Sustainable agricultural practices are crucial for meeting global food demands while preserving soil fertility. Most studies evaluating agricultural sustainability focus on either soil element cycling or crop yields, with few comprehensively assessing both over multiple decades. Long-term field experiments offer invaluable insights into the long-term effects of tillage, fertilizer application, and biomass removal on soil properties and yields. Previous long-term studies have shown varying effects of fertilizer addition on soil organic carbon (TOC) stocks; some reported decreases irrespective of fertilization, while others showed increases in fertilized treatments. The effects of inorganic phosphorus (P) application on TOC stocks are particularly ambivalent, with some studies reporting no significant impact while others observed decreases due to potential P-induced SOM desorption. This study aimed to investigate the combined effects of eight decades of agricultural production and P addition on soil organic matter (SOM) stocks, stoichiometry, soil P pools, and SOM decomposition in a temperate cropland. Using an 80-year-old long-term cropland experiment in southern Sweden, we analyzed the development of soil organic carbon, SOM stoichiometry, soil P pools, crop yields, and the Δ14C of soil TOC to understand long-term dynamics.

Existing literature reveals inconsistent results regarding the impact of long-term agricultural practices and phosphorus fertilization on soil organic carbon (TOC) stocks. Some studies report decreases in TOC stocks over decades regardless of fertilizer type (organic or inorganic), attributing this to land-use changes and the disruption of soil aggregates caused by repeated plowing. Other studies show TOC stock decreases only in control treatments without nutrient addition, with increases observed in fertilized plots. The effect of inorganic P addition on TOC is also debated; while increased plant productivity from P fertilization could lead to higher TOC stocks, especially when combined with nitrogen, some experiments show no significant impact or even decreases in TOC. This decrease may be due to phosphate-induced desorption of adsorbed SOM, making it more readily available for microbial decomposition. P application can also alter SOM stoichiometry and affect different soil P pools, potentially influencing plant-available P concentrations and organic P stocks. The existing literature lacks long-term observation-based studies combining yield and biogeochemical perspectives; therefore, there is a need for long-term field experiments to provide a basis for sustainable agricultural practices.

This study utilized data from an 80-year-old long-term cropland experiment in southern Sweden (58°20′49.9"N and 13°07′36.1"E), one of the oldest in Europe. The experiment comprises three treatments: a control (no P addition), a P treatment (1.75 g P m⁻² yr⁻¹), and a P6 treatment (same P amount but concentrated every six years). All treatments received Ca(NO₃)₂ annually. Crop yields were recorded almost annually since the experiment's beginning. Soil samples (0-20 cm) were collected every six years (with additional sampling in 2021) and analyzed for various parameters. Total organic carbon (TOC), total nitrogen (TN), and total organic phosphorus (TOP) were measured using a LECO CNS-2000 analyzer. Plant-available P (P-AL) was determined using acid ammonium lactate extraction. Phytate-P was analyzed using a hypobromite oxidation method. Soil pH was measured after suspending soil in water. The ¹⁴C:¹²C ratio of TOC was analyzed using accelerator mass spectrometry (MICADAS). Soil texture analysis was performed using wet sieving and sedimentation. A two-pool carbon model with connection in series was developed using the SoilR and FME R packages to simulate TOC dynamics based on the time series of annual yields and atmospheric bomb curve data. The model parameters were optimized using inverse data assimilation, estimating the decomposition rates of fast and slow carbon pools, carbon transfer rate between pools, and the proportion of annual yield contributing to soil carbon input. Statistical analysis, including ANOVA and linear and exponential modeling, was performed using R (version 4.1.1).

Over 53 years (1968-2021), soil TOC, TN, and TOP decreased linearly in the uppermost 20 cm, regardless of the fertilization treatment. TOC stocks decreased by 13.8%, TN by 13.8%, and TOP by 11.6%. The annual TOC stock decrease (0.14 t C ha⁻¹) was consistent with findings from a similar long-term experiment in Switzerland. A two-pool carbon model indicated a mean carbon transit time below 10 years and a median below 7.5 years, highlighting rapid carbon respiration. The model showed an overall increase in soil organic carbon turnover rate. Despite increased crop yields (from ~220 g dry weight m⁻² to >500 g dry weight m⁻²), there was no significant difference in soil TP and TOP stocks between the control and P addition treatments after 80 years, suggesting P uptake from below 20 cm. P addition significantly increased yields by 26-30%, considerably higher than the average increase reported in other long-term P addition experiments. The TOC:TOP ratio was significantly low (139), likely due to the high clay content (45%), which preferentially stabilizes organically phosphorylated compounds.

The findings demonstrate a continuous decrease in soil organic matter (SOM) despite increased crop yields, suggesting unsustainable agricultural practices. The rapid turnover of carbon, with a mean transit time below 10 years, indicates that a majority of carbon inputs are quickly respired, and that there is a large carbon loss. The lack of a significant difference in TP and TOP stocks between the control and P-fertilized treatments after 80 years points to substantial P uptake from the subsoil, challenging the notion that topsoil is the primary source of P for plants. The high yield response to P fertilization might be attributed to the high clay content and relatively low historical P fertilization at the study site. The low TOC:TOP ratio, also attributed to high clay content, suggests that organic P is effectively stabilized by adsorption to clay minerals. The absence of a positive effect of increased plant productivity (from P addition) on SOM stocks likely reflects the dominance of clay-mediated stabilization of organic matter over P-induced desorption. The continuous SOM decline, although not currently impacting yields, could compromise soil fertility and other soil properties in the future.

This 80-year study reveals the unsustainable nature of current agricultural practices, showing continuous SOM depletion despite increased plant productivity. High yields in the control treatment are supported by P uptake from the subsoil, and P addition significantly boosts yields. The rapid carbon turnover and low TOC:TOP ratio highlight the complex interplay between SOM dynamics and soil properties, particularly in clay-rich soils. Future research should explore strategies to enhance SOM stabilization and optimize P fertilization to achieve sustainable agricultural systems. This study emphasizes the crucial role of long-term experiments in understanding long-term element cycling in agroecosystems.

The study focuses on the uppermost 20 cm of soil, potentially overlooking deeper soil processes. Changes in bulk density over time weren't considered in stock calculations. The model's assumptions regarding constant or variable carbon input may affect the interpretation of transit time results. The high clay content at the study site may limit the generalizability of the findings to other soil types.