Mario B. Pesendorfer, Institute of Forest Ecology, Department of Forest and Soil Sciences, BOKU – University of Natural Resources and Life Sciences, Vienna, Austria, discusses his article: Positive spatial and temporal density-dependence drive early reproductive economy-of-scale effects of masting in a European old-growth forest community
In temperate forests, tree populations often produce large bumper crops of seeds in some years followed by years in which few or no seeds can be found. Such interannual variation in reproduction tends to be synchronized across spatial scales, ranging from stands or populations to whole continents. Called “mast-seeding” or “masting”, this behavior plays an important role in forest dynamics – from populations to communities and whole ecosystems – as the resource pulses reverberate across trophic levels, from seed-consuming rodents to their predators, parasites, and even the transmission of zoonotic diseases from ticks to humans. While masting has received lots of attention from ecologists in recent years, the fitness benefits of such reproductive synchrony are not fully understood.
Understanding tree behavior takes time
Studying masting requires commitment because, like many ecological processes, it unfolds over decades rather than the usual 3-year duration of research funding. Thus, long-term studies, such as the one initiated at Rothwald forest in 2004 by Georg Gratzer, Associate Professor for Mountain Forest Ecology at BOKU University, contribute invaluable data to masting research. The 400-ha IUCN-1a protected area located in the Dürrenstein-Lassingtal Wilderness Area in Lower Austria is the only remaining montane primeval forest in the Alps, and thus provides a unique window into the ecological and evolutionary drivers and consequences of masting behavior of local trees. In two 1-ha plots dominated by European beech, Norway spruce, and silver fir, all adult trees have been mapped repeatedly since the 1960s. Since 2004, 81 seed traps were placed to provide a high-resolution data set of spatial and temporal variation in seed rain, measured as the number of seeds per m2. Using this data set, we are now able to answer questions about the effects of masting on seed fate and plant recruitment.
Several hypotheses about the fitness benefits of masting involve “economies of scale”, the notion that fitness costs per individual unit decline (or inversely, benefits increase) as more seeds are produced. In fact, there are several steps at which such density-dependent mechanisms may apply to shape ultimate seed fate: pollination, pre-dispersal predation, seed dispersal, and post-dispersal predation. Because seed rain varies both between years as well across space within years, for example because of size or site differences among trees, there may be an interaction between the temporal and spatial axes of density dependence.
Seed traps provide insights into predispersal seed fate
For the Rothwald seed traps from 2008 to 2022 (14 years total), seeds of each dominant tree species were sorted into three categories: full seeds, seeds with insect damage (i.e., predispersal predation), and hollow seeds indicative of pollen limitation. For each species, we examined the effect of spatial and temporal variation in overall seed density on pollination efficiency, using the proportion of non-hollow seeds, and on relative seed predation, using the proportion of non-hollow seeds with insect damage. Because spatial and temporal variation in seed rain are inherently linked, we analyzed the data on multiple levels: a) by comparing plot averages across years (2008-2022), b) by comparing cumulative totals (over the whole study period) of each seed category among traps within plots, and c) by scaling within-year variation (to an index between 0 and 1) and simultaneously modeling the effects of spatial and temporal variation in seed rain at the trap-level.
In all three species, we found that the proportion of fully formed seeds was higher in years of high seed production than in years of low seed production. Importantly, we found that this increase only occurred below a threshold in seed density (about 20% of the maximum), above which additional increases in seed density resulted in no further benefits for seed fate. These temporal economy-of-scale effects were strongest in European beech, the species with the highest density in the study area, and weaker in Norway spruce and silver fir, which both occur at low densities. The spatial comparison of cumulative seed rain in traps showed no such relationship, illustrating that there are no distinct areas in which seeds are generally better off because they are consistently produced at high density.
Economies of scale provide fitness benefits in masting species
The detailed analyses at the trap-level showed that in European beech, pollination efficiency only varies with the temporal axis of seed rain density variation, but not with the within-year variation in a plot. In Norway spruce, we found that both axes of variation affected pollination efficiency, while no effects were found for silver fir. Because beech occurs at high density, we did not expect pollen limitation to vary in space. In the low-density Norway spruce, we did find spatial patterns, perhaps because clusters of trees can likely cross-pollinate, while isolated individuals lack pollen donors.
In contrast, we expected both spatial and temporal effects of reproductive effort on predation, because predators are mobile and thus tend to move to areas where they find more prey. In European beech, we indeed found a combined effect of spatial and temporal variation in seed density on predispersal predation. In Norway spruce, the temporal effect on seed predation was strong, and differed between plots, but not with spatial variation within plots. In silver fir, seed predation was generally low and showed no spatial or temporal effects of seed rain variation.
In summary, our results illustrate that there are positive density-dependent effects of synchronized reproduction. In years of high seed production, larger proportions of seeds are pollinated and fewer are depredated by insects in the crown. Spatial variation within years can affect pollination or seed predation, but did not result in cumulative spatial differences over the 14-year study period. Therefore, masting leads to more economy-of-scale effects over time than in space.
While tree growth and mortality and their contribution to forest dynamics have been studied extensively, seed rain, seedling recruitment, and the subsequent life history stages are still poorly understood. Our work suggests that positive density dependence mediates seed fate dynamics (i.e., when there are lots of seeds, there are proportionally better outcomes), while seedling survival and growth are dominated by negative density dependence, for example because of Janzen-Connell effects. The fact that density dependence can at first be positive, as illustrated here for pollination and insect predation, and then be negative in the next life history stage, as is the case for seedlings, illustrates the complexity of forest dynamic processes that determine population structure and community composition.
Learn more about Mario B. Pesendorfer and his work at his website and social media.
