrix, 1999
; Ramsey and Vaughton, 2000
; see Caruso, 2002
). Another risk for sparsely distributed plants is that opportunistic pollinators may be ineffective. Insects that find isolated flowers may be incapable of transferring pollen (Stout et al., 1998
; see Waites and Ågren, 2004
), clog stigmas with foreign pollen, or waste pollen of focal plants by carrying it to flowers of other species (Kunin, 1993
; Groom, 1998
; Zorn-Arnold, 2005
). Details of mating system also affect pollination success. Self-compatible plants may fail to be pollinated at low densities if male and female flowers are separated in time or space. In such circumstances, the reproductive assurance normally associated with selfing is compromised because spatial or temporal constraints keep pollinators from connecting potential mates.
We explored the effects of plant density on seed production as a function of pollinator activity in Penstemon digitalis Nutt. (Plantaginaceae), a self-compatible herbaceous perennial with separation of male and female function in time (Dieringer and Cabrera, 2002
). The species is extensively visited by long- and short-tongued bumblebees (Bombus) (Clinebell and Bernhardt, 1998
; Mitchell and Ankeny, 2001
; Dieringer and Cabrera, 2002
). Tongue length and associated body size define the morphologies of two bee guilds that have different foraging behaviors and energy demands (Macior, 1974
; Heinrich, 1976a
, b
; Inouye, 1980
; Harder, 1983
, 1985
). We asked whether abundance or activity of long- and short-tongued bee guilds affects seed production in Penstemon clusters of different densities. We used an existing experimental design and flower manipulations to test whether (1) bumblebee visits reflect plant density; (2) Penstemon seed production differs with plant density; (3) plant density influences use by long- vs. short-tongued bumblebees; and (4) differences in long- and short-tongued bumblebee visitation patterns influence Penstemon seed production. The overarching question is whether Penstemon density determines the use patterns of members of the bumblebee assemblage and in doing so influences seed production.
MATERIALS AND METHODS
Focal plant species and botanical context
Penstemon is a large North American genus with approximately 270 described species (Wolfe et al., 2002
). The genus has conventionally been placed in the Scrophulariaceae, but genetic evidence that the family is polyphyletic puts taxonomy of the group and its presumed relatives into turmoil (Olmsted et al., 2001
). The current status, if not consensus, is that molecular evidence places the genus in a much-expanded family Plantaginaceae (Albach et al., 2005
). The genus Penstemon itself is a coherent taxon within the Tribe Cheloneae.
The genus Penstemon has mixed pollination systems. Most members of the genus are pollinated by insects, 39 species have adaptations for hummingbird use, and a number are pollinated by both (Wilson et al., 2006
). Penstemon digitalis is primarily bee-pollinated, although wasps and butterflies (Clinebell and Bernhardt, 1998
) and hummingbirds (unpublished observation, H.F. Howe) sometimes use the flowers to unknown effect.
Penstemon digitalis (hereafter Penstemon) occurs in moist to mesic prairies throughout the eastern tallgrass region of central North America (Straw, 1956
). The self-compatible, hermaphroditic flowers are protandrous, with 1 d in the male phase then 2 d in the female phase (Dieringer and Cabrera, 2002
). Deep, white tubular corollas are marked with purplish lines inside the throat. Flowers of both male and female stages can be open simultaneously on the same raceme, indicating that geitonogamous selfing is possible. In our sample, ovules per ovary average 168.7 ± 17.9 SE (N = 30). At our site, Penstemon blooms from late May through June. The site is an experimental tallgrass planting (see Experimental design section) occupying one unmowed ha of a 5-ha old field, that is mowed annually.
Experimental design
This pollination study was conducted in the year 2000, using replicated prairie plantings designed to test rodent, planting season, and seeding density effects on tallgrass vegetation (Howe and Brown, 1999
; Howe et al., 2002
). The experiment consisted of 24 plots that were established on a backhoed and disced hayfield at the Morton Arboretum, Lisle, Illinois in 1997. Each of the 14 x 14 m plots was divided into four 7 x 7 m subplots (total of 96 subplots). The vole exclusion experiment had no direct effect on bee activities but over time indirectly favored Penstemon at the expense of competitors that the rodents preferred to eat (Howe et al., 2006
).
