Hot regions of labile and stable soil organic carbon in Germany – Spatial variability and driving factors

Atmospheric carbon dioxide levels can be mitigated by sequestering carbon in the soil. Sequestration can be facilitated by agricultural management, but its influence is not the same on all soil carbon pools, as labile pools with a high turnover may be accumulated much faster but are also more vulnerable to losses. The aims of this study were to (1) assess how soil organic carbon (SOC) is distributed among SOC fractions on a national scale in Germany, (2) identify factors influencing this distribution and (3) identify regions with high vulnerability to SOC losses. The SOC content and proportion of two different SOC fractions were estimated for more than 2500 mineral topsoils (< 87 g kg−1 SOC) covering Germany, using near-infrared reflectance spectroscopy. Drivers of the spatial variability in SOC fractions were determined using the machine learning algorithm cforest. The SOC content and proportions of fractions were predicted with good accuracy (SOC content:R2= 0.87–0.90; SOC proportions: R2= 0.83; ratio of performance to deviation (RPD): 2.4–3.2). The main explanatory variables for the distribution of SOC among the fractions were soil texture, bulk soil C /N ratio, total SOC content and pH. For some regions, the drivers were linked to the land-use history of the sites. Arable topsoils in central and southern Germany were found to contain the highest proportions and contents of stable SOC fractions, and therefore have the lowest vulnerability to SOC losses. North-western Germany contains an area of sandy soils with unusually high SOC contents and high proportions of light SOC fractions, which are commonly regarded as representing a labile carbon pool. This is true for the former peat soils in this area, which have already lost and are at high risk of losing high proportions of their SOC stocks. Those “black sands” can, however, also contain high amounts of stable SOC due to former heathland vegetation and need to be treated and discussed separately from non-black sand agricultural soils. Overall, it was estimated that, in large areas all over Germany, over 30 % of SOC is stored in easily mineralisable forms. Thus, SOC-conserving management of arable soils in these regions is of great importance.


37
There is increasing interest in soil organic carbon (SOC) in agricultural soils, as it contributes to soil 38 fertility and also to mitigation of climate change when organic carbon (OC) sequestration is enhanced 39 (Post and Kwon, 2000). The pathway of atmospheric carbon to SOC is controlled by land use and 40 agronomic management. However, SOC comprises a large range of compounds, ranging from 41 recently added organic matter, such as root litter and exudates, to highly condensed and 42 transformed organic matter that may even be derived from the geogenic parent material. These 43 different compound classes are stabilised in different ways and therefore have different turnover 44 times (Lehmann and Kleber, 2015). Although SOC is now considered as having a continuum of 45 turnover times, it is mostly described and modelled as consisting of different pools that vary in their 46 turnover time (e.g. labile pool, intermediate pool and stabilised pool). The effects of land use and 47 management are not the same for all soil organic matter compounds, however, but differ between 48 SOC pools (Cardinael et al., 2015;Chimento et al., 2016). This is why the different SOC pools need to 49 be assessed separately from the bulk SOC when discussing the influence of land use and 50 management on stabilisation and storage of SOC. 51 One method for experimental quantification of the distribution of SOC among different SOC pools is 52 fractionation. Various fractionation procedures for quantifying SOC fractions have been developed, 53 mostly aiming at isolating fractions with differing turnover times (Poeplau et al., submitted). 54 Determining the distribution of SOC among fractions with assumedly different carbon turnover times 55 is one step towards understanding the factors influencing SOC stabilisation. 56 Some impact factors are consistently reported as being important at site scale for the distribution of 57 SOC among different fractions or pools, one of which is land use. In croplands and grasslands, a 58 similarly large share of bulk SOC is attributed to fractions regarded as stable, while in forest soils a 59 larger proportion of SOC is attributed to more labile SOC fractions (John et  As the oPOM fraction generally contained a small proportion of total SOC, it was combined with the 166 fPOM fraction to give a 'light fraction' for the purpose of prediction. Soil samples were dried at 40°C, 167 sieved through a 2 mm sieve and finely milled in a rotary mill. Before analysis, the samples were dried 168 again at 40°C and equilibrated to room temperature for a few minutes in a desiccator. The soil 169 samples were scanned with spot size 4 cm diameter in a Fourier-Transform near-infrared 170 spectrophotometer (FT-NIRS, MPA -Bruker Optik GmbH, Germany). Spectral data were measured as 171 absorbance spectra (A) according to A = log (1/R), where R is the reflectance expressed in wave 172 number from 11000 to 3000 cm -1 (NIR region) with 8 cm -1 resolution and 72 scans. The final spectrum 173 was obtained by averaging two replicates. 174 To improve the model accuracy a spectral pre-treatment was applied, using Savitzky-Golay first 175 derivative smoothing (3 points) and wavelength selection from 1330 to 3300 nm, since these regions 176 contain the main absorbance information. The calibration set consisted of the 145 calibration site 177 samples, and the remaining samples were used for prediction. Partial least squares regression (PLSR) 178 was performed in the pls package (Mevik et al., 2015), based on near-infrared (NIR) spectra and 179 reference laboratory data. A cross-validation was applied using leave-one-out to avoid over-and 180 under-fitting. To obtain the carbon fractions and ensure that the sum of light and heavy fractions was 181 equal to total SOC content, the log ratio of the light and heavy fraction was predicted (Jaconi et al., in 182 prep.). Model performance was evaluated using the root mean square error of cross-validation 183 (RMSECV), Lin's concordance correlation coefficient (ρc) and the coefficient of determination (R²) 184 between predicted and measured carbon content in the fractions. In addition, residual prediction 185 deviation (RPD) was calculated, using the classification system devised by Viscarra Rossel et al.

