موقع شبرون للتقنية والأخبار- متابعات تقنية:
You’ve probably heard the stories: that through an intricate network of underground fungi, trees send nutrients and warning signals back and forth to one another. In Pulitzer Prize-wining novels, New York Times feature articles, PBS documentaries, and TED talks, there have been ample mentions in recent years about the “woodwide web,” or the fungus-mediated connections that supposedly help forests thrive. But that concept may not be all it’s cracked up to be.
Every so often in science, a revision is in order. A prevailing idea inflates to inaccurate proportions. A set of experiments is taken out of context. Uncertainty gets ignored in favor of the most interesting explanation. Something snowballs from one small over-emphasis to the whole story. Through these multiple avenues of misinterpretation and more, existing research might not actually support the importance of fungal connections between trees for forest health, according to a new analysis. In other words, the “woodwide web” could be canceled.
A sweeping review study, published this month in Nature Ecology and Evolution, presents a counter-narrative to the popular ideas that have come to define our understanding of underground fungi in forests. In the review, three ecologists looked back at all of the published studies they could find on these fungal webs, called common mycorrhizal networks (‘mycorrhizal’ means fungal root).
They searched for evidence to back up three common stories, repeated across media: that common mycorrhizal networks are widespread, that they result in shared tree resources and improve seedling performance, and that trees communicate defense signals through the underground web of fungi. Yet in lieu of convincing evidence, what the biologists found was a compounding series of thinly supported claims and experiments that might not show what scientists previously thought. This forest internet system, built by fungus and establishing a direct line from tree-to-tree, could be no more than a myth.
Through their analysis of hundreds of previously published papers, the ecologists noted that there’s very little evidence to suggest that common mycorrhizal networks are widely occurring. In fact, there are only five total maps of these fungal webs in just two different forest types—not nearly enough to support the assumption that these webs of connections between trees, mediated by fungus, are present in forests around the world.
Related story: Is Plant ‘Intelligence’ Just a Human Fantasy?
The scientists also found that, in studies of resource sharing and tree growth, possible alternate explanations make it impossible to say if trees are passing nutrients to one another via fungi. Carbon and minerals could simply be traveling through soil, not along fungal filaments between trees. Even if mycorrhizal networks are helping trees send support to one another, there’s no evidence that it happens in large enough quantities to boost forest health. The handful of lab experiments have yielded a confusing mix of results, wherein sometimes trees seem to benefit from access to underground fungal connections and sometimes they fare worse.
G/O Media may get a commission
Additionally, the researchers noted that there hasn’t been a single peer-reviewed study demonstrating that trees share defense signals via common mycorrhizal networks in an actual forest setting. In one of the most referenced lab-based experiments, the presence of nearby roots seemed to cancel out the benefit of mycorrhizal connections, a set-up that would never occur in a natural forest.
Finally, the ecologists looked beyond the individual studies to the connections between them. There, they noted it’s not just in popular media that unsupported ideas spread. Scientific research of common mycorrhizal networks has selectively and repeatedly cited a small number of studies that show positive correlations between tree health and mycorrhizal connections, while largely ignoring the rest. Whether inadvertent or intentional, scientists have demonstrated similar biases as the media and the public, according to the researchers.
It seems that we all wanted to believe in the “woodwide web.” To find out why we should be more skeptical, I spoke with the three authors of the new review: University of Alberta ecologist Justine Karst, University of British Columbia biologist Melanie Jones, and University of Mississippi biologist Jason Hoeksema. Below is our conversation, lightly edited for clarity and length.
Lauren Leffer, Gizmodo: Can you explain what you actually found in your review, in the simplest terms possible?
Justine Karst: Well, I guess I would say in very plain words: The [media] story is ahead of the science. That’s, I think, the basic message. But it involves scientists as well: That, perhaps without intention, scientists have contributed to this mis-characterization of common mycorrhizal network function in forests.
Gizmodo: So you’re saying that underground mycorrhizal networks aren’t important for forest or tree health?
