The best predictors of soil feedback were selected on the basis of Akaike information criterion (AIC) from five groups of variables: plant traits, soil abiotic properties, and soil fungal community properties (pathogens, AM fungi, and saprotrophs). Each variable was tested in a separate linear model as a predictor of soil feedback, and predictors yielding the lowest AIC scores within each variable set were retained for use in the path analysis to explore possible paths by which plant traits and soil properties jointly affect soil feedback (table S1). Additional variables explaining more than 10% of variation in the feedback index, and not strongly correlated with the best predictor within each variable group, were also included in the path analysis. Within the set of abiotic soil properties, many variables were highly correlated and aligned closely with the first principal component. Hence, the latter was included in the path analysis to reflect the major axis of variation in abiotic soil properties. For the analysis of specific feedback, AIC scores with a sample size of 55 were calculated, as our dataset contained 55 mesocosms in the conditioning stage, representing true replicates in plant trait and soil measurements.

Path analysis was used to test whether fungal communities (putative pathogens, AM fungi, and saprotrophs) in conspecific and heterospecific soil affected plant-soil feedback and whether soil abiotic properties and plant traits have a direct or microbially mediated indirect effect on plant-soil feedback. Plant traits could be affected by soil abiotic conditions [i.e., phenotypic plasticity (20)]. In addition to direct effects of traits and abiotic conditions on saprotroph abundance and diversity, saprotrophs could also be affected by pathogens and AM fungi via changes in the rate of tissue death and litter properties as well as direct biotic interactions between different fungal guilds (3234). Paths with P values higher than 0.1 were sequentially dropped (table S4).

For specific plant-soil feedback, the effects of plant traits and soil abiotic and biotic properties were explored separately for soils conditioned by conspecifics and heterospecifics (table S5). For conspecific soil properties, paths significant at P < 0.1 were retained. For heterospecific soil properties, no significant paths were identified and paths with a significance of P < 0.4 were retained to illustrate the contrast in the strength of paths in conspecific versus heterospecific soils. In the final model, all retained properties from the conspecific and heterospecific soil models were combined into a single path analysis.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.