The Effects of Land Management on Fire Regimes and Landscape-Scale Vegetation Patterns in the Mountain Ranges of Southern California

南加州的陆地生态系统受到周期性野火的影响，但火灾制度在历史上已经被扑灭，从夏季的自然闪电点火到极端火灾天气期间人为引起的点火源和灾难性火灾的转变。火情的改变对植被结构、组成和景观格局产生了深刻的影响，并对人类的野火危害产生了深远的影响。在这项研究中，一个模拟景观动态的计算机模型（LANDIS，由David Mladenoff等人开发）将被用于预测南加州区域尺度上火灾频率、强度和大小变化对植物群落演替的影响，以及由此产生的植被组成和结构模式。这些火灾状态的变化可能是由灭火、人为点火、土地利用变化、气候变化或这些因素之间的相互作用造成的。它们可能导致当地植物物种或群落的丧失（否则被认为是广泛存在的，而不是罕见的），栖息地模式的变化，对传播能力有限或有特定栖息地要求的动物产生潜在的负面影响，并反馈给火灾状况（或多或少易燃的景观）。模拟建模用于预测在空间和时间尺度上的变化，因为实验是不可能的，监测数据在空间和时间上是稀疏的。通过模拟实验将解决以下问题：火灾间隔的增加是否会导致森林和灌木林物种优势和景观年龄结构的大规模变化？预测的变化发生在哪里？模型结果将通过以下方式进行评估：a)测试由重复模型运行产生的景观格局指标差异的显著性；B)与火灾历史地图的比较；c)与文献报道的和通过重新调查1930年首次调查的植被样地确定的植被动态的点特异性格局进行比较。模式预测可以为未来生态系统动态的地理分层长期监测提供建议。

The terrestrial ecosystems of southern California are subject to periodic wildfires, but fire regimes have been altered historically by fire suppression, and a shift from natural lightning ignitions in summer to human-caused ignition sources and catastrophic fires during extreme fire weather. An altered fire regime has a profound influence on vegetation structure, composition and landscape pattern, and wildfire hazards for humans. In the proposed research, a computer model that simulates landscape dynamics (LANDIS, developed by David Mladenoff and others) will be used to predict the effects of changes in fire frequency, intensity and size on plant community succession, and on the resulting patterns of vegetation composition and structure at a regional scale in southern California. These changes to the fire regime may result from fire suppression, anthropogenic ignition, land use change, climate change, or interactions among these. They could result in local loss of plant species or communities (that are otherwise thought of as widespread, not rare), changes in habitat pattern with potentially negative effects on animals with limited dispersal ability or specific habitat requirements, and feedbacks to the fire regime (a more, or less, flammable landscape).Simulation modeling is used to predict changes over spatial and temporal scales for which experimentation is impossible and monitoring data are sparse in space and time. The following will be addressed through modeling experiments: Do increased intervals between fires lead to large-scale changes in forest and chaparral species dominance and landscape age structure? Where do the predicted changes occur? The model results will be evaluated by a) testing for the significance of the difference in landscape patterns metrics resulting from replicated model runs; b) comparison with fire history maps; and c) comparison with site-specific patterns of vegetation dynamics reported in the literature and determined through re-survey of vegetation plots first surveyed in ca. 1930. Model predictions can yield recommendations for geographicallystratified long-term monitoring of future ecosystem dynamics.