JK: No, no, we would not say that. So, we have to back up. We do not dispute the importance of mycorrhizal fungi in forests and ecosystems—forming soils, all of that. There is a lot of evidence to support their importance in forests and other ecosystems as well. What we focused on were these connections: Fungi physically connecting roots of two different trees or seedlings. We were really poking around to find out, ‘what do we know about those connections?’ ‘What are their ecological relevance?’ ‘How do they influence forests?’
And one of the things that came out of this is we actually don’t really know much about the role of the fungi in this story. But I wouldn’t go as far to say that common mycorrhizal networks are not important in forests. I would say that these popular claims are not supported by evidence.
Jason Hoesksema: I think the jury is still out on how important common mycorrhizal networks are in forests, per se. To reiterate, that doesn’t mean that mycorrhizal fungi are not important. We know a great deal about the importance of mycorrhizal symbiosis for trees. But these physical connections between trees—it really still remains to be seen if they’re important in forests.
Gizmodo: What prompted this review?
JK: I felt there were a lot of extraordinary claims about mycorrhizal networks and forests out there. And to the point that I didn’t necessarily remember some of the studies, or I thought maybe I wasn’t on top of the literature.
I thought maybe I was missing some papers, so I wanted to go back and check, well, what is the support for these claims? Am I missing something? And that’s when I contacted Melanie, my former PhD advisor, because I thought, ‘well, Melanie must know all these papers that are coming out.’ And so we chatted about it. And then it kind of goes from there.
Melanie Jones: I’ve been aware for a while that some of the papers, some of the early papers, weren’t always getting cited correctly. Maybe in the excitement of the early days, things were twisted slightly to put a more positive spin on the results. Then of course, that can get amplified over time and get locked into a story that there were only positive results, or that the results were strong. The questions that are raised in the papers, or the alternative explanations drop away. I was aware of that for the early papers, but I hadn’t done a deep dive into some of the more recent ones. So this was a really good excuse to do that.
JH: For me, it was a very similar process. I was hearing these extraordinary claims in the media and wondered whether I was missing something. And I think there were a couple of instances in particular where I was pushed over the top, to dive back into it. Some of my graduate students and I started reading some of these recent papers more closely. We found that the evidence didn’t match up with some of the claims.
Gizmodo: When you all started this, what were you expecting to find?
JK: There were so many surprises on the way. I was not expecting to see the citation bias. That was a huge surprise for me. I was not expecting to find that, for example, there’s no evidence or testing of the claim that big old trees are sending defense signals to kin in a forest.
JH: I expected to find a lot of variability in results, a lot of variability and outcomes. I already had a feel for that. And I knew that the outcomes of these experiments were highly variable in what I had read before. But what I did not expect was, the deeper we dug, how the results were sometimes contradictory to the current narratives. And also some of the real limitations of the experiments that came out in our group discussions, that was a surprise to me as well.
Gizmodo: Do you have any idea why those three ideas you examined in your review seem to have taken off in such a big way?
MJ: I think especially the ideas of sharing or moving material between trees, and especially among related individuals—it’s a very heartwarming story. And these days, I think people are looking for heartwarming stories. There’s enough bad negative stuff, people in-fighting, all that kind of thing out in the world. People want to hear something happy.
And, you know, I just want to say that, in this system, we are talking about a symbiotic system, where fungi and plants are growing together. And although they’re under their own natural selection and they’re functioning as individuals. there’s absolutely no question that the fungus depends on the trees to get its carbon, to get its energy, and to be able to reproduce. And that when we try to plant trees in places where there aren’t these mycorrhizal fungi, they don’t thrive.
This is still a story that talks about organisms that have some mutual benefit to each other, especially over the course of their lives. It’s just the tree-to-tree part that we really need to look at a lot more.
Gizmodo: Can you walk me through each of those three ideas you examined?
