Improving the sustainability of phosphorus use in arable farming – ‘Targeted P’


Cereals & Oilseeds
Project code:
01 August 2010 - 31 July 2015
AHDB Cereals & Oilseeds.
AHDB sector cost:
Total project value:
Project leader:
Roger Sylvester-Bradley1 , Paul Withers2 , Alison Rollett3 , Pete Talboys2 , Robin Walker4 , Tony Edwards4 , Sevil Payvandi5 , James Heppell5 , Tiina Roose5 & Davey Jones2 . 1 ADAS Boxworth, Battlegate Road, Boxworth, Cambridgeshire, CB23 4NN 2 Bangor University, College Road, Bangor LL57 2DG 3 ADAS Gleadthorpe, Meden Vale, Mansfield, NG20 9PD 4 SRUC Aberdeen Campus, Craibstone Estate, Aberdeen AB21 9YA 5 University of Southampton, Highfield, Southampton SO17 1BJ


pr569_final-project-report_summary pr569_final_project_report

About this project


A collaborative project conducted from 2010 to 2015 involved 17 partners from government, industry and academia. It explored technologies to support a more efficient and sustainable strategy for phosphorus (P) use on UK arable land, dubbed ‘feed the crop, not the soil’.  Through (i) a review, (ii) experiments on crop performance and P run-off, both in pots and the field, and (iii) modelling, the project explored how the farming industry might make the transition to this new approach.

Fifty years ago the current strategy for P use (as set out in the Fertiliser Manual, RB209; Defra, 2010) replaced an even less efficient one of relying entirely on fresh P fertilisers to support crop production.  Soil tests for available P (STP) were then introduced so that fertiliser applications could be adjusted according to the estimated P supplying capacity of each soil.  However, this ‘soil P storage’ approach is now becoming increasingly costly, amid concerns over finite and affordable global phosphate supplies.  Also, the now-large reserves of P in most UK soils have become a significant cause of inland and coastal eutrophication, so threatening the achievement of good ecological status under the EU Water Framework Directive.

Two field studies, and a meta-analysis of published data, showed direct positive relationships between soil P (by Olsen’s method) and dissolved P in surface runoff and drain-flow; these indicated that reducing STP from 25 to 10 mg kg‑1 could help substantially in achieving the challenging target concentrations set for P in UK water bodies under the Water Framework Directive.  Along with environmental benefits, large economic savings could also be made if the 80% of UK arable land now at P Index 2 (or more) were run down to P Index 1.  However, reducing soil P levels will be incompatible with enhancing crop yield levels until (i) fertiliser P efficiencies can be markedly improved, and (ii) crop P sufficiency can be monitored more certainly than just with STP.

Ten field experiments on English or Scottish sites with low STP showed crop capture of soil-derived P to average 80% of estimated crop requirements (range: 42% to >100%) with no fertiliser applied.  However, crop responses to fresh triple superphosphate (TSP) broadcast and incorporated in the seedbed were generally very small, as judged by the difference method, with extra P in the initial crop averaging only 4% of the P applied.  The best way of enhancing initial recovery of fertiliser P was to use struvite (with slower P release than TSP) and place it close to the seed, but best initial recoveries of P were still less than 10%.

Experiments in pots showed the potential of P seed dressings and/or foliar sprays to supply P directly to crop plants, as well as the potential of struvite granules to act as a slow-release P fertiliser.  However, two further field experiments at Ropsley (Lincs.), where soils ranged from P Index 0, failed to detect any positive responses to small amounts of seed-dressed P or foliar P, even though crops responded by ~2 t ha‑1 grain as soil P increased to Index 2.  The exact fates and mechanisms of seed-dressed and leaf-applied P now need to be investigated in the field.

A new mathematical model, devised to simulate P transfers between fertiliser, soil, root and shoot, and crop yield effects, demonstrated high sensitivity of outcomes to the P buffering capacity (= fixation capacity) of the soil.  However, disappointingly, AVAIL®-treatment of TSP to reduce soil fixation of added P, showed no consistent benefits either in pots or in the field.

It is concluded that the current strategy for P use (of ‘feeding the soil to feed the crop’) should be maintained for the time being, but that strong environmental imperatives should drive further intensive research into ‘feeding the crop, not the soil’, and a transition to this new strategy should be planned.  Four ‘P run-down sites’ were established for this purpose.  STP monitoring of these sites over 5-6 seasons demonstrated scope for arable farms to exploit soil-stored P whilst testing for more efficient P nutrition approaches, but uncertainties in determination and interpretation of STP were also shown.  It is recommended that industry initiatives to improve P fertilising efficiencies should be monitored through careful routine use of STP, augmented by additional routine analysis of crop P.  Critical P contents of harvested biomass (e.g. grain) are proposed to support researchers and farmers in their quest for new practices and products that will improve the sustainability of P use in future arable farming.