Thesis (Ph. D.)--University of Wisconsin, 1965. / Includes bibliographical references.

Thesis (PhD)--Stellenbosch University, 2014. / ENGLISH ABSTRACT: Insect pollination contributes in various degrees toward the production of a variety of agricultural crops that ensure diversity and nutritional value in the human diet. Although managed honeybees (Apis mellifera L.) are still the most economically valuable pollinators of monoculture crops cultivated globally, wild pollinator communities can contribute substantially toward crop pollination through pollination ecosystem services sourced from neighbouring natural habitats. Pollination ecosystem services are thus valuable and can motivate for the protection of natural ecosystems hosting diverse insect pollinator communities. F1 onion hybrid seed production is entirely dependent on high insect pollinator activity to ensure cross pollination, seed set and profitable seed yields. Data was collected on 18 onion hybrid seed crops grown in the semi‐arid Klein Karoo and southern Karoo regions of the Western Cape, South Africa. These two main production regions are located within the Succulent Karoo biome, recognized as a global biodiversity hotspot of especially high plant diversity. It is also habitat to the indigenous Cape honeybee (Apis mellifera capensis Esch.). Sites selected varied in the percentages of available natural habitat and managed honeybee hives stocking density. Diverse anthophile assemblages were sampled with pan traps within all the onion fields, regardless of the percentage of available natural habitat near the crop. Crop management practices significantly affected the diversity of anthophile species caught within onion fields, although less than 20% of this diversity was observed actually visiting onion flowers. The honeybee (managed and wild) was by far the most important pollinator because of its high visitation frequency and regular substantial onion pollen loads carried on their bodies. Honeybee visitation significantly increased onion hybrid seed yield, while anthophile diversity and non‐Apis visitation had no effect on seed yield. Neither managed hive density, nor percentage natural habitat were important in determining honeybee visitation or seed yield. Total annual rainfall was the only significant factor determining honeybee visitation. Secondary factors caused by rainfall variability, such as wild flower abundance or soil moisture, may have significantly affected honeybee visitation. In addition, the positive correlation between honeybee visitation and the diversity of hand‐sampled insects from onion flowers; indicate that either or both onion varietal attractiveness and/or pollinator population size may have had significant effects on overall insect visitation. Honeybees showed marked discrimination between hybrid onion parental lines and preferred to forage on one or the other during single foraging trips. Hybrid onion parents differed significantly in nectar characteristics and onion flower scent which would encourage selective foraging through floral constancy. Interspecies interactions were insignificant in causing increased honeybee pollination because of the scarcity of non‐Apis visitors. Most farming practices are subjected to favourable environmental conditions for successful production. However, and especially in the South African context, the dependence of onion hybrid seed crops on insect pollination for successful yields, increase its reliance on natural ecosystem dynamics that may deliver abundant wild honeybee pollinators, or attract them away from the crops. Nevertheless, this dependence can be mitigated effectively by the use of managed honeybee colonies to supplement wild honeybee workers on the flowers. / AFRIKAANSE OPSOMMING: Insek bestuiwing dra in verskillende grade by tot die produksie van landbou gewasse wat variteit en voedingswaarde in die mens se dieet verseker. Al is die heuningby (Apis mellifera L.) steeds die waardevolste ekonomiese bestuiwer van verboude enkelgewasse, kan wilde bestuiwers wesenlik bydra tot gewasbestuiwing deur middel van ekosisteem dienste afkomstig van natuurlike habitatte. Bestuiwing ekosisteem dienste is daarom waardevol en kan dus die bewaring van natuurlike ekosisteme, wat diverse gemeenskappe huisves, regverdig. F1 basterui saadproduksie is totaal afhanklik van hoë insek‐bestuiwer aktiwiteit om kruisbestuiwing, saadvorming en winsgewende saadopbrengste te verseker. Data is ingesamel op 18 basterui saad aanplantings in die half‐droë Klein Karoo en suid‐Karoo streke van die Weskaap, Suid‐Afrika. Hierdie twee hoof produksie streke is