Carbon sequestration potential and fractionation in soils after conversion of cultivated land to hedgerows

Hedgerows supply multiple ecosystem services in agricultural landscapes and have been advocated to provide a relevant carbon sink for climate change mitigation. Using a space-for-time approach, we investigated carbon sequestration and fractionation in soils beneath hedgerows 1 to 70 yr after planting on cultivated land in the humid continental climate of Eastern Austria. We sampled 54 pairs of hedgerow and adjacent cultivated soil volumetrically to 40 cm depth and analysed them for bulk soil ( < 2 mm) mass, SOC fractions of differential stability and related soil properties. Total SOC stocks to 40 cm depth increased significantly (p < 0.05) by 34.7 ± 4.70 Mg C ha (cid:0) 1 to 119 ± 6.77 Mg C ha (cid:0) 1 beneath 31 – 70 yr old hedgerows, and by 15.6 ± 5.94 Mg C ha (cid:0) 1 to 106 ± 8.98 Mg C ha (cid:0) 1 beneath 1 – 30 yr old hedgerows. Carbon sequestration was limited to the organic and mineral topsoil layer (0 – 20 cm), SOC changes ( Δ SOC) at 20 – 40 cm depth were small and insignificant (p < 0.05). Linear regression of Δ SOC on hedgerow age (1 – 70 yr) yields a mean SOC sequestration rate of 0.65 ± 0.10 Mg ha (cid:0) 1 yr (cid:0) 1 down to 40 cm depth. Combining our data with published work we found hedgerow age explaining ~65 % of the variation of Δ SOC in topsoils, and ~68 % in shallow subsoils; Δ SOC increases with hedgerow age in topsoils, but declines in subsoils. Annual sequestration rates decline with hedgerow age in both soil layers. Up to 30 years after conversion, SOC is preferably sequestered in labile particulate organic matter (POM; > 20 µ m), and sta-bilised in the mineral-associated organic matter (MAOM; < 20 µ m fraction) thereafter. In the bulk soil and MAOM fraction, SOC strongly increases with oxalate-extractable Al (Al o ). We conclude that SOC sequestration beneath hedgerows offers a relevant CO 2 sequestration potential which is mainly limited to topsoils and primarily controlled by hedgerow age and Al o .


Introduction
Soils store the largest terrestrial carbon pool (~1500 Pg in upper 100 cm) with fundamental importance for food security and the global carbon cycle (Batjes, 1996).Increasing soil organic carbon (SOC) stocks enhances soil fertility and protection (e.g., against erosion) while simultaneously mitigating climate change by sequestering atmospheric CO 2 into SOC (Lal, 2004).The importance of SOC sequestration is increasingly recognized in the international community (e.g., '4 per 1000 initiative' and EU Soil Strategy for 2030) stressing that SOC enhancing management will be a key solution for climate change mitigation and adaptation (European Commission, 2021a;Minasny et al., 2017).
Generally, conversion of land into agricultural production has depleted SOC compared to natural baselines (on average by approximately 8%) depending on various factors, including land management and initial SOC stocks (Sanderman et al., 2017).Another effect of agricultural expansion is a reduction in local terrestrial biodiversity, estimated by Newbold et al. (2015) to approximately 10% globally in total abundance for land-use change, where agriculture is a major driver.Coupled with the vast spatial extent of agriculture production, agroecosystems exhibit large potentials for SOC sequestration and increased biodiversity.Adopting measures that promote biodiversity and SOC sequestration is therefore paramount if agriculture is to contribute to sustainable development.
Planting hedgerows, i.e., establishing linear features of woody vegetation on field boundaries is an example of one such promoted measure.Hedgerows provide multiple ecosystem services (Holden et al., 2019;Weninger et al., 2021) such as increasing populations of birds and insects (Laura et al., 2017;Heath et al., 2017), providing soil erosion reduction (Haddaway et al., 2018) and acting as both above-and belowground C sinks (Drexler et al., 2021;Mayer et al., 2022).Consequently, hedgerow planting has been incorporated in agrienvironmental policy (European Commission, 2021b).
SOC sequestration occurs when inputs of organic matter exceed outputs to the soil system, however, when input and output rates stabilize, SOC stocks will reach a steady state (Chenu et al., 2019).Thus, soils are not infinite carbon sinks.In addition, the potential for soil sequestration depends on multiple factors including management type, climate, soil mineralogy, nutrient status as well as interactions between mineral soil particles and their properties, and the soil biome (roots, microbes, and fauna) (Wiesmeier et al., 2019).
Evidence suggests that residence time of carbon as a constituent in soil organic matter (SOM) can vary three orders of magnitude depending molecular size and density (von Lützow et al., 2006).When soil fauna and microbes sequentially break down SOM into ever-smaller fragments, adsorption and occlusion to soil minerals and aggregates protect SOM from further decomposition (Lehmann and Kleber, 2015).One approach to distinguish between SOM components of variable persistence and functioning is the separation of particulate (POM) and mineral-associated organic matter (MAOM) (Lavallee et al., 2020).POM represents relatively large (>20 µm), plant-derived fragments of unprotected organic matter, while MAOM is a small-sized (<20 µm) and persistent fraction of microbial and plant origin.Monitoring changes in both POM and MOAM fractions enables an analysis that is more sensitive to the sustainability of long-term SOC sequestration.
Previous studies report that hedgerows increase SOC stocks when compared to adjacent soils under cultivation (Chiartas et al., 2022;Udawatta et al., 2022;Viaud and Kunnemann, 2021), however, the number of studies is limited.To our knowledge, no studies have presented a comprehensive dataset with evenly distributed hedgerow ages over a sufficient age-span that also considers SOC fractionation.
This study aims at enhancing the database and our understanding of SOC sequestration and allocation in fractions of differential stability in cultivated soils after conversion to hedgerows.Specific research questions that are addressed include: • How long does it take for the hedgerow soils to approach a new steady state for SOC? • Quantification of SOC stocks, annual sequestration rates and their change with hedgerow age; • How is the sequestered carbon distributed to topsoils and shallow subsoils?• To which SOC fractions (POM, MAOM) is the carbon allocated in hedgerow soils of differential age? • Which soil characteristics are primary controls of SOC in the soils of the study area?
Employing a space-for-time substitution approach, we investigate the potential for long-term  hedgerow-induced SOC sequestration including 54 pairs of hedgerows and cultivated soils in the north-eastern lowlands of Austria.We hypothesize that (1) SOC stocks increase concurrently with hedgerow age but annual sequestration rates level off as a new steady state is approached; (2) sequestration of the additional C input predominately occurs in the POM fraction in young hedgerows, while, based on the scarce available information, we do not know if the MOAM fraction increases or decreases following conversion of cultivated land to hedgerows; (3) based on published data the contribution of subsoils remains unclear but based on the study of Liebmann et al. (2022), we expect limited transfer of SOC to subsoil layers; (4) apart from hedgerow age, SOC and MAOM are largely controlled by soil properties (fine mineral fraction < 20 µm, hydrous oxides of Fe and Al, pH) that determine the soil's capacity to stabilise organic compounds.

