How should we measure fungal diversity?

(a quick review of Frøslev et al. 2019)
https://doi.org/10.1016/j.biocon.2019.02.038

That methodologies are complementary – and not competitive – in science is important to keep in mind, but often viewed more in terms of one “being the best”. However, the scientific method depends upon the assurance of results across methodologies and studies, so why denigrate others’ research that does not follow one’s own methods? Isn’t that contrary to what we should do? And let’s be honest: the more research that is conducted, the more jobs that there are for research, and the more scientists that are employed. This apparent discrepancy between science-in-theory versus science-in-action has been a topic of interest to me since my PhD years.

Mycologists, especially, I have worried can be myopically obsessed with the notion of one experimental method winning out over another, to the point of rejecting science that is conducted in a different way. As if too many hours staring into the microscope or pipetting in the lab may have opened up an amazing world of fungi or molecular technology, but prevented re-focus on the surrounding environment of science.

Frøslev et al. (2019) effectively demonstrate the parallel importance of two main methodologies to assess fungal diversity: (1) via comprehensive but time-consuming and laborious fruit body surveys, requiring taxonomically trained specialists, whom are dwindling with time, and (2) by samples of soil and state-of-the-art molecular analyses of the fungal DNA contained within (eDNA), which has increased with time.

The fungal molecular revolution of the ’90s inadvertently served, in ways, as a backlash to fruit body collections in the later decades: we could see that species were identified in soil that never created fruit bodies, and we knew that fungal species can be rather fickle in terms of when they do fruit, hence, eDNA must be the most accurate methodology to use to identify fungal species present in the environment. The dogma has been, to greater or lesser degrees depending on the mycologist: don’t trust fruit body studies, only eNDA. Not every fungus fruits, but every fungus has DNA, hence, “DNA all the way”! However, as the authors explain in the introduction, two sides of a coin (fruit body or eDNA) actually combine to make the worth of something, in this case the value of fungal diversity and not the penny, pence, krone, franc, or other currency. You can look at the one side or the other of the coin, but the value actually remains relatively the same.

Of course there are times when each method is most important, and results via each are certainly not directly comparable across all metrics in ecology. However, the authors turn the question to ask if either method may serve as a substitute for the other, in terms of the correlation of community measures across environmental gradients. Their primary interests are in conservation biology, where to be able to identify a species is the first important step in a staircase down a pathway to an ocean of determining if a species is of conservation threat, especially due to global change and/or land-use and management.

Fungal species data only historically exist in a morphological form (fruit body species), and there is a wealth of fruit body data available. The authors recognized this asset of the past, and ask if we could not merge fruit body and eDNA methods of the current, while taking important historical data as possible to calibrate how fungal species have changed, for example, due to climate change and in terms of conservation? Wouldn’t it be nice if we could stop having to support the use of historical fruit body records, and instead focus on topics of greater ecological interest – with a mind to the future? Yes, it would. And, yes, we can!

For a few years, the same 40m x 40m plots (130 in total), were surveyed simultaneously for fruit bodies and eDNA produced via soil sampling across the major environmental gradients of Denmark. Denmark, as once described to me by a Norwegian, is the glacial remains of Norway. It contains substantial arable land, also marshlands grading between mostly deciduous forests, with lesser extents of conifer representation.

There are many details regarding the sampling and processing of the data that had to be considered. Ultimately, the authors chose to keep taxa most uniform between the two forms of data, and to investigate direct comparisons between the methods. I agree with their choices, but it should be noted if one includes the taxa only found as eDNA, or lesser but possible, fruit bodies, this could conceivably impact some of the results. They provide some comparisons to help with this fact. It seems results are fairly robust, hence, not too worrying.

Not all species that were found fruiting were also found in the eDNA, and species amounts from eDNA far exceeded those species that fruited. This is the norm we find when comparing the two methods. Importantly, the richness values (diversity) extracted from the two methods was highly correlated (r = 0.39 to 0.80). Meaning, when comparing richness in a site for a method, and relative to the other sites, one may expect similar results based on fruit body or eDNA samples. Low fruit body richness ≈ low eDNA richness. Keep in mind that the exact richness value, however, is almost always higher with eDNA than fruit bodies. Hence, a richness of 20 fruit body species at a site may be, for example, 25 eDNA species. A relatively low value is the same between methods, as would be a relatively higher value. Since comparison is most often of greater importance in research than absolute values, these results can be viewed positively for conservation science.

Interestingly, the authors conducted an *a posteriori* cost analyses between the two methods: they were about the same.

Fruit bodies captured more red listed species (i.e., identified as of conservation concern and priority) than did eDNA. Red lists were built off of fruit body surveys, so that to an extent this is not ultimately too surprising – except, wait a minute, this means that eDNA is not without its own faults: we don’t always detect species in soil samples, even if they are obviously present due to fruit body records. The notion of eDNA as the most “true” method to find fungal species, then, is not really correct.

In terms of the ecological information that can be linked to fruit body and eDNA data, the authors found relatively similar information related to environmental gradients. Variation in community composition of species based on eDNA was explained more by environmental variables than for fruit body communities, but this should not be surprising as the environmental variables linked directly to soil attributes. More important is that the variability between fungal communities across the environmental gradients was very similar and strongly correlated (r = 0.67) for fruit body and eDNA data.

The authors credibly defend the use of fruit body surveys (hence, citizen science initiatives) to find and monitor species for fungal conservation – sometimes they do a better job than soil sampling eDNA can, although the latter gains every year. Additionally, and as mentioned by the authors, soil sampling does a rotten job of finding fungi that grows in or on wood. Polypores and lichenized fungi, then, are grossly under-represented. We really do need to employ multiple methods to ascertain fungal species of conservation concern.

So what does this all mean? It means, don’t discount fruit body data. It has disadvantages, namely it cannot taxonomically represent all of fungi, but it “holds its own” in terms of being able to support ecological research in global change and conservation sciences, and is our only historical source. It can even be used to find fungi of conservation priority that the lens of eNDA cannot yet focus upon. It should be possible to compare results between the data sources. Thus, we should retain both methodologies, and be aware of what each can scientifically inform upon. As the authors also suggest, we should consider fruit body and eDNA as truly complementary research.

Flip a coin and catch the breeze of climate change: it will land the same irrespective of which side faces up.