Looking back to move forward: evolutionary history and plant-soil feedbacks

This #psejclub post is by Michael Van Nuland and Rachel Wooliver. If you’d like to write a guest post, please get in touch! Get involved with the discussion of this post using the comments section below, or on Twitter or Facebook using #psejclub.

One of the most fundamental questions in ecology is to better understand the distribution and structure of biodiversity in nature. Plant-soil feedback (PSF) studies have brought lots to the table in this regard over the last decade by expanding what we know of plant community dynamics and patterns of succession. However, our understanding of these scenarios is rarely the complete picture (is anything?). Incorporating evolutionary ideas into plant-soil linkage studies offers an important and fruitful research direction because the current situation of species and communities can be as much a reflection of their evolutionary history as their contemporary interactions. In other words, species in the ecological theatre may play out on the evolutionary stage (Post & Palkovacs 2009), but everyone has a story.

The novelty of the recent Anacker et al. (2014) study is in using species evolutionary relatedness to understand plant-soil feedbacks in the context of whole communities. This work used a combination of previous studies (notably Klironomos et al. 2002, 2003), mixed with some phylogenetic inference (more on that below), to evaluate similarity in PSFs among ~60 old-field plant species. One of the most interesting approaches in the study comes from comparing phylogenetic signals in PSF to that of plant abundance in the field (i.e., bringing things back down to those basic questions in ecology).

University of Guelph Arboretum, home of the long-term mycorrhizal research station used in the Klironomos 2002, 2003 studies (photo: Gardens of Canada)

The major take-home message is that evolutionary history can leave plant species sharing similar interactions with their soil microbes, which turns out to be a key predictor of plant abundance in these old-field communities. This was particularly clear for the net whole-soil and conspecific AMF feedback responses, but not so much for the soil treatment with the generally abundant AMF Glomus etunicatum. Given the diversity and complexity of soil communities, it makes sense that one particular member might not have a clear or dominant influence on the direction of PSFs across an entire community of plant species. However, in terms of predicting abundance in the field, all soil biotic treatments (net whole-soil, conspecific AMF, and G. etunicatum) were retained as critical factors in the best models, indicating that the whole suite of belowground interactions likely has important consequences for structuring plant abundance.

That the authors have examined PSFs in a phylogenetic framework is what makes this study exciting for moving these ideas forward. Their results provide a mechanism for explaining plant abundance in the field: as a plant, you and your relatives interact the same way you do with soil communities. In human terms, an analogy would be that you and your siblings, cousins, parents, and Uncle Eddy have similar careers, shop at similar stores, and eat at the same restaurants (i.e., construct and interact with a similar environment), all of which shapes how big your family is in relation to everyone else in town.

This begs the question of why you are so similar to your family. Is it simply because you are related? To an extent, it might be expected that you are more similar to your family given the nature of evolution (i.e., inheritance and descent with modification). Or is it because you are somehow constrained (e.g., due to lack of variation or strong trade-offs that lead to the extinction of anyone who is different) to keep what you’ve inherited? Constraints are the essence of phylogenetic conservatism, and there are several evolutionary processes that can lead to this pattern (Crisp & Cook 2012). Yet, in beginning to answer this question, it is important to clarify what is meant by the magical term “phylogenetic conservatism”.

Some evolutionary and ecological causes that can strengthen (or weaken) a pattern of phylogenetic niche conservatism (from Crisp & Cook 2012)

The most thorough explanation that we’ve come across has been Losos (2008), a perspectives piece motivated by the widespread misinterpretation of (and lack of clear distinction between) the terms phylogenetic signal and phylogenetic conservatism. Essentially, phylogenetic signal is a pattern “in which ecological similarity between species is related to phylogenetic relatedness”. Given the phylogenetic methods used in this particular study, a ‘significant’ Blomberg’s K value would indicate phylogenetic signal. Phylogenetic conservatism, however, is a pattern in which the “similarity of closely related species is significantly greater than would be expected based on phylogenetic relatedness” (Losos’s emphasis). In terms of K values, only those values that are greater than one indicate phylogenetic conservatism. All reported K values in the Anacker et al. study are significant, but less than one, which is indicative of phylogenetic signal but not phylogenetic conservatism. In other words, you and your family may all be doctors, but little Johnny is not constrained to this and instead pursues a related career in dentistry. The point here is that just because plants are similar doesn’t mean that they are constrained to be similar – it is simply because they have inherited similar traits from their recent common ancestors.

