Soil physico-chemical changes half a century after drainage and cultivation of the former Antela lake (Galicia, NW Spain)

Catena 217 (2022) 106522 Available online 18 July 2022 0341-8162/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Soil physico-chemical changes half a century after drainage and cultivation of the former Antela lake (Galicia, NW Spain) Serafín Gonz´ alez-Prieto * , Marc Romero-Estonllo Misi´ on Biol´ ogica de Galicia, Sede de Santiago, CSIC, Apartado 122, E-15780 Santiago de Compostela, Spain ARTICLE INFO Keywords: Wetland Drainage Carbon Nitrogen Acidification Subsidence ABSTRACT Wetlands in Mediterranean climates undergo large fluctuations in water levels that lead to wet-dry cycles in lake beds, with important effects on sediments and water biogeochemistry. Thus, wetland drainage and ploughing have different effects on soil properties that depend on previous waterlogging periods. However, current knowledge of this topic remains scarce. Fifty-six years after drainage of the Antela lake (Galicia, NW Spain) and characterisation of the main soil properties, in 2019 a stratified sampling - using elevation as proxy for waterlogging length - was done. We checked three hypotheses: a) that changes in the soil pH and organic C were higher in the lake core (permanently waterlogged) than in sediment samples representing four different periods of waterlogging; b) that 13 C and 15 N signatures are useful for tracing differences in the soil C and N dynamics related to the prior duration of waterlogging; and c) that land subsidence in the area previously occupied by peat was higher than predicted immediately after drainage. As hypothesized, we found that drainage and ploughing triggered acidification and C mineralization processes, which were more intense in the previously most waterlogged soils (-0.8 pH H2O units; −0.5 pH KCl units; 20–70% of C lost). In the 2019 samples: a) the highest soil N content (values up to 4-fold higher) and lowest C/N ratio (up to 4 units lower) were recorded in the area that was previously permanently waterlogged; b) soil δ 15 N increased significantly (around 3‰) with elevation (i.e. shorter waterlogging period), while soil δ 13 C did not vary; and c) peat subsidence ranged from 17 to 32 cm (3.0–5.7 mm y −1 ). The main consequence of draining the Antela lake and cultivating the reclaimed land was the opening of water and nutrient cycles, drainage effects being greater than the cultivation ones, and land subsidence 6–8 times higher than initially predicted. 1. Introduction Although wetlands are among the most productive and biodiverse ecosystems and provide invaluable services, these habitats are still being drained, degraded and converted to other uses (Costanza et al., 2014; Ramsar_Convention_on_Wetlands, 2018). During the 20th and early 21st centuries, the rate of wetland loss around the world increased almost fourfold, leading to the disappearance of 70% of the wetlands that existed in 1900 (Davidson, 2014). Throughout Europe, many wetlands have been transformed for different purposes: agriculture, forestry, mining (for peat, sand, gravel) and establishment of industrial and/or urban settlements (ˇ Cíˇ zkov´ a et al., 2013). In Spain around 60–70% of wetlands were lost between 1900 and the 1980s (MITECO, 2020), and coastal and inland wetlands are currently the most deteriorated ecosystems in the country (Spanish_National_Ecosystem_Assessment, 2013). In addition to causing loss of biodiversity, wetland transformation worldwide has also led to intense pressure on water, C and N cycles, leading to loss of soil organic C, increased emissions of greenhouse gases (CO 2 ; NO x ) and land compaction and subsidence, especially during the first years after transformation and in the most disturbed soils (Asada et al., 2005; ˇ Cíˇ zkov´ a et al., 2013; Peacock et al., 2019; Rodriguez et al., 2021; S¨ aurich et al., 2019). The degradation processes are especially worrying in the case of peatlands, which are spatially the most efficient C stores in terrestrial ecosystems, as they contain around 30% of global soil C in only 3% of global land area (Krause et al., 2021; Schwieger et al., 2021). The enhanced mineralisation and transformation processes triggered by drainage have led to the irreversible conversion of peatland Abbreviations: AEMET, Agencia Estatal de Meteorología; GPS, Global Positioning System; IAEA, International Atomic Energy Agency; IMP, Instituto de Meteorología de Portugal; MITECO, Ministerio para la Transici´ on Ecol´ ogica; SOC, soil organic carbon; USAF, US Air Force; WHC, water holding capacity. * Corresponding