Benefit accrual: Quantifying the non-linear relationship between pollinator visits and seed production

Pollinator populations are declining worldwide with uncertain downstream impacts to the plants that they pollinate. Most flowering plants require an insect pollinator to reproduce, receiving many visits from many distinct pollinator species before setting fruit. Given that pollinator populations are declining and visitation rates are changing, how will seed production change? Linear relationships between pollinator visits and seed production (as is often assumed in theoretical population models) would result in commensurate decreases or increases in seed set from pollinator declines, whereas saturating benefits (more biologically realistic given finite ovule numbers per flower) would result in little change to seed sets until a critical inflection point has been crossed. Despite some preliminary work in this space, the relationship between individual flower pollinator visits and seeds produced remains challenging to quantify. To characterize the benefit that plants accrue through pollinator visits in wild populations of wildflowers we used a novel camera-trap system coupled with neural net based object-detection algorithms to collect information on the full set of pollinator visits to individual flowers. We then counted resulting seed sets and fit mechanistic models of benefit accrual. While we are still collecting and analyzing data, preliminary results show support for a saturating benefit accrual functional form with some limited evidence for “over-visitation” costs, i.e. a point at which additional visits actually decrease seed set. Understanding the functional form of pollinator benefit accrual relationships could help predict impacts of ongoing pollinator declines, and additionally different functional forms may have distinct stability implications for plant-pollinator networks.