Because the paper could not offer more definitive conclusions on phylogenetic niche conservatism, we wish the authors had speculated or explored more about how conservation of traits might be related to feedback responses among groups. We know that plant traits determine PSF by shaping interactions with soil communities and controlling nutrient processes, and we thought this topic deserved more attention given the implications of this work. For example, are there traits that are more important for predicting feedback responses when incorporating species’ evolutionary histories (i.e., is there a phylogenetically conserved suite of traits that underlie the observed phylogenetic effects on PSF)? Is it possible for PSFs to promote the conservation of certain traits, and if so how might you test that? Are the traits always organism-level (e.g., r- or k-selected growth strategies), or is it possible to use more specific shoot/root traits with phylogenetic inference to predict PSFs and their consequences well? Is this even a reasonable approach to take, or should we be interpreting phylogenetic signal differently when thinking about PSF given the cyclical nature of the relationship?

We also appreciated the section on potential limitations of the study, because their conclusions did encompass multiple independent experiments that were not quite replicated in the same manner or for the same amounts of time. This is especially important because we know PSF studies can be finicky based on experimental designs (Brinkman et al. 2010, van der Voorde et al. 2012) – one obvious example being their drastically different results between net whole-soil and conspecific AMF feedback patterns. However, one concern they did not address is the potential over interpretation of species-level effects when the comprising studies largely did not account for within-species variation. There are exceptionally few studies (let alone PSF experiments) that span individual to phylogenetic levels, but this is still worth considering given that the majority of experiments finding interspecific differences in PSF typically do not appropriately considered intraspecific variation.

The future of community and ecosystems ecology is with an evolutionary perspective, and overall we recommend this paper since it provides an important step in that direction. We wonder how widespread these trends of phylogenetic signal on feedback responses are across more than just old-field communities, and if we can begin to identify broader patterns of important traits or selective pressures that are related to niche conservatism in PSFs. There’s a whole other barrel of fun with their native/exotic comparison, which we won’t go into explicitly but would love to hear your thoughts on as well.

References

Anacker BL et al. (2014) Phylogenetic conservatism in plant-soil feedback and its implications for plant abundance. Ecology Letters, doi: 10.1111/ele.12378.

Brinkman EP et al. (2010) Plant-soil feedback: experimental approaches, statistical analyses and ecological interpretations. Journal of Ecology, doi: 10.1111/j.1365-2745.2010.01695.x.

Crisp MD & Cook, LG (2012) Phylogenetic niche conservatism: what are the underlying evolutionary and ecological causes?. New Phytologist, 196, 681-694.

Klironomos JN (2002) Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature, 417, 67-70.

Klironomos JN (2003) Variation in plant response to native and exotic arbuscular mycorrhizal fungi. Ecology, 84, 2292-2301.

Losos, JB (2008) Phylogenetic niche conservatism, phylogenetic signal and the relationship between phylogenetic relatedness and ecological similarity among species. Ecology letters, 11, 995-1003.

Post DM & Palkovacs, EP (2009) Eco-evolutionary feedbacks in community and ecosystem ecology: interactions between the ecological theatre and the evolutionary play. Philosophical Transactions of the Royal Society B, 364, 1629-1640.

van de Voorde TF et al. (2012) Soil inoculation method determines the strength of plant–soil interactions. Soil Biology and Biochemistry, 55, 1-6.

PSE journal club: Vegetation exerts a greater control on litter decomposition than climate warming in peatlands

A bright, blustery day in England’s north Pennines. Fluffy cotton grass heads bob and bounce on the breeze. On the side of a hill, a strange array of hexagonal, knee-high structures glint and sparkle. Four figures, hunched against the wind, move methodically along jaunty wooden boardwalks, which rest on the blanket bog, crouching at each hexagon in turn. Welcome to Moor House National Nature Reserve, the site of an experiment designed to investigate how a warmer climate will affect the speed with which plant litter is recycled back into the soil, and ultimately the atmosphere.

