Cover crops and soil loosening are key components for managing P and C stocks in agricultural soils

Agricultural soils have accumulated phosphorus (P) and lost carbon (C) in the recent decades. Simultaneously, soil structure has been degraded by the large increase in machinery weight. High wheel loads compact the subsoil, resulting in reduced root growth, decreased yields, and hence decreased C inputs and P removals. To reduce the accumulated excess P stock, the P balance should be kept strongly negative for decades, which requires high biomass production. P accumulation in poorly yielding field parts has created local hotspots, often characterized by soil compaction and poor drainage. They may be only 1%–6% of the agricultural landscape area but present a high risk for waterbodies. Identifying and targeting these hotspots for P removal can be a key strategy for mitigating P emissions. Achieving the P removal is most likely to require soil loosening to repair the damage caused by compaction and use of cover crops to mine out the accumulated P. This soil improvement strategy can also be beneficial for C sequestration and crop productivity. This commentary highlights the current status and recent developments in the interactions between compaction, cover crops and P loss to provide a research basis for developing landscape‐level strategies.

the chemistry of C and P, but the management of both elements using two agricultural practices: cover crops and removing soil compaction by subsoiling.Both have been initially adopted for soil erosion control and improving soil quality, but they also present clear benefits for soil C sequestration and P emission reduction (Blanco-Canqui et al., 2015).
The article first outlines the changes in agricultural soils from 1960 to 2020 and the current status of European soils with regard to three indicators: C balance, P balance and soil compaction.The focus is on European soils, but the findings can be generalized to areas of China and United States with mechanized farming and high fertilizer application rates.Then, the focus shifts to within-field variation and the P loss hotspots created by decades of different P and C balances.Finally, the role of cover crops and removing compaction with subsoiling is presented in terms of their potential to reduce P emissions while building soil C.

