The Peel–Harvey estuarine system in Western Australia's south-west is affected by poor water quality, algal blooms, and fish kills. Phosphorus (P) discharge from agricultural activities is the main source of poor water quality. The catchment’s soils are naturally infertile, but P application has increased P fertility. This paper draws on and undertakes a meta-analysis of 20 200 surface (0–10 cm) and profile (to 100 cm depth) soil samples collected in the period 1983–2018. Soil P content was high, with 70% of samples with Colwell P content in excess of agronomic requirements; Production is more likely limited by low soil pH(CaCl2) and low K (92% and 67% of paddocks respectively). Strong P stratification in the soil is evident, particularly topsoil; sandy soils are saturated to depth; and clay soils show signs of P saturation in the topsoil. Management of P in sandy soil near the estuary is a high priority as is P stratification in highly P retentive soil. Soil P stocks increased since clearing compared with uncleared soils (1221 kg ha–1 m–1 and 285–694 kg ha–1 m–1, respectively). Thirteen percent of samples had P content in excess of agronomic requirements in 1983, rising slowly to 69% in 2018. Landholder practices need to be analysed in detail to confirm if this accumulation occurs everywhere or is only confined to actively farmed land.

Context: Leaching of phosphorus (P) from water soluble agricultural fertilisers applied to sandy soil can adversely impact downstream water quality. Less soluble fertilisers may reduce P leaching and increase production. However, repeated application of low water soluble P (LWSP) fertiliser with high citrate soluble P (CSP) has potential to increase P leaching and P accumulation. Methods: We examined the effect of LWSP fertilisers (single superphosphate, SSP; fertilisers low in water soluble P, WSP) on leaching losses in soil with low P retention/low P status; with/without bauxite residue amendment and with/without plants. Key results: On low P retentive soils with a low P status, P leaching was reduced by reducing the WSP of fertiliser. Plants and soil amendment reduced P leaching further, but P loss remained proportional to WSP applied. Under field conditions, a subset of LWSP fertilisers greatly increased pasture dry matter production and increased soil test P values. Soil test P increases were positively correlated to fertiliser citrate soluble P content. Soils already containing at least 80% of the required plant available P did not require a WSP fertiliser to provide early season P. Conclusions: LWSP fertilisers and soil amendment have potential to reduce P leaching in low P retentive soils. Fertiliser formulations targeting reduced WSP and high CSP require attention. Implications: LWSP fertilisers may increase soil P residues reducing future P needs; however, ongoing soil testing is required for assessment of soil P accumulation, requirements and leaching potential.

Context: Phosphorus (P) use in agriculture can lead to eutrophication. Agronomic soil tests such as Colwell P and P buffering index (PBI) define critical soil P levels for pasture production. These tests have potential for re-use as environmental risk indicators of dissolved reactive P (DRP) loss from paddocks but are constrained because a 0–10 cm sample does not necessarily align with the dominant hydrological loss pathways of runoff or leaching. Aims: To identify influences on the benchmark environmental measure of DRP (CaCl2-extractable P or CaCl2-P) by agronomic-based measures such as PBI, Colwell P and depth, and Colwell P to PBI ratio (P environmental risk index; PERI). To estimate CaCl2-P at any depth from a 0–10 cm sample, and the potential for change in DRP loss risk through the adoption of evidence-based fertiliser management based on soil testing. Methods: Archives of 692 0–10-cm soil samples, along with 88 sites sampled at 0–10 cm and 0–1, 1–2, 2–5, 5–10, 10–20, and 20–30 cm were analysed for Colwell P, PBI, CaCl2-P, PERI, and P fertility index (PFI). Derived relationships between CaCl2-P and Colwell P for different PBI were applied to 30 981 0–10-cm samples to estimate the potential for DRP reduction resulting from the adoption of evidence-based fertiliser management. Key results: CaCl2-P, Colwell P, PERI, and PFI decreased with depth, with an associated increase in DRP loss risk from surface soil. The CaCl2-P decreased with increasing PBI. The CaCl2-P, Colwell P, PERI, and PFI could be estimated at any depth from a 0–10 cm sample, with r2 > 0.77. The CaCl2-P was estimable from PERI, and soils with low PBI or with high PFI had high DRP loss risk. The CaCl2-P was positively correlated with Colwell P, with the slope decreasing with increasing PBI and becoming invariant when PBI > 100. When applied to the current soil Colwell P and estimated current CaCl2-P and compared to CaCl2-P at the critical Colwell P for different relative yields (RYs), DRP loss risk could be reduced by 24% for a RY target of 95%, and 59% for a RY target of 80%. Conclusions: Because current Colwell P levels in soils exceed critical values, DRP loss risk can be substantially reduced by adopting evidence-based fertiliser management with little or no loss of utilised pasture. Implications: Fertiliser management based on evidence of P requirements determined from soil testing has a significant role in reducing DRP loss risk.

Phosphatic fertilisers have made grazing in the south-west of Western Australia (WA) viable. However, there is evidence that a large proportion of pasture paddocks exceed soil test critical values at which 95% of maximum yield is achieved as identified in the national Better Fertiliser Decisions for Pasture (BFDP) project. Of 22 000 soil samples collected between 2009 and 2020, 56% exceeded the critical value for phosphorus (P), although there were constraints to potassium (K) and sulfur (S) and from soil acidity. Soils with available P exceeding the critical value are expected to lead to excessive losses of P to waterways, resulting in eutrophication. A trial program was established to validate the critical P values from BFDP so that concerns can be addressed about the relevance of these critical P values to WA conditions and to contemporary pasture varieties. Measured relative yields for 19 trials in the first year were mostly within 10% of that predicted from BFDP for soils with a P buffering index (PBI) >10. Soils with PBI <10 had measured relative yields up to 25% greater than predicted by BFDP, suggesting response calibrations for low PBI soils may require adjustment in the BFDP dataset. Some pasture yield gaps occurred when soil pH and P were low. Application of nitrogen (N), K and S almost doubled the yield when P was limiting or sufficient. Agronomic advice and practice should seek to optimise these multiple inputs, thereby optimising P use rather than applying P to levels above the critical value.

Examples of sediment budgets are needed to document the range of budget types and their controls. Sediment budgets for three small agricultural catchments (7.6 to 15.6 km²) in southwestern Australia are dominated by channel and gully erosion, with sheet and rill erosion playing a subordinate role. Erosion was increased by clearing naturally swampy valley floors and hillslopes for agriculture and grazing, and episodic intense rainstorms. The proportion of sediment from channel and gully erosion in the sediment budget appears to be determined by the depth of alluvial fills. Dryland salinization caused by clearing native vegetation has connected hillslopes to channels across narrow floodplains, increasing the Sediment Delivery Ratio (SDR). Yield and SDR are found to be insensitive to major in-catchment changes of vegetation cover after initial clearing, the ratio of sheet and rill erosion/channel and gully erosion, and sediment storage masses. This supports the idea that yield alone is often a poor indicator of the impact of land use and land management change. Riparian vegetation would reduce sediment yield but not phosphorus yield. This study demonstrates the value of mixed methods where field observations and chemical analysis are combined with information from local people.