Study area
The study area is located in the political districts Gänserndorf, Hollabrunn and Mistelbach in the northern and eastern parts of the Weinviertel region in Lower Austria, and belongs to the north-eastern lowlands according to the Austrian classification of agri-environmental regions.The regional mean annual temperature (MAT) is 9.1 • C, mean annual precipitation (MAP) varies between 477 mm in the northern and 528 mm in the eastern parts.According to the World Reference Base for Soil Resources (IUSS Working Group WRB, 2014), the soil cover is mainly composed of a mosaic of Phaeozems and Chernozems, and to a lesser extent Cambisols.
Since 1950, the Lower Austrian District Authority has been supporting the establishment of hedgerows.Between 1950 and 2019, 2911 ha (3656 km) of hedgerows have been established in in the province of Lower Austria, 1475 ha (1753 km) thereof in the study area.Generally, the hedges are composed of a mix of 15 to 20 native, regional tree and shrub species.In the study area, important tree species include Acer campestre, Carpinus betulus, Fraxinus excelsior, Populus tremula, Prunus padus, Quercus cerris, Q. petrea, Q. robur, Salix caprea, Tilia cordata, T. platiphyllos and Ulmus minor.Large shrub species planted are Corylus avellana, Prunus mahaleb and Sambucus niger, important smaller shrubs include Cornus mas, C. sanguinea, Euonymus europea, Frangula alnus, Ligustrum vulgare, Rhamnus cathartiga, Rosa canina and Viburnum opulus.Some hedges are amended by local fruit trees such as Juglans regia, Prunus avium, Prunus domestica and P. insititia.The vegetation structure and species composition are subject to changes related to time elapsed since conversion, and therefore vary among the sites.The hedgerows established in the study area have an average width of 8.42 m, and cover an area of 0.29 ha.

Experimental design, field work and soil characteristics
We obtained information about locations, hedgerow age and species composition from the database of the Lower Austrian District Authority.Within the study area, we selected 54 sampling locations (Fig. 1) based on the following criteria: • To keep the variability of soil properties among the sampling sites relatively small, selection of sites was limited to the soil groups Chernozem and Phaeozem (IUSS Working Group WRB, 2014) on loess, tertiary or fluviatile fine-textured sediments; for selection we used the information provided by the Austrian Digital Soil Map (Bundesamt für Wald, 2019); • To allow for assessment of carbon stock changes over time, we ensured even distribution of the 54 hedgerows across the available age span, i.e., 1-70 yr after establishment (space-for-time approach).
At each selected location, we collected pairs of soils beneath the hedgerow and under cultivation in autumn 2019.The adjacent cultivated soil was located ~10 m outside the hedgerow thus minimizing both differences in soil properties and hedgerow effect on the control (Viaud and Kunnemann, 2021).The selection of the side of the hedgerow sampled for cultivated controls was based on information obtained from the Austrian Digital Soil Map, to ensure comparable soil type and properties.We sampled the mineral soils volumetrically to 40 cm depth using the split-tube sampler set 04.17 (Ejkelkamp, Giesbeek, The Netherlands) with an inner diameter of five centimetres.The soil core was split in two depth increments (0-20, 20-40 cm); if present, we W.W. Wenzel et al. sampled organic layers volumetrically using a humus frame.For transport to the laboratory, we transferred the soil material of each layer quantitatively to polyethylene bags.
At each sampling location, we also collected soil blocks from the two depth increments of the mineral soil using a spade to determine soil textural and structural characteristics according to the German field manual (Ad-hoc Arbeitsgruppe Boden, 1994).Subsequently, we used the soil texture classes to determine the textural qualifiers according to the World Reference Base for Soil Resources (IUSS Working Group WRB, 2014), indicating a dominance of clayic and siltic textural qualifiers in topsoils and subsoils (Fig. S1).
We selected a subset of 21 locations to determine carbon fractions of differential stability.These locations are evenly distributed across the investigated age span of hedges (1-70 yr).Using a subset of 8 locations of relatively uniform hedgerow age (1-20 yr), we explored the effect of soil characteristics including hydrous oxides of Fe and Al and the fine mineral fraction < 20 µm on SOC and MAOM in the hedgerow and adjacent cultivated soils.The hedgerow age span was kept in this narrow range to minimise the variation due to hedgerow age.