Drivers of soil organic carbon distribution in fractions 199
A total of 75 potential drivers of differences in carbon proportions in different fractions was compiled 200 from the soil analysis data, complemented with data from a farm survey and geographical data (for a 201 complete list of predictors, see Table S2). The farm survey related to management practices 202 implemented over the 10 years prior to sampling. Using this information, carbon and nitrogen inputs 203 and outputs were calculated for the sites. When data were missing in the survey responses, yields 204 were calculated using regional yield estimates. Carbon and nitrogen inputs through mineral or 205 organic fertiliser were also calculated based upon the survey data. Climate and site data acquired 206 from GIS data layers completed the set of predictor variables (climate data from Deutscher 207 Wetterdienst, normalised difference vegetation index (NDVI) data from ESA, elevation data from 208

10
The conditional inference forest algorithm (cforest; Hothorn et al., 2006), was used to identify the 213 most influential drivers of SOC distribution among the different fractions. Cforest is an ensemble 214 model and uses tree models as base learners that can handle many predictor variables of different 215 types and can also deal with missing values in the dataset (Elith et al., 2008). The cforest algorithm is 216 similar to the better known random forest algorithm, a non-parametric data mining algorithm that 217 uses recursive partitioning of the dataset to find the relationships between predictor and response 218 variables (Breiman, 2001). 219 Bootstrap sampling without replacement was carried out in order to prevent biased variable 220 importance (Strobl et al., 2007). Ten cforest models were created, each containing 1000 trees and 221 using different random subset generators. From these models, the variable importance of predictors 222 was extracted and the relative variable importance was calculated and averaged over all 10 models. 223 Variables were considered important when their relative variable importance was higher than 100/n, 224 where n is the number of predictors in the model. This is the variable importance that each variable 225 would have in a model where all variables are equally important (Hobley et al., 2015). It should be 226 noted that the relative variable importance value obtained from the cforest algorithm does not 227 necessarily imply direct relationships between the proportion of SOC in the light fraction and the 228 main drivers, as the algorithm also takes into account interaction effects between the variables. 229 Model performance was assessed by the coefficient of determination ( 2 ), as defined by the 230 explained variance of out-of-bag estimates, which represent a validation dataset: 231 where is the mean squared error of out-of-bag estimates and Var z is the total variance in 232 the response variable. The fPOM fraction contributed an average of 23% to bulk SOC (23% ±2.36 (mean ± standard error 244 (SE)) in croplands and 25% ±3.79 in grasslands (Fig. 1). The oPOM fraction accounted for an average 245 of 4% of SOC (3% ± 0.49 in croplands, 8% ±1.26 in grasslands) across all calibration sites (Fig. 1). The 246 heavy fraction contributed the largest proportion to bulk SOC (73% in all soils, 73% ± 2.46 in 247 croplands and 68% ± 4.43 in grasslands). The differences between land uses were not significant. 248 There was great variation in the carbon distribution between the fractions, with the fPOM fraction 249 contributing between 3 and 99% to bulk SOC. The absolute carbon content (g kg -1 ) of the fractions 250 was significantly different for the heavy fraction, with grasslands having significantly higher heavy 251 fraction carbon content than croplands (31 g kg -1 compared with 13 g kg -1 ). 252 There were significant differences in the contribution of the different fractions to bulk SOC 253 depending on the main soil texture class (Fig. 2). In sandy soils, the fPOM fraction contributed 254 significantly more and the heavy fraction contributed significantly less to bulk SOC than in other soils. 255 For the oPOM fraction, the difference between sandy soils and clayey, silty and loamy soils was not 256 significant. The absolute SOC content (g kg -1 soil) was significantly higher in the heavy fraction of 257 clayey soils than in the heavy fraction of all other soil textures and it was significantly higher in the 258 oPOM fraction of sandy soils than in the fPOM fraction of all other soils. 259

