“Positive Citation Bias and Over-Interpreted Results Lead to Misinformation on Common Mycorrhizal Networks in Forests”, 2023-02-13 ():
[author commentary] A common mycorrhizal network (CMN) is formed when mycorrhizal fungal hyphae connect the roots of multiple plants of the same or different species below-ground. Recently, CMNs have captured the interest of broad audiences, especially with respect to forest function and management.
We are concerned, however, that recent claims in the popular media about CMNs in forests are disconnected from evidence, and that bias towards citing positive effects of CMNs has developed in the scientific literature.
We first evaluated the evidence supporting 3 common claims:
The claims that ‘CMNs are widespread in forests’ and
that ‘resources are transferred through CMNs to increase seedling performance’ are insufficiently supported because results from field studies vary too widely, have alternative explanations or are too limited to support generalizations.
One common assertion is that seedlings benefit from being connected to mature trees via CMNs. However, across the 28 experiments that directly tackled that question, the answer varied depending on the trees’ species, and on when, where, and in what type of soil the seedling is planted. In other words, there is no consensus. Allowed to form CMNs with larger trees, some seedlings seem to perform better, others worse, and still others seem to behave no differently at all. Field experiments designed to allow roots of trees and seedlings to intermingle—as they would in natural forest conditions—cast still more doubt on the seedling hypothesis: In only 18% of those studies were the positive effects of CMNs strong enough to overcome the negative effects of root interactions. To say that seedlings generally grow or survive better when connected to CMNs is to make a generalization that simply isn’t supported by the published research.
The claim that ‘mature trees preferentially send resources and defence signals to offspring through CMNs’ has no peer-reviewed, published evidence.
We next examined how the results from CMN research are cited and found that unsupported claims have doubled in the past 25years; a bias towards citing positive effects may obscure our understanding of the structure and function of CMNs in forests.
We conclude that knowledge on CMNs is presently too sparse and unsettled to inform forest management.
…One of us (Jones) was involved in the first major field study on CMNs, published more than 25 years ago. That study found evidence of net carbon transfer between seedlings of two different species, and it posited that most of the carbon was transported through CMNs, while downplaying other possible explanations. This is what’s known as “confirmation bias”, and it is an easy trap to fall into. As hard as it is to admit, it was only due to our skepticism of the recent extraordinary claims about the wood-wide web that we looked back and saw the bias in our own work.
Over decades, these and other distortions have propagated in the academic literature on CMNs, steering the scientific discourse further and further away from reality, similar to a game of “telephone.” In our review, we found that the results of older, influential field studies of CMNs have been increasingly misrepresented by the newer papers that cite them. Among peer reviewed papers published in 2022, fewer than half the statements made about the original field studies could be considered accurate. A 2009 study that used genetic techniques to map the distribution of mycorrhizal fungi, for instance, is now frequently cited as evidence that trees transfer nutrients to one another through CMNs—even though that study did not actually investigate nutrient transfer. In addition, alternative hypotheses provided by the original authors were typically not mentioned in the newer studies.
As these biases have spilled over into the media, the narrative has caught fire. And no wonder: If scientists themselves could be seduced by potentially sensational findings, it is not surprising that the media could too.