Analytical procedures
After air-drying, we passed the mineral soils through a 2-mm screen and recorded the masses of coarse fragments (>2 mm) and fine earth (<2 mm).We calculated bulk density by relating the total soil mass of each layer to the corresponding sample volume (392.5 cm 3 ), and the oven-dry factor as the difference between air-dry soil and soil dried to constant mass at 105 • C. All data is presented for oven-dry (105 • C) material.
After milling, we measured organic soils and subsamples of mineral soils for organic and inorganic carbon.To this end, we weighed aliquots (15-200 mg) of milled soil in steel vessels and determined the carbon concentrations sequentially at 400, 550 and 900 • C using a soli TOC cube® carbon analyser (Elementar, Langenselbold, Germany).The sum of the fractions up to 550 • C represents soil organic carbon (SOC), the carbon released between 550 and 900 • C the inorganic carbon.Using the stoichiometric ratio, we converted the inorganic carbon to the CaCO 3 equivalent.
Using a subset of 21 locations, we measured carbon fractions of differential stability using a modification of the method published by Spielvogel et al. (2006).The method physically separates three fractions according to particle size by employing sequential ultrasonic energy and sieving steps.The selection of ultrasonic energy outputs aims at maximising dispersion of aggregates (macroaggregates at 60 J mL − 1 ; microaggregates at 440 J mL − 1 ) while minimising redistribution of organic matter among the fractions (Amelung et al., 1998;Amelung and Zech, 1999;Just et al., 2021;Thuriès et al., 2000).The >200-µm fraction is considered to represent coarse particulate organic matter (coarse POM), either free or associated with macroaggregates, the 200-20 µm fraction fine POM (free or associated with microaggregates), the fraction < 20 µm is interpreted as mineral-associated organic matter (MAOM).The stability is deemed to increase in the order coarse POM < fine POM < MAOM.
For breaking up macroaggregates (>200 µm) we placed 30 g sieved (<2 mm), air-dry soil into 200-mL polyethylene containers and added 150 mL de-ionised water before ultrasonic treatment (Branson W-250 D, Emerson Technologies GmbH & Co. OHG, Dietzenbach, Germany; oscillation 250 Hz, maximal power 200 W) at an output energy of 60 J mL − 1 .The ultrasonic probe was immersed two centimetres into the suspension.Subsequently, we subjected the suspension to wet sieving using a vibration sieving system (Retsch AS 200, 200-µm steel screen; Retsch AG, Haan, Germany) at an amplitude of 50% and ~250 mL deionised water to separate the coarse POM fraction.The particles retained in the sieve were rinsed with deionized water until the washing solution was clear, transferred in porcelain dishes and dried to constant weight for at least 24 h at 105 • C. The suspension (<200 µm) passing the sieve was collected in a 2-L glass container.To reduce the volume for the subsequent ultrasonic treatment, the coarser particles were allowed to settle down before the overlying suspension was pumped into a 2-L Schott bottle, and combined with the rinsing solutions.The suspension was passed through a 20-µm screen to separate fine POM, fine sand and silt fractions with an equivalent diameter of 200-20 µm.The material retained in the sieve was combined with the sediment of the suspension obtained in the first sieving step, transferred into a clean 250-mL bottle and filled with deionised water to 150 mL; the resulting suspension was cooled down to 5 • C in an ice-filled container to avoid overheating during the subsequent ultrasonic treatment.The output energy of this treatment was set to 440 J mL − 1 to destroy all microaggregates while avoiding breakdown of silt-sized organic particles.Subsequently, the suspension was subjected to wet sieving through a 20-µm screen.The suspension passing the screen was combined with the one previously pumped into the 2-L Schott bottle, after reducing the volume of the latter.The material retained in the 20-µm sieve as well as the suspension < 20 µm were, after reducing the volume, each transferred to porcelain dishes and dried to constant weight for at least 24 h at 105 • C, and the mass of all fractions (>200 µm, 200-20 mm < 20 µm) was recorded.The < 200-µm fractions were homogenised with a pestle, the > 200-µm fraction was milled using a Retsch MM400 mill employing 30 oscillations per second.Subsequently, we measured organic and inorganic carbon in each fraction using a soli TOC cube® carbon analyser as described above.
The recovery of the soil mass after fractionation was on average 97.4 % (SD = 1.95 %; n = 106); the recovery of the organic carbon determined as the sum of OC in the fractions compared to direct OC measurements of the fine earth (<2 mm) was 96.2 ± 7.3% (n = 73).
Using a subset of 8 sites with hedges of similar age (5-20 yr), we determined ammonium oxalate extractable Al and Fe in the mineral soil layers of the hedgerow and adjacent cultivated soils using a modification of the Loeppert and Inskeep (1996) procedure.After homogenization with a mortar, we placed 0.5 g soil into 100 mL shaking bottles wrapped with aluminium foil, and added 30 mL of 0.175 M ammonium oxalate [(NH 4 ) 2 C 2 O 4 )] + 0.1 M oxalic acid (H 2 C 2 O 4 ) solution adjusted with HCl to pH 3. The samples were shaken for 2 h in the dark using a rotary shaker and filtered through a filter paper (Grade 1290, Ahlstrom-Munksjö, Helsinki, Finland).The filtrate was acidified by the same volume of 4 % HNO 3 and stored at 4 • C. For measuring ) using ICP-OES (OPTIMA 8300, Perkin Elmer, Rodgau-Jügesheim, Germany), we added 1 mL of Y dissolved in 2 % HNO 3 as an internal standard to the vials containing 10 mL of the acidified filtrate.The reference soil SO26 (subsoil of EUROSOIL 7; Weissteiner et al., 1999) was used as external control.