Influences on soil organic carbon distribution among fractions (calibration and prediction sites) 260
With the machine-learning algorithm cforest, 75 variables that may act as drivers for the regional 261 distribution of SOC fractions were evaluated (Fig. 3a). For the 'normal' soils (non-black sands) 262 dataset, soil texture had the highest explanatory power in predicting the contribution of the light 263 fraction to bulk SOC (Fig. 4) The analysis of historical land use data of northern Germany confirmed that the former peatland, 270 heathland and grassland sites had significantly higher ((p<0.01) proportions of bulk SOC in the light 271 fraction than sites used as cropland in the same period (Fig. 5a). These historical peatland, heathland 272 and forest sites also had significantly higher (p<0.05) C/N ratio than the historical cropland and 273 grassland sites (Fig. 5b). Regarding the total SOC content, historical peatland and grassland sites had 274 significantly higher (p<0.001) values than historical croplands (Fig. 5c). 275 For the black sands dataset, bulk soil SOC content was the most important driver of SOC distribution 276 in the fractions (Fig. 3b), followed by C/N ratio, soil temperature in summer and soil bulk density. The 277 SOC content had a positive relationship with percentage of SOC in the light fraction, and with C/N 278 ratio (Fig. 4). For soil temperature there was no clear relationship. There was a negative relationship 279 between SOC proportion in the light fraction and soil bulk density. 280

Distribution of soil organic carbon into fractions across Germany 281
Regions featuring high proportions of SOC in the light fraction (over 40%) nearly all lie in northern 282 Germany (Fig. 7). fractions from such regional inventories. In a previous study using paired land-use change sites, the 293 POM proportion was found to be twice as high in grasslands as in croplands (Poeplau and Don, 294 2013b). Even though the fraction distribution did not differ significantly between croplands and 295 grasslands in the present study, there was a trend for slightly higher fPOM content in grasslands than 296 in croplands. The significant differences observed in the SOC content of fractions between different 297 land uses were to be expected, as grassland soils in Germany contain on average more than twice as 298 much SOC in the upper 10 cm as cropland soils (42±16 g kg -1 compared with 17±9 g kg -1 ). 299

Black sands in Germany 300
All samples with medium or high proportions of SOC in the light fraction were found to originate 301 from northern Germany. This is the area in which the black sands are present, which store large parts 302 of their SOC in the light fraction. Springob & Kirchmann (2002a) examined the presence of black 303 sands in Lower Saxony in Germany and linked it to the land-use history. In Ap-horizons of soils 304 formerly used as heathland or plaggen, they found a high fraction of SOC resistant to oxidation with 305 HCl. This HCl-resistant fraction was positively correlated with the total SOC content, but soil microbial 306 biomass carbon content showed a negative relationship with total SOC and, when incubated, the 307 specific respiration rates were lowest for the soils with the highest SOC content (Springob & 308 Kirchmann, 2002a). Those authors concluded that a large proportion of the organic matter in the 309 former heathland soils is resistant to decomposition and suggested that low solubility of the SOC 310 wider range of C/N ratios than control soils without a heathland history. Certini et al. (2015) showed 313 that SOC under heathlands is rich in alkyl C and contains high contents of lipids, waxes, resins and 314 suberin, all of which hinder microbial degradation. This confirms the claim that sandy soils under 315 former heathland and contain high contents of stable SOC even though they also contain a large 316 amount of POM. In such soils, the POM fractions may not be directly linked to higher turnover rates 317 and lower stability. 318 "Historical" peatlands may have lost much of their former carbon stocks due to a number of reasons: 319 Drained peatlands emit huge amounts of CO 2 (German grasslands on average 27.7 to CO 2 ha -1 yr -1 , 320 (Tiemeyer et al., 2016)) until the peat has virtually vanished. There might have also been peat 321 extraction, and the remaining peat layer might have been mixed with underlying sand. Finally, former 322 peatland soils were often mixed with large amounts of sand in order to make them usable for arable 323 cultivation, but still often contain substantial proportions of (degraded) peat and therefore have 324 relatively high SOC content, with a large part of the SOC in the light fraction. It has been found 325 elsewhere (Bambalov, 1999;Ross and Malcolm, 1988;Zaidelman and Shvarov, 2000) that the SOC 326 content in sand-mix cultures declines rapidly after mixing with sand and that the decline increases 327 with increasing intensity of mixing. In a 15-year long-term trial, Bambalov (1999)  peatlands and heathlands are not necessarily distinguishable due to their SOC content and C/N ratio, 343 so that knowledge on the land use history is necessary. In some cases, however, even the distinction 344 on site can be difficult, e.g. on dry peatlands with heath vegetation (Calluna, Erica). In future studies 345 it would therefore be interesting to incubate pairs of former heathland and peatland in order to be 346 able to make accurate claims on the vulnerability of the light fraction SOC in these soils. 347 The presence of black sands poses a problem for interpretation of the SOC fractions. In most cases, 348 the SOC in the light fraction (fPOM + oPOM fractions) is seen as representing a labile carbon pool 349 with short turnover times. Therefore sites with high proportions of bulk SOC in the light fraction 350 would be seen as being at risk of losing this substantial part of their SOC stock quite rapidly and 351 easily. For the black sands, however, their former heathland land use history has led to quite stable 352 and not easily degradable POM (Overesch, 2007;Sleutel et al., 2008;Springob and Kirchmann, 2002), 353 while for former peatland that was drained and possibly mixed with sand the classification of the 354 light fraction into a labile SOC pool may well be justified (Leiber-Sauheitl et al., 2014). This implies 355 that the results need to be interpreted in a different way for black sands than for other soils. 356