geleë binne die Sukkulente Karoo bioom wat erken word as ʼn globale biodiversiteits “hotspot” met hoë plant diversiteit. Dit is ook die habitat van die inheemse Kaapse heuningby (Apis mellifera capensis Esch.). Aanplantings is gekies om verskillende grade van beskikbare natuurlike habitat en bestuurde heuningby korf digthede te verteenwoordig. Diverse versamelings blom‐besoekers is versamel met water‐wippe in al die aanplantings, ongeag die persentasie natuurlike habitat beskikbaar by elke aanplanting. Gewas bestuurspraktyke het die diversiteit van blombesoekers betekenisvol beïnvloed. Tog is minder as 20% van hierdie diversiteit as aktiewe besoekers op die uiekoppe waargeneem. Heuningbye (bestuur of wild) was oorwegend die belangrikste bestuiwers as gevolg van hoë besoek frekwensies en wesenlike ladings uiestuifmeel op hulle liggame. Heuningby besoeke het saadopbrengs betekenisvol verhoog, maar blom‐besoeker diversiteit en nie‐Apis besoeke het geen effek op saadopbrengs gehad nie. Bestuurde korf digtheid en persentasie natuurlike habitat was nie belangrik in die bepaling van heuningby besoeke of basterui saadopbrengste nie. Totale jaarlikse reënval was die enigste betekenisvolle faktor wat heuningby besoeke bepaal het. Sekondêre faktore wat versoorsaak word deur reënval veranderlikheid, soos veldblom volopheid of grondvog, kon betekenisvolle effekte op die aantal heuningby besoeke gehad het. Bykomend, dui die positiewe korrelasie tussen heuningby besoeke en die diversiteit van hand‐versamelde insekte vanaf die uiekoppe op die moontlike betekenisvolle effek van elk of beide basterui variteit aantreklikheid en/of bestuiwer populasie grote op algehele insek besoeke. Heuningbye het noemenswaardige diskriminasie getoon tussen die basterui ouerlyne en het verkies om op een of die ander te wei tydens enkele weidingstogte. Basterui ouerlyne het betekenisvol verskil in nektar eienskappe en blomgeur wat die selektiewe weiding van heuningbye, toegepas deur blomkonstantheid, sal aanmoedig. Tussen‐spesie interaksies was onbetekenisvol in die verhoging van heuningby bestuiwing omdat nie‐Apis besoekers baie skaars was. Meeste boerdery praktyke is onderhewig aan gunstige omgewings toestande vir suksesvolle produksie. Maar, en veral in die Suid‐Afrikaanse konteks, omdat basterui saad aanplantings afhanklik is van insek bestuiwing vir suksesvolle opbrengste, word daar meer staat gemaak op natuurlike ekosisteem dinamika wat volop wilde heuningby bestuiwers kan voorsien, of selfs bestuiwers van die aanplanting kan weg lok. Nietemin, hierdie afhanklikheid kan effektief verlaag word deur die gebruik van bestuurde heuningby kolonies om die aantal wilde heuningby werkers op die blomme aan te vul.

California's almond industry, valued at $2.3 billion per year, depends on the pollinator services of honey bees, although pollination by other insects, mainly solitary wild bees, is being investigated as an alternative because of recent declines in the number of honey bee colonies. Our objective is to model the movements of honey bees and determine the conditions under which they will forage in less favorable areas of a tree and its surroundings when other pollinators are present. We hypothesize that foraging in less favorable areas leads to increased movement between trees and increased cross pollination between varieties which is required for successful nut production. We use the Shigesada-Kawasaki-Teramoto model (1979) which describes the density of two species in a two-dimensional environment of variable favorableness with respect to intrinsic diffusions and intra- and interspecific interactions of species. The model is applied to almond pollination by honey bees and other pollinators with environmental favorableness based on the distribution of flowers in trees. Using the spectral-Galerkin method in a rectangular domain, we numerically approximated the two-dimensional nonlinear parabolic partial differential system arising in the model. When cross-diffusion or interspecific effects of other pollinators was high, honey bees foraged in less favorable areas of the tree. High cross-diffusion also resulted in increased activity in honey bees in terms of accelerations, decelerations, and changes in direction, indicating rapid redistribution of densities to an equilibrium state. Empirical analysis of the number of honey bees and other visitors in two-minute intervals to almond trees shows a negative relationship, indicating cross-diffusion effects in nature with the potential to increase movement to a different tree with a more favorable environment, potentially increasing nut production.