Penstemon abundance varied across the site, offering the opportunity to study the association of bumblebee visitation and Penstemon seed production in a simplified set of communities that lacked flowering competitors for bee attention. Of the 96 subplots, the 27 containing Penstemon were used in this study. The nearest area with other plants flowering at this time that were attractive to bumblebees was >1 km away.
Penstemon abundance, neighbor distance, and reproduction
Penstemon abundance is defined as the number of rosettes with at least one (the usual) raceme within each 7 x 7 m subplot. Patches are defined as discrete groupings (>4 m apart) created within subplots (5 x 5 m) when vegetation along fences (>1 m) was cut to control weeds. Groupings were centered within subplots and used by bees as separate patches; related bees prefer to fly <1.5 m between plants (Waser, 1982
).
Penstemon density ranged from four to 41 individuals per 5 x 5 m patch (0.16 to 1.64 plants/m2) plants/m2. There were two patches of six, 15, and 19 plants, and four of seven plants; no other patch sizes were duplicated. Penstemon plants in each patch were individually numbered and mapped to coordinates on a Cartesian plane. A total of 912 seedlings were mapped; 650 reached reproductive status and were used in this study. The X, Y coordinates of adult Penstemon plants were used to determine the mean distance between conspecific neighbors (the mean of all pair wise distances between Penstemon plants) within each patch, and distance to first, second, and third nearest neighbor of all plants within a patch using the general distance formula:
|
| (1) |
Penstemon bloomed from the end of May to mid-June, producing a mean of 137 ± 8.8 flowers per plant in the year 2000 (N = 650 plants). All open flowers from all plants were counted on pollinator observation days either before or after observations. Seeds were collected in July and August as fruits matured. For natural pollination, five fruits were randomly collected from each of five randomly selected plants per patch for a total of 25 fruits per patch, or 675 fruits in all. For the smallest patch (four plants), we collected five fruits from each of four plants. For hand-pollination, five flowers from each of five additional randomly selected plants (when possible) per patch received pollen from plants at least 10 m away.
Hand-pollinations were accomplished using a fine paintbrush saturated with pollen and applied to receptive stigmas. Because early and late-forming flowers generally have low seed set, the flowers and fruits selected for this study excluded the first 15 and last 15 flowers on racemes. By selecting the 100 or so flowers on a raceme with optimal performance capability, we isolated pollinator effects likely to be associated with density dependence from various factors associated with naturally low flower production common in early and late stages of flowering phenology. We averaged seeds per fruit by plant and then by patch. Seeds were removed from their capsules by hand and counted using a Contador Seed Counter (Pfeuffer GmbH, Kitzingen, Germany). Fruit set was determined by counting fruits from all 650 plants. The number of fruits per plant was averaged by patch size. Percent fruit set was determined as (total number of fruits)/(total number of flowers) x 100.
To quantify pollen limitation at the flower level, we compared the number of naturally pollinated seeds (NP) with the number of hand-pollinated seeds (HP) for each selected flower per plant per patch.
|
| (2) |
Bumblebee features
Pollinator species were short-tongued Bombus bimaculatus, B. impatiens, and B. griseocollis and long-tongued B. fervidus, B. pennsylvanicus, and the large carpenter bee Xylocopa virginica, all common in Illinois. Nests of B. bimaculatus, B. fervidus, and B. griseocollis were found on the site; the other species had nests in the adjacent old field. Bees were identified using keys from Heinrich (1979)
and Day (1993)
and were all worker bees with a life expectancy of about 2 wk (Heinrich, 1979
). Queen bees emerged in late March but were not present during this study. Bees foraged only on nectar and coincidentally carried pollen. Bumblebees were the only pollinators observed visiting Penstemon flowers at this site.
At the end of the season, 15 bees of each species were caught in jars and placed on ice to slow metabolism; tongue length was measured while bees were alive. Bees were then euthanized, stored on ice, and later refrigerated. We measured body mass (without nectar content) within 8 h and wingspan and wing area within 72 h. We used derivative measurements of aspect ratio (wingspan2/wing area) and wing loading (body mass/wing area) to assess bee flight efficiency (Norberg, 2002
). A high aspect ratio (long and narrow wings) indicates slow flight and low energy expenditure. High wing loading (large body relative to wing area) indicates long flight capability but poor maneuverability.
Foraging observations
Bees within the 27 flowering patches were observed during peak blooming time from 31 May to 16 June 2000 on sunny days with winds <15 k/h (McCall and Primack, 1992
). Bees in flowering patches were observed for 9 d over a 2-wk period from 0900 to 1500 hours. Each patch received 2 h of observations with equal but randomly assigned morning and afternoon observations.
