Evidence for and Testing the Positive Feedback Hypothesis
We suggest that the outcome of the combined evolutionary-ecological process outlined above will be to cause seasonal differences in the balance between nectar supply and demand, with one or more seasons characterized by relative nectar abundance in which plants chase scarce pollinators, and one or more seasons by relative nectar shortage in which pollinators chase scarce nectar. These imbalances would arise via positive feedback exacerbating pre-existing, but smaller imbalances. Thus, a pre-existing situation in which many plants bloom in spring in a temperate-climate region could lead to a situation in which spring-blooming flowers produce more nectar than summer-blooming flowers. Is there any evidence for seasonal imbalance? In addition, how can the positive feedback hypothesis be more directly tested?
Several lines of evidence indicate that nectar is considerably more available in spring than summer, suggesting a seasonal imbalance. For example, honey bee (Couvillon et al. 2014b) and bumble bee (Pope and Jha, 2018) foraging distances are greatest during summer. Honey bee colonies gain more weight during spring (Balfour et al., 2017; Garbuzov et al., 2020) and a greater proportion of foragers returning with empty crops in summer than spring (Couvillon et al. 2014a). In addition, standing crops of nectar in lavender flowers are greater in spring than summer (Wignall et al., 2020). Summer is also the season of intense nectar competition among flower-visitors (e.g. Inouye, 1978; Weatherwax, 1986; Balfour et al., 2015) during which foragers compete for flowers containing minute quantities of nectar (c. 0.1 μl; Williams, 1998; Balfour et al., 2015) and nectar robbing among honey bee colonies occurs (e.g. Sakofski et al., 1990). Furthermore, it is likely that there are simply more flower-visiting insects on the wing in summer (Balfour et al., 2018) available to pollinate. For example, two major pollinators, honey bees and bumble bees, will be most abundant in summer as by then their colonies have reached maximum populations (Balfour et al., 2018).
In parallel with these findings, previous research in temperate-climate areas shows that species blooming earlier in the year tend to have lower seed or fruit set (e.g. Motten et al., 1981; Thomson, 2010; Kudo and Ida, 2013). It has been suggested this is due to low pollinator activity at this time of year. However, a formal measure of the seasonal ratio of pollinators to floral rewards is still lacking.
The positive feedback hypothesis needs to be tested, ideally by testing predictions that arise from it. Here we propose two tests, and encourage other biologists to devise more. The first test uses phylogenetically independent contrasts to compare the nectar secretion rates of native plants growing in a particular region in which there is existing background ecological information to suggest suitable seasons to compare. For the situation in southern England this would be spring versus summer (e.g., March-May v. July-August), with the a prioriprediction being that spring-flowering plants produce more nectar per flower than summer-flowering plants. Specifically, we propose comparing the nectar secretion rates of sequentially blooming species within multiple families across seasons.
One advantage of flowering plants as a study group is that they are diverse. As a result, obtaining an adequate number of species to study should be possible. We propose that study species should be native and bloom exclusively or predominately in one of the designated seasons. Additional variables that are likely to affect per-flower nectar secretion such as flower size, plant type (herb, shrub, tree), and habitat can be included as additional variables and controlled for. In principle, any species that are not wind pollinated and for which their phylogentic position can be determined could be studied. In practice, focusing on particular taxa could give practical advantages. In the Lamiaceae, the flowers of many species are large enough to allow nectar volumes in individual flowers to be quantified using capillary tubes (Balfour et al. 2015). In addition, Lamiacae produce little pollen, so the main reward is nectar. Although the Asteracecae are highly-attractive to insects (e.g. Warzecha et al., 2018), their flowers are aggregated into dense inflorescences and it is difficult to accurately quantify their nectar volumes (Dósa, 2008).
Our second suggested test is experimental/manipulative and also uses native plant species that bloom exclusively or predominantly in one of the two seasons to be compared. Plants of multiple species would be reared under controlled conditions to retard or advance their bloom (Wignall et al. 2020). We predict that plants that normally bloom in the pollinator-scarce season will attract more pollinators per flower per unit time when caused to bloom in the pollinator-abundant season andvice versa .