Our understanding of the chemistry of soil and fertilizer phosphorus (P) and its relationship to chemical extractability and plant availability, have changed substantially over the last 50 or so years.

Prior to the 1950s, the forms of inorganic P in soils were usually described imprecisely but it was generally believed that phosphate ions were adsorbed at soil surfaces. Work at TVA in the 1950s and 1960s, often involving the addition of concentrated phosphate solutions to soil minerals, gave rise to the thinking that discrete-phase iron, aluminum, and calcium phosphates controlled the solubility and availability to plants of soil inorganic P. This led to the development of fractionation schemes for characterizing the forms of inorganic P in soils. It was commonly held that P was irreversibly ‘fixed' in soils because of precipitation reactions and as a consequence, the efficiency of phosphate fertilizer was low---of the order of 20 per cent. Such thinking persisted for many years and was commonly used to justify continuing research on soil and fertilizer P, in addition to supporting the promotion and sales of fertilizer P.

Subsequent developments in our understanding of the reactions of inorganic phosphate ions at the surfaces of soil components, often involving the use of radio-isotopes, new instrumental techniques such as NMR, and also modeling, gradually led to the thinking that inorganic P was not precipitated or irreversibly adsorbed and that over time, the recovery of added P could be quite high. There was limited evidence to support this thinking, largely because the link between soil chemistry and agronomy was not being made.

Our recent and ongoing analysis of experimental data from those long-term field experiments, which permit an adequate evaluation of the efficiency of soil and fertilizer P, support the thinking that the recovery of added P over time can be high, even very high in some cases, thus supporting the concept of reversibility, given adequate time. This new paradigm has major implications for fertilizer use in the context of both agronomy and the environment.

Armed with this new approach, the challenge is to establish the research protocols and methodologies required to elaborate the new thinking and evaluate its implications. For example, it could be argued that our inability to adequately describe the reactions of phosphate ions at soil surfaces limits our understanding of soil phosphate equilibria, i.e., the relationship between phosphate on the solid phase and phosphate ions in solution---the major control on the plant availability of P in terrestrial and aquatic systems. This points to the need for a major, focused research effort in soil phosphate chemistry.

Soil phosphorus research must be revisited because of (i) issues relating to phosphate loss from soil to the aquatic environment and concerns about eutrophication and (ii) the limited global reserves of P and the consequent need to use P effectively. There is an urgent need to rethink our definition of efficiency in relation to P use both for fertilizer and also for manures. This will require us to engage in some lateral and innovative thinking to link soil science research with agronomic and environmental issues. We must respond to this challenge. Can we?