One of the inherent difficulties associated with upland experiments.

If you’re wondering about the shift in tone in the opening paragraph of this #psejclub post compared to some of the others, it’s because I was there. I helped to set up and sample the aforementioned experiment, which is located a few hundred metres away from where I did my PhD fieldwork, in exchange for help with my own work, so I have to admit to having a degree of personal bias! I do think that the work is of general interest, though, and conveys some important findings about litter decomposition in peatlands that will help us to build a picture of how these key processes might change as plant communities shift in response to climate change.

The paper, currently a preprint in ESA Ecology, describes an experiment that uses a combination of open-top, passive warming chambers and plant removal treatments to investigate how the presence of certain plant functional types and warmer temperatures affect rates of litter decomposition. The authors used litter bags, filled with the litter of each plant functional type and buried in plots beneath the plant removal and warming treatments. In all, there were eight treatments (combinations of graminoid, shrub and bryophyte removal, a bare plot and a control with no plants removed), replicated over four blocks, with half the plots warmed by the passive chambers.

So how do peatland processes respond to warming?

The main finding of the paper is that presence or absence of plant functional groups had a stronger effect on peatland litter decomposition than the warming of approximately 1°C achieved by the passive chambers. Removing the shrubs from the peatland resulted in faster decomposition of graminoid and bryophyte litter, after two years. Litter identity was also important – in the first year of the experiment this was the main factor controlling rates of litter decomposition, with bryophyte litter decomposing most slowly, followed by shrubs and graminoids. After two years, the live plants present in the plot (i.e. presence of shrubs) were more important than the litter identity. Warming affected the composition of the bacterial community, while the fungi responded more strongly to the presence of shrubs.

While these results are compelling, they should be taken with caution, as the authors suggest: since the duration of the experiment was two years, further interesting effects resulting from the decomposition of shrub and bryophyte litter, which happens more slowly than in graminoids, might not have been captured. After four or five years, the decomposition of more recalcitrant litter could reveal more interesting effects. The same is true for the warming treatment: given a longer study period, the effect of 1°C warming on plant litter decomposition might become more important. It is, however, easy to write ‘more long-term experiments needed!’ while, in reality, the amount of effort required to maintain the plant removal treatments and warming chambers in the harsh upland environment of the Pennines represents a considerable hurdle. And one has to motivate one’s volunteers to see past the inherent absurdity of weeding a moorland!

Overall then, what’s the message? Dead or alive, whether you’re graminoid, shrub or bryophyte seems to exert much stronger control on litter decomposition rates in peatlands than temperature. While warming doesn’t have much of an effect on plant decomposition, it does affect the bacterial community in the peat, which might have important implications for graminoid decomposition, since bacteria are well-equipped to munch through labile substrates. Fungi respond when you take away the shrubs, which provide the more recalcitrant litter they’re specialised for dealing with, and might therefore moderate the response of the peatland carbon cycle to warming. Given a longer time period, more effects may emerge from this experiment.

I’m interested to know about the responses of decomposition processes in other ecosystems to warming and plant functional group removal. In grasslands, for instance, what happens when you increase the ambient temperature and remove the nitrogen fixers? Of course, these sorts of studies require a sufficiently long period to become stable and start producing results. In a time of apparently perpetually shrinking budgets, what’s next for long-term field experiments in ecology and biogeochemistry? What are the current barriers to their deployment, or do we in fact have enough to answer our more pertinent questions? Let me know what you think – get involved by commenting on this post, and posting on Twitter and Facebook using the #psejclub tag. I’m looking forward to hearing from you!

Soil biodiversity and soil community composition determine ecosystem multifunctionality

This paper, by Cameron Wagg et al., which was published online early in PNAS last month, describes the results of a very interesting experiment in which the authors manipulated soil biodiversity and measured the effect of these manipulations on a range of ecosystem functions.

