Models of the Dynamics of Mosaic Landscapes

Abstract

Upon reflection, it seems ecological research often involves either dismantling ecosystems into their components or assembling “parts” into ecosystems. Thankfully, some ecologists do both. For example, A. S. Watt upon his return from the “Great War” produced a pair of wonderful papers deconstructing the components and processes of British beech woods (Watt 1923, 1925). These papers outlined the microsites favored by regeneration of different tree species, documented experiments on excluding seed predators and seedling grazers, and investigated how such processes influenced the regeneration patterns of the trees in the canopy gaps produced by tree falls of different sizes. A second “Great War” later, he drew this together (Watt 1947) as a towering synthesis paper, which theorized that vegetation pattern arises from the local spatial dynamics produced by deaths of dominant individual plants and the consequent recovery process. Summation across many mosaic elements with their unstable, quasiperiodic dynamics produced the on-average predictable spatial pattern of the vegetation —an ecosystem quantum mechanics with canopy elements as the quanta. Watt’s paper trail is but one of the threads that interweaves with others to produce the fabric of an ongoing synthesis of plant communities as working mechanisms. Consider Joseph Grinnell’s niche theory (Grinnell 1917a, b). The son of an Indian Agency physician, Grinnell grew up among the Oglala Lakota (Sioux) and was befriended by their chief, Red Cloud (Maȟpiya Luta, 1822–1909) in what is now the Pine Ridge Indian Reservation in South Dakota. Grinnell’s concept of the ecological niche is a huntergatherer’s concept, “What conditions (climate, soil, other associated species, etc.) determine where a particular species of plant or animal will be found?” Other threads include increasing quantification and observational capacity in ecology: remote sensing of large-scale patterns of ecosystems; systems ecology with dynamic equations to formalize theories on change; multivariate statistics to explore and interpret large complex data sets; increased computational power coupled with diminished per calculation cost. These come to together in the formulation, testing, and analysis of computer models of Watt’s mosaic elements based on simulating the birth, death, and growth of individual plants. Called individual-based models (Huston et al. 1988), these algorithms channel Grinnell’s niche theory to determine conditions that might produce successful plant regeneration. With a dollop of ecophysiology to simulate individual plant growth and death, one form of these models, called “gap models” (Shugart and West 1980), essentially reproduce the dynamics that Watt postulated to underlie vegetation pattern. Collectively, gap models (Botkin et al. 1972, Shugart and West 1977) produce a theoretical basis for forest dynamics.