Movements and time length of each bumblebee foraging bout were recorded by B. Zorn-Arnold. Key data were the individual plants that a bee visited, the number of flowers visited per visited plant, time spent foraging per flower per plant and all subsequent visits to neighboring plants until the bee left the patch or on occasion disappeared from sight. A total of 527 foraging bouts at 5658 individual flowers were noted. We averaged the measures of visitation behavior per bee by bout and by patch.
We defined the effective bumblebee density as the number of bees per plant per patch. Both the number of plants visited and the number of flowers visited per visited plant in each patch were first averaged by the observed bee species and then averaged by guilds reflecting bee morphology (long- or short-tongued bees). Both the proportion of plants visited per patch and the proportion of flowers visited per visited plant per patch allow an indirect assessment of whether bees promote outcrossing or selfing. We assume that bee movement between individual plants increases the likelihood that outcrossed pollen will be distributed to receptive stigmas of other individuals even if some male and female flowers are open simultaneously on a single plant. Time spent per flower per patch reflects the level of nectar reward received per patch (see Harder, 1983
).
Statistical analyses
Relationships between number of plants per patch, bee visitation patterns, and variables associated with seed production were examined using a series of linear and nonlinear regressions.
First, we used separate regressions to test for relationships between patch size and bee response variables, and between bee visitation patterns and Penstemon fecundity. Nonlinear results from quadratic fits are presented if values of higher order terms (X2) are significant in a generalized regression equation y = ß0+ ß1X + ß2X2. No third-order terms were used. Values of r2 roughly reflect the importance of variation in one or more dependent variables attributable to the relationship indicated. Illustrated relationships with untransformed data were confirmed by regressions with transformed data and are summarized in the text. For purposes of visual clarity and thoroughness, figures contain fitted lines and are accompanied by key statistical values from untransformed data.
Second, we used patch size and bee visitation variables that are positively associated with Penstemon fecundity in a multiple regression. Stepwise backward elimination regression with transformed data assessed partial correlation coefficients and corresponding F statistics, which allowed us to measure the strength of the linear relationship between two variables after we controlled for the effects of other variables (Kleinbaum et al., 1998
). This allowed us to determine which variables are most critical to Penstemon seed set at the flower level and also to distinguish between bee variables that actually affected fecundity from variables that influence bee behavior (i.e., patch size). Non-interaction terms retained in the final model are then evaluated for confounding by assessing whether the ß1X coefficient changes meaningfully when potential confounders are removed from the model (Kleinbaum et al., 1998
). Meaningful changes in the ß1X coefficient would indicate the presence of confounding.
Because statistical comparisons of bee variables on seed set involved varied distributions and some nonlinear terms, all bee variables in the regressions were transformed to meet assumptions of linearity. Data positively associated with plant patch size were log transformed (long-tongued bee density, seed set per flower), data negatively associated with plant patch size were square transformed (short-tongued bee density), and proportions were arc sine square-root transformed (proportions of flowers and plants visited, and pollen limitation). Summarized regression equations and the stepwise multiple regression were generated with transformed data and presented in the text. Removal of three outliers did not meaningfully alter results. We used SYSTAT 11 (Systat Software Inc, CA, USA) for all analyses.
RESULTS
Bumblebee morphology
Bee size and morphology fall into two distinct groups (Appendix). Bombus bimaculatus, B. griseocollis, and B. impatiens are small, short-tongued species with mean body masses <65 mg and tongue lengths <31% of body length. Bombus fervidus, B. pennsylvanicus, and Xylocopa virginica are large, long-tongued species with mean body masses >130 mg and tongue lengths >50% of body length. Aspect ratio (wingspan2/wing area) ranges from 3.2 to 3.5 for short-tongued bees and from 1.7 to 2.5 for long-tongued bees. Wing loading (body mass/wing area) ranges from 38.7 to 44.4 for short-tongued bees and from 45.9 to 55.3 for long-tongued bees. See the Appendix for details of bee morphology.