The importance of European honey bees in improving fruit set, yield and fruit weight of kiwifruit on the central east coast of Australia was investigated. Field investigations were carried out using different bee saturations and different types of male pollen parents. These investigations confirmed the importance of honey bees in kiwifruit fruit set, yield and fruit weight. However, the results suggested that increasing bee activity alone may not increase pollination of kiwifruit by honey bees. Many factors need to be understood before introducing bees into the orchard. Bees were more effective during the early part of the flowering period, and bee activity varied according to the sex of the vine, planting design and the time of day. The type of male pollen parents also influenced fruit size and quality. Flowers pollinated by different pollen parents were assessed for pollen tube growth and histochemical changes. The resulting fruit were also examined for weight and seed numbers. Honey bees play the major role in the size and yield of kiwifruit, but the design of male vines, their age and type of male pollen may also contribute. The kiwifruit pistil also possesses important features that can be considered as adaptations to insect pollination. / Doctor of Philosophy (PhD)

Aciphylla species (wild spaniard/speargrass) are an iconic component of the Australasian high country flora, but their reproductive system is enigmatic. They are insect-pollinated dioecious mast seeders (synchronous highly variable seed production), which seems maladaptive. The resource supply to pollinators is highly variable, yet dioecious plants are dependent on pollinators, and dioecious masting requires male and female plants to flower synchronously. Floral display in Aciphylla is relatively large, with tall inflorescences bearing thousands of flowers, suggesting that plants would not have the resources to produce such large stalks every year. But why do they have such huge inflorescences in the first place? I tested whether pollinator attraction is providing an economy of scale which favours intermittent production of very large inflorescences, by manipulating floral display size during a high-flowering year and measuring insect visitation rates and seed set (female reproductive success). Using space-for-time substitution and selective removal of male inflorescences, I also tested whether female seed set was affected by distance to flowering male plants (i.e. changes in local pollen availability) to see if flowering asynchrony would reduce pollination success. Bags were used to exclude pollination by insects and test for wind pollination, and hand pollination was done to test for pollen limitation. Insect surveys suggest that Aciphylla has a generalist pollination system (to avoid satiating a specialist pollinator during 'mast' years'). Male inflorescences received significantly more visits than females, and some seeds were set inside bags (although only 20-30%), suggesting wind pollination may occur at low levels. Seed set rate was higher for taller inflorescences with greater flowering length in A. aurea but tall inflorescences with excess flowers led to a decrease in seed set rates in A. scott-thomsonii. Hand pollination significantly increased seed set rates although these effects were not as large as expected (e.g. 10% increases from natural to hand-pollinated inflorescences were typical). There was no evidence for resource limitation in any species. Female plants in dense flowering populations had higher seed set rates, and individual floral display size in females was particularly important when females were 'isolated' from males. Insect visitation rates were generally higher on inflorescences with a larger floral display, suggesting that display size is important for pollinator attraction. Overall, these results suggest that the pollinator-attraction benefits of such a large floral display (at both the plant and population level) are possibly providing an economy of scale, although the relative effects are small.

