Selective benefits of masting
An economy of scale (EOS) is required for mast seeding to be selectively advantageous (Janzen 1978, Norton & Kelly 1988). If an EOS is operating, reproduction is more efficient in occasional large reproductive efforts than the same resources invested in smaller more regular efforts. Economies of scale are reviewed by Norton and Kelly (1988) and Kelly (1994). There are two specific mechanisms that frequently provide an EOS associated with mast seeding:
-Wind Pollination
The pollination efficiency hypothesis states that masting should be strongly selected in species that can achieve greater pollination efficiency through synchronized above-average flowering effort. This efficiency is most likely to be seen in wind-pollinated plants because they are at less risk of saturating the pollinators (Janzen 1978, Smith et al. 1990, Sork 1993, Kelly et al. 2001). Essentially, the plants have a few large flowering efforts rather than many small ones, because in each small flowering effort, the density of pollen in the air is too low for most stigmas to be pollinated, causing wasteful losses from unpollinated flowers. In a large flowering effort, higher pollen densities ensure higher levels of pollination, and thus lower losses.
The pollination efficiency hypothesis depends on the size of the current flowering effort, but the sequence of reproductive efforts among years is irrelevant; for example, having successive mast years would not be disadvantageous, as seen in rimu (Norton & Kelly 1988). Plants in productive habitats could reach a high enough level of reproductive effort for efficient pollination every year, so they could reproduce constantly and avoid the negative consequences of masting. Plants in unproductive habitats are unable to reach this level every year, so they accumulate reserves to be expended in occasional large efforts. For a detailed summary of this topic, see Kelly Hart and Allen (2001).
Effect of increasing levels of mast seeding on relative pollination benefits in five plant species: Nothofagus solandri, Nothofagus menziesii, Dacrydium cupressinum, Betula alleghaniensis and Chionochloa pallens.
-Predator Satiation
Salisbury (1942) first pointed out that in European beech (Fagus) and oaks the only seeds to escape predation were produced in mast years, and that if a species had a constant seed crop its natural enemies could increase in number until all seeds were destroyed every year.[photos of oak and squirrel here?] The predator satiation hypothesis was more fully developed by Janzen (1971) and states that seed predators cause selection for masting when larger seed crops synchronized among individuals experience lower percentage seed predation. Predator satiation favours masting when variation in seed crops satiates seed predators in high-seed years. Predator satiation requires interannual variation in seed crops, but it is unclear whether selection acts directly to favour gaps between mast years by starving predators in low-seed years, or whether gaps are an inevitable consequence of selection for larger crop size (Janzen 1978, Silvertown 1980).
One very well documented example of predator satiation is in Chionochloa pallens in New Zealand (Kelly and Sullivan 1997). Improved pollination efficiency provides very little benefit from mast seeding in C. pallens, mainly because only in the very lowest flowering years does pollination success decrease. Pollination seems to be efficient even at low flowering densities because plants are capable of self-pollination within a solitary inflorescence. In contrast, Chionochloa has been shown to gain major benefits from predator satiation. There are three different native seed predators (see the section on Chionochloa below). Each of the three predatory insects alone can destroy more than 50% of florets. The three species often co-occur, and total levels of predation can exceed 90% (Kelly et al. 1992; Kelly and Sullivan 1997; Sullivan and Kelly 2000). However, mast flowering is effective in reducing levels of predation in C. pallens from around 80% losses in years with lower flowering effort than the previous year, down to less than 10% predation in years with much higher flowering effort than the previous year (Kelly and Sullivan 1997).
The effect of masting intensity (variation of seed output) on predation rates in Chionochloa pallens from 1986 to 2005.
Note however that more variable seed crops can not only reduce mean predation (as shown below for the American Betula alleghaniensis for predation by insects), but can also in some cases increase it (as shown below for predation by birds, which are attracted to feed on larger seed crops and thus eat a higher percentage of them). For more detail, see Kelly Hart & Allen (2001) and Koenig et al. (2003).
The effects of the degree of masting on reproductive efficiency in Betula alleghaniensis, resulting in changes in percentage of flowers that are wind pollinated, and the percentage of seeds that escape insect and bird predation.
