To understand impacts of post-disturbance assembly mechanisms on the functional diversity (FD) of plant communities, it is necessary to determine how the environment drives their functional trait composition. In the boreal forest, post-fire abiotic filters may control community assembly by selecting plants with specific traits. Ericaceous heaths are characterized by low FD and are thought to be subject to such filters. We hypothesized that soil parameters select for a specific suite of traits and act as a secondary abiotic filter in post-fire ericaceous heath and contribute to the observed reduction of FD. We measured six soil parameters, five functional traits, and plant species abundances in eight post-fire heath and four regenerating forest sites in Eastern Canada. We conducted a combined analysis of RLQ (R-table Linked to Q-table) and fourth-corner methods to examine the links between plant traits and plot-level soil parameters, mediated by species abundances. Only below ground traits were significantly linked to soil variables. Specific root length and ericoid mycorrhizal associations were negatively linked to total soil nitrogen, available ammonium, and pH. Post-fire heath soils favour a specific suite of species traits. Only a portion of the regional species pool possesses the above-mentioned traits, and when they are favoured by habitat conditions, they assemble into a community with low FD. The novelty of our study is here we show how the relationship between traits and soil chemistry can act as a secondary filter and exert community-level trait changes responsible for the low functional diversity observed in heaths.

要了解干扰后组装机制对植物群落功能多样性（FD）的影响，有必要确定环境如何驱动其功能性状组成。在北方森林中，火灾后的非生物滤池可以通过选择具有特定性状的植物来控制群落的组装。含碳荒地的特点是FD较低，并且被认为会受到此类过滤器的影响。我们假设土壤参数选择了一组特定的性状，并在火灾后的泥质荒地中充当次生非生物过滤器，并有助于观察到的FD降低。我们在加拿大东部的八个火后荒地和四个再生林地中测量了六个土壤参数，五个功能性状和植物物种丰度。我们对RLQ（与Q表关联的R表）和第四角方法进行了组合分析，以检查由植物物种丰富度介导的植物性状与样地土壤参数之间的联系。仅地下特征与土壤变量显着相关。具体的根长和类胡萝卜素菌根协会与土壤总氮，有效铵和pH呈负相关。火灾后的荒地土壤有利于特定的一组物种性状。只有一部分区域物种库具有上述特征，并且当它们受到栖息地条件的青睐时，它们聚集成一个FD较低的群落。