Historical application of phosphorus (P) fertilisers has enabled grazing agriculture in south-west Western Australia but has led to excessive soil P levels that pose a eutrophication risk. However, adoption of critical soil-test P values for pastures determined from the Better Fertiliser Decisions for Pastures (BFDP) meta-analysis of historical Australian trials has been poor because of perceived lack of relevance to the soils and contemporary pasture species in the region.

We aimed to validate critical Colwell P values from the BFDP project for soils and contemporary pasture species of south-west Western Australia.

Fifty P-rate response trials were performed with contemporary pasture species, and the results compared with BFDP.

Trial results were consistent with BFDP and fell within 95% prediction intervals of response calibrations of Colwell P and relative yield pairs for different P buffering index (PBI) ranges. Soils with PBI <15 required lower critical Colwell P values. Responsiveness to P application was 8% (±4%), consistent with a 5% response expected for Colwell P levels supporting 95% relative yield. Application of nitrogen, potassium and sulfur significantly increased yield while increasing basal cover of annual ryegrass (Lolium multiflorum) and reducing that of clover (Trifolium spp.) in the sward. Response calibration coefficients and critical Colwell P values before and after addition of the trial data to BFDP showed a strong correlation (r² >0.99), although critical values were slightly lower for lower soil PBI.

Critical Colwell P values from BFDP are relevant and applicable to soils and contemporary pasture species of south-west Western Australia.

Agronomic advice and application of P should be based on interpretation of Colwell P values, with P applied only when levels are below the critical value. Pasture composition should also be considered when interpreting Colwell P values. This approach minimises risks of P loss and improves economic outcomes for growers.

历史上磷（P）肥料的施用使西澳大利亚西南部的放牧农业得以实现，但导致土壤磷含量过高，造成富营养化风险。然而，由于人们认为与该地区的土壤和当代牧场物种缺乏相关性，根据澳大利亚历史试验的更好牧场施肥决策 (BFDP) 荟萃分析确定的牧场关键土壤测试 P 值的采用效果不佳。

我们的目的是验证 BFDP 项目中西澳大利亚西南部土壤和当代牧场物种的关键 Colwell P 值。

对当代牧草品种进行了 50 次 P 率响应试验，并将结果与​​ BFDP 进行了比较。

试验结果与 BFDP 一致，并且落在 Colwell P 响应校准和不同 P 缓冲指数 (PBI) 范围的相对产量对的 95% 预测区间内。PBI <15 的土壤需要较低的临界 Colwell P 值。对磷施用的响应率为 8% (±4%)，与支持 95% 相对产量的 Colwell P 水平预期的 5% 响应一致。施用氮、钾和硫显着提高了产量，同时增加了一年生黑麦草（Lolium multiflorum）的基部覆盖度并减少了草地中三叶草（Trifolium spp.）的基部覆盖度。将试验数据添加到 BFDP 之前和之后的响应校准系数和临界 Colwell P 值显示出很强的相关性 ( r ²  >0.99)，尽管较低土壤 PBI 的临界值略低。

BFDP 的临界 Colwell P 值与西澳大利亚西南部的土壤和当代牧场物种相关并适用。

P 的农艺建议和应用应基于 Colwell P 值的解释，仅当水平低于临界值时才应用 P。在解释 Colwell P 值时还应考虑牧场组成。这种方法最大限度地降低了磷损失的风险，并改善了种植者的经济成果。