author. E-mail address: [email protected] (S. Gonz´ alez-Prieto). Contents lists available at ScienceDirect Catena journal homepage: www.elsevier.com/locate/catena https://doi.org/10.1016/j.catena.2022.106522 Received 8 February 2022; Received in revised form 29 June 2022; Accepted 8 July 2022 Catena 217 (2022) 106522 2 into moorsh (secondary transformation of peat) (Szajdak et al., 2020). As in other ecosystems, useful information about the study of the processes underlying organic matter degradation and preservation can be obtained from δ 13 C and δ 15 N isotopic signatures (Andersson et al., 2012; Harris et al., 2018). Besides, as δ 15 N is an integrator of the N cycle (Robinson, 2001), soil δ 15 N can provide useful information on the openness of the N cycle in the soil–plant system. Peatland subsidence is primarily due to settling and secondarily to organic matter oxidation, the former being twice the latter (Ewing and Vepraskas, 2006). Subsidence depend on peat type and density, as well as drainage intensity and land use, although peat loss in drained sites has been observed to be rather similar in permanent grassland and arable cropland (planted with cereals and potatoes) and mainly dependent on drainage status (Kandel et al., 2018). Subsidence rates range from 0.4 to 5.3 cm y −1 around the world [see Grzywna (2017) and references therein] and from 0.2 to 3.7 cm y −1 in temperate zones, with a mean value of around 1.5 cm y −1 (Ikkala et al., 2021). Subsidence rates reported for former muddy lake bottoms range from 1.3 to 5.0 cm y −1 [see Ikkala et al. (2021)]. In wetlands in regions where precipitation is more or less regular over the year, fluctuations in water levels are relatively small, while wetlands in Mediterranean regions are characterized by wide, or even extreme, fluctuations in the water level (ˇ Cíˇ zkov´ a et al., 2013; de Vicente, 2021). As a consequence, large areas of lake beds in Mediterranean wetlands are subjected to wet-dry cycles of variable extent, the effects of which on sediments and water biogeochemistry have scarcely been studied (de Vicente, 2021). Similarly, the effects of drainage and cultivation on soil properties will probably vary depending on the length of the waterlogging period prior to drainage. However, despite the interest of this topic, scarce information is available. Occupying an area of up to 3600–4200 ha in winter-spring and around 1100–1300 ha in summer and of mean depth varying between 0.5 and 2 m, the ancient Antela lake (Ourense, NW Spain; Fig. 1) was one of the largest fresh water lakes in the Iberian Peninsula, as well as the jewel in the crown of the Limia-Antela wetland complex, which covered an area of around 8000 ha in winter and 1500 ha in summer (Villarino et al., 2017). The Antela lake survived several failed drainage attempts throughout the 19th century as well as the Camb´ o Law, of 24 July 1918, which promoted “the drainage and sanitation of lakes, marshes and swampy or flooded land, provided that the area drained or sanitized was greater than100 ha”. However, finally, the “Law of 27 December 1956 on drainage and colonisation of the lake of Antela” declared the area of High National Interest and ordered “the conquest for agricultural and livestock production of the lands of the Antela Lake“. The work began in September 1958, leading to the drainage of the Antela wetland in May 1962, although the channelling works were not finished until 1963 (Zapata Tejedor, 1967). Shortly after the channelling works in the Antela lake (the exact date is not known but is probably in 1962–63 considering the characteristics of the area and the number of samples), the Instituto Nacional de Colonizaci´ on conducted intensive sampling of the topsoil and determined the main soil characteristics: soil type, texture, pH and organic C content (Contreras-Brotons et al., 1969; Fern´ andez-Lavandera, 1967). Largely contrasting with subsidence rates around the world, the preliminary assays conducted shortly after drainage of the Antela lake suggested (total) subsidence “not higher than 4% of peat thickness due to cultivation” (Fern´ andez-Lavandera, 1967), i.e. no more than 2–5 cm considering the reported peat thickness of 40–120 cm. However, no subsequent studies were carried out and therefore there is no information about the changes in the soil properties or land subsidence that occurred after drainage of the area. As both crop rotation and agricultural management are similar throughout the area of the former Antela lake, we hypothesized that changes in soil properties would decrease from the previous lake core (permanently waterlogged before drainage) to the areas of lake beds with increasingly longer and stronger dry periods. To check this hypothesis, we established a total of 106 sampling points in five sectors with different waterlogging periods and we measured the soil variables studied in 1962–63: soil pH in H 2 O and KCl, organic C and, within the former peatland area, the thickness of the remaining peat/moorsh; soil texture was not assessed as no measurable changes were expected in such a flat area. Besides, we also measured soil WHC, organic N, as well as 13 C and 15 N isotopic signatures aiming: a) to evaluate the usefulness of 13 C and 15 N isotopic signatures to monitor differences in the C and N cycles related to the length of the waterlogging period; and b) to Fig. 1. Situation of the drained Antela lake in the centre of the Ourense province (Galicia, NW Spain, SW Europe). Note: a detailed recent topographic map and a bathymetric chart of the former Antela lake, both with the grid of soil sampling points, are provided as Supplementary material. S. Gonz´ alez-Prieto and M. Romero-Estonllo Catena 217 (2022) 106522 3 establish baseline values of soil WHC, N content, C/N ratio and δ 13 C and δ 15 N for future studies on the evolution of the Antela wetland soils. 2. Material and methods 2.1. Study area and experimental design The drained Antela lake was located at an elevation of 615–621 m a. s.l., in the NE sector of the A Limia alluvial floodplain (Ourense, NW Spain), between latitude 42◦04′46′’ and 42◦09′44′′ North and longitude 7◦37′56′′ and 7◦44′40′′ West (Fig. 1). The A Limia floodplain has ‘Csb’ temperate climate with dry and warm summer, also called Mediterranean oceanic climate, according to K¨ open-Geiger classification (AEMETIMP, 2011; Beck et al., 2018). Mean annual temperature is 11.5 ◦C, with wide differences between the coldest (January: mean 5.2 ◦C; mean of the minimum −0.2 ◦C; mean of the maximum 10.5 ◦C) and the warmest month (July: mean 19.4 ◦C; mean of the minimum 11.2 ◦C; mean of the maximum 27.6 ◦C). There are 81 d y −1 with minimum temperature ≤ 0 ◦C and 83 d y −1 with maximum temperature ≥25 ◦C. Mean annual precipitation is 952 mm, with 60.0% of the rain between October and February, being December the rainiest month (140 mm) and July and August the driest (23 mm) (https://agroclimap.aemet.es/#). In 1962–63, the Instituto Nacional de Colonizaci´ on performed systematic sampling (n =543) of the topsoil (0–30 cm) in the Antela wetland, following a grid of around 250 ×250 m. The equivalence in the World Reference Base for soil resources (IUSS Working Group, 2014) of the catena of soil types mapped in the ancient lake by Fern´ andez-Lavandera (1967) ordered by decreasing altitude (and increasing waterlogging period) is the follows: Leptic Podzols (316 ha), Gleyic Fluvisols (745 ha), Gleyic Podzols (949 ha) and Histosols (Subaquatic Histosols, 969 ha; Dystric/Eutric Histosols, 264 ha). Most of the soils were sandy (66.4%), and the other were quite similarly distributed between silty (17.9%) and clayey (15.7%) textures (Contreras-Brotons et al., 1969). The latter authors highlighted the fact that the soils were acidic (pH H2O : mean 5.81, range 4.70–7.70; pH KCl : mean 4.68, range 3.92–6.67) and had a very variable organic C content (mean value: 8.36% C; range: 0.28–37.52% C; measured with the Walkley-Black method which underestimate the soil organic C content). After examination of the aerial photographs taken by the USAF before the channelling works (the so-called “1956-57 American fly”, https://mapas.xunta.gal/visores/descargas/), the bathymetric chart of the Antela lake (with 0.5 m depth contours) constructed for the drainage project (Anonymous, 1958) and the map of soil types (Fern´ andez-Lavandera, 1967), Villarino et al. (2017) concluded that the level of permanent waterlogging was around 617.1–617.2 m a.s.l. (where most paths disappeared), while that of prolonged waterlogging was around 618.0 m and the rest of the surface was only temporarily waterlogged. Accordingly, we established five intervals of increasing height above sea level (elevation) and decreasing duration of waterlogging: <617.0 m, 617.0–617.5 m, 617.5–618.0 m, 618.0–619.0 m and >619.0 m. Although these five altitude intervals are small, they are strongly related with the previous lengths of the waterlogging period and, consequently, with the five main soil types mapped by Fern´ andez-Lavandera (1967). There is not detailed information available on management and fertilization for the plots sampled, but the rotation (cereal-potatoes) is similar throughout the study area