You are here

Harvard Forest >

Harvard Forest Symposium Abstract 2012

  • Title: Interspecific interactions in changing environment: How do community context and abiotic conditions shape the outcome of plant-granivore association?
  • Primary Author: Rafa Zwolak (Adam Mickiewicz University)
  • Additional Authors: Elizabeth Crone (Tufts University)
  • Abstract:

    Objectives.—Certain interspecific interactions cannot be easily classified: they involve both costs and benefits, and the balance of these determines where the net outcome falls between mutualism and antagonism. Plant-granivore interactions are particularly well-suited for investigations of such context-dependent interactions. Many vertebrate granivores consume some seeds, but cache others in favorable microsites, and their ultimate impact on plant recruitment may vary with environmental conditions. We will experimentally test the influence of biotic (masting and community context) and abiotic (weather conditions affecting seedling recruitment) factors on the outcome of interactions between oaks and rodent acorn consumers in two complementary study systems.



    First, we will investigate the influence of beech masting on pedunculate oak recruitment in western Poland. In our previous research we documented strong impact of beech masting on rodent abundance and behavior (Zwolak et al., unpublished data). Populations of wood mice and bank voles reach extremely high abundances one year after masting and crash in the next year. Relative beech seed removal and pilferage rates are low in mast years and high in non-mast years (Zwolak et al., unpublished data). The numerical and behavioral changes in rodent populations could influence the recruitment of co-occurring oaks, possibly switching the outcome of rodent-oak interaction from mutualism in beech mast years (when acorns are cached but rarely retrieved) to antagonism in post-mast years (when acorns are consumed on the spot or after caching). If the hypothesized beech-mediated switch in the outcome of oak-rodent interaction does exist, it would provide an outstanding example of mast-mediated context-dependency in interactions between plants and animals.



    Second, we will evaluate the effects of eastern chipmunks on interactions between white-footed mice in northern red oak stands of eastern USA. Eastern chipmunks primarily larderhoard seeds in deep burrows, thus their net effects on plants are thought to be decidedly negative. In contrast, white-footed mice scatterhoard seeds in shallow caches, which might benefit plant recruitment. When chipmunks are abundant, seeds removed by mice might have higher survival than seeds left on surface, where they can be easily found by the strict seed predators. Thus, the consequences of scatterhoarding by mice depend on whether seed germination from caches is greater than the probability of seed germination by seeds left on the surface multiplied by the probability that they are not removed by chipmunks. In habitats or periods with low chipmunk abundance, seed germination from mouse caches would have to be higher than germination of seeds left on the surface. Therefore, the mutualism between mice and oaks would be conditional upon the presence of third party players: eastern chipmunks.



    In both study systems, the biotic effects will be investigated with respect to weather variation. If the above-described interactions are shaped solely by biotic components, their outcomes are relatively predictable. However, predictions become much more difficult when the outcome is also mediated by stochastic fluctuations in abiotic environment. The outcome of plant-scatterhoarder interactions depends on the benefits of seed burial (germination of buried seeds relatively to germination of seeds left on surface). Field results (Zwolak et al., unpublished data) suggest that the germination probabilities (and therefore the advantage of caching) vary widely from year to year because of changing weather conditions. If this finding holds and plant-granivore interactions are strongly influenced by varying abiotic conditions, their dominating outcomes might change e.g. in the consequence of long-term trends in global weather patterns.



     

    Methods.—We will quantify seed rain (beech and oak in Poland; oak in the US) to identify potential mast years, conduct small mammal trapping to gauge how changes in seed production translate into changes in small mammal abundance, carry out acorn removal & radioactive tracking experiments to assess the influence of beech masting (in Poland) and granivore community composition (in the US) on the scatterhoarding behavior of mice, conduct seedling recruitment experiments to measure the magnitude of benefits associated with acorn caching (by comparing seedling recruitment of acorns sown on surface and acorns experimentally buried) and to quantify oak recruitment with and without rodents (by comparing seedling emergence in control recruitment cages that allow rodent access and exclusion cages, where acorns are protected from rodents), conduct chipmunk removal experiments (only in the US) to analyze their role in oak recruitment and their impact on oak-mouse interactions, and conduct artificial masting experiment: masting events are unpredictable, thus we cannot be certain that during the course of the study beech masting will occur in western Poland and oak masting in Harvard Forest. If natural mast does not happen, we will conduct artificial masting to evaluate how beech crop influences oak-rodent interactions in Poland and how acorn abundance affects tripartite oak-chipmunk-mouse interactions in eastern US.



    Impact of results.—Our research will evaluate the qualitative outcome of plant-granivore interactions, in terms of whether granivore activity provides a net benefit or cost to plant populations. This goal differs from past studies, many of which have tended to assume seed-caching granivores are beneficial or detrimental to plants, then quantify the magnitude of the assumed cost or benefit. In addition to this specific goal, our research will also address the very general question of whether we can build community models from pairwise interactions. Finally, we will evaluate whether plant-granivore interactions change in predictable ways in response to changes in the biotic and abiotic environments.



    Understanding how plant-granivore interactions change under different ecological scenarios is particularly timely because mast-consumer interactions may be especially likely to change in novel environments. Forest ecologists tend to focus on growth and survival in relation to environmental change, but changes in distribution ultimately depend on reproduction and dispersal. In large-seeded plants (e.g. most Fagaceae: beeches, oaks, and chestnuts) both reproduction and dispersal are strongly influenced by vertebrate seed consumers. In part, reproduction is likely to change because patterns of masting may be mathematically chaotic, i.e, extremely sensitive to small changes in the environment. Even if it is not chaotic, reproduction is likely to be sensitive to environmental change because survival and growth are more important for fitness in steady-state environments, and less important vital rates tend to vary more. In addition, consumer communities do not necessarily respond to the same cues as plants. In the current environment, mast production is the primary determinant of mouse population sizes, but if the environment changes, rodents could be limited by other factors, such as snow cover, predator abundance, or disease.

  • Research Category: Biodiversity Studies
    International Research Projects