Collaborative Research: Landscape Modeling: The Synthesis of Ecological Processes Over Large Geographic Regions and Long Time Scales

这项研究项目将测量、描述和模拟大面积地理区域的生态模式和过程的变化，以测试景观方法在分析和预测生态过程(即生产力、演替)变化、优化资源管理和评估人类影响方面的有效性。这项工作将通过测试不同空间模拟技术的相对性能以及将新的计算机技术与数据丰富的沿海研究地点相结合，帮助制定开发和测试空间生态系统模型的指导方针。能够预测景观格局变化方式的方法对景观生态学的发展至关重要，但还没有出现连贯的预测工具。研究人员将综合当前的生态数据，并收集新的数据，以纳入预测景观格局变化的空间模型。从基于过程的模拟到转移概率模型，一系列空间模型所包含的理论将使用描述性统计和模式匹配技术系统进行测试，该系统将作为本研究的一部分进行开发和评估。将为三种不同的沿海景观开发模型，对这些景观已经有了大量的历史数据和正在进行的长期研究。结果将提供：(1)增加对控制景观变化的过程的理解；(2)调整空间和时间尺度以优化模型的可预测性的原则；(3)景观数据集的比较；(4)确定特定目标、尺度和分辨率的最佳模型类型的原则；以及(5)检验预测和数据之间的拟合度的新方法，这些数据适用于空间生态数据，并且需要一定程度的空间模式识别。最后，在可能的范围内，该项目将尝试将这些原理和方法综合成用于预测景观动态的普遍理论。调查人员在方法上是创新的。制度支持和设施都很好。这项工作的成果将对建立新的基本生态研究方向以及土地利用规划和资源管理领域具有重要意义。

This research project will measure, describe, and model change in ecological patterns and processes over large geographical areas in ways that will test the utility of the landscape approach for analyzing and predicting changes in ecological processes (i.e. productivity, succession), for optimizing resource management, and for evaluating human impacts. The work will help establish guidelines for the development and testing of spatial ecosystem models by testing the relative performance of different spatial simulation techniques and by combining new computer technologies with data-rich coastal study sites. Approaches that can predict the way landscape patterns change are crucial to the development of landscape ecology, yet no coherent predictive tools have emerged. The investigators will synthesize current ecological data and gather new data for incorporation into spatial models that predict change in landscape patterns. The theories embodied within a spectrum of spatial models, ranging from process-based simulation to transition probability models, will be tested using a system of descriptive statistics and pattern matching techniques which will be developed and evaluated as part of this research. Models will be developed for three different coastal landscapes, for which much historical data and ongoing long term research already exists. Results will provide: (1) increased understanding of the processes controlling changes in landscapes; (2) principles for adjusting spatial and temporal scales to optimize predictability in models; (3) comparison of landscape data sets; (4) principles for determining the optimal type of model for a particular set of objectives, scale and resolution; and (5) new methods for examining the goodness-of-fit between predictions and data that re appropriate for spatial ecological data and which require a degree of spatial pattern recognition. Finally, to the extent possible, the project will attempt to synthesize the principles and methods into a generalized theory for predicting landscape dynamics. The investigators are innovative in their approach. Institutional support and facilities are excellent. Results from the work should be of significance to the establishment of new basic ecological research directions as well as to the areas of land use planning and resource management.