Connor Fitzpatrick is a graduate student based at the University of Toronto Mississauga. The article below gives us a background into his work and recent Journal of Ecology paper; Phylogenetic relatedness, phenotypic similarity and plant–soil feedbacks.

This project started from the simple idea that closely related organisms will interact more strongly due to their similarity in traits that determine ecological function. A lot of people are excited about this idea because it provides an evolutionary context for the sometimes unpredictable universe of ecology. Explaining contemporary interactions between species and the resultant community and ecosystem processes with evolution is an inspiring goal and is what propelled our project.

Connor and co-author Laura Gehant

Plant-soil feedbacks have a rich, but somewhat recent, history in ecology. Traditionally, ecologists thought that plant distributions were shaped by soil properties and climate; relatively unchanging features of the abiotic environment. But in the last 30 years, ecologists found that plants can actually alter the properties of their surrounding soil and that these alterations can affect the growth of plants in future generations. Interestingly, humans have known for hundreds of years that soil conditioning by a given plant can affect the growth of another, for example, farmers using crop rotation or different combinations of crops to provide long-term soil and plant health.

We now have a lot of compelling evidence that plant-soil feedbacks play an important role in plant populations and entire terrestrial ecosystems. Yet I think we’re still puzzled by why plant species respond differently to soil conditioned by another. In other words why does plant species A perform well in soil conditioned by plant species B, but plant species C does not?

We wanted to know if soil feedbacks between pairs of species are influenced by: 1) phylogenetic relatedness, 2) overall phenotypic similarity, and 3) plant traits. We took common and co-occurring species (49 species across the major North American plant families) found at the Koffler Scientific Reserve, north of Toronto, and performed a classic plant-soil feedback experiment.

First step: grow soil conditioning plant species in a homogeneous soil. Second step: grow focal plant species in soil harvested from step 1. Third step: analyse the effect of soil conditioning from step 1 on the growth of plants in step 2.

Soil conditioning phase. We grew 49 species in homogeneous field soil for one growing season. We measured phenotypic traits and harvested soil from each.

We observed strong negative and positive effects of soil conditioning on plant growth in the second generation but of course, this varied with soil conditioning and focal plant species. We also found that our experimentally measured soil feedback between 2 species predicted, to an extent, their co-occurrence in the field.

How about the effect of phylogenetic relatedness, phenotypic similarity, and plant traits? It’s complicated. Despite no overall effect of phylogenetic relatedness and overall phenotypic similarity on the strength of soil feedback between species we did find strong effects of individual plant traits. Specific leaf area was related to both the response of focal plant species to, and the effect of soil conditioning plant species on, plant-soil feedbacks. Furthermore, some plant traits, but not all, exhibited strong phylogenetic signal – i.e. close relatives were more similar to one another.

The focal plant growth phase. We grew 9 focal species in soil harvested from phase 1.

We think the key to explaining these results is the effect of trait differences between species. Large differences in relevant traits between species can sometimes lead to competitive exclusion but they can also lead to stable coexistence, it all depends on what roles those traits have in the ecology of the organism. We think the same applies to plant-soil feedbacks. The soil feedback between plant species will be shaped by differences in relevant traits and the importance of any given trait may vary across different pairs of plants.

Connor Fitzpatrick
Graduate Student, University of Toronto Mississauga