Aciphylla species (wild spaniard/speargrass) are an iconic component of the Australasian high country flora, but their reproductive system is enigmatic. They are insect-pollinated dioecious mast seeders (synchronous highly variable seed production), which seems maladaptive. The resource supply to pollinators is highly variable, yet dioecious plants are dependent on pollinators, and dioecious masting requires male and female plants to flower synchronously. Floral display in Aciphylla is relatively large, with tall inflorescences bearing thousands of flowers, suggesting that plants would not have the resources to produce such large stalks every year. But why do they have such huge inflorescences in the first place? I tested whether pollinator attraction is providing an economy of scale which favours intermittent production of very large inflorescences, by manipulating floral display size during a high-flowering year and measuring insect visitation rates and seed set (female reproductive success). Using space-for-time substitution and selective removal of male inflorescences, I also tested whether female seed set was affected by distance to flowering male plants (i.e. changes in local pollen availability) to see if flowering asynchrony would reduce pollination success. Bags were used to exclude pollination by insects and test for wind pollination, and hand pollination was done to test for pollen limitation. Insect surveys suggest that Aciphylla has a generalist pollination system (to avoid satiating a specialist pollinator during 'mast' years'). Male inflorescences received significantly more visits than females, and some seeds were set inside bags (although only 20-30%), suggesting wind pollination may occur at low levels. Seed set rate was higher for taller inflorescences with greater flowering length in A. aurea but tall inflorescences with excess flowers led to a decrease in seed set rates in A. scott-thomsonii. Hand pollination significantly increased seed set rates although these effects were not as large as expected (e.g. 10% increases from natural to hand-pollinated inflorescences were typical). There was no evidence for resource limitation in any species. Female plants in dense flowering populations had higher seed set rates, and individual floral display size in females was particularly important when females were 'isolated' from males. Insect visitation rates were generally higher on inflorescences with a larger floral display, suggesting that display size is important for pollinator attraction. Overall, these results suggest that the pollinator-attraction benefits of such a large floral display (at both the plant and population level) are possibly providing an economy of scale, although the relative effects are small.

Doctor of Philosophy (PhD) / A renewed focus on generalised pollinator systems has inspired a conceptual framework which highlights that spatial and temporal interactions among plants and their assemblage of pollinators can vary across the individual, population, regional and species levels. Pollination is clearly a dynamic interaction, varying in the number and interdependence of participants and the strength of the outcome of the interaction. Therefore, the role of variation in pollination is fundamental for understanding ecological dynamics of plant populations and is a major factor in the evolution and maintenance of generalised and specialised pollination systems. My study centred on these basic concepts by addressing the following questions: (1) How variable are pollinators in a generalised pollination system? To what degree do insect visitation rates and assemblage composition vary spatially among populations and temporally among flowering seasons? (2) How does variation in pollinators affect plant reproductive success? I chose to do this using a model system, Trachymene incisa subsp. incisa (Apiaceae), which is a widespread Australian herbaceous species with simple white flowers grouped into umbels that attract a high diversity of insect visitors. The Apiaceae are considered to be highly generalist in terms of pollination, due to their simple and uniform floral display and easily accessible floral rewards. Three populations of T. incisa located between 70 km and 210 km apart were studied over 2-3 years. The few studies investigating spatial and temporal variation simultaneously over geographic and yearly/seasonal scales indicate that there is a trend for more spatial than temporal variation in pollinators of generalist-pollinated plants. My study showed both spatial and temporal variation in assemblage composition among all populations and variation in insect visitation rates, in the form of a significant population by year interaction. However, removing ants from the analyses to restrict the assemblage to flying insects and the most likely pollinators, resulted in a significant difference in overall visitation rate between years but no difference in assemblage composition between the Myall Lakes and Tomago populations. These results indicate more temporal than spatial variation in the flying insect visitor assemblage of T. incisa. Foraging behaviour provides another