In many plant populations, reproductive output fluctuates dramatically over time.  These fluctuations, often called mast-seeding, directly affect population dynamics of plants, florivores, and seed consumers, including dynamics and stability of economically important wildlife and pest species.  Although many studies have explored the ultimate, evolutionary advantages of mast-seeding, empirical tests of the proximate causes of mast-seeding are virtually nonexistent. In some cases, researchers have not been able to identify environmental cues for mast years, even after screening multiple weather and climate variables.  In contrast, recent theoretical models have been strikingly good at explaining mast-seeding based on dynamics of stored resources within plants, as a function of past reproduction, environmental conditions, and pollen availability.  Here, we show that this “pollen coupling” hypothesis explains alternate-year flowering in Astragalus scaphoides, a perennial wildflower. We conducted a four-year experiment to test the predictions of pollen coupling models by removing inflorescences in 2005 to desynchronize reproduction in two ways: “press” removal of inflorescences for three years, 2005-2007, and “pulse” removal in 2005 only. Destructive measurements of nonstructural carbohydrates were conducted in 2006 and 2007 in an ancillary experiment with three groups of plants: flowering and non-flowering plants harvested early in the growing season; plants harvested after fruits matured in August; and plants harvested in August but with inflorescences removed to prevent fruit set (2007 only).

Results/Conclusions

Higher than average flowering occurred in 2005 and was followed by alternate-year flowering in subsequent years. As predicted, pulse and press plants flowered more in 2006 and 2007 than control plants. Fruit set in control and pulse plants was very low in 2006, a low flowering year, allowing pulse plants to flower more again and set more fruit in 2007 relative to control plants. In 2008, a very low flowering year, flowering was highest in press plants, although fruit set was similar among treatments likely due to pollen limitation.  Our results show that fruit set, depleted mobile carbohydrate stores, and decreased flowering the next year.  This resource depletion caused individual plants to flower in alternate years.  During low-flowering years when pollen is limited, fruit set decreased and allowed plants to flower the next year.  Density-dependent pollen limitation, coupled by fluctuations in resource availability, synchronizes alternate-year flowering among plants within populations. Therefore, synchronous mast-seeding can be seen as a consequence of physiological costs of reproduction, as well as an adaptive strategy to attract pollinators or satiate seed consumers.