Nitrogen Cycling and the Control of Chaos in a Boreal Forest Model

Abstract

The cycling of nitrogen — a growth limiting nutrient — between plants and soils is a strong feedback in forest ecosystems. The rate by which nitrogen is released from decaying plant litter controls its supply rate to plants with which they can fix carbon through photosynthesis; the decay rate of litter in turn is controlled by its carbon chemistry, which varies among species. Here, we present a well-validated forest ecosystem model that simulates this feedback and its interaction with competing plant species and climate. We show, using spectral analysis and the Grassberger-Procaccia algorithm, that under climates typical of northern hardwoods the model produces a periodic attractor but under cooler climates typical of boreal forests the output is a strange attractor. We further show that these attractors are governed by this feedback — the chaotic behavior of the boreal forest disappears when the assumption that plants are limited by nitrogen is relaxed. We then augment (by fertilization) or decrease (by leaching or harvesting) the amount of nitrogen that is cycling through the forest to determine if we can control this behavior. Augmenting the supply of nitrogen to the plants increased productivity, but also increased variance of model output, essentially increasing the scaling region of the attractor. Decreasing nitrogen availability decreased variance, but also decreased productivity. This suggests that increasing productivity by fertilization may come at the expense of long-range predictibility. If sustainability of production implies long range predictibility, then it appears that increases in production are not sustainable. 1.