Calculations and statistics
We calculated organic carbon stocks in the mineral soil using the equivalent soil mass (ESM) approach according to Juvinyà et al. (2021).For each soil profile, we fitted the plots of cumulative organic carbon masses (Y; Mg ha − 1 ) versus the cumulated mineral fine earth masses (X; Mg ha − 1 ) using a modified fitting procedure: where a and b are dimensionless fitting constants.Mineral fine earth was calculated from the fine earth masses by deducting the organic matter content; we obtained the latter by multiplication of the organic carbon concentrations with the Van-Bemmelen factor (1.724).Using Eq. ( 1) and the fitted constants we recalculated the SOC stocks for reference soil masses assigned to each depth layer which we determined as the means of the soil masses in each mineral soil layer under cultivation in the study area, yielding 2540 ± 48.1 Mg ha − 1 in topsoils, and 2890 ± 42.1 Mg ha − 1 in the shallow subsoils.The variability measures denote standard errors of the mean.The ESM approach corrects for biases due to bulk density changes during SOC accumulation in the hedgerows.We estimated the air capacity of the soils from bulk density, soil textural class and SOC using tabulated pedotransfer functions (Ad-hoc Arbeitsgruppe Boden, 1994), adjusted for the content of rock fragments.Air capacity is defined as the share of macro-pores > 50 µm.
We grouped the organic carbon data (concentrations or stocks) according to soil layers (organic layer, topsoil 0-20 cm, shallow subsoil 20-40 cm, mineral soil 0-40 cm, and whole soil, i.e., organic layer plus mineral soil) into two categories with similar number of observations, i. e., 1-30 (n = 24) and 31-70 (n = 30) years each for hedgerow and cultivated soils, and tested them separately for normal distribution using the Kolmogorov-Smirnov test.The null hypothesis, i.e., normal distribution was confirmed for all presented data.
We evaluated the significance (p < 0.05) of differences of SOC concentrations or stocks, and of air capacity between cultivated and hedgerow soils within each age category using paired t-tests.
We also calculated the paired differences of SOC stocks (ΔSOC) between hedgerows and cultivated soils, and performed regression and correlation analysis between hedgerow age (independent variable) and ΔSOC.The goodness of the fits was evaluated by the coefficient of determination (R 2 ), the predictive power by the root squared mean error (RSME) and prediction bands.
We compiled published data on SOC stocks, ΔSOC and annual sequestration rates in soils beneath hedgerows and shelterbelts from the northern hemisphere.The selection of the publications was based on a literature search in the SCOPUS database using the keywords "organic carbon" and "agroforestry", "hedgerow", or "shelterbelt".As agroforestry systems largely differ in their SOC sequestration potential, with mean sequestration rates of 0.67 Mg C ha − 1 yr − 1 beneath hedgerows, and 0.21 Mg C ha − 1 yr − 1 in agroforestry (Mayer et al., 2022), we extracted only data for hedgerows and shelterbelts.The number of studies selected was further narrowed down by the requirement to compare pairs of hedgerows with cultivated soils, and a detailed evaluation of the availability of the relevant information on climate (Koeppen classification, MAT, MAP), soils, hedgerow age, tree species, SOC stocks and sequestration rates.Finally, eight studies remained that fulfil these requirements to a large extent.For comparison with these published data, we calculated annual SOC sequestration rates as the ratio between mean ΔSOC and the mean hedgerow age (Drexler et al., 2021;Juvinyà et al., 2021;Mayer et al., 2022) for the two hedgerow age groups (1-30 and 31-70 yr) of our study.Using this dataset, we performed regression and correlation analysis between hedgerow age and mean ΔSOC or annual sequestration rates as described before.
Using a subset of eight sites of relatively uniform age (5-20 yr) we investigated the effect of physical and chemical soil characteristics on SOC concentrations in cultivated and hedgerow mineral soils by multiple regression analysis.Based on previous work (Fukumasu et al., 2021;Mayer et al., 2022;Wiesmeier et al., 2019) we chose the fine mineral fraction mass (<20 µm) along with ammonium oxalate-extractable Al (Al o ) and Fe (Fe o ) as independent variables to predict the SOC fractions < 2 mm (SOC in bulk soil) and < 20 µm (SOC in MAOM).As additional independent variables the mean of the soil depth increment (10 cm for topsoils 0-20 cm; 30 cm for subsoils 20-40 cm), and land use category (cultivated land and hedgerows) were included in the regression analysis after converting the latter to dummy-coded variables with cultivated soils as reference.In initial model runs we included all variables.Employing a top-down approach, we subsequently excluded variables with insignificant (p < 0.05) regression coefficients (mineral fraction mass < 20 µm and Fe o ) to obtain the final model.