Effective bumblebee density and patch size
Long- and short-tongued bumblebees responded differently to patch size. Effective density of long-tongued (Lt) bees increased with patch size (PS) up to about 20 plants, then saturated as patch size approached 40 plants (r2 = 0.55, F1,23 = 28.4, P < 0.0001, Lt bee density = –5.08 + 1.23PS, Fig. 1). Effective density of short-tongued (St) bees declined asymptotically with patch size [r2 = 0.95, F2,24 = 211.0, P < 0.0001, St bee density = 2.8 – 1.48PS + 0.20(PS x PS)]. Long-tongued bees were absent from the two smallest patches (four and five plants). Short-tongued bees were present in all patches but did not visit any flowers in the fifth largest patch (19 plants). For short-tongued bee density, the smallest patch and largest patches (four and 41 plants) were outliers, perhaps reflecting long-tongued bee activity.
|
|
|
|
The proportion of flowers visited per plant (PF) by bees also influenced seed set per flower. Seed set declined when long-tongued bees visited proportionately more rather than fewer flowers per plant (r2 = 0.62, F1,23 = 37.1, P < 0.0001, seed set = 5.72 – 1.56Lt PF, Fig. 5), probably because they transferred more pollen within plants the more flowers on a plant they visited. Similarly, pollen limitation increased when long-tongued bees visited more rather than fewer flowers per plant (r2 = 0.62, F1,23 = 37.6, P < 0.0001, Poll = –0.29 + 1.27Lt PF). Seed set and pollen limitation were unaffected by the proportion of flowers visited by short-tongued bees (in both cases, r2 < 0.02, P > 0.05).
|
DISCUSSION
Interactions between pollinators and plants are contingent on responses of flower visitors to plant attributes and to the ecological context in which plants occur. For much of the modern history of pollination biology, the focus has been on interactions of pollinators with individual flowers or with flower displays on individual plants (Heinrich and Raven, 1972
). It is now clear that pollinators respond to factors out of the physiological or evolutionary control of individual plants, including weather, cover, spatial arrangement of plants of the same species, and relationships of a focal species of interest with competitors of other species (McCall and Primack, 1992
; Kunin, 1993
; Molano-Flores and Hendrix, 1999
). With recognition of dynamic conditions for response and of the loyalty of flower visitors is the recognition of the pervasive potential for pollen limitation (Larson and Barrett, 2000
; Knight et al., 2005
). Exploration of the effects of plant density on pollinator assemblages is a novel extension of the general perception that spatial effects matter in dynamic interactions among insects and flowers.
Density-dependent seed production
Flower and plant density affects seed production in self-incompatible species from a wide variety of habitats, including temperate understory forest herbs (Tomimatsu and Ohara, 2002
; Knight, 2003
; Zorn-Arnold, 2005
), tropical and Mediterranean shrubs (Silander, 1978
; Feinsinger et al., 1991
; Metcalfe and Kunin, 2006
), grassland perennials (Kéry et al., 2000
; Moody-Weis and Heywood, 2001
; Wagenius, 2006
), a desert perennial (Roll et al., 1997
) and a montane herb (Bosch and Waser, 2001
). Mounting evidence links low seed production in isolated plants and populations to pollinator abundance. Small plant populations do not attract or maintain adequate levels of pollinators to avert serious pollen limitation, with adverse consequences for seed production.
Here we show that density-dependent seed production also occurs in a self-compatible species. In our study, P. digitalis seed set per flower is 376% higher in the largest as compared with the smallest patches. These positive effects on seed set reflect three aspects of relationships between activity of long-tongued bees and increasing patch size: (1) higher effective bee density as patch size increases, (2) greater proportions of plants visited as patch size increases, and (3) smaller proportions of flowers visited per visited plant as patch size increases. Short-tongued bees have little or no role in promoting Penstemon seed set in large patches but may be important for pollination in small patches ignored by long-tongued bees. Our Penstemon system extends well-recognized patterns of bee resource partitioning among species to changes in seed set and flowering plant patch size in a self-compatible prairie perennial.
Pollinator behavior and plant density
Species partition resources by interference or exploitation competition (Miller, 1967
). Unlike some stingless bees and hummingbirds that aggressively displace competitors at flowers (Johnson and Hubbell, 1974
; Hubbell and Johnson, 1978
; Feinsinger, 1976
), bumblebees appear to partition Penstemon patch size by body and wing size, and associated aerodynamic attributes that affect exploitation. Long-tongued bees take more time to extract more nectar from flowers in larger than smaller patches, thereby reducing flight costs and potentially depleting nectar that short-tongued bees might use. Long-tongued bees are more than double the mass of short-tongued bees and have higher total energy demands (Heinrich, 1979
). In our study, aspect ratios of long-tongued bees are 28–46% lower than those of short-tongued bees, while wing loading for long-tongued bees is 16–20% higher. These indicate that long-tongued bees have substantially higher flight costs and lower maneuverability than short-tongued species (see Norberg, 2002
). The large bees apparently have trouble maneuvering around close flowers on a raceme, making small floral patches worth less to visit than large ones where their foraging constraints are not disadvantageous.
