尺度扩展（Scaling up）过程中的植物类群合并问题是生态系统建模领域的难点之一。该文采用机理性生理生态学模型对内蒙古典型草原9种植物净光合速率和气孔导度与环境因子的关系进行了分析，净光合速率模型和气孔导度模型分别能够平均解释生长季内78.19%和55 .87%的净光合速率和气孔导度的日变化。在此基础上根据拟合得到的8个植物生理参数进行聚类分析，将内蒙古典型草原9种植物分为3个植物功能类群：克氏针茅（Stipa krylovii ）、阿尔泰狗哇花（Heteropappus altaicus）、冷蒿（Artemisia frigida）、银灰旋花（ Convolvulus ammannii）和小叶锦鸡儿（Caragana microphylla）抗旱性强，生物化学光合能力中等，归为一类称为强抗旱中光合植物功能类群；羊草（Leymus chinensis）、芨芨草（Achnatherum splendens）和马蔺（Iris lactea）抗旱性中等，生物化学光合能力较强， 归为一类称为中抗旱高光合植物功能类群；串铃草（Phlomis mongolica）抗旱性和生物化学光合能力都比较低，称为低抗旱低光合植物功能类群。在对多种植物存在的自然生态系统 进行模拟时可以按此方法将植物分成若干具有相似特点的功能类群，而不必对每一种植物都作模拟。这种处理方法可以降低模型复杂性和节省运算时间，较之于只用优势种来代替所有物种的模拟也更加接近实际情况。这将为生态系统模型尺度扩展过程中如何合理有效合并植物类群，从而正确判别植物功能型提供一种可行的方法。
Aims The issue of rationally classifying plant species into plant functional types at different scales has been a major challenge in ecosystem sciences, especially ecosystem simulation. A typical steppe in Inner Mongolia of China was chosen for study. We asked: 1) Can plant species be classified into several plant functional groups according to their ecophysiological characteristics of stomatal conductance and net photosynthesis? 2) Are there common ecophysiological traits of each plant functional group? 3) What are the advantages and disadvantages of this classification method in ecological modeling?
Methods Diurnal stomatal conductance and net photosynthetic rate of nine plant species were measured in the field during May, July, and late August 2005. Ecophysiological characteristics of these species were quantified by applying models of stomatal conductance and net photosynthesis to the field data. The models were fitted to the data to obtain model parameters for each species. The analysis showed that the model explained up to 55.87% and 78.19% of the variation in the stomatal conductance and net photosynthetic rate, respectively. Cluster analysis was then applied to identify plant functional types on the basis of the model parameters, which are regarded as ecophysiological traits of plant species.
Important findings Nine plant species were classified into three plant functional groups: 1) highly drought-resistant plants with moderate photosynthetic efficiencies, including Stipa krylovii, Heteropappus altaicus, Artemisia frigida, Convolvulus ammannii and Caragana microphylla;2) medium drought-resistant plants with high photosynthetic efficiencies, including Leymus chinensis, Achnatherum splendens and Iris lacteal;3) low drought-resistant plants with low photosynthetic efficiencies, including Phlomis mongolica. This study suggests that plant species in natural ecosystems can be classified into several plant functional groups using our methods. Therefore, the complexity of ecological models and calculation times can be reduced by substituting plant functional groups for individual species. Our approach can be an effective way to quantitatively distinguish plant traits, thus contributing to scaling up of ecosystem simulation.