JK: Our first claim we were looking at was whether common mycorrhizal networks are widespread in forests. And to answer this question, ideally, we could dig in the soil and trace out this network happening or present between trees. But we can’t do that, because as soon as you dig in the soil, it basically destroys the network. So we use indirect tools to map these networks, and mapping.
Mapping is very hard work. It’s very tedious. What they’re doing is they’re genotyping roots of trees, genotyping particular species of fungi, at discrete locations, and then inferring whether there’s connections between roots of those trees by a particular genotype. It’s challenging work, technically.
We found that there are only five such maps done in the world. Four of those maps were done in a single forest type. So these are interior Douglas fir forests. And there was another map done in a pine forest in Japan. Of course, there are many, many, many different forest types in the world. We only have any data for two.
Then, of those five maps, just two were done in a way so that we could be very confident [in the fungal connections]. So, to say that common mycorrhizal networks are widespread in forests, we just don’t actually have the evidence to be making that claim. We think it’s very likely that trees are connected below ground. But again, when we’re really pushing for the evidence, and there’s not very much there.
The second claim that we investigated was whether seedlings benefit from plugging into these mycorrhizal networks, and if those benefits are through resource transfer. To evaluate that claim, we dug up all the field studies we could find. What we found is that there’s actually no conclusive evidence that resources are moving through these mycorrhizal networks. And that’s because the experiments are not designed so that they can pinpoint that resources are actually moving through a continuous fungal link.
There is good evidence that there are small amounts of resources like carbon moving below ground. You can imagine a tree photosynthesizing and taking up some carbon, then some of it seems to be leaking through mycorrhizal roots into the soil, picked up by another mycorrhizal root of a different tree, and then moved into that second tree. So there’s this little bit of below ground transfer happening that doesn’t actually require a mycorrhizal network. But there is no conclusive evidence that resources are moving through mycorrhizal networks.
The other part of this claim is evaluating whether seedlings benefit from access to these mycorrhizal networks. For a variety of reasons, the experiments we use have these confounding effects that make it difficult to, again, pinpoint the effects of access to a mycorrhizal network versus other factors that could be influencing seedling growth. We found that in the majority of cases, seedlings show no response to the potential access to mycorrhizal networks.
Gizmodo: And the third point, about signal sharing through fungal root networks?
JH: We found no peer-reviewed published evidence from studies and forests. We paid close attention to a couple of highly cited laboratory experiments, greenhouse experiments, that are often mentioned in association with these claims.
There’s one greenhouse experiment for example, that tested whether, when Douglas fir is either clipped to simulate insect damage or actually has insects attacking it, there’s potential for it to signal to a ponderosa pine seedling neighbor. In that experiment, they found some really intriguing results that the Ponderosa pine defensive enzymes were up-regulated when the Douglas fir was under attack.
But they also included another treatment in that experiment that allowed not only mycorrhizal networks, but also allowed the neighboring seedlings to have their roots intermingle naturally as would occur in a wild forest. And when they allowed roots to intermingle, the apparent signaling effect went away. The recipient seedlings were no longer up-regulating their defensive enzymes in response to the target plants being attacked. That is a really fascinating result, and we need to understand that kind of variability in the potential effects of mycorrhizal networks.
There’s another greenhouse experiment that tested whether adjacent seedlings that were close relatives of each other, whether resources would be transferred between them more readily than between unrelated seedlings. They included a series of treatments, some of which allowed mycorrhizal network connections to form by themselves, some of which also allowed transfer of resources through other pathways, for example, through the soil. They found evidence that there was more carbon moving between closely related seedlings than between unrelated seedlings.
A really intriguing result, absolutely. But in that experiment, the evidence suggests that the carbon movement was not specifically mediated by mycorrhizal fungi. Instead, a large portion of it may have been moving just through the soil without the aid of mycorrhizal networks.
Gizmodo: The thing that I’m getting from all of these examples is that something is, in fact, going on with trees underground—some transfer is happening, either of information or nutrients. But there’s no solid evidence that these fungal networks are responsible.
