At a recent meeting (December 2014), we discussed the dependence of the relationship between soil P availability and plant response on the soil P test that is used.
Identifying appropriate methods for quantifying P, in addressing different research questions, remains a challenge. Comparing between soil P analysis methods, in addition to soil properties and experimental variables, limits the progress we can make towards understanding how different conditions affect P dynamics. Whilst a globally-applicable test to quantify each P fraction is unattainable, questioning our rationale for the measures we select is essential. Furthermore, it is useful to consider what an alternative measure would mean in terms of the results we acquire and how we interpret them.
We discussed the following paper:
Six L, Smolders E, Merckx R. 2013. The performance of DGT versus conventional soil phosphorus tests in tropical soils—maize and rice responses to P application. Plant and Soil, 366, 49–66.
We focused the conclusion:
For predicting yield response to applied P, an intensity measure (DGT) was most effective for maize, whereas conventional quantity measures (e.g. Olsen P) were most effective for rice.
This suggested that rice does not depend on diffusion of P in the soil (as measured by DGT). Compared to maize, rice has a greater ability to acquire P, via: secretion of organic acids to solubilise non-labile P, symbiotic associations with mycorrhizal fungi, and a more efficient root system for P uptake. Thus availability of soil P depends on the crop species.
In response, we considered the question:
- How to identify the most representative soil P tests for specific soil-plant systems, without laborious preliminary experimentation or requiring numerous (costly) analyses?
Our main points were:
- The standard test for plant-available P for UK soils is Olsen P. How did this arise? From a comprehensive review of the available tests and selection of the most replicable and appropriate measure? Assessed for what purpose and soil-plant system?
- Six et al. (2013) raised three criteria for soil P tests. If these are satisfied by numerous different tests, does it matter which we select? This paper and our discussion suggested it does, in which case, what other criteria are we applying to our judgement?
- How to rectify the perhaps contradictory aims for greater harmonisation of methods (at laboratory group/ university/ regional/ global scales) and for soil-plant system specific measures?