due to soil and climatic restrictions for agriculture; only in the last three-four years potatoes were replaced by maize as summer crop in some plots. Both mineral and organic fertilizers (mostly poultry manure and swine slurry) are applied; the usual doses are 30 m 3 ha −1 for organic fertilizers and 180 kg N ha −1 , 100 kg P 2 O 5 ha −1 and 300 kg K 2 O ha −1 for mineral fertilizers (L´ opez-Mateo, 2007). By combining the map derived from the 1962–63 soil sampling (Contreras-Brotons et al., 1969) with the bathymetric chart of Antela (Anonymous, 1958), we found the following distribution of soil samples per height interval: <617.0 m, 37.8%; 617.0–617.5 m, 19.9%; 617.5–618.0 m, 15.6%; 618.0–619.0 m, 11.9%; and >619.0 m, 14.8%. These intervals were used in September 2019 for sampling 106 plots similarly distributed per height interval as in the 1960 sampling (See Table S1 and Fig. S1 in the Supplementary material). 2.2. Soil sampling and chemical analyses Soil sub-samples (0–30 cm depth layer) were obtained in September 2019 with a stainless steel probe (4 cm internal diameter) placed at least 10 m away from the plot borders in the four corners of a 10 ×10 m quadrat. The centre of the quadrat was geo-referenced with a GPS (Supplementary Material, Table S1). The sub-samples were combined to make a composite sample, which was thoroughly homogenised, airdried and sieved (<2 mm). In the 23 sampling points within the former peatland area, the same stainless steel probe was used to assess the thickness of the remaining peat/moorsh and to compare it with the pre-drainage values [40–120 cm thick (Fern´ andez-Lavandera, 1967)]. The dry matter content was determined by oven-drying sub-samples of the soil at 105 ◦C for 5 h, while the soil WHC was measured in a Richards’ membrane-plate extractor (Soil Moisture Equipment, USA) at a pressure of 10 kPa. Soil pH in H 2 O and 2 M KCl (1:2.5 w/v ratio) was measured in air-dried samples with a pH meter (Metr¨ ohm, Switzerland). Aliquots of the composite samples were finely ground (<100 μ m) in a planetary ball mill (Retsch PM100, Germany) and weighed into tin capsules for elemental (organic C and N) and isotopic (natural abundances of 13 C and 15 N in δ notation) in an elemental analyser (CNS 1508, Carlo Erba Instruments, Italy) coupled on-line to an isotopic ratio mass spectrometer (Finnigan Mat, delta C, Germany). To check the accuracy of the determinations, an elemental reference material (Soil-3 from Eurovector, Italy) and two isotopic standards (IAEA-CH-6 and IAEA-CH7 for δ 13 C; IAEA-N1 and IAEA-N2 for δ 15 N, from the International Atomic Energy Agency, Austria) were included in each set of 10 samples. When necessary, drift correction was made against internal standards during the run. The isotopic values are expressed as relative deviations from the respective international standard (VPDB, for 13 C; N 2 -air for 15 N) in δ-notation. 2.3. Statistical analyses Initially, exploratory data analysis was carried out to detect outliers and anomalies that could affect the results, as well as to check the normality of the data distribution (Shapiro-Wilk’s W test) and equality of variance among groups (Levene’s test); when these assumptions were not fulfilled, the original data (Supplementary Material, Table S1) were transformed by Tukey’s ladder of powers. Data on pH H2O , pH KCl and organic C content were analysed by two-way ANOVA, with sampling year (1962–63 or 2019) and elevation (<617.0 m, 617.0–617.5 m, 617.5–618.0 m, 618.0–619.0 m and >619.0 m) as factors. Data on WHC, total N, δ 13 C and δ 15 N (only recorded in 2019) were analysed by one-way ANOVA, with elevation as the factor. Significant differences (at P <0.05) between the group means were established using the Bonferroni test for multiple comparisons. The proportion of the variation accounted for each factor or interaction in the ANOVA was determined by the partial eta-squared ( η p 2) statistic. Statistical procedures were performed with SPSS 25.0 for Windows. 