JK: That, and we don’t know if those transfers have any influence on seedling performance.
Gizmodo: You mention in your paper that some of your own research included a few of these problems we’re talking about, like misused citations or even that you came to conclusions that, revisiting them now, you aren’t necessarily so confident in. Coming from that perspective, how did this happen in science and in this field?
JK: Yeah, well, I’ll put it out there that it hurts a little bit. When I revisited those papers and looked at them with fresh eyes, I think for me, the biggest thing was recognizing the confounding effects as part of those experiments. Once I was convinced that, ‘oh, yeah, there are some confounding effects,’ I couldn’t unsee it.
It was then that I started looking at other experiments, and I was just like, ‘oh, no, yeah, I see what the problems are.’ Some of those confounding effects, like for example, the one that I described earlier with the differences in pathogens across treatments, I was just not thinking about that when I set up that experiment. I was so focused on the common mycorrhizal networks. And then, when we realized and identified these other factors, it’s like, as any scientist should, you go back and you look at these previous experiments, and decide ‘okay, these are not actually conclusive or they’re not showing what I thought they were.’ So it was an exercise of having to update my knowledge, update my interpretations, and being open to some limitations that I did not see before. And now, like I said, I can’t unsee them.
MJ: I think a lot of it really comes down to confirmation bias. We’ll make a hypothesis that carbon is moving through common mycorrhizal networks. And therefore, if we do these treatments, and we see this result that supports that hypothesis, then we assume that’s what happened.
But as Justine said—even though generally, we know as scientists, there can always be some other explanation—we didn’t really think about the other explanations at the time. And because it’s such an exciting idea, it just grew and grew and grew.
Gizmodo: That brings me to this question of alternate explanations. What are some of the biggest alternative explanations that you came to in doing this review?
JH: I’ll just quickly mention the one from my paper. When you do these experiments in the field, you have to physically manipulate the potential for these mycorrhizal networks to form. It’s very easy, in doing that, to alter other things in the system, like the type of fungi that are encouraged to grow in a particular treatment. There are diverse fungal species that can be involved in these treatments, and by putting in barriers or removing barriers to mycorrhizal networks, we’re changing the potential for different kinds of growth—both beneficial and pathogens.
In fact, in four different studies that have tested this, they have found common experimental treatments alter the fungal community, and those different fungi may have different traits that influence the growth of plants. That’s a potential confounding factor that you can’t easily tease apart from the effect of being connected to a mycorrhizal network, per se.
MJ: One of the studies that I was involved in is a really well known one. Suzanne Simard is the first author on it, and it was published in Nature in 1997. The conclusions for that result relied on control treatment that we had, where there was a tree that could not form a mycorrhizal network. So we had two tree species, birch and Douglas fir that could form a network, and a third one, western red cedar, that had a different type of mycorrhiza and couldn’t join in.
Our conclusion was the carbon must be moving via the mycorrhizal network because the cedar received much less carbon than the other two did. But we’ve learned since that these [types of fungi the cedars do form] are much more diffuse and are [worse at uptaking nutrients]. Conversely, the [birch and fir] fungi are better at taking up exudates that might leak out of another mycorrhizae or out of roots. So, our results could’ve been explained just by the fact that two of the species were forming the one type of mycorrhiza that have more hyphae and can slurp up the exudates better than the other type.
Gizmodo: So it maybe wasn’t that the cedar couldn’t key in with that existing network between the other two species. It was just that it doesn’t use ectomycorrhizal networks at all.
MJ: Yeah, it’s the type of mycorrhizal association. It has a very different group of fungi. Very common, but a very different group of fungi. They don’t seem to be as good at taking up organic nutrients from the soil.
JK: Then I guess there’s one last sort of confounding effect that we discussed in that paper. It has to do with these mesh bags. So the mesh bags, we use them and they have different pore sizes that either allow fungal hyphae to travel through, prevent them, or allow roots to travel through or prevent them and in different combinations.
