Constant Swinkels

Introduction

In recent decades, it has become clear that flower-visiting insects, or pollinators, are declining in abundance. This is concerning, as pollinators are essential for ecosystem functioning. Approximately three-quarters of all plant species depend on insect pollination for their reproduction, and many agricultural crops also benefit from their activities. A major driver of pollinator decline is the loss of suitable habitats and the reduction of floral resources in the landscape, making food sources increasingly scarce.

To protect pollinators, many conservationists and landscape managers therefore focus on increasing the availability of flowers. Well-known examples include flower strips in agricultural landscapes and flower-rich roadside verges. A large body of research shows that such measures can be effective, leading to higher pollinator abundance and increased species richness.

However, there are important caveats. Although pollinators often benefit from increased floral resources, not all species respond in the same way. From an ecological theory perspective, this is not surprising. Populations are constrained by limiting factors. For pollinators, flowers can be such a limiting factor, but once sufficient floral resources are available, other factors may become more important. These include a lack of suitable nesting sites, exposure to pesticides, or the quality of the surrounding landscape. This thesis therefore addresses the question of to what extent pollinator communities are actually limited by floral resource availability.

Pollinators on dikes

To address this question, the first part of this thesis examines how pollinator communities change along gradients of floral richness. By comparing sites ranging from flower-poor to highly flower-rich, it is possible to assess how pollinators respond to increasing floral availability. This research was conducted in dike grasslands. The Netherlands has almost 18,000 kilometres of dikes, making these grasslands an extensive and potentially important habitat for pollinators.

Chapter 2 shows that dike grasslands harbour surprisingly rich bee communities. More than 150 wild bee species were recorded across the studied sites, including many species listed on the Dutch Red List. Moreover, flower-rich dike grasslands appear to be important not only as foraging habitat, but likely also as nesting habitat for many species.

Pollinators and floral richness

Chapters 2, 3 and 4 of this thesis together show that the response of pollinator communities to floral richness is complex. On the one hand, both pollinator abundance and species richness increased strongly with increasing numbers of flowers and flower species. However, this increase is not unlimited: at relatively low levels of floral richness, a clear saturation occurs, after which additional flowers contribute only marginally to higher abundance or species richness. This suggests that floral limitation is alleviated for a large part of the pollinator community relatively quickly, and that other factors subsequently become more important in shaping community composition.

At the same time, not all pollinators respond in the same way. Rare and threatened species, in particular, are more frequently found at the most flower-rich sites. In addition, the way pollinators use floral resources also changes. As more flower species become available, pollinators collect resources from a smaller number of plant species, and niche overlap between species decreases. As a result, different pollinators appear to interfere less with one another. Thus, additional flowers do not necessarily lead to larger communities, but may contribute to more stable and more specialised pollinator communities.

Pollinators and management

In the second part of this thesis, the question of floral limitation is addressed from a management perspective. This involves examining how pollinators respond to different management interventions, which together provide insight into the role of floral resources in restoration practices.

Mosaic mowing

Chapter 5 presents the results of an experiment on the effects of mosaic mowing in dike grasslands. In the experimental sites, part of the vegetation remained unmown while the rest was mown. Changes in these strips were compared to the situation before mowing and to unmown control plots.

This shows that a sudden reduction in floral resources was associated with an increased concentration of pollinators in the remaining refuges, indicating increased competition for limited resources. Subsequently, both pollinator abundance and species richness in these refuges declined to below the level of the unmown controls. This underlines that, despite the relatively rapid saturation of floral richness observed in Chapters 2 and 3, pollinator communities remain strongly dependent on sufficient floral resources at the right time of the season.

Ecosystem restoration

Another increasingly common form of management focuses on steering ecosystem processes rather than targeting individual taxonomic groups. It is therefore relevant to ask whether this approach can also support pollinator communities, even without directly increasing floral resources. Chapter 6 presents the results of an analysis of two long-term monitoring datasets from the Bargerveen nature reserve. In recent decades, extensive measures have been taken there to restore hydrological conditions. The analysis shows that, in contrast to many national trends, pollinator abundance and species richness remained remarkably stable. Even trends in rare species showed no decline. This suggests that restoring ecosystem processes as a whole can be an effective strategy for conserving pollinators.

Garden varieties

In restoration and sowing projects, seed mixes often include non-native plant species or cultivated varieties of native species. Although these may appear visually attractive, it remains unclear to what extent pollinators actually use these novel floral resources. Chapter 7 examines flower visitation on different cornflower types: the native form and three cultivated varieties. Pollinators were given a choice between these four options in field experiments.

The results show that wild-type plants are significantly more attractive to pollinators than cultivated ornamental varieties. This pattern was consistent in both urban and rural environments and across different pollinator groups, indicating a robust effect. This highlights that not only the quantity of flowers matters, but also which plant species are used.

Conclusions

The studies in this thesis show that the relationship between pollinators and flowers is not linear. Pollinators are undeniably dependent on flowers as a food resource, but this dependence appears to saturate quickly: beyond a certain level of floral richness, community size and species richness increase only marginally. At the same time, interactions within communities change, with reduced niche overlap and increased specialisation.

These complex relationships are also reflected in the management experiments. The sudden reduction in floral resources through mosaic mowing was associated with a decline in pollinator abundance in the remaining vegetation. This suggests that pollinators in dike grasslands are adapted to a high and seasonal availability of flowers, and that disruption of established management can lead to temporary negative effects on community composition. In contrast, management focused on restoring ecosystem processes, as in Chapter 6, may contribute to the long-term stability and persistence of pollinator communities.

Together, these results show that pollinator communities benefit from a high availability of flowers, but are at the same time strongly dependent on the context in which these flowers occur. Communities appear to adapt over time to local and consistent management, highlighting the importance of stable management regimes. In addition, landscape context plays an important role: pollinators require more than flowers alone to persist in the long term.