We hypothesize that long-tongued bee species select large plant patches to minimize flight costs between patches and further that they visit fewer flowers per plant because moving up and down densely clustered flowers on a single plant is more difficult than moving between flowers on neighboring plants. Per flower visits by long-tongued bees continue to decline at high plant densities, likely in response to resource depletion and ease of moving to neighboring plants rather than struggling with packed flowers on a raceme. For Penstemon, this means more plants visited per patch in large patches and fewer flowers visited on each raceme. Both attributes are statistically associated with higher seed set per flower.
Smaller, short-tongued bees have different opportunities and face different challenges. Aspect ratios indicate that short-tongued species are much more maneuverable, suggesting a mechanism for the apparent ease with which they use flowers on the same raceme. These species are most active in small patches ignored by long-tongued bees. The question arises why short-tongued species do not often use large Penstemon patches in the absence of direct interference from larger competitors. The answer may be that heavy use of large patches by long-tongued species could sufficiently reduce the predictability of nectar rewards to make their use uneconomic by smaller species. In a classic experiment with artificial flowers of different colors, Real (1981)
found that the short-tongued species B. sandersoni was exceedingly sensitive to variance in nectar rewards, switching quickly from the preferred yellow to blue flowers if uncertainty of rewards increased in yellow flowers, even as the mean reward stayed the same in the two "morphs." Our short-tongued species probably avoided large patches where heavy use by larger bees increased the uncertainty of rewards. We hypothesize that short-tongued bees are important pollinators in small Penstemon patches but are unlikely to be important in large patches frequented by other species, including long-tongued competitors.
Implications of density-related seed set
Differences in seed production affect communities by influencing population growth. For long-lived trees and shrubs, high variance in seed production and seedling establishment make differences in seed production much less influential in population growth than death of adults (e.g., Miriti et al., 2001
; Miriti 2006
). For short-lived plants or those in successional transition, seed production directly relates to population growth (e.g., the Mexican rainforest herb, Calathea ovandensis; Horvitz and Schemske, 1995
; see Silvertown et al., 1993
). Allee effects in pollination systems are likely to be demographically influential in plants, like Penstemon, that occupy transitional or periodically disturbed habitats.
Thresholds and local extinction
Pollen limitation has direct and indirect effects. Populations experiencing Allee effects may have critical extinction density thresholds, below which consistent pollination failure causes local extinction of the plant species and flower visitors that depend on it (Lande at al., 1998
; Amarasekare, 2004
). Demographic forces among species experiencing such thresholds brought on by low seed production may doom sparsely distributed plants by attrition without replacement. Pollen limitation may also have indirect effects. Chronic declines in seed production and seedling establishment from sexual reproduction may give an advantage to species that are not reproductively limited (i.e., autogamous selfers, rhizomatous cloners), resulting in asymmetrical competition that shifts species composition (Ashman et al., 2004
; also see Zorn-Arnold et al., 2006
). Direct effects from low fecundity on populations can be further exacerbated by indirect effects of competitive displacement.
Conclusion
Self-compatible species, like many self-incompatible species, may also experience low seed production at low plant densities and high seed production at higher densities. In our Penstemon system, total pollinator abundance is a small part of the reason why positive density dependence in fecundity occurs. For seed set to increase with plant density, pollinators need to visit more plants in a patch and visit proportionately fewer flowers per visited plant, thereby distributing pollen between rather than within individuals. In this experimental population, long-tongued bumblebees are responsible for positive density-dependent effects on seed production, leaving short-tongued bees as the important pollinators in low-density patches. The dramatic differences in the ways in which guilds of the pollinator assemblage respond to plant density underscore the importance of understanding pollen limitation in plant species that differ markedly in patch density and isolation, either naturally or as a consequence of habitat fragmentation.
|
1 The authors thank L. Borghesio, N. Cordeiro, T. Golubski, M. Jorge, A. Taylor Sullivan, and A. Tietmeyer Kramer for helpful comments on the manuscript, and the staff of the Morton Arboretum for enthusiastic support of the project. The authors are grateful for financial support from the University of Illinois Campus Research Board and from the National Science Foundation (DEB 9815289, 9907873). 