Soil characteristics beneath hedgerows and cultivation
Table 1 compares the means of relevant soil characteristics in mineral topsoils and shallow subsoils under cultivation and hedgerows.The soils are moderately calcareous, the content of coarse fragments (>2 mm) is generally small (<3.02 % m/m).Both bulk soil and fine earth densities are by ~30 % and ~15 % larger in cultivated as compared to hedgerow soils.Organic carbon concentrations are by ~54 % larger in topsoils, and ~15 % larger in subsoils under hedgerows as compared to cultivated soils (Table 1).

Carbon concentrations in bulk soil and fractionation
The conversion of cultivated land to hedgerows generally resulted in increased SOC concentrations in topsoils (Table 1), however, data compiled in Table 2 indicate that the means of SOC increased only by ~20 % (not significant; p = 0.05) during the first 30 years but significantly by ~85 % beneath hedgerows 31-70 yr after their establishment.In shallow subsoils, the observed increases of SOC concentrations are significant but generally smaller, with ~16-17 % in both age categories.

Carbon stocks
In response to the conversion of cultivated land to hedgerows, SOC stocks in mineral topsoils (reference soil mass 2540 ± 48.0 Mg ha − 1 in the 0-20 cm layer) increased on average by 7.98 ± 4.23 Mg ha − 1 (~16 %) after 1-30, and 28.8 ± 3.34 Mg ha − 1 (~64 %) after 31-70 yr (Fig. 3).Note that the variability measures refer to the standard error of the mean.These changes are relevant, but sensu stricto only significant for hedgerows established > 30 yr ago as the p value for the age class 1-30 yr is 0.072 and thus slightly above the pre-defined significance level of 0.05.Carbon accumulation in the mineral soils beneath hedgerows was complemented by the formation of an organic surface layer, resulting in SOC stocks of 2.86 ± 0.42 Mg ha − 1 1-30 yr, and 4.46 ± 0.64 Mg ha − 31-70 yr after conversion.As subsoil (20-40 cm) SOC stocks remained virtually unchanged, the increase in whole soil C stocks to 40 cm closely reflects the changes observed in mineral topsoils and organic layers, with sequestration of 15.6 ± 5.94 Mg C ha − 1 (~ +17 %) and 34.7 ± 4.69 Mg C ha − 1 (~ +41 %) after 1-30 and 31-70 yr, respectively.The corresponding relative increases in the entire mineral soil (0-40 cm) are ~15 and ~38 %, indicating that the contribution of organic layers is relatively small (Fig. 3).
We further explored the effect of time elapsed since the establishment of hedgerows on the accumulation of SOC stocks by correlation and regression analysis (Fig. 4).The calculated differences of SOC stocks (ΔSOC) between the pairs of hedgerows and cultivated soils are fairly strong and positively correlated with hedgerow age, explaining 42 % of the variation in whole soils (including organic layers), and 50% in topsoils, respectively.The regression coefficients suggest a mean annual increase of SOC stocks by 0.53 ± 0.07 (standard error) in topsoils, and 0.65 ± 0.10 Mg ha − 1 in whole soils for hedges up to 70 years after their establishment, but the predictive power is limited as indicated by the wide prediction bands and the root mean square errors (RMSE) (Fig. 4).Similarly, we observe a mean annual increase of the SOC stocks associated with mineral-associated organic matter (MAOM, < 20 µm) by 0.31 g ha − 1 , with hedgerow age explaining 70 % of the variation (Fig. 4).

Soil physico-chemical controls of soil organic carbon fractions
Using a subset of eight sites of relatively uniform age (5-20 yr) we investigated the effect of physical and chemical soil characteristics on SOC in cultivated and hedgerow mineral soils by multiple regression analysis as detailed in section 2.4.The regression models for MAOM (<20-µm) and bulk soil (<2 mm) are highly significant with a large share of the variation (>70 %) explained by the predictor variables land use category, soil depth and Al o (Table S1).The effect of the soil characteristics is visualised in the scatterplots for each land usedepth increment category (Fig. 5), highlighting the strong influence of Al o on both SOC fractions, whereas Fe o and the mass fraction < 2 mm are only loosely related.Al o explains > 70 % of the variation in both SOC fractions except for SOC <2mm in the hedgerow topsoils (52 %), which is in line with the larger contribution of POM (>20 µm) (Fig. 2) that is not associated with the fine mineral phase.

Potential carbon sequestration in hedgerow soils
We show that after conversion of cultivated land to hedgerows, SOC sequestration in the soils of the study region continues for at least years (Fig. 4), implying that steady state is not reached within this period.Over the entire age span, the potential mean sequestration rate beneath hedgerows in the whole soil including the organic layer is 0.65 Mg ha − 1 yr − 1 , with most of the carbon sequestered (0.53 Mg ha − 1 yr − 1 ) in the topsoil (0-20 cm) of which more than half (0.30 Mg ha − 1 yr − 1 ) is allocated to MAOM.Our findings indicate a relevant sequestration potential per unit area hedgerow established, with a considerable share of SOC being stabilised.
A comparison of potential sequestration rates at early (0-30 yr; mean age 14.5 yr) and later stages (31-70 yr; mean age 49.3 yr) of hedgerow development reveals that topsoil sequestration rates remain almost unchanged at ~0.55 Mg ha − 1 yr − 1 , whereas those of organic layers decline with hedgerow age from 0.20 to 0.09 Mg C ha − 1 yr − 1 (Table 3).We also observe a decline of the sequestration rates in whole soils from 1.08 Mg

