|dc.description.abstract||Flower constancy (the tendency of pollinators to restrict visits to one floral species) provides obvious benefits to plants, however, the benefits to pollinators remain unclear. For example, specialisation to a single species in an environment where many floral resources are available would likely result in the exclusion of food sources that are energetically superior. The hypotheses put forward as potential explanations of this behaviour can be separated into two schools of thought; those that argue flower constancy arises due to constraints of pollinator cognition (cognitive constraint hypotheses) and those that argue the behaviour evolved to maximise foraging success (adaptive hypotheses). Despite a growing number of papers reporting evidence against cognitive constraint hypotheses, they continue to be cited as offering the most popular explanation of why flower constancy arises. However, to date there are no known attempts to quantify the level of support for each hypothesis within the literature.
In chapter one, a literature review first summarises flower constancy hypotheses. A brief bibliometric study was then conducted to gauge the level of support present for each hypothesis category (cognitive constraint or adaptive). Several factors including the study type (lab, field or semi field), methods used, number of citations gained, and the hypothesis supported were recorded for papers published on flower constancy. The bibliometric study identified that cognitive constraint hypotheses do indeed offer the most popular explanation of why flower constancy arises. However, this may be attributed to the setting in which these studies were conducted, an overwhelming majority of papers reporting evidence for cognitive constrains hypotheses were conducted under laboratory conditions using artificial flowers. To date there have been no known studies that have attempted to replicate laboratory studies with the use of natural flowers. Before a study of this kind can be conducted, natural flowering species must first be identified.
Chapter 2 aims to determine flowering plant species suitability for use in a greenhouse (semi field) experiment on flower constancy. Five flowering plant species were grown from seeds and seedlings. Flowering plant species were judged to be suitable if three or more colour morphs could be reliably grown, were consistently visited by bees, provided a constant food source of similar quality and were not greatly affected by disease or pests. Snapdragon (Antirrhinum majus) and cornflower (Centaurea cyanus) were identified as the flowering plant species to best meet this selection criteria. Further analyses (flower colour spectrum, pollen quality and nectar quantity) were conducted on the chosen flowering plant species. Both cornflower and snapdragon were found to produce three and four colours distinguishable within bee colour space respectively, produce lipid predominant pollen (high energetic value to insect pollinators), and an equal nectar quantity (less than 1 micro-litre) within and across species. A key observation of this study was the difference in behaviour between pollen and nectar foraging bumble bee (Bombus terrestris) workers. Whether these differences should be considered when quantifying flower constancy is discussed.
The aim of chapter 3 was to determine whether predictions made by cognitive constraint hypotheses are met within a semi field environment using natural flowers. The flower constancy of bumble bee workers was measured across simple and complex environments, in arrays of morphologically complex and simple flowers, and compared between nectar bouts requiring simple handling skills and pollen bouts requiring complex handling skills. No significant differences were observed across any of the experimental treatments (environmental array, flower morphology complexity or handling skill). This could be explained by the criteria used to define complexity being subjective, differences between treatments not being large enough or cognitive constraint hypotheses not being the main driving force of bumble bee constancy within a semi field environment. A key observation was nectar foragers exhibiting higher levels of constancy than pollen foragers (although this difference was not significant). This observation provides some evidence to the idea that constancy may arise due to reward thresholds, as nectar (containing high sugar content) is easier to attribute value to than pollen (carbohydrates). This has potential implications to pollination efficiency although further studies need to be conducted to substantiate this claim.||