Editor’s Choice
 



July 2011 (99:4)


Masting, the synchronous production of large seed crops at irregular intervals, has fascinated ecologists and evolutionary biologists alike. From an evolutionary perspective, masting is problematic as it is a form of reproductive delay, and so masting individuals, which reproduce less often, suffer a demographic cost as a result of mortality between the mast events.

Explaining masting in Southeast Asia
The key hypotheses proposed to explain occurrence of masting involve benefits of scale that accrue either as a result of predator satiation or improved wind pollination. In Southeast Asia, masting is particularly interesting as entire communities reproduce in synchrony over huge areas resulting in a carpet of seeds on the forest floor. Predator satiation is one of the key hypotheses used to explain masting in Southeast Asia, but is it enough?
This is the question tackled by Visser et al. (2011) in their paper on masting in Shorea leprosula (Strict mast fruiting for a tropical dipterocarp tree: a
demographic cost–benefit analysis of delayed reproduction and seed predation
).

Asessing fitness consequences in a long-lived tree species
So how then do you assess the fitness consequences of masting in a long-lived tree species (reproductive individuals live for on average ~100 years) where growth, survival and fecundity all depend on tree size? Several approaches are possible using ad hoc indices of reproductive success, such as total seed production over some interval, but these are difficult to justify from a population dynamic perspective. To account for the size-dependence of demography Visser et al. built a stochastic matrix model and compared strategies using the long-term stochastic growth rate, a quantity that defines the rate at which new mutations spread through the tree population. If the masting strategy has a higher long-term stochastic growth rate than the annual reproductive strategy, then it will be favoured by evolution. Note, this analysis allowed Visser et al. to test if the current strategy can be invaded by a new strategy with a different probability of reproducing (having a mast year), without specifying where density dependence operates in the system. This is because the estimated demographic parameters implicitly include the effects of density dependence. In systems where density dependent processes are quantified a more refined evolutionarily stable strategy (ESS) approach can be used to predict the frequency of reproduction.

Are the results robust?
As expected, the best strategy, when there is no seed predation, is annual reproduction, although the effects of delaying are slight because annual survival is high (survival probability per year 0.98). Incorporating the effect of seed predators that have a maximum seed consumption changes this result dramatically, with masting being favoured over the annual reproductive strategy over a wide range of predation levels. Comparing the current masting strategy with an annual reproductive strategy suggests a fitness difference of ~5%. So this is very encouraging, but, as ever, the devil is in the details and all models make assumptions, so are the results robust? The basic model assumed plants could 1) efficiently store an unlimited amount of resources when they refrain from reproduction, 2) there was no density-dependent mortality following masting, and 3) producing a large seed crop did not affect adult demography. With a parameterised model Visser et al. explored how adding each of these complexities changed the models predictions.


Of these three key assumptions, it turns out that two are important, and that 1) limiting the amount of resources a plant can store (so resources that could have been used to produce seeds are lost when plants do not reproduce), and 2) incorporating density-dependent seedling survival can dramatically influence the advantages that accrue from masting. Curiously, when masting results in no growth in mast years there are still substantial advantages to masting. This result might in part reflect the model structure; there are only 2 reproductive size categories, and again the sensitivity of this model prediction can be explored, say with an integral projection model (Ellner & Rees 2006).
The tools now exist for plant ecologists to explore a wide range of ecological and evolutionary problems as Visser et al.’s study clearly demonstrates.

Mark Rees
Editor, Journal of Ecology

References


  • Ellner, S. P. & Rees, M. (2006) Integral projection models for species with complex demography. American Naturalist, 167, 410-428.
  • Visser, M. D., Jongejans, E., van Breugel, M., Zuidema, P. A., Chen, Y.-Y., Rahman Kassim, A. & de Kroon, H. (2011) Strict mast fruiting for a tropical dipterocarp tree: a demographic cost–benefit analysis of delayed reproduction and seed predation. Journal of Ecology, 99, 1033–1044.