基于生态信息诊断框架的杉木人工林生物量生态预测研究

森林生物量会直接影响森林生态系统服务的评估。如何运用景感生态学，准确预测区域尺度下森林生物量的时空演变趋势，是关乎国家重大方针政策制定和生态产业体系建设的关键性战略课题。本研究目的是构建一套生态信息诊断框架，优化趋善化模型（3PG2模型）结构，解决由于模型结构设计所导致在森林景感营造过程中生态预测的不确定性。以杉木林分布广泛的福建南靖县为研究区域，选择合适的阈值范围和空间统计分析识别出模拟生物量的不确定性区域，构建包含Geogdetector软件、遗传技术和计算机程序3个部分组成的生态信息诊断框架，使用Geogdetector软件阐明多重因素交互作用对模型模拟的影响及机理，采用遗传技术优化模型结构以提升模拟精度，运用计算机程序和3PG2模型准确预测区域尺度杉木林生物量的时空演变趋势。结果表明：林龄是导致3PG2模型生物量模拟结果不确定性的主导因素。通过景感生态学（谜码数据和趋善化模型）构建的生态信息诊断框架可以准确预测森林生物量，实现区域尺度上的可持续森林管理。

Prediction of the biomass of a Chinese fir plantation based on an ecological information diagnosis framework

Forest biomass directly affects the value of forest ecosystem services. Determining how to apply new technologies of landsenses ecology to accurately predict the spatiotemporal evolution of forest biomass at the regional scale is a key strategic issue for the formulation of major national policies and ecological industry systems. The purpose of this study was to construct a set of ecological information diagnosis frameworks to optimize the structure of the 3PG2 forest growth model and to solve the problem of uncertainty in an ecological prediction the defects in the model structure during the process of forest landscape construction. Nanjing County, Fujian Province was selected as the study area, where the Chinese fir forests are widely distributed. A threshold scale was selected, and spatially statistical analysis was used to identify the uncertainty of the biomass simulation results. The Geogdetector software was used to construct an ecological mechanism of multi-factor interactions on the model simulation, and genetic technology was used to optimize the model structure to improve the simulation accuracy. The computer program (python) and the 3PG2 model were used to accurately predict the spatiotemporal evolution trend of the Chinese fir forest biomass on a regional scale. The results showed that forest age is the dominant factor that leads to uncertainty in the biomass simulation results of the 3PG2 model. We conclude that sustainable forest management and ecological information at the regional scale requires the accurate prediction of forest biomass, and can be achieved through the fusion of mix-marching data and optimizing models.