This letter, by Yann Hautier and many others, appeared in Nature ten days ago. Its focus is grasslands, and what happens to their [above-ground] diversity when you add fertiliser. I was drawn to this paper by the simple message communicated in the title, and the universality implicit in ‘natural grasslands’. Excellent; a succinct study of the effects of excessive nutrients on the modulation ecosystem functioning by diversity – read on!
It’s immediately apparent that the strength of this work lies in the size and global reach of its dataset. The authors utilised an established network of experiments looking at the relationships between fertilization, diversity and production in grasslands: the Nutrient Network (NutNet). This allowed the authors to address their hypotheses using data collected from 41 different grasslands, spanning five continents. One of the major advantages of using sites in an established network is that methods are broadly standardised; this goes some way towards ensuring that, while the sites encompass a wide range of variation in grassland types, the data are comparable between sites. In the extended methods section, the authors describe the sensitivity analyses they performed to check that distinctive sites (subject to strong seasonality, or anthropogenic influence, for example) did not unduly influence their results.
So what did the authors measure? They chose above-ground net primary productivity (ANPP) as their response variable, which they estimated by measuring above-ground live biomass at each site. The authors also introduced two related concepts: the stability of ANPP and species asynchrony; stability of ANPP increases as species asynchrony increases and the productivity of individual species fluctuates in response to the environment at different rates. This is a neat idea: in an asynchronous community, a decline in the productivity of one species is more liekly to be compensated for by another, and so the productivity of the community as a whole is more likely to remain stable.
To arrive at their finding that the application of fertiliser weakens the stabilising effect of diversity, the authors observed that, while the temporal variation of ANPP decreased at higher diversities in the unfertilised communities, it increased at higher diversities in the fertilised communities. The increased variation in ANPP, combined with lower species asynchrony in fertilised communities, led the authors to conclude that [over]-fertilisation of grasslands can reduce the stability of their productivity. Importantly, the authors showed that the loss of diversity caused by fertilisation didn’t affect species asynchrony.
While I initially found this article difficult to get to grips with, due to the introduced concepts and number of scatterplots and lines to compare, but now that I get it, I agree with the results. One of the best aspects of this study is that it draws its conclusions from a controlled experiment, rather than observational correlations. I think the work also raises a number of further questions, which would be interesting to address using NutNet or a similar project.
- The authors used species richness as their measure of diversity. What about functional diversity? Perhaps some groups (grasses, forbs, legumes) respond to fertilisation more than others?
- While ANPP is certainly likely to be related to below-ground productivity, it would be interesting to see whether stability of below-ground productivity showed the same patterns in response to fertilisation. Are there other ecosystem functions that are worth measuring? What about carbon and nitrogen cycling? These functions obviously come with the caveat that they are much more difficult to measure (you can’t just see them, unlike plants), but if it was feasible, what patterns might we expect to see?
- The amount of fertiliser applied in the study was quite high. Do we see the same results at lower concentrations?
- There were no sites in South America – do we think this is likely to have biased the result significantly?
Now that I’ve had my say, it would be great to hear yours; did you find this paper interesting, do you think its conclusions are valid? How might you have done things differently? Do you think there are specific lessons to be learnt? Please get in touch by commenting on this post, or on Twitter or Facebook using the hashtag #psejclub. I’m looking forward to hearing from you!