3. Results Both the sampling year and elevation had a significant influence (p <0.001) on soil pH H2O , but only explained 6.7–6.8% of the variance, i.e. less than the 9.4% explained by the year*elevation interaction. Shortly after the lake was drained, the highest pH H2O was recorded in soils at an elevation of 617.0–617.5 m (mean value pH =6.0). The pH H2O decreased steadily with elevation up to 0.5 pH units, whereas in 2019 the pH H2O increased with elevation between 617.0 m and 619.0 m, and stabilized thereafter (Fig. 2A). Fifty-six years after drainage of the lake and cultivation of the reclaimed land, soil pH H2O varied only little in S. Gonz´ alez-Prieto and M. Romero-Estonllo Catena 217 (2022) 106522 4 plots located at >618.0 m a.s.l., while below this elevation significant acidification of up to 0.8 pH units occurred (Fig. 2A). The soil pH KCl followed a very similar trend to that of pH H2O , except for the lower proportion of the variance explained by the sampling year (2.7%) and the lower acidification (up to 0.5 pH units), which was restricted to soils below 617.5 m (Fig. 2B). The analysis of outliers conducted for the present study showed that the soils with extremely low organic C contents in the 1960s sampling were located close to the main drainage channels and were probably disturbed by the channelling works and mixture with sandy subsoil sediments; after excluding these outliers, the soil C content ranged from 0.72 to 37.52% (n =531). The ANOVA-2 showed that both studied factors and their interaction significantly influenced the soil organic C content (p <0.01 to p <0.001). Sampling year explained 3.6%, elevation 22.0%, and the year*elevation interaction 2.6% of the total variance. Immediately after drainage of the Antela lake, the soil organic C content was highest in plots below an elevation of 617.0 m a.s.l., with a mean value around 140 g kg −1 , i.e. 2–3 times higher than in the other soils which did not differ significantly from each other (Fig. 2C). Between the early 1960s and 2019, the soil organic C content decreased significantly in all soils below 618.0 m a.s.l. (-21 to −70%), but it did not decrease significantly in plots at elevations of 618.0–619.0 m a.s.l. and did not vary in the other plots (Fig. 2C). Two soil groups were differentiated by the total N content in the soil in 2019: those from elevations below 617.0 m a.s.l., with around 8 g kg −1 , and the others, with values four times lower (p <0.05 to p < 0.001) and relatively homogeneous (Fig. 3A), with elevation explaining 35.4% of the variance in total N. In 2019, the soil C/N ratio increased steadily with plot elevation, which explained 34.8% of the variance in the C/N ratio, with highly significant differences (p <0.001) between the soils below 617.5 m (C/ N ratio: 12–14) and those above 619.0 m a.s.l. (C/N ratio: 16–18; Fig. 3B). Regarding the δ 13 C isotopic signature in the soils sampled in 2019, the values ranged from −26.7 to −28.2‰ with no differences between soils grouped according to elevation (Fig. 3C). Conversely, the δ 15 N isotopic signature increased significantly (p <0.001) from soils located at <617.0 m (δ 15 N around 4‰) to soils located at 618.0 m a.s.l. (around 5.4‰) and those at 619.0–620.0 m a.s.l. (around 6.7‰), with elevation explaining 21.2% of the variance in δ 15 N (Fig. 3D; raw data available in Table S1). In the 2019 sampling, elevation explained 43.4% of the variance in soil water holding capacity, which was significantly higher (p <0.001) in the soils at the lowest altitude (680 g H 2 O kg −1 d.w.) than in the other soils (around 240–300 g H 2 O kg −1 d.w.) (Fig. 3E). In the area previously occupied by peat, the thickness of the organic layer currently ranges from 23 to 88 cm (mean =43 cm; median =43 cm; st.dev. =14 cm; kurtosis =3.246; skewness =1.289; n =23). As there are no data available on bulk density of soils sampled in 1962–63, accurate assessment of soil organic C losses triggered by drainage and cultivation is not possible. This is especially true for the former peatland area, as peat is very sensitive to compaction and subsidence. However, for the other soil types (sandy and with much lower organic C content) to assume no significant changes in bulk density between both sampling dates could be acceptable, allowing tentative estimates for the amounts of C lost: 166 Mg ha −1 in soils at 617.0–617.5 m, 134 Mg ha −1 in soils at 617.5–618.0 m, 63 Mg ha −1 in soils at 618.0–619.0 m, and 5.5 Mg ha −1 in soils above 619.0 m). 4. Discussion The small change in the pH of soils above 617.5–618.0 m a.s.l. (i.e. those previously with the shortest period of waterlogging) and the significant decrease (up to 0.5 pH KCl units and 0.8 pH H2O units) at lower elevations between the 1960s and 2019 clearly suggest that drainage was a more important factor than cropping in relation to soil Fig. 2. Mean pH in H 2 O (A) and KCl (B), and mean organic C content (C) of soils grouped according to elevation and sampling year. Different lowercase letters show significant differences between sampling years for the same elevation, while different uppercase letters show significant differences among elevations for the same sampling year (p <0.05). S. Gonz´ alez-Prieto