LOSS OF C IN INCREASINGLY COMPACTED EUROPEAN FARMLAND SOILS
The last decades of the 20th century saw, among others, three major changes in European farmland soils: accumulation of P, loss of C and increased compaction from machinery (Figure 2).All three changes are because of shifts in the balances.The P balance is defined as the difference in the inputs and outputs of P to the soil.In a simplified form (based on Panagos et al., 2022): ΔP = P fertiliser + P manure − P yield − P loss .
Conversely, the C balance is defined as the difference in the inputs and outputs of C to the soil.In a simplified form (based on Chenu et al., 2019): ΔC = C abovegroundresidue + C belowgroundresidue + C manure − C decomposition − C erosion .The two balances are linked through the residue (related to yield), manure and erosion.Soil compaction is not described as a dynamic balance but is caused by soil stress from machinery exceeding soil strength.When the stress exceeds soil strength, the soil compresses (deforms) until it can match the stress, resulting in long-lasting deformation (Keller et al., 2019).This section reviews the development of these three balances over the last 60 years in Europe.
The use of mineral P fertilizers increased exponentially after World War II (Ashley et al., 2011).Europe used increasing amounts of P fertilizers from 1960 to 1990, and in the 1980s, European P fertilizer use was half of the world total usage (FAOSTAT, 2023).The collapse of the Soviet Union in 1990 decreased P usage, especially in Eastern Europe, but the balance remained positive.The P balance in the EU27 region was +25 kg P/ha/year in 1960-1990, +10 kg P/ha/year in 1990-2010 and + 5 kg/ha/ year in 2010-2020 (FAOSTAT, 2023).The use of mineral P has decreased, but the accumulated P stock still cycles in the agricultural system: half of the P input in 2019 was from manure (Panagos et al., 2022).The annual average loss to waterbodies is only 0.35 kg P/ha/year (Panagos et al., 2022), indicating that the soils have retained much of the excess P balance.The cumulative sum of the P F I G U R E 1 Soil C and P balances can be controlled by soil management that targets vegetative cover and soil structure through cover crops and subsoiling.et al., 2005;McGrath et al., 2023)) and (c) were compacted by increasingly heavier machinery (Keller & Or, 2022).balance has accumulated 400-1500 kg/ha of P (Figure 2a), which can be both a resource for food production (Noë et al., 2020) and a long-term threat for waterbodies (Schulte et al., 2010).The availability of the accumulated legacy P can be estimated using the P saturation concept (Nair et al., 2020).In this concept, each soil has a limited P storage capacity, which is determined by Fe and Al (and Ca in alkaline conditions), and the filling of the capacity (P saturation) can be calculated by Psat = P/(Fe + Al) (Nair et al., 2020).Using a common agricultural soil test (Mehlich-3), high P availability soils and sediments can be delineated with a threshold of 10% Psat (Dari et al., 2018;Mattila & Ezzati, 2022).This threshold can be used to estimate both the amount of P that could be mined from the soil without affecting crop yields and the amount of P that needs to be removed to decrease dissolved P emissions.The amount of legacy P exceeding the Psat threshold found in agricultural soil surveys matches the scale of accumulated P in the P balance (Figure 2a) 600-1000 kg P/ha (Dari et al., 2018;Mattila & Rajala, 2020;Saarela, 2002).However, the distribution of accumulated excess P in soils is not uniform; commonly, only a small fraction of the surveyed soils have high amounts of leachable P, highlighting the need for identification and site-specific management (Mattila & Ezzati, 2022;Mattila & Rajala, 2020;Nair et al., 2020).Removing legacy P from a high emission-risk site will require mining out excess P with plant roots and exporting it through yield.Maintaining a negative P balance (−7-30 kg P/ha/year) for 10-20 years may be necessary to reduce emissions from high P soil (Schulte et al., 2010).Just reducing the P fertilizer amounts is not adequate, as the stored P can cycle for decades, especially in a larger region (Noë et al., 2020).
Maintaining a strongly negative P balance requires high yields and good P use efficiency.Historically, grain yields doubled from 1960 to 1990, but the yield increases stagnated, increasing only 23% from 1990 to 2020 (FAOSTAT, 2023).The causes of the yield stagnation are a combination of climate, socio-economics and political decisions (Ray et al., 2012), but it also co-occurred with increasing widespread soil compaction (Keller et al., 2019).A single soil compaction event can reduce yields by 20%-30%, with the soil recovering gradually, but not reaching previous productivity (Etana & Håkansson, 1994).Repeated compaction events can result in a downward spiral in which the soil is not recovered from previous compaction when it is repeatedly recompacted (Keller et al., 2019).Annual compaction events became more common, as field machinery mass tripled from 1960 to 1990 and then again doubled from 1990 to 2020 (Keller & Or, 2022) (Figure 2c).This resulted in soil stress that exceeded soil strength deeper in the soil profile, causing subsoil compaction (Keller & Or, 2022).Accumulated compaction limits P removal from the fields by reducing both yields and crop root growth (Colombi & Keller, 2019).The contribution of subsoil P can be considerable for plant nutrient supply, but it is also seriously constrained if the subsoil is compacted (Ma et al., 2022).Reduced yield and root growth also have serious negative consequences for building soil C as well.
Plant roots, root exudates and the unharvested part of aboveground residues are the main C input to soils.The belowground C inputs (root fragments and rhizodeposition) are especially important because, as they are stabilized in the soil with higher efficiency than aboveground inputs (39% vs. 17% in Kätterer et al., 2011; 190% more stabilized with belowground inputs in Sokol & Bradford, 2019).The allocation of belowground C between different soil layers can be highly variable (subsoil fraction of total rhizodeposition varied 11%-42% even in a single field experiment) (Liang et al., 2022).This can be expected because, as fields are highly variable in their compaction distribution (tyre tracks, headlands and low strength areasNaderi-Boldaji et al., 2013) and roots respond strongly to compaction-induced environmental changes (Colombi & Keller, 2019).In a single field monitored over time, gradual soil compaction reduced root elongation by 50% from 1960 to 2010 (Keller et al., 2019).The reduced root growth from poor soil structure can control C sequestration rates more than the amount of aboveground C inputs (Colombi et al., 2019).Soil compaction results in soils that accumulate less C and leak more P.
Over time, European croplands have lost considerable amounts of C. Figure 2b presents the accumulated C budget combined from two sources.The regression equation between starting C and annual C loss from Bellamy et al. (2005) was applied for 1960-1990 (starting at 3.5% C), and results from two C cycle models were used for 1990-2020 (Bellamy et al., 2005;McGrath et al., 2023).Two models were used to represent the model uncertainty in estimating C balances from crop production, weather and soil data.The combined data set suggests a cumulative average C loss of 3000-4000 kg C/ha from 1960 to 2020.Measured time series from North Europe suggest roughly triple the loss rates (200 kg C/ha/year between 1974 and 2010 (Heikkinen et al., 2013)) because of higher starting C concentrations.This aligns with modelling studies, which show the highest C loss in North and West Europe (McGrath et al., 2023).The regions also have the highest P accumulation (Figure 2a) and soil compaction risks (Keller & Or, 2022).Overall, these three developments have created a combined problem of decreasing C, increasing compaction and accumulating but leaky P. The problem is not evenly distributed regionally, with most of the problem accumulated in North and West Europe.The problem is also not evenly distributed locally, which is the topic of the next section.