source of variation in plant-pollinator interactions. Trachymene incisa exhibits umbels that function as either male or female at any one time and offer different floral rewards in each phase. For successful pollination, pollinators must visit both male and female umbels during a foraging trip. Insects showed both preferences and non-preferences for umbel phases in natural patches where the gender ratio was male biased. In contrast, insects showed no bias in visitation during a foraging trip or in time spent foraging on male and female umbels in experimental arrays where the gender ratio was equal. Pollinator assemblages consisting of a mixture of different pollinator types coupled with temporal variation in the assemblages of populations among years maintains generalisation at the population/local level. In addition, spatial variation in assemblages among populations maintains generalisation at the species level. Fire alters pollination in T. incisa by shifting the flowering season and reducing the abundance of flying insects. Therefore, fire plays an important role in maintaining spatial and temporal variation in this fire-prone system. Although insect pollinators are important in determining the mating opportunities of 90% of flowering plant species worldwide, few studies have looked at the effects of variation in pollinator assemblages on plant reproductive success and mating. In T. incisa, high insect visitation rates do not guarantee high plant reproductive success, indicating that the quality of visit is more important than the rate of visitation. This is shown by comparing the Agnes Banks and Myall Lakes populations in 2003: Agnes Banks received the highest visitation rate from an assemblage dominated by ants but produced the lowest reproductive output, and Myall Lakes received the lowest visitation rate by an assemblage dominated by a native bee and produced the highest seedling emergence. Interestingly, populations with different assemblage composition can produce similar percentage seed set per umbel. However, similar percentage seed set did not result in similar percentage seedling emergence. Differences among years in reproductive output (total seed production) were due to differences in umbel production (reproductive effort) and proportion of umbels with seeds, and not seed set per umbel. Trachymene incisa is self-compatible and suffers weak to intermediate levels of inbreeding depression through early stages of the life cycle when seeds are self-pollinated and biparentally inbred. Floral phenology, in the form of synchronous protandry, plays an important role in avoiding self-pollination within umbels and reducing the chance of geitonogamous pollination between umbels on the same plant. Although pollinators can increase the rate of inbreeding in T. incisa by foraging on both male and female phase umbels on the same plant or closely related plants, most consecutive insect movements were between plants not located adjacent to each other. This indicates that inbreeding is mostly avoided and that T. incisa is a predominantly outcrossing species, although further genetic analyses are required to confirm this hypothesis. A new conceptual understanding has emerged from the key empirical results in the study of this model generalised pollination system. The large differences among populations and between years indicate that populations are not equally serviced by pollinators and are not equally generalist. Insect visitation rates varied significantly throughout the day, highlighting that sampling of pollinators at one time will result in an inaccurate estimate and usually underestimate the degree of generalisation. The visitor assemblage is not equivalent to the pollinator assemblage, although non-pollinating floral visitors are likely to influence the overall effectiveness of the pollinator assemblage. Given the high degree of variation in both the number of pollinator species and number of pollinator types, I have constructed a model which includes the degree of ecological and functional specialisation of a plant species on pollinators and the variation encountered across different levels of plant organisation. This model describes the ecological or current state of plant species and their pollinators, as well as presenting the patterns of generalisation across a range of populations, which is critical for understanding the evolution and maintenance of the system. In-depth examination of pollination systems is required in order to understand the range of strategies utilised by plants and their pollinators, and I advocate a complete floral visitor assemblage approach to future studies in pollination ecology. In particular, future studies should focus on the role of introduced pollinators in altering generalised plant-pollinator systems and the contribution of non-pollinating floral visitors to pollinator assemblage effectiveness. Comparative studies involving plants with highly conserved floral displays, such as those in the genus Trachymene and in the Apiaceae, will be useful for investigating the dynamics of generalised pollination systems across a range of widespread and restricted species.

