Cui Yang, Wang Silong, Yan Shaokui
Institute of Applied Ecology, Chinese Academy of Sciences, Shenyang, China.
University of Chinese Academy of Sciences, Beijing, China.
PLoS One. 2016 Mar 1;11(3):e0150380. doi: 10.1371/journal.pone.0150380. eCollection 2016.
Phi coefficient directly depends on the frequencies of occurrence of organisms and has been widely used in vegetation ecology to analyse the associations of organisms with site groups, providing a characterization of ecological preference, but its application in soil ecology remains rare. Based on a single field experiment, this study assessed the applicability of phi coefficient in indicating the habitat preferences of soil fauna, through comparing phi coefficient-induced results with those of ordination methods in charactering soil fauna-habitat(factors) relationships. Eight different habitats of soil fauna were implemented by reciprocal transfer of defaunated soil cores between two types of subtropical forests. Canonical correlation analysis (CCorA) showed that ecological patterns of fauna-habitat relationships and inter-fauna taxa relationships expressed, respectively, by phi coefficients and predicted abundances calculated from partial redundancy analysis (RDA), were extremely similar, and a highly significant relationship between the two datasets was observed (Pillai's trace statistic = 1.998, P = 0.007). In addition, highly positive correlations between phi coefficients and predicted abundances for Acari, Collembola, Nematode and Hemiptera were observed using linear regression analysis. Quantitative relationships between habitat preferences and soil chemical variables were also obtained by linear regression, which were analogous to the results displayed in a partial RDA biplot. Our results suggest that phi coefficient could be applicable on a local scale in evaluating habitat preferences of soil fauna at coarse taxonomic levels, and that the phi coefficient-induced information, such as ecological preferences and the associated quantitative relationships with habitat factors, will be largely complementary to the results of ordination methods. The application of phi coefficient in soil ecology may extend our knowledge about habitat preferences and distribution-abundance relationships, which will benefit the understanding of biodistributions and variations in community compositions in the soil. Similar studies in other places and scales apart from our local site will be need for further evaluation of phi coefficient.
Phi系数直接取决于生物的出现频率，已在植被生态学中广泛用于分析生物与地点组的关联，以表征生态偏好，但它在土壤生态学中的应用仍然很少。基于一项单一的田间试验，本研究通过将phi系数得出的结果与排序方法在表征土壤动物-栖息地（因子）关系方面的结果进行比较，评估了phi系数在指示土壤动物栖息地偏好方面的适用性。通过在两种亚热带森林之间相互转移去除动物的土壤核心，实现了八种不同的土壤动物栖息地。典型相关分析（CCorA）表明，分别由phi系数和从偏冗余分析（RDA）计算出的预测丰度所表达的动物-栖息地关系和动物类群间关系的生态模式极为相似，并且在两个数据集之间观察到高度显著的关系（Pillai迹统计量=1.998，P = 0.007）。此外，使用线性回归分析观察到phi系数与螨类、弹尾目、线虫和半翅目的预测丰度之间存在高度正相关。通过线性回归还获得了栖息地偏好与土壤化学变量之间的定量关系，这与偏RDA双变量图中显示的结果类似。我们的结果表明，phi系数在局部尺度上可用于评估粗分类水平上土壤动物的栖息地偏好，并且phi系数得出的信息，如生态偏好以及与栖息地因子的相关定量关系，将在很大程度上补充排序方法的结果。phi系数在土壤生态学中的应用可能会扩展我们对栖息地偏好和分布-丰度关系的认识，这将有助于理解土壤中的生物分布和群落组成变化。除了我们当地的地点之外，还需要在其他地点和尺度上进行类似研究，以进一步评估phi系数。