MAKES P HOTSPOTS WORSE
The previous section examined P and C balances aggregated over fields and regions.In practice, the balance varies considerably even within a single agricultural field.If the P fertilization rate is the same across the field, but the yield varies, this will result in spots of negative and positive P balances (Figure 3).Repeated over time, this results in the formation of different fertility zones in the field.The same applies to C, increasing variability in the field and making it difficult to track changes in C stock over time (Stanley et al., 2023).Management zone-based precision agriculture has been used to address this problem for fertilizer application (Hedley, 2015), but the concept is also valuable for sequestering C and preventing P leakage.This section examines the localized within-field variation in P accumulation and risk to waterbodies and then considers how C sequestration practices could help mitigate these problems.
Recent research into P transport to waterbodies has highlighted the role of hotspots ('critical source areas') (Gonzales-Inca et al., 2018;Thomas et al., 2016).These hotspots can be created even with a neutral overall P balance.Figure 3 presents an example of such a scenario, with soil compaction limiting yields and resulting in the build-up of P in high-risk zones.This is an illustrative example, drawn based on the distribution of high-resolution compaction and yield measurements (Naderi-Boldaji et al., 2013) and typical topographic wetness index (TWI) flow paths from another study (Riihimäki et al., 2021).Assuming an average yield of 5 t/ha and fertilization rate of 16 kg/ha P, the P balance is +1 kg/ha, or about half of the European current average (Panagos et al., 2022).However, the compacted zones may have only 1-2 t/ha yield, whereas the best yielding zones can have 8-10 t/ha yields (Naderi-Boldaji et al., 2013).Compaction has been shown to seriously limit yields in several field surveys (Keller et al., 2012;Mueller et al., 2009), and limited rooting depth can result in 3 t/ha yield loss in non-irrigated crops (White et al., 2015).The differences in yield but constant application rate of P fertilizer result in a deficit of −11 kg P/ha/year for the high yielding zone and a surplus of +13 kg/ha/year for the low yielding zone.Repeated for 10 years, these differences result in mining out 110 kg/ha P or building extra reserves of 130 kg P/ha.Therefore, field zones can accumulate or remove legacy P, even when the overall field P balance is neutral.
The accumulation of P would be of little consequence if the parts where P accumulates were only weakly connected to waterbodies.However, compaction commonly occurs in field margins and in low-lying, wetter field regions (Alaoui & Diserens, 2018;Naderi-Boldaji et al., 2013).Compaction in these strongly waterbody-connected regions increases runoff and creates perfect conditions for P emissions.Indeed, recent research has highlighted that only 1%-6% of agricultural landscapes are high risk for legacy P emissions to waterbodies (Thomas et al., 2016).High-risk areas are characterized by connectivity to a waterbody (through runoff) and accumulated P (Thomas et al., 2016).Identifying and targeting hotspots for improved management may be an effective way to control both total and dissolved reactive P emissions at the landscape level (Speir et al., 2022).
The high-risk zones could be covered with perennial vegetative filter strips, but these strips do little to remove the accumulated P (Ramler et al., 2022).If they are not carefully managed, P may even accumulate on the soil surface of the strip (Ramler et al., 2022).Woody filter strips can be effective in accumulating P into aboveground plant parts, which can then be harvested (Neilen et al., 2017).Harvesting biomass is critical for both wooded and grassed filter strips to avoid P build-up (Ramler et al., 2022).As soil compaction can limit yields (Section 2), it will slow down the removal process and hinder P scavenging by roots.To achieve strongly negative P balances, the growing conditions need to be improved (improving soil structure, solving nutrient limitations, improving drainage, possibly liming), and the accumulated P has to be mobilized to become plant available, without losing it to the waterbody.Fortunately, cover crops can help with both problems, which is further elaborated in the following section.-