More specifically, they created a gradient of reduced soil biodiversity (including a range of faunal and microbial groups) by sieving the soil through a number of decreasing mesh sizes, adding the fraction that passed through the sieve to sterilized soil, while also adding the sterilized fraction that remained on top of the sieve. They then grew plant communities consisting of common grasslands species in the soil for 14 and for 24 weeks, in two separate experiments. At the end of the first experiment, and after 12 and 24 weeks of the second experiment, they measured plant diversity and productivity, carbon sequestration, litter decomposition, nitrogen turnover, N₂O emission, phosphorus and nitrogen leaching as ecosystem functions, and fungal and bacterial diversity (by TRFLP), mycorrhizal root colonization (microscopically), and nematode abundance (microscopically).

They then used these data to relate the ecosystem functions measured to the soil biodiversity treatments. In addition, they calculated z-scores for the range of ecosystem functions measured as well as for all groups of organisms quantified, and regressed these against each other to answer the question whether ecosystem multifunctionality is related to soil biodiversity. This approach, of summarising a number of ecosystem processes into one ecosystem multifunctionality index, has been used previously by Maestre et al. (2012).

Their findings are very interesting and will make a lot of soil ecologists very happy: they find that a number of the individual ecosystem functions are reduced with declining biodiversity, but also that ecosystem multifunctionality is positively correlated with overall soil biodiversity.

When taking a closer look at the data, it becomes clear that the reduction in soil biodiversity varies between groups and isn’t linear with the decreasing mesh sizes – mycorrhiza and nematodes drop down sharply after the third ‘dilution’, whereas the other parameters show a more gradual decline. The authors have taken this into account by not only relating ecosystem functioning to the diversity treatments, but also to the abundance and diversity of individual groups. When taking a closer look at this, it becomes clear that the microbial properties measured have a far stronger effect than nematode abundance. In addition, the effect of reduced soil biodiversity on a range of functions is indirect, through effects of plant productivity and diversity.

Of course, it is very easy to criticise aspects of this study. You can question whether bacterial and fungal diversity, microbial biomass, mycorrhizal colonization, and nematode abundance together are a realistic representation of soil biodiversity. For example, why was nematode diversity not assessed? And why not higher trophic levels, such as Collembola and mites? Microbes and nematodes are only a fraction of the soil food web (Fig. 1). With the current analyses, the title ‘Soil microbial diversity and community composition determine ecosystem multifunctionality’ might have been more appropriate.

A (simplified) example of a soil food web, with the groups measured by Wagg et al. (2014) indicated by the dashed line.

Also, it would have been interesting to see root biomass in addition to mycorrhizal colonisation – a number of recent papers point to the importance of roots for ecosystem functioning (e.g. Orwin et al. 2010, Grigulis et al. 2013)

A more technical comment relates to the measurement of nitrogen turnover – this was assessed by measuring the uptake of ¹⁵N from Lolium multiflorum litter into aboveground L. multiflorum biomass. So, this measurement might be a proxy for L. multiflorum biomass, which decreases with decreasing soil biodiversity, rather than for nitrogen turnover.

On another note, and I would be very interested in other people’s opinion, I am wondering about the value of using an index for ecosystem multifunctionality. True, this averages across ecosystem functions and can therefore inform management to optimize overall ecosystem functioning. However, are the ecosystems that have the greatest average functioning really the most sustainable, and thus, desirable ecosystems? Are all ecosystem functions equally important? There might be trade-offs between different ecosystem functions – for example between crop yield and nitrogen retention, or between decomposition and carbon sequestration. We might want to optimize a certain function in a certain area, of which we already know that it has potential in delivering a certain function, rather than promoting multifunctionality across the board. For example, peatlands store large amounts of carbon because of their low decomposition rates, and agricultural production systems have high yields but low carbon sequestration.

However, in this paper, the multifunctionality index serves the purpose of summarizing overall ecosystem functioning, which shows a strong and positive relationship with soil biodiversity. Done like this, it summarizes a range of measurements that non-specialists might struggle to interpret – thus, it simplifies and reinforces the message of the paper that soil biodiversity determines ecosystem functioning.

Experiments like this require an enormous amount of work, and you simply can’t include everything. It is incredibly difficult to modify soil biodiversity without simultaneously changing soil properties, and the authors of this paper have achieved this by used an elegant method of reducing soil biodiversity. Thus, in contrast to many earlier studies, they were truly able to mechanistically elucidate the role of groups of soil organisms in ecosystem functioning.