Table 1
Arithmetic means and standard error of the mean of mineral topsoils (0-20 cm) and shallow subsoils (20-40 cm) under cultivation and hedgerows in the study area (n = 54 for each land use category).C ha − 1 yr − 1 to 0.70 Mg C ha − 1 yr − 1 .Our findings suggest that the sequestration potential beneath hedgerows in the study area is largely limited to topsoils.It appears that at early stages a larger proportion of carbon cannot be incorporated in the mineral soil, probably because soil faunal communities responsible for bioturbation require some time to adopt to the changed land use regime.
Comparison of our results with published data is limited by different age spans, stock calculation methods (ESM versus fixed depth method), soil depth sampled, and requires consideration of the environmental context (e.g., climate, soil type) and management.A summary of relevant literature data along with key results of our study is compiled in Table 3, considering only studies where SOC stocks beneath tree or hedgerows were compared to those in adjacent cultivated fields.
Reported SOC stocks in the organic layer vary between 3.1 and 14.8 Mg C ha − 1 (Table 3), however, it appears that in some studies the organic layer, if present, was not sampled (Chiartas et al., 2022;Mayer et al., 2022;Wiesmeier et al., 2018;Seitz et al., 2017).The SOC stocks accumulated in the organic layer in our study region are close to the lowest values reported in the literature.The SOC stocks in the cultivated mineral topsoils (0-20 cm) of our study amount to 49.8 (1-30 yr) and 44.9 Mg C ha − 1 (31-70 yr) which compares to a range of 26.4 to 88.4 Mg C ha − 1 reported in the literature (Table 3).Only few studies report data for deeper soil layers, with SOC stocks ranging from 14.1 to 34.1, and 10.8 to 39.7 Mg C ha − 1 for cultivated and hedgerow soils, respectively (Chiartas et al., 2022;Juvinyà et al., 2021;Seitz et al., 2017).This compares to SOC stocks of about 40 to 45 Mg ha − 1 in cultivated and hedgerow subsoils of our study (Table 3).Overall, the comparisons indicate stronger allocation of SOC to the shallow subsoil layer in the soils of the Weinviertel as compared to most other regions.
The annual sequestration rates, calculated as the ratio between ΔSOC and hedgerow age vary between 0.1 and 2.4 Mg C ha yr − 1 in topsoils, and − 0.1 and 0.8 Mg C ha yr − 1 in shallow subsoils, with those observed in the Weinviertel region falling within these ranges (Table 3).
We further explored the data compiled in Table 3 to determine the effect of hedgerow age on SOC accumulation and the annual SOC sequestration rate (Fig. 6).As expected, we find the differences between SOC stocks beneath broadleaved hedgerows and adjacent cultivated topsoils (ΔSOC) increasing with hedgerow age.The hedgerow age explains ~65 % of the variation even though the site and soil characteristics span across different climate and soil conditions (Table 3).Similarly, it explains ~68 % of ΔSOC in the subsoils, however, the SOC stocks are slightly decreasing with hedgerow age (Fig. 6).The regression analysis of annual SOC sequestration rates as affected by hedgerow age shows declining sequestration for topsoils and subsoils.Hedgerow age explains substantial shares of the overall variation, with ~48 % in topsoils and ~87 % in subsoils (Fig. 6).As indicated by the logarithmic regression curve, the decline of annual sequestration rates is fast during the first two decades after hedgerow establishment and slows down thereafter.
One site with hedgerows dominated by coniferous species (Sauer et al., 2007) is not included in the regression analysis as, in line with results of a meta-analysis of temperate agroforestry systems (Mayer et al., 2022), it shows considerably smaller SOC accumulation in the mineral topsoil compared to the broad-leaved hedges of similar age.This could be related to the better faunal turnover and microbial degradability of organic compounds derived from broad-leaved species, resulting in enhanced formation of stable organo-mineral complexes (Mayer et al., 2022).
In line with the findings in our study region, the linear increase of ΔSOC (Fig. 6) in hedgerow topsoils compiled from different studies suggests that hedgerow systems do not reach steady state even > 50 yr after establishment.However, annual sequestration rates are slowing down, indicating a larger SOC sequestration potential during the early stage of hedgerow development (Fig. 6).
The small sequestration rates in shallow subsoils observed in our study and other regions (Table 3; Fig. 6) are in contrast to claims that deep-rooting, permanent woody species could contribute to a relevant extent to SOC sequestration in hedgerows and similar systems such as forests and agroforestry.While the role of roots is hardly explored, transport and stabilisation of litter-derived organic carbon appears to be limited by multiple process constraints even though the physical capacity of subsoils to store additional MAOM is large (Liebmann et al., 2022).Experimental evidence for temperate forests suggests that only ~0.5 % of the litter-derived carbon is transported to forest subsoils where quick mineralisation is more pronounced than deemed in earlier studies (Liebmann et al., 2022).Some studies, at least for certain sites, report SOC sequestration in deeper soil layers following conversion of cultivated land to hedgerows (Cardinael et al., 2017;Chiartas et al., 2022;Meyer et al., 2021;Seitz et al., 2017) but these findings are limited to hedgerows in their early stage of development (<20 yr; Table 3).As shown in Fig. 6, it seems that older hedgerow subsoils even tend to lose SOC, albeit the number of observations is limited.Given the typical tillage depth of 25-30 cm, this could be related to the formation of biopores and transformation of compacted plough pans in the shallow subsoil layers to well-aggregated soil with lower bulk density.As a consequence, improved soil aeration could enhance mineralisation of SOC in shallow subsoils.This hypothesis is supported by decreased bulk density (Table 1), improved air capacity (Fig. S2) and the conversion of compacted platy to porous subangular blocky and granular structure observed in the subsoils of our study.