This project examined multiple aspects of the reproductive biology of the wetland invasive species, purple loosestrife (<i>Lythrum salicaria</i> L.), in central Saskatchewan. An examination of insect taxa visiting the three floral morphs of <i>Tristylous</i> L. <i>salicaria</i>, as well as a ranking of the pollination efficiency of individual insect species, an apparent first for L salicaria, was undertaken. Surface features of the floral nectary of L. <i>salicaria</i>, as well as floral nectar secretion dynamics, were also investigated. This project also re-visited some of the previous work done on this invasive species, including various floral organ morphometrics in relation to heterostyly, and aspects of the tristylous breeding system including self-fertilization, and fertilization potential of both illegitimate pollination and legitimate pollination.<p>The trimorphic nature of the sexual floral organs of L. <i>salicaria</i> were well defined in Saskatchewan. Significant differences in length (long-, intermediate- and short-style lengths) exist between all three floral morphs. Lengths of the staminal filaments (long, intermediate, and short) were also significantly different. Also the floral nectary in L. <i>salicaria</i> is located in a depression formed at the interface of the hypanthium and the gynoecium. Several stomata are located at regular intervals along the nectary surface, and may constitute the escape route for floral nectar. No morphological differences in nectary structure were apparent among the three floral morphs.<p>Nectar secretion dynamics of L. <i>salicaria</i> were examined between the three floral morphs throughout two summer days in 2006. Peak average nectar volumes and nectar sugar quantities were detected at 3:00 pm, and, interestingly, no significant differences were detected between floral morphs, in accordance with nectary morphology. The estimated secretion rates for L. <i>salicaria</i> ranged from 61 83 µg of nectar sugar per flower per hour.<p>Hand-pollination experiments carried out over the summers of 2006 and 2007 at three field sites in and around Saskatoon have verified the strong self-incompatibility in the breeding system of this tristylous species. Intramorph pollination, using illegitimate pollen, did not result in fertilisation, whereas legitimate hand-pollination experiments yielded multiple pollen tubes at the style base, without exception.<p><i>Lythrum salicaria</i> in central Saskatchewan was visited by several bee taxa including honeybees (<i>Apis mellifera</i> L.), bumblebees (Bombus spp.), leafcutter bees (Megachile spp.), and sweat bees (Lasioglossum spp.). A single visit by <i>Anthophora furcata</i> (Panzer) was also recorded in 2007. Generally, bee visits led to high levels of pollination success as determined by fluorescence microscopy of pollen tubes following single insect visits to previously-unvisited flowers. However, most visits by hoverflies (Syrphidae) were non-pollinating. Visits by Pieris rapae (L.), yellowjacket wasps (Vespidae) and some non-syrphid flies (Diptera) also yielded no pollen tubes at the style base.<p>A study of the ultrastructure and development of the floral nectary of the purple loosestrife cultivar Morden Gleam (<i>Lythrum virgatum</i> L. x L. alatum Pursh.) showed that starch build up in pre-secretory nectary tissues declined throughout secretion, and is virtually absent in post-secretory nectary tissues. The lack of a direct vascular supply to the floral nectary suggests that the starch breakdown products likely make up most of the floral nectar carbohydrates. Surface features of the floral nectary in Morden Gleam closely resembled those of L. salicaria, located in the valley formed between the hypanthium and gynoecium. Nectary stomata, occasionally in pairs, likely serve as outlets for nectar in this cultivar.