This paper adds to the growing body of literature that soil biodiversity plays a crucial role in ecosystem functioning, and highlights the importance of conserving, and promoting, soil biodiversity. That’s what I like to hear!

Interactions with soil biota shift from negative to positive when a tree species is moved outside its native range

Pinus contorta. Image by Rudi Riet, Washington, DC, United States [CC-BY-SA-2.0], via Wikimedia Commons

This week’s paper is a Rapid Report by Gundale et al which appeared in New Phytologist in late January.  It also happens to be the paper we discussed at the Hawkesbury Institute for the Environment’s student journal club this week.

This study was a really good example of a home-away growth experiment, and showed some really striking results.  Seeds from Pinus contorta trees of known provenance were collected in Sweden, and then grown in soil from their native Canadian sites and their introduced

Swedish sites to look for differences in aboveground and belowground biomass.  Sounds simple?  It would be, but they didn’t just compare fresh soils containing whatever biota could contend with the 2mm sieve.  They also used sterilised soils from both countries, sterilised soils which had been cross-inoculated with soils from the opposite country, and included a fertilisation element to make it more interesting.

So what was the take-home message?  The clue is in the title, but the short version is that it’s all about the soil biota.  Seedlings grown in Swedish soils had much higher biomass (around 43% higher) than those grown in native their native Canadian soils, despite the Swedish soils having lower pH and nutrient availability.  When seedlings were grown in sterilised soils, the soil origin had no effect on biomass.  When seedlings were grown in sterilised soil from either country, but inoculated with soil biota from Sweden, biomass was again much greater than when sterilised soils were inoculated with Canadian soil biota.  The effect of soil biota was even greater than the effect of fertiliser on biomass.

Considering these different trials together, it becomes clear that not only did the Swedish biota have a strong positive effect on P. contorta seedling biomass, but that the Canadian biota actually had a negative effect on biomass. This provides evidence that better growth in soils from outside of the native range is probably down to a combination of pathogen release and positive biotic associations.

However, this evidence also highlights the gap in this paper – analysis of the soil biota itself.  We see the effects, but what are the actual differences in the bacterial, fungal, and/or mesofaunal communities?  Are there differences in community structure? Abundances or biomass? Activity?  All of these?  It would be great to see data on this, especially regarding known pathogens or ectomycorrhizal fungi associates.  I hope/suspect there may be another paper on the way exploring some this.

Overall, this was a really nicely designed project that asked interesting questions and addressed them in a very straightforward way.  The paper itself was well written.  It flows nicely and is easy to follow and understand even on a quick reading.  The methods made no attempt to disguise the logistical issues associated with transporting soils half way around the world, and the data analysis and presentation was simple and direct with no unnecessary frills or risk of misinterpretation.  All of these are elements that we should expect of any paper, but sometimes experiments are complex and difficult to describe, methodological detail gets glossed over in an attempt to meet word counts, and data are not easy to interpret.  While it is not possible to answer all questions with a glasshouse study, or to present all data with bar graphs, it is a nice reminder of the clarity with which we should try to communicate our work.

So now that I’ve had my say, exposing my soil ecology and science communication biases, we’d love to hear what you thought about the paper.  Was there anything you would have done differently?  What do you think the next steps should be?  What did you think of the tree provenance element of the study?  It didn’t show a significant effect here, but would it be worth pursuing in other systems?  How well do you think the results of this glasshouse study represent what’s happening in the natural communities?

You can let us know your take on this either in the comments or via twitter using the hashtag #psejclub.  We’d also love to hear from you If you have a suggestion for a paper you’d like us to discuss, or if you want to write a post yourself!

Pinus contorta native range. Image by Elbert L. Little, Jr., of the U.S. Department of Agriculture, Forest Service, USGS

P. contorta range in Sweden. Image by Philippe Rekacewicz, UNEP/GRID-Arendal

#psejclub discusses: plants, fungi, competition and carbon storage

It’s nearly two weeks since Franciska wrote the first post for our journal club (#psejclub), about a paper published in Nature, entitled ‘Mycorrhiza-mediated competition between plants and decomposers drives soil carbon storage‘. Her thoughts on the paper prompted an interesting discussion, which you can catch up on here.

Keep your eyes peeled for a new #psejclub post in the next couple of days!