2 Author for correspondence (hfhowe{at}uic.edu
; phone: 312-996-0666; fax: 312-413-2435
LITERATURE CITED
Ågren J.. 1996. Population size, pollinator limitation, and seed set in the self-incompatible herb Lythrum salicaria. Ecology 77: 1779-1790..[CrossRef][ISI]
Albach D. C. Meudt H. M. Oxelman B.. 2005. Piecing together the "new" Plantaginaceae. American Journal of Botany 92: 297-315..
Allee W. C.. 1951. The social life of animals, 2nd ed. Beacon Press, Boston, Massachusetts, USA..
Allee W. C. Emerson A. E. Park O. Park T. Schmidt K. P.. 1949. Principles of animal ecology, 2nd ed. W.B. Saunders Co., Philadelphia, Pennsylvania, USA..
Amarasekare P.. 2004. Spatial dynamics of mutualistic interactions. Journal of Animal Ecology 73: 128-142..[CrossRef][ISI]
Ashman T. Knight T. M. Steets J. A. Amarasekare P. Burd M. Campbell M. Dudash M. R. Johnston M. O. Mazer S. J. Mitchell R. J. Morgan M. T. Eilson W. G.. 2004. Pollen limitation of plant reproduction: ecological and evolutionary causes and consequences. Ecology 85: 2408-2421..[CrossRef][ISI]
Bosch M. Waser N. M.. 2001. Experimental manipulation of plant density and its effect on pollination and reproduction of two confamilial montane herbs. Oecologia 126: 76-83..[CrossRef][ISI]
Caruso C. M.. 2002. Influence on plant abundance in pollination and selection on floral traits of Ipomopsis aggregata. Ecology 83: 241-254..[ISI]
Clinebell R. R. Bernhardt P.. 1998. The pollination ecology of five species of Penstemon (Scrophulariaceae) in the tallgrass prairie. Annals of the Missouri Botanical Garden 85: 126-136..[CrossRef][ISI]
Day L.. 1993. Bombus identification. M.S. Thesis, University of Illinois, Chicago, USA..
Dieringer D. Cabrera R. L.. 2002. The interaction between pollinator size and bristle staminode of Penstemon digitalis (Scrophulariaceae). American Journal of Botany 89: 991-997..
Feinsinger P.. 1976. Organization of a tropical guild of nectarivorous birds. Ecological Monographs 46: 257-291..[CrossRef][ISI]
Feinsinger P. Tiebout H. M. III Young B. E.. 1991. Do tropical bird-pollinated plants exhibit density-dependent interactions? Field experiments. Ecology 72: 1953-1963..[CrossRef][ISI]
Forsythe S. A.. 2003. Density-dependent seed set in the Haleakala silversword: evidence for an Allee effect. Oecologia 136: 551-557..[CrossRef][ISI][Medline]
Groom M. J.. 1998. Allee effects limit population viability of an annual plant. American Naturalist 151: 487-496..[CrossRef][ISI]
Harder L. D.. 1983. Flower handling efficiency of bumblebees: morphological aspects of probing time. Oecologia 57: 274-280..[CrossRef][ISI]
Harder L. D.. 1985. Morphology as a predictor of flower choice by bumblebees. Ecology 66: 198-210..[CrossRef][ISI]
Heinrich B.. 1976a. Resource partitioning among some eusocial insects: bumblebees. Ecology 57: 874-889..[CrossRef][ISI]
Heinrich B.. 1976b. The foraging specializations of individual bumblebees. Ecological Monographs 46: 105-128..[CrossRef][ISI]
Heinrich B.. 1979. Bumblebee economics, 1st ed. Harvard University Press, Boston, Massachusetts, USA..
Heinrich B. Raven P. H.. 1972. Energetics and pollination ecology. Science 176: 597-602..