THROUGH COVER CROPS AND COMPACTION MANAGEMENT
The dissolved reactive P emission reduction by cover crops can be considerable (31%-88%) at the catchment scale (Speir et al., 2022).The benefits of cover crops are linked to improvements in site hydrology.Figure 4 is a conceptual model of the key differences between a compacted low-vegetation cover soil and well-structured high-vegetation cover soil, as documented by recent cover crop research (Basche & DeLonge, 2017;Blanco-Canqui et al., 2015;Hanrahan et al., 2021).Crude approaches to soil C storage, such as manure application (importing C) or avoiding tillage (to decrease decomposition), may increase P emissions by concentrating excess P to the soil surface.Stratifying P in an already high P situation, as in the field margin of Figure 3, can result in major increases in P emissions (Baker et al., 2017).However, increasing vegetative cover through cover crops, intercrops and crop rotations, which include perennial grasses, has a much broader effect on soil properties than only stratifying P (Basche & DeLonge, 2017;Hanrahan et al., 2021).At the landscape level, increased vegetation cover can reduce both total and dissolved P emissions by reducing water flows during peak flow events and by cover crop uptake of dissolved P (Hanrahan et al., 2021).
In compacted, shallow-rooted and sparsely vegetated soil, much of the rainfall is directed to surface runoff (Figure 4).Some of the water enters soil macropores (cracks and earthworm burrows) and is directed to drainage pipes via preferential flow.Improving the structure and adding vegetative cover change the water cycle at many levels (Blanco-Canqui et al., 2015).Increased vegetation cover intercepts more rainfall before it reaches the soil surface.Although the interception is only a few millimetres per rainfall event, it can have major effects on the annual runoff and especially peak runoff during storms (Kozak et al., 2007;Speir et al., 2022).After entering the soil surface, water either infiltrates or leaves via surface runoff.Cover crops increase soil porosity and water holding capacity (Basche & DeLonge, 2017;Blanco-Canqui et al., 2015), which again decreases runoff and increases infiltration.Root channels from the cover crop can result in tripled infiltration rates even months after the cover crop has been terminated (Haruna et al., 2022).A porous and root-filled soil profile conducts the infiltrated water through gradual percolation instead of preferential flow (Mossadeghi-Björklund et al., 2016), allowing time for water to be held in the soil profile and for P to react with soil Fe and Al surfaces (Nair et al., 2020).The changes in soil properties result in less drainage runoff (ca.27 mm or 10%-20% reduction Meyer et al., 2018).In some cases, the runoff reduction can be 80% (Blanco-Canqui et al., 2015).More importantly, increased vegetative cover reduces the peak runoff events, which produce the most P emission load to waterbodies (Speir et al., 2022).
In addition to reducing P emissions, cover crops can improve the P use efficiency of cropping (P harvested in yield per unit of fertilizer P).Cover crops can mine out P (7-10 kg P/ha/year) from accumulated soil reserves and supply them to the following crop (Hallama et al., 2019).The P in the cover crop is as available for the crop as mineral fertilizer P (Maltais-Landry & Frossard, 2015), presenting opportunities for reducing P inputs while maintaining crop yields.If the 7-10 kg P/ha/year mined out by the cover crops could be taken up by the subsequent crop and exported from the field, this would maintain the P balance in the range necessary to reduce P emissions in the medium term (Schulte et al., 2010).
In some situations, cover crops can improve soil structure by themselves (Blanco-Canqui et al., 2015;Hao et al., 2023).Where compaction is severe, biological-mechanical soil loosening is more reliable (Colombi & Keller, 2019).Creating cracks in vegetation-covered soil with a low-disturbance subsoiler or a mole plough and then allowing roots to stabilize the restructured soil can be both fast and long-lasting interventions to improve soil structure (Colombi & Keller, 2019;Schneider et al., 2017).Improving the structure via mole drainage reduces surface runoff and allows time for water to react with soil surfaces, in some cases reducing both P and dissolved P emissions considerably (18%-24%) that are compacted and suffer from low yields (Schneider et al., 2017;Figure 3).Stabilizing the soil after loosening is critical (Colombi & Keller, 2019), and cover crops can help in making the soil less prone to subsequent compaction (Blanco-Canqui et al., 2015).
Increasing living vegetation cover is a key strategy for soil C sequestration (Chenu et al., 2019;Paustian et al., 2019).The recent findings reviewed in this section also present it as a way to both reduce P emissions and to mine out accumulated excess legacy P stocks.It can be a rare win-win-win situation for climate mitigation, eutrophication control and agricultural productivity improvement.However, accumulated and widespread soil compaction can limit the success of this strategy because compacted soil is a hostile environment for plant roots (Colombi & Keller, 2019).Soil management in the 20th century produced agricultural soils that are rich in P, poor in C and seriously compacted.The challenge of the 21st century is to turn the inherited legacy P problem into a legacy C resource.Cover crops, site-specific management and careful soil loosening are effective solutions for addressing this challenge.

ACKNO WLE DGE MENTS
A previous version of this manuscript was commented by Jussi Heinonsalo, Helena Soinne, Risto Uusitalo and Petri Ekholm.That manuscript had an overly ambitious scope, covering also the chemical interactions of Figure 1 and various C sequestration practices in addition to cover crops.Thank you also to three anonymous peer-reviewers for improving the clarity of the terminology throughout the manuscript.

F
I G U R E 3 Within-field variation can create large differences in soil phosphorous stock.(Illustrative example based on maps from Naderi-Boldaji et al., 2013 and Riihimäki et al., 2021.The blue lines represent flowpaths and the colour shading is yield.).
(Valbuena-Parralejo et al., 2019).Such interventions should prioritize areas F I G U R E 4 Soil structure and vegetation cover control runoff and P emissions.