This project examined multiple aspects of the reproductive biology of the wetland invasive species, purple loosestrife (<i>Lythrum salicaria</i> L.), in central Saskatchewan. An examination of insect taxa visiting the three floral morphs of <i>Tristylous</i> L. <i>salicaria</i>, as well as a ranking of the pollination efficiency of individual insect species, an apparent first for L salicaria, was undertaken. Surface features of the floral nectary of L. <i>salicaria</i>, as well as floral nectar secretion dynamics, were also investigated. This project also re-visited some of the previous work done on this invasive species, including various floral organ morphometrics in relation to heterostyly, and aspects of the tristylous breeding system including self-fertilization, and fertilization potential of both illegitimate pollination and legitimate pollination.<p>The trimorphic nature of the sexual floral organs of L. <i>salicaria</i> were well defined in Saskatchewan. Significant differences in length (long-, intermediate- and short-style lengths) exist between all three floral morphs. Lengths of the staminal filaments (long, intermediate, and short) were also significantly different. Also the floral nectary in L. <i>salicaria</i> is located in a depression formed at the interface of the hypanthium and the gynoecium. Several stomata are located at regular intervals along the nectary surface, and may constitute the escape route for floral nectar. No morphological differences in nectary structure were apparent among the three floral morphs.<p>Nectar secretion dynamics of L. <i>salicaria</i> were examined between the three floral morphs throughout two summer days in 2006. Peak average nectar volumes and nectar sugar quantities were detected at 3:00 pm, and, interestingly, no significant differences were detected between floral morphs, in accordance with nectary morphology. The estimated secretion rates for L. <i>salicaria</i> ranged from 61 83 µg of nectar sugar per flower per hour.<p>Hand-pollination experiments carried out over the summers of 2006 and 2007 at three field sites in and around Saskatoon have verified the strong self-incompatibility in the breeding system of this tristylous species. Intramorph pollination, using illegitimate pollen, did not result in fertilisation, whereas legitimate hand-pollination experiments yielded multiple pollen tubes at the style base, without exception.<p><i>Lythrum salicaria</i> in central Saskatchewan was visited by several bee taxa including honeybees (<i>Apis mellifera</i> L.), bumblebees (Bombus spp.), leafcutter bees (Megachile spp.), and sweat bees (Lasioglossum spp.). A single visit by <i>Anthophora furcata</i> (Panzer) was also recorded in 2007. Generally, bee visits led to high levels of pollination success as determined by fluorescence microscopy of pollen tubes following single insect visits to previously-unvisited flowers. However, most visits by hoverflies (Syrphidae) were non-pollinating. Visits by Pieris rapae (L.), yellowjacket wasps (Vespidae) and some non-syrphid flies (Diptera) also yielded no pollen tubes at the style base.<p>A study of the ultrastructure and development of the floral nectary of the purple loosestrife cultivar Morden Gleam (<i>Lythrum virgatum</i> L. x L. alatum Pursh.) showed that starch build up in pre-secretory nectary tissues declined throughout secretion, and is virtually absent in post-secretory nectary tissues. The lack of a direct vascular supply to the floral nectary suggests that the starch breakdown products likely make up most of the floral nectar carbohydrates. Surface features of the floral nectary in Morden Gleam closely resembled those of L. salicaria, located in the valley formed between the hypanthium and gynoecium. Nectary stomata, occasionally in pairs, likely serve as outlets for nectar in this cultivar.

Doctor of Philosophy (PhD) / A renewed focus on generalised pollinator systems has inspired a conceptual framework which highlights that spatial and temporal interactions among plants and their assemblage of pollinators can vary across the individual, population, regional and species levels. Pollination is clearly a dynamic interaction, varying in the number and interdependence of participants and the strength of the outcome of the interaction. Therefore, the role of variation in pollination is fundamental for understanding ecological dynamics of plant populations and is a major factor in the evolution and maintenance of generalised and specialised pollination systems. My study centred on these basic concepts by addressing the following questions: (1) How variable are pollinators in a generalised pollination system? To what degree do insect visitation rates and assemblage composition vary spatially among populations and temporally among flowering seasons? (2) How does variation in pollinators affect plant reproductive success? I chose to do this using a model system, Trachymene incisa subsp. incisa (Apiaceae), which is a widespread Australian herbaceous species with simple white flowers grouped into umbels that attract a high diversity of insect visitors. The Apiaceae are considered to be highly generalist in terms of pollination, due to their simple and uniform floral display and easily accessible floral rewards. Three populations of T. incisa located between 70 km and 210 km apart were studied over 2-3 years. The few studies investigating spatial and temporal variation simultaneously over geographic and yearly/seasonal scales indicate that there is a trend for more spatial than temporal variation in pollinators of generalist-pollinated plants. My study showed both spatial and temporal variation in assemblage composition among all populations