Horvitz C. C. Schemske D. W.. 1995. Spatiotemporal variation in demographic transitions of a tropical understory herb: projection matrix analysis. Ecological Monographs 65: 155-192..[CrossRef][ISI]
Howe H. F. Brown J. S.. 1999. Effects of birds and rodents on synthetic tallgrass communities. Ecology 80: 1776-1781..[CrossRef][ISI]
Howe H. F. Brown J. S. Zorn-Arnold B.. 2002. A rodent plague on prairie diversity. Ecology Letters 5: 30-36..[CrossRef][ISI]
Howe H. F. Sullivan A. T. Zorn-Arnold B. Brown J. S.. 2006. Massive and distinctive effects of meadow voles on grassland vegetation. Ecology 87: 3007-3013..[CrossRef][ISI][Medline]
Hubbell S. P. Johnson L. K.. 1978. Comparative foraging behavior of six stingless bee species exploiting a standardized resource. Ecology 59: 1123-1136..[CrossRef][ISI]
Inouye D. W.. 1980. The effect of proboscis and corolla tube lengths on patterns and rates of flower visitation by bumblebees. Oecologia 45: 197-201..[CrossRef][ISI]
Johnson L. K. Hubbell S. P.. 1974. Aggression and competition among stingless bees: field studies. Ecology 55: 120-127..[CrossRef][ISI]
Kéry M. Matthies D. Spillman H.. 2000. Reduced fecundity and offspring performance in small populations of the declining grassland plants Primula veris and Gentiana lutea. Journal of Ecology 88: 17-30..[CrossRef][ISI]
Kleinbaum D. G. Kupper L. L. Muller K. E. Nizam A.. 1998. Applied regression analysis and other multivariable methods, 3rd ed. Duxbury Press, Pacific Grove, California, USA..
Knight T. M.. 2003. Floral density, pollen limitation, and reproductive success in Trillium grandiflorum. Oecologia 442: 557-563..[CrossRef]
Knight T. M. Steets J. A. Vamosi J. C. Mazer S. J. Burd M. Dampbell D. R. Dudash M. R. Johnston M. O. Mitchell R. J. Ashman T.. 2005. Pollen limitation of plant reproduction: pattern and process. Annual Review of Ecological Systems 36: 467-497..[CrossRef]
Kunin W. E.. 1993. Sex and the single mustard: population density and pollinator behavior affects seed set. Ecology 74: 2145-2160..[CrossRef][ISI]
Lande R. S. Engen S. Sather B. E.. 1998. Extinction times in finite metapopulation models with stochastic local dynamics. Oikos 83: 3060-3072..
Larson B. M. H. Barrett S. C. H.. 2000. A comparative analysis of pollen limitation in flowering plants. Biological Journal of the Linnean Society 69: 503-520..[CrossRef][ISI]
Macior L. W.. 1974. Pollination ecology of the front range of the Colorado Rocky Mountains. Melanderia 15: 1-59..
McCall C. Primack R. B.. 1992. Influence of flower characteristics, weather, time of day, and season on insect visitation in three plant communities. American Journal of Botany 79: 434-442..[CrossRef][ISI]
Metcalfe D. B. Kunin W. E.. 2006. The effects of plant density upon pollination success, reproductive effort and fruit parasitism in Cistus ladanifer L. (Cistaceae). Plant Ecology 185: 41-47..[CrossRef][ISI]
Miller R. S.. 1967. Pattern and process in competition. Advances in Ecological Research 4: 1-74..
Miriti M. N.. 2006. Ontogenetic shift from facilitation to competition in a desert shrub. Journal of Ecology 94: 973-979..[CrossRef][ISI]
Miriti M. N. Wright S. J. Howe H. F.. 2001. The effects of neighbors on the demography of a dominant desert shrub (Ambrosia dumosa). Ecological Monographs 71: 491-509..[ISI]
Mitchell R. J. Ankeny D. P.. 2001. Effects of local conspecific density on reproductive success in Penstemon digitalis and Hesperis matronalis. Ohio Journal of Science 2: 22-27..
Molano-Flores B. Hendrix S. D.. 1999. The effects of population size and density on the reproduction output of Anemone canadensis L. (Ranunculaceae). International Journal of Plant Sciences 160: 759-766..[CrossRef][ISI]
Moody-Weis J. M. Heywood J. S.. 2001. Pollination limitation to reproductive success in the Missouri evening primrose, Oenothera macrocarpa (Onagraceae). American Journal of Botany 88: 1615-1622..
Norberg U. M. L.. 2002. Structure, form, and function of flight in engineering and the living world. Journal of Morphology 252: 52-81..[CrossRef][ISI][Medline]
Olmsted R. G. Depamphilis C. W. Wolfe A. D. Young R. D. Elisons W. J. Reeves P. A.. 2001. Disintegration of the Scrophulariaceae. American Journal of Botany 88: 348-361..
