In the new Oikos paper (now on Early View), “Simulated responses of moose populations to browsing-induced changes in plant architecture and forage production”, John Pastor and Nathan R. de Jager present a model examining how tree crown architecture affects moose populations. Here, they give a background to the study:
“In the recently published paper, “Simulated responses of moose populations to browsing-induced changes in plant architecture and forage production”, we report the results from a model we developed nearly a decade ago as part of Nate’s Master’s thesis at the University of Minnesota-Duluth. The model examines the feedback effects of moose browsing-induced changes in plant architecture on moose population dynamics. We were able to construct the model because of a very well thought out and executed experiment in northern coastal Sweden (Persson et al. 2005 a, b). Inga-Lill Persson and her colleagues annually removed plant tissue from study plots in proportion to different moose population densities and also added corresponding amounts of urine and fecal material. By measuring the architectural responses of different tree species to simulated moose densities (De Jager and Pastor 2008, 2010) we were able to ask a very simple question: Can the forage produced by trees that have been previously browsed in proportion to known moose densities support the same moose population densities over the long-term? Our main finding of the field study was that some properties of the crown architecture of deciduous trees, such as fractal dimension and twig density, responded quadratically to increased moose population density. At intermediate moose densities, these properties more than compensate for reductions in twig size, leading to small increases in forage production.
Our approach to constructing the model was extremely simple, assembling these equations for the architectural responses of plants to known moose population densities and the winter food requirements of moose, but ignoring other known factors that influence moose population dynamics (e.g. animal feeding rates, population demographics) and other known effects of moose on ecosystems (e.g. changes in soil fertility). The point was to see if these architectural responses in isolation could in principle determine moose population densities and dynamics. In fact, it was the simplicity of the model that kept us from submitting it as a manuscript for several years. John developed a renewed interest in the model after receiving positive comments from colleagues in Sweden following a presentation in 2010. Indeed, the editorial reviewers at Oikos liked the model and our paper because of its simplicity, not in spite of it.
It turns out that the architectural responses of plants that we measured can produce realistic moose population densities for northern Sweden (an average of ~10-15 moose/1000 ha). But these population densities were only sustainable at the sites with the highest productivity and with species compositions heavily weighted toward deciduous trees, which can overcompensate for lost tissue due to moose browsing. One of the new things we found was the quadratic responses of plant architecture to moose population density, especially those of birch, produced oscillations in moose populations on highly productive sites. The lessons we learned from this model were, first, that architectural responses of plant crowns to browsing may play a more important role in regulating moose population density than previously suspected and, second, that these architectural responses might cause complex population dynamics such as population cycles.”