and variation in insect visitation rates, in the form of a significant population by year interaction. However, removing ants from the analyses to restrict the assemblage to flying insects and the most likely pollinators, resulted in a significant difference in overall visitation rate between years but no difference in assemblage composition between the Myall Lakes and Tomago populations. These results indicate more temporal than spatial variation in the flying insect visitor assemblage of T. incisa. Foraging behaviour provides another source of variation in plant-pollinator interactions. Trachymene incisa exhibits umbels that function as either male or female at any one time and offer different floral rewards in each phase. For successful pollination, pollinators must visit both male and female umbels during a foraging trip. Insects showed both preferences and non-preferences for umbel phases in natural patches where the gender ratio was male biased. In contrast, insects showed no bias in visitation during a foraging trip or in time spent foraging on male and female umbels in experimental arrays where the gender ratio was equal. Pollinator assemblages consisting of a mixture of different pollinator types coupled with temporal variation in the assemblages of populations among years maintains generalisation at the population/local level. In addition, spatial variation in assemblages among populations maintains generalisation at the species level. Fire alters pollination in T. incisa by shifting the flowering season and reducing the abundance of flying insects. Therefore, fire plays an important role in maintaining spatial and temporal variation in this fire-prone system. Although insect pollinators are important in determining the mating opportunities of 90% of flowering plant species worldwide, few studies have looked at the effects of variation in pollinator assemblages on plant reproductive success and mating. In T. incisa, high insect visitation rates do not guarantee high plant reproductive success, indicating that the quality of visit is more important than the rate of visitation. This is shown by comparing the Agnes Banks and Myall Lakes populations in 2003: Agnes Banks received the highest visitation rate from an assemblage dominated by ants but produced the lowest reproductive output, and Myall Lakes received the lowest visitation rate by an assemblage dominated by a native bee and produced the highest seedling emergence. Interestingly, populations with different assemblage composition can produce similar percentage seed set per umbel. However, similar percentage seed set did not result in similar percentage seedling emergence. Differences among years in reproductive output (total seed production) were due to differences in umbel production (reproductive effort) and proportion of umbels with seeds, and not seed set per umbel. Trachymene incisa is self-compatible and suffers weak to intermediate levels of inbreeding depression through early stages of the life cycle when seeds are self-pollinated and biparentally inbred. Floral phenology, in the form of synchronous protandry, plays an important role in avoiding self-pollination within umbels and reducing the chance of geitonogamous pollination between umbels on the same plant. Although pollinators can increase the rate of inbreeding in T. incisa by foraging on both male and female phase umbels on the same plant or closely related plants, most consecutive insect movements were between plants not located adjacent to each other. This indicates that inbreeding is mostly avoided and that T. incisa is a predominantly outcrossing species, although further genetic analyses are required to confirm this hypothesis. A new conceptual understanding has emerged from the key empirical results in the study of this model generalised pollination system. The large differences among populations and between years indicate that populations are not equally serviced by pollinators and are not equally generalist. Insect visitation rates varied significantly throughout the day, highlighting that sampling of pollinators at one time will result in an inaccurate estimate and usually underestimate the degree of generalisation. The visitor assemblage is not equivalent to the pollinator assemblage, although non-pollinating floral visitors are likely to influence the overall effectiveness of the pollinator assemblage. Given the high degree of variation in both the number of pollinator species and number of pollinator types, I have constructed a model which includes the degree of ecological and functional specialisation of a plant species on pollinators and the variation encountered across different levels of plant organisation. This model describes the ecological or current state of plant species and their pollinators, as well as presenting the patterns of generalisation across a range of populations, which is critical for understanding the evolution and maintenance of the system. In-depth examination of pollination systems is required in order to understand the range of strategies utilised by plants and their pollinators, and I advocate a complete floral visitor assemblage approach to future studies in pollination ecology. In particular, future studies should focus on the role of introduced pollinators in altering generalised plant-pollinator systems and the contribution of non-pollinating floral visitors to pollinator assemblage effectiveness. Comparative studies involving plants with highly conserved floral displays, such as those in the genus Trachymene and in the Apiaceae, will be useful for investigating the dynamics of generalised pollination systems across a range of widespread and restricted species.