ABSTRACT
Many fields of science have experienced a replication crisis, where results from experiments with low statistical power published in the literature cannot be replicated. Ecology so far has not been drawn into this crisis, but there is no reason to think that this problem is absent in our field. Here, we originally attempted to replicate findings that showed pea (Pisum sativum L.) roots had strong differences in growth in the presence or absence of neighbours. Our original goal was just to develop a simple model system for studying how plant roots respond to competition from neighbours.
In an attempt to replicate previous findings, we performed four separate experiments with 480 individual plants, across three years. Each time plants were grown in the full factorial combination of above and belowground competition. In addition, pea has been studied in similar experiments across six additional studies. Thus, we used meta-analysis to combine previous findings with our new findings.
We were unable to replicate previous findings, and in all four experiments plants grew the same whether there were neighbours or not. Despite variability in individual studies, meta-analysis revealed that pea has no growth responses to neighbours and grows the same whether there is or is not below ground competition.
Synthesis: Many other fields have gradually been drawn into a growing replication crisis, that is thought to be the result of low statistical power. Even though this is just one case study where a somewhat controversial result could not be reproduced, there is no reason to think ecology is immune from the replication crisis. We suggest that solutions developed in other fields might pre-emptively ward off similar problems. These include stricter cut-offs for statistical significance, a growing use of large replicated studies, and considering avenues for pre-registration of methods.
Competing Interest Statement
The authors have declared no competing interest.