Carbon fractionation in hedgerow soils
We show that during the first decades after conversion of cultivated land to hedgerows, most C in topsoils is sequestered in labile (POM) fractions, followed by sequestration in the fine mineral fractions, resulting in stabilisation in MAOM (<20 µm) in the long term.In older hedgerows (>30 yr), the SOC concentration in MAOM is substantially

Table 2
Descriptive statistics for soil organic carbon (SOC; mg kg − 1 ) concentrations in topsoils and shallow subsoils under cultivation and hedgerows in the study area.Data is grouped by age of the hedgerows (n = 24 for age category 0-30 yr; n = 30 for age category 31-70 yr).According to a paired t-test differences between means of hedgerow and adjacent cultivated soils within each age category (p < 0.05) are significant, except for topsoils beneath 0-30 yr old hedges.larger than in adjacent cultivated topsoils (Fig. 2).Generally, changes of SOC fractions in subsoils after conversion of cultivated land to hedgerows are small compared to those observed in topsoils (Fig. 2).Significant (p < 0.05; paired t-test) changes are limited to increased SOC concentrations (+0.44 g kg − 1 ; +139 %) in the > 200µm fraction and the fine earth (<2 mm) beneath 1-30 yr old hedgerows (Fig. 2).We could not find any similar studies presenting subsoil fractionation data.
Published studies of SOC fractionation in soils beneath hedgerows or similar woody vegetation are scarce.The few studies available do not support a consistent understanding of the response of SOC fractionation to hedgerow establishment on cultivated land.In north-western France, Viaud and Kunnemann (2021) compared C fractionation in topsoils (0-30 cm) close to 20 to 120 yr old hedgerows (1 m distance) and in adjacent cultivated fields or grassland.They found larger C concentrations in POM close to the hedgerows in all three study sites but significantly larger C concentrations in MAOM only at one site.In topsoils (0-10 cm) beneath hedgerows (mean age 59.5 yr) in Moldova, Wiesmeier et al. (2018) found considerable C accumulation in POM fractions but no significant change of the C concentration in the fine fraction (<20 µm).In Armagh County (Northern Ireland), Fornara et al. (2018) found increased C concentrations associated with the fine fraction (<53 µm) of topsoils (0-20 cm) beneath silvopastoral trees ~30 yr after establishment as compared to adjacent grassland, albeit not significant.Similarly, Baah-Acheamfour et al. ( 2014) report significantly (p < 0.10) larger C concentrations in the fine fraction (<53 µm) of topsoils (0-10 cm) beneath 40 to 100 yr old hedgerows compared to adjacent grassland and cultivated sites in Alberta (Canada) but no significant differences for coarser fractions.In response to afforestation (1-30 yr after establishment) of loess-derived soils in Huachi county (Gansu province, China), carbon sequestration in the heavy fractions was larger in the 0-10 cm mineral topsoil layer beneath older forest stands, indicating long-term C stabilisation in MAOM (Jiang et al., 2019) as observed in our study.In a similar loess plateau study (Yongshou county, Shaangxi province, China), Zhang et al (2020) could not confirm this finding for woodland 20-35 yr after conversion.In both Chinese studies, most of the SOC was sequestered in lighter (POM) fractions.Our study is the first to clearly show a sequential incorporation of the additional C inputs in topsoils beneath hedgerows in POM during the initial phase after hedgerow establishment, followed by increasing stabilisation in MAOM after few decades.

Soil characteristics controlling carbon in bulk soil (<2mm) and MAOM
Using multiple regression analysis, we show that Al o clearly outperforms other soil characteristics (mineral fraction < 20 µm, Fe o ; Wiesmeier et al., 2019) expected to control to SOC storage and stabilisation in soil (Table S1, Fig. 5).A strong influence of Al o on SOC concentrations has been shown before (Fukumasu et al., 2021;Wiesmeier et al., 2019).Our findings indicate that Al o could serve as predictor of the soil's capacity to sequester and stabilise carbon in a given ecological region and land use regime.

Implications for climate change mitigation potentials in the study area
In 2019, hedgerows established by the programme of the Lower Austrian District Authority covered 1,475 ha in the study area, and 2,911 ha in the whole province of Lower Austria.Agriculture, i.e., grassland and cultivated land covered in 2019 ~878,480 ha of Lower Austria (Bundesministerium für Landwirtschaft, Regionen und Tourismus, 2020) with only ~0.33 % being planted with hedgerows (mean age 29.4 yr) through the Lower Austrian District Authority programme.In the study area, the agricultural area covers ~250,000 ha dominated by cultivation, with a share by hedgerows of ~0.58 %.The recently launched Biodiversity Strategy of the European Commission suggests a target of 10 % of the European agricultural landscapes to be covered by hedgerows and other landscape elements supporting high biodiversity by 2050 (European Commission, 2021b;Finn and ÓhUallacháin, 2020).Applied to Lower Austria, this translates to an area of ~88,000 ha in the   whole province, and ~25,000 ha in the study area.Assuming an average SOC sequestration rate beneath hedgerows of 0.65 Mg C ha − 1 yr − 1 for the whole soils down to 40 cm (Fig. 4), planting additional hedgerows to reach this target would allow for sequestration of ~15 Gg C yr − 1 in the study area, and 55 Gg C yr − 1 in the whole province of Lower Austria.As we could show that sequestration continues over several decades, during a period of 50 yr the SOC stocks beneath hedges could build up to ~765 Gg in the study area, and ~2.8 Tg in the entire province.This compares to current SOC stocks of ~24.4 Tg in cultivated soils (678,549 ha in 2019) and 10.8 Tg in grassland soils of entire Lower Austria (Wenzel et al., 2022).In this scenario, establishing hedgerows according to the target of the Biodiversity Strategy could offset ~1.4 % of the annual CO 2 emissions of Lower Austria which amount to ~15 Tg yr − 1 (Anderl et al., 2017) containing ~4 Tg C. As the implementation of the 10%-target cannot be achieved immediately, a realistic estimate of the annual sequestration rate will be smaller, whereas the period of relevant sequestration would expand.Note that we did not assess C sequestration in the hedgerow biomass.Based on published data (Kurganova et al., 2015;Morris et al., 2007) we estimate the sequestration potential of hedgerow biomass to exceed that of the soils by 5 to 6 times.Overall, these figures suggest that establishing hedgerows on 10 % of the agricultural land of Lower Austria would be a relevant contribution to measures of climate change mitigation in the region.