'Normal' agricultural soils (non-black sands) 358
The most important driver for the SOC distribution among the fractions in 'normal' soils was the soil 359 texture (Fig. 3a). This is well in line with the frequently reported relationship between clay content 360 and mineral-associated (heavy fraction) SOC, whereby clayey soils can stabilise SOC through 361 mechanisms that protect it against microbial decay by absorption or occlusion (v. Lützow et al., 362 2006). The SOC that is bound to the mineral phase is mostly assigned to a conceptual stable SOC 363 pool. The negative relationship between SOC content and percentage of SOC in the heavy fraction 364 (Fig. 4) may indicate SOC saturation of the mineral fraction at rising SOC content, so that excess SOC 365 can only be stored as particulate organic carbon. 366 The positive correlation between C/N ratio and C proportion in the light fraction (Fig. 4) is related to 367 the inherent higher C/N ratio of the light fraction compared with the heavy fraction, so that a higher 368 share of light-fraction C leads to a higher C/N ratio of the total soil. Thus C/N ratio may be useful as 369 an indicator of SOC stability in 'normal' agricultural soils in Germany. 370 The fact that land use is an important driver for the distribution of SOC among the fractions is mainly 371 due to the fact that topsoils under grassland store a significantly higher share of SOC in the light 372 fraction than topsoils under cropland. This is in line with higher inputs of roots, which make up part 373 of the light fraction, into grassland topsoils. The higher proportion of SOC in the light fraction was 374 also noted in the calibration dataset, but the difference was not significant in that case. 375 Most arable topsoils in Germany do not contain carbonate. The 9% of arable soils that contained 376 over 5% carbonate in this study consistently had a high proportion of heavy-fraction carbon and were 377 therefore classified as containing mainly stabilised SOC (Fig. 4). Calcium bridges may foster 378 absorption of SOC onto mineral surfaces and, via an active soil fauna, high pH enhances the turnover 379 and transformation of SOC from recently added biomass to mineral-associated SOC that can be 380 stabilised via absorption (Oades, 1984). In general, there was a trend for a higher proportion of SOC 381 in the light fraction with lower pH (Fig. 4), which is well in line with the finding by Rousk et al. (2009)

18
The influence of soil type is mainly due to the Podzol soils storing a much higher proportion of bulk 385 SOC in the light fraction than all other soil type classes (Fig. 6). Podzols often develop on sandy soils 386 and therefore do not have a high capacity for SOC stabilisation in the heavy fraction. 387

Black sands 388
In the dataset containing only the black sands, soil total SOC content was the most important driver 389 for the SOC distribution among the fractions, with increasing light fraction with increasing SOC 390 content (Fig. 4). On the one hand, this could indicate saturation of the heavy fraction at high SOC 391 contents, which would lead to further storage in the light fraction only, as already mentioned above 392 for 'normal' soils. Another possible explanation is that those soils with the highest SOC content in the 393 dataset are degraded peatlands, in which a high percentage of the SOC ends up in the light fraction. 394 On former heathlands, the soil total SOC content is also quite high compared with that in other sandy 395 soils and the light fraction is mainly built up from Calluna vulgaris litter, since Calluna vegetation 396 dominates on many heathlands. Calluna litter contains very stable SOC due to high contents of lipids, 397 long-chain aliphatics and sterols, and may persist in the light fraction of soil for decades or even 398 centuries (Sleutel et al., 2008). 399 There is a close link between land-use history as peatland and heathland and soil C/N ratio, with high . Therefore it is evident that land-402 use history is a main driver for the high proportions of bulk SOC found in the light fraction in these 403 soils. This is well in line with the significantly higher C/N ratios reported for soils in Lower-Saxony and 404 Mecklenburg-Western Pomerania, which were under heathland or peatland more than 100 years ago 405 (Fig. 5). The influence of land-use history reinforces the relationship between C/N ratio and the light 406 fraction. 407 In black sands, there was a significant negative relationship between soil temperature and the light-408 fraction SOC proportion, but this was not found for the other soils (Fig. 4) Fig. 1: a)