Ramsey M. Vaughton G.. 2000. Pollen quality limits seed set in Burchardia umbellate (Colchiceae). American Journal of Botany 87: 845-852..
Rathcke B.. 1983. Competition and facilitation among plants for pollinators. In L. A. Real, [ed.], Pollination biology, 305-329. Academic Press, New York, New York, USA..
Real L.. 1981. Uncertainty and pollinator–plant interactions: the foraging behavior of bees and wasps on artificial flowers. Ecology 62: 20-26..[CrossRef][ISI]
Roll J. Mitchell R. J. Cabin R. J. Marshall D. L.. 1997. Reproductive success increases with local density of conspecifics in a desert mustard (Lesquerella fendleri). Conservation Biology 11: 738-746..[CrossRef][ISI]
Salisbury E. J.. 1942. The reproductive capacity of plants. Bell, London, UK..
Sih A. Baltus M.. 1987. Patch size, pollinator behavior, and pollinator limitation in catnip. Ecology 68: 1679-1690..[CrossRef][ISI]
Silander J. A. J.. 1978. Density-dependent control of reproductive success in Cassia biflora. Biotropica 10: 292-296..[CrossRef][ISI]
Silvertown J. Franco M. Pisanty I. Mendoza A.. 1993. Comparative plant demography—relative importance of life-cycle components to the finite rate of increase in woody and herbaceous perennials. Journal of Ecology 81: 465-476..[CrossRef][ISI]
Steffan-Dewenter I. Tscharntke T.. 1999. Effects of habitat isolation on pollinator communities and seed set. Oecologia 121: 432-440..[CrossRef][ISI]
Stout J. C. Allen J. A. Goulson D.. 1998. The influence of relative density and floral morphological complexity on the behavior of bumblebees. Oecologia 117: 543-550..[CrossRef][ISI]
Straw M.. 1956. Adaptive morphology of the Penstemon flower. Phytomorphology 6: 112-118..
Tomimatsu H. Ohara M.. 2002. Effects of forest fragmentation on seed production of the understory herb Trillium camchatcense. Conservation Biology 16: 1277-1285..[CrossRef][ISI]
Wagenius S.. 2006. Scale dependence of reproductive failure in fragmented Echinacea populations. Ecology 87: 931-941..[CrossRef][ISI][Medline]
Waites A. R. Ågren J.. 2004. Pollinator visitation, stigmatic pollen loads and among-population variation in seed set in Lythrum salicaria. Journal of Ecology 92: 512-526..[CrossRef][ISI]
Waser N. M.. 1982. A comparison of distances flown by different visitors to flowers of the same species. Oecologia 55: 251-257..[CrossRef][ISI]
Wilcock C. Neiland R.. 2002. Pollination failure in plants: why it happens and when it matters. Trends in Ecology and Evolution 7: 270-277..[CrossRef]
Wilson P. Castellanos M. C. Wolfe A. D. Thomson J. D.. 2006. Shifts between bee and bird pollination in Penstemons. In N. Waser, J. Ollerton, [eds.], Plant–pollinator interactions: from specialization to generalization, 47-68. University of Chicago Press, Chicago, Illinois, USA..
Wolfe A. D. Datwyler S. L. Randle C. P.. 2002. A phylogenetic and biogeographic analysis of the Cheloneae (Scrophulariaceae) based on ITS and matK sequence data. Systematic Botany 27: 138-148..[ISI]
Zorn-Arnold B.. 2005. Positive density dependence and pollination ecology of three bee-pollinated perennials. Ph.D. dissertation, University of Illinois, Chicago, Illinois, USA..
Zorn-Arnold B. Brown J. S. Howe H. F.. 2006. Obvious and cryptic vole suppression of a prairie legume in experimental restorations. International Journal of Plant Sciences 167: 961-968..[CrossRef][ISI]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) vs. short-tongued bees (
) observed visiting Penstemon patches. Long-tongued bee density is positively correlated with patch size (r2 = 0.74, P < 0.0001), whereas short-tongued bee density is negatively correlated with patch size (r2 = 0.82, P < 0.0001). For visual clarity, fitted lines and corresponding statistical values are from untransformed data. Regressions on analyses using transformed data are given in the text.