and M. Romero-Estonllo Catena 217 (2022) 106522 5 acidification. This result agrees with the fact that agricultural practices do not differ between altitude sectors, so no differences should be expected among sectors in the input–output budget of basic cations (and related acidification) with crops. Conversely, the effect of drainage on soil hydrology was expected to differ between altitude sectors because the permanently waterlogged area of the Angela lake was previously a semi-endorheic basin (Villarino et al., 2017), while the temporarily flooded belts were increasingly exorheic with altitude. Consequently, the increase of basic cations leaching (and subsequent acidification) triggered by drainage was likely higher in soils from the permanent lake than in the seasonally flooded ones. The absolute (g kg −1 d.w.) and relative (% of initial C) decreases in the organic C content were highest in the Antela soils located at an elevation of 617.0–617.5 m a.s.l., followed by soils at elevations below 617.0 m and those at 617.5–618.0 m, with non-significant or zero depletion of soil organic C above 618.0 m. However, it must be highlighted that the 1962–63 measurements were based on the WalkleyBlack method, which is known to underestimate the soil organic C Fig. 3. Mean organic N content (A), C/N ratio (B), δ 13 C (C), δ 15 N (D) and water holding capacity (E) of soils sampled in 2019 grouped according to elevation. Different letters indicate significant differences among groups (p <0.05). S. Gonz´ alez-Prieto and M. Romero-Estonllo Catena 217 (2022) 106522 6 content (Chen et al., 2015; Mikhailova et al., 2003). Therefore, the observed differences in organic C content are conservative estimates and may actually be larger than inferred from the comparison between old and new analyses. This trend is consistent with the fact that wetland drainage for agricultural purposes usually leads to strong mineralisation of soil organic C (Bini and Zilocchi, 2004; Grzywna, 2017; Kandel et al., 2018; Krause et al., 2021; S¨ aurich et al., 2019). With the available historical information, it is not possible to elucidate whether the lower C losses in soils below 617.0 m than in those between 617.0 and 617.5 m a. s.l. are real or apparent. On the one hand, most soils below 617.0 m are Histosols that had an organic layer 40–120 cm thick before draining; thus, C losses could be (partially) masked until peat thickness fell below the sampling depth of 30 cm. On the other hand, soils below 617.0 m are still temporarily waterlogged during rainy periods in winter-spring, reducing organic C losses by mineralization because soil respiration rates and peat loss mainly depend on the drainage status (Kandel et al., 2018). Moreover, relative to soils below 617.0 m, those between 617.0 and 617.5 m will be more degraded after drainage and, therefore, specific basal respiration will be higher, as previously observed in other drained organic soils used for agricultural purposes (S¨ aurich et al., 2019). The lack of historical data on N content and WHC of the Antela soils precludes any comparison with the present values and analysis of the changes after drainage and cultivation. As usual in sandy soils (Gonz´ alez-Prieto and Villar, 2003), the WHC is strongly correlated with the soil organic C (r 2 =0.83) and, thus, the observed trend in the studied soils was very similar to that previously discussed for C. Present levels of soil organic N in the area occupied by Histosols are at the lower end of the range reported for peatlands (0.9–2.7% d.w.) which depends on vegetation assemblages and nutrient status during peat formation (Andersson et al., 2012). The organic N contents of soils from other areas of the Antela lake system are well within the range reported for agricultural soils in NW Spain (Gonz´ alez-Prieto et al., 1992). The increase of up to 4 units in the soil C/N ratio with elevation is consistent with the changes in organic C levels between the 1960s and 2019, as already mentioned, and also with the fact that drainage enhances C rather than N mineralization (S¨ aurich et al., 2019). The low C/ N ratio (12–14) in the plots from the area characterised by Histosols contrasts with the range (20–60) in relatively undegraded peats (Andersson et al., 2012) and again indicates intense mineralization triggered by the hydrological and agricultural management. The narrow range of δ 13 C isotopic signature (-26.7 to −28.2‰) in all 2019 soil samples from the ancient Antela lake contrasts with the wide range (-20 to −35‰) found in terrestrial C3 plants (Dawson and Siegwolf, 2007) and also in peats (−