Conclusions
• Establishing hedgerows on cultivated land in the study region resulted in long-term SOC sequestration (at least 70 yr) at an average rate of 0.65 Mg ha − 1 yr − 1 in the soil layers down to 40 cm depth, with most of the sequestration occurring in the mineral topsoil (0-20 cm) and, at early stage of hedgerow development, the organic layer; • This long-term sequestration potential is consistent with information derived from a compilation of literature data on comparable hedgerow systems in the northern hemisphere; • Whereas steady state is not reached even after seven decades, annual sequestration rates overall decline with time elapsed since establishment of hedgerows; • Based on data from the study region and our literature compilation, shallow subsoils appear to provide only small sequestration potential, seemingly limited to younger hedgerows;       3. Dashed lines represent regression lines/curves.

Fig. 1 .
Fig. 1.Map of the sampling sites in the north-eastern lowlands of Lower Austria.
W.W.Wenzel et al.

Fig. 2 .
Fig. 2. Carbon concentrations of different size fractions in the soil mass < 2 mm up to 30 yr (n = 13), and 31 to 70 yr (n = 8) after conversion of cultivated soils to hedgerows.Significant differences (p < 0.05) according to a paired t-test between cultivated and adjacent hedgerow soils within each age group are indicated by capital (1-30 yr) and lower case (31-70 yr) letters.

Fig. 3 .
Fig. 3. Soil organic carbon stocks up to 30 yr, and 31 to 70 yr after conversion of cultivated soils to hedgerows.SOC stocks for mineral soil layers were calculated according to the equivalent soil mass method.Reference soil masses are 2540 ± 48.1 Mg ha − 1 in topsoils, and 2890 ± 42.1 Mg ha − 1 in the shallow subsoils (variability measures refer to the standard error of the mean).Significant differences (p < 0.05) according to a paired t-test between cultivated and adjacent hedgerow soils within each age group are indicated by capital (1-30 yr) and lower case (31-70 yr) letters.

Fig. 4 .
Fig. 4. Changes of soil organic carbon (ΔSOC) stocks with hedgerow age relative to the adjacent cultivated fields.Upper panel: bulk soil (<2mm) of whole soils; lower panel: bulk soil of topsoils; lower panel: mineral-associated organic matter (MAOM, < 20 µm) in topsoils.The solid line represents the linear regression, dashed lines the upper and lower prediction bands.The regression models (F-test) and the regression coefficients (t-test) are highly significant (p < 0.001).
W.W.Wenzel et al.

Fig. 5 .
Fig. 5. Scatterplots of SOC concentrations in the bulk soil (SOC <2mm ) and MAOM (SOC <20µm ) on ammonium oxalate extractable Al (Al o ) and Fe (Fe o ), and the mass fraction < 20 µm.Solid linear regression lines represent shallow subsoils, dashed lines topsoils.Regression lines are only drawn for relations with r 2 > 0.5.
presented in original paper.b Translated to WRB soil group based on national classification provided in the original paper.cClimateclassification according to Koeppen-Geiger: Bsk (cold semi-arid); Cfa (humid subtropical climate); Cfb (oceanic); Csa (hot summer Mediterranean); Dfa (humid continental hot summers with year around precipitation); Dfb (humid continental, mild summers, wet all year); Dfc (subarctic with cold summers and year around rainfall).dFD (fixed depth approach); ESM (equivalent soil mass approach).eVariability measures refer to the standard error of the mean.

Fig. 6 .
Fig. 6.Changes of soil organic carbon (ΔSOC) stocks relative to the adjacent cultivated fields (upper panel) and annual SOC sequestration rates (lower panel) with hedgerow age for topsoils and shallow subsoils.Data are taken from Table3.Dashed lines represent regression lines/curves.
• A substantial share of carbon in hedgerow soils is allocated to MAOM and thus considered stable;• Amorphous hydrous aluminium oxides (Al o ) are identified as important control of the soil's sequestration potential; • Expanding hedgerows in the study area from < 1% to 10% as advocated by the European Biodiversity Strategy could offset ~1.4 % of the annual CO 2 emissions of Lower Austria by carbons sequestration in soil, and probably five to sixfold more if carbon storage in biomass is considered.

Table 3
Summary of SOC stocks and differences (ΔSOC stock) beneath hedgerows and adjacent cultivated soils, calculated sequestration rates and related information about site characteristics and SOC stock calculation methods for this study and relevant literature data for comparison.
(continued on next page) W.W.Wenzel et al.