/Here is a guest post written by Vincent Maire for his new paper in which Ülo Niinemets is one of the coauthors./
There is a longstanding debate among ecosystem ecologists to conclude if the soil has an independent impact on ecosystem productivity at global scale (Wang, Prentice and Davies 2014; Huston 2009, 2012). Some argue that climate and soil co-vary and there is no need to consider soil properties in dynamic global vegetation models to simulate the growth primary productivity, ‘only climate matters’ (while they recognize that soil may have an impact on NPP; Wang, Prentice and Davies 2014). Others reckon that soil age, which is influence both by climate and geology, is the primary control of net ecosystem productivity (NPP mainly, but this is not clear for GPP; Elser et al 2007; Huston 2009). Taking this into context, we investigated at global scale the independent and joint impacts of soil and climate on leaf photosynthetic rate, which is an important component of ecosystem primary productivity.
In my opinion, the paper published in Global Ecology & Biogeography (link to full text here) provides several interesting messages. Firstly, we clearly demonstrate that soil is crucial factor for understanding the variation of soil photosynthetic rate at global scale, while the climate does not play such an important role. Still we cannot conclude if this is true for ecosystem productivity but this is a first step. In addition, it is still unclear what the drivers of soil development are in our study, i.e. the ones independent of climate (microclimate, vegetation, parent rock, topography, time).
Second important message is that we provide additional statistical evidences that two independent strategies (nitrogen strategy vs. water strategy) can be used by species to process photosynthesis. Nitrogen and water can be substitutable resources for photosynthesis (Wright et al 2003), and species strategy can be delineated in function of the preferred resource to increase photosynthetic rate (Prentice et al 2014). Approaching leaf photosynthesis optimization as the relative cost between nitrogen and water acquisition and utilization opens interesting avenues to understand the impact of environment on leaf photosynthesis variation.
Thirdly, nitrogen or water strategy are independently promoted by soil pH and soil available phosphorus, respectively. Hence, we show that soil fertility is multidimensional, while each dimension can have potentially independent impact on vegetation. This contrasts with the unidimensional view of soil fertility, which proposes that all soil variables co-vary with soil age for instance. A significant future challenge to this work will be to better disentangle the co-varying physiological, ecological and evolutionary mechanisms that underpin trait-environment relationships that has been observed in that work.
While I could discuss more about the theoretical aspect of this work, I found interesting to provide additional details on the methods. We used an existing trait dataset that comprised leaf photosynthetic traits (leaf N and P content, SLA, stomatal conductance, leaf photosynthetic rate) for thousand species and from many sites around the world. The dataset has been collated for many years by Ian Wright and coauthors. For the sites where plant traits where measured, we attributed climate and soil data by using soil and climate worldwide dataset as well as bioclimatic models. Attributing climate to each plant trait site was an easier task compared with soil data. CRU and Worldclim website are easily searchable and applicable. Radiation data were derived from a recent bioclimatic model that has been implemented by Colin Prentice and his colleagues (Wang et al 2014; Gallego-Sala et al 2010).
Regarding soil data, it was more of a patchwork. When we started this work, two main worldwide soil databases were on the market with a recent version in 2012 (HWSD and ISRIC, the previous version was from 2002) published by a similar scientist network, including complementary and coherent data but at different resolution scales (10’ and 30’’). At this time, we contacted Niels Batjes, who was one of the important figures behind creating and developing soil worldwide databases. He happily joined the team and collaborated actively on soil methods and the development of the theoretical model of soil fertility. In 2013, Shangguan et al published a new soil worldwide dataset (GSDE) at high resolution (30’’) that includes all variables present in the precedent soil datasets and, in addition, phosphorus data (both total and available), which is quite a holy grail to understand and represent in biogeosciences (Selmants & Hart 2010; S Yang & Post 2011, Yang et al 2013). However, when we compared both GSDE dataset and HWSD, we surprisingly founded no match for all common variables in both datasets (similar result with ISRIC) among our sites. We kept using HWSD and ISRIC for several reasons. The expertise in such work has been built for longer time by soil scientists like Niels Batjes. Poor signal was detected between soil variables and leaf traits with GSDE. For few sites, we were able to compare actual values of soil variables (soil pH, total carbon content, sand, etc.) with the ones derived from soil datasets, and the match was clearly better with HWSD.
Finally, in 2014, Hengl et al published an updated and combined version of HWSD and ISRIC at 30’’ resolution scale, which is called ‘soilgrids’ and has been finally used in the paper. Soilgrids include few variables (soil pH, soil bulk density, soil texture, soil organic matter) but with a high level of confidence in value accuracy. However, soilgrids does not include soil phosphorus data, something that we were strongly interested in getting accurate values for our sites. But this is another story that we could probably discuss in future. To conclude this methodology discussion, I would like to say that soil worldwide datasets have improved considerably the last 10 years and this is an important development to follow if we really want to progress in the debate of the relative edaphic and climatic control on primary productivity at the global scale as well as the multidimensionality of soil fertility. To do so, I provide here an interesting website that compile many of the soil dataset over the globe: http://www.css.cornell.edu/faculty/dgr2/research/sgdb/sgdb.html
Hope you will enjoy the reading of the article. Feel free to contact me at email@example.com if you want to discuss further the paper or this methodology discussion.
Full citation: Maire, V., Wright, I. J., Prentice, I. C., Batjes, N. H., Bhaskar, R., van Bodegom, P. M., Cornwell, W. K., Ellsworth, D., Niinemets, Ü., Ordonez, A., Reich, P. B. and Santiago, L. S. (2015), Global effects of soil and climate on leaf photosynthetic traits and rates. Global Ecology and Biogeography, 24: 706–717