Ideally, what we’re doing with those mesh bags is we’re manipulating whether the seedling in the bag has access or not, to be able to form common mycorrhizal networks or to interact with neighboring roots. But one confounding effect that could happen is that the amount of soil volume available for these foraging hyphae could be affected by the mesh treatment.
If you imagine a seedling growing in a pot, we want that pot to be big enough so that the roots are not going to become restricted by the pot because if they become restricted, the seedling doesn’t grow so well. The seedlings in our mesh bags are of course colonized by mycorrhizal fungi. And the fungi are growing out to forage for nutrients, water, and everything else that they provide the seedling.
But the bags that we use with the smallest pores that don’t allow any fungi to grow in or out—those roots and fungi could become restricted because now they can’t access all of the soil volume. If they’re restricted in the soil volume that they can forage, it might be that the seedling then is restricted in the resources that it’s accessing through those mycorrhizal fungi.
The mesh bags could also be changing the amount of soil volume available to these mycorrhizal fungi. We might interpret that to be, ‘Oh, it’s access or not to mycorrhizal networks.’ But it’s like, no, it doesn’t have anything to do with that. It’s all about the changes in soil volume that the seedlings have access to.
Gizmodo: Got it. So it’s not even network access. It’s just the basic soil nutrient access.
MJ: Pretty much. Or, it could be! We don’t know.
Gizmodo: Moving into the future, what are you suggesting and hoping that researchers take away from your review and do?
JK: Our intent is not to put a chill on research on common mycorrhizal networks and forest. If anything, I mean, we’re quite encouraged and inspired about the next generation of experiments.
I think we’re hoping to reset and reorient the field in terms of what happens next. The other really important component is the citation bias. It’s acknowledging that, okay, there is some bias going on in the scientific literature. It’s not just out there in the public media.
Then, there’s this name for mycorrhizal networks. We call it the ‘woodwide web.’ I mean, everyone knows it by that name. But it’s a bit of a misnomer, because it treats fungi as just these cables that are moving stuff around for trees. But it ignores that they’re individual organisms—they’re not just these passive conduits for information and resources. Fungi are trying to maximize their own growth and survival.
Gizmodo: Do you have a proposed alternative to the name that’s equally snappy?
JK: I’ve been trying to think of this. I mean, woodwide web has so much cache. It is so catchy. And it’s shorthand for so much. It’d be hard to take it away from the public. But I think if we could think about ways to sort of put guardrails on it, some constraints. But no, I’m not sure if we’ll ever get rid of it.
MJ: You think of something, Lauren. Forest and fungi, there’s got to be something that works in there. Nature magazine came up with that original ‘woodwide web.’ It wasn’t the scientist.
Gizmodo: Well, I’ll certainly be workshopping headlines with my editor. Fingers crossed we strike gold.
You say that you want this study to be something of a wake-up call for the field. Do you think that this problem is widespread, beyond just this particular area?
JK: Anecdotally, yes. Before the paper came out, we each did seminars in different places. And one thing that struck me was, after giving a seminar of our study, people came up to me from different fields and said, ‘we’ve got a similar problem.’ I think that speaks to the pressure put on scientists to hype results, for a lot of reasons: to get grants, to get papers, click baits, whatever it is. It seems to be that science has kind of been twisted in a way that’s not actually good for science.
MJ: What journals tend to look for is something new. You know, ‘are they changing the paradigm?’ ‘Are they doing something that’s really changing the field?’ Unfortunately, there’s not a lot of reward for some of the types of experiments we’re asking for in our paper.
Gizmodo: One final question: Is it possible there’s actually some alternate explanation or confounding effects you haven’t considered that better explain all of the things you found in your review?
JK: Maybe there’s another explanation for all of this. We’re excited to see it and to keep an open mind for that alternative explanation.
JH: One alternative is that there could very well be important effects of common mycorrhizal networks in forests. And we’re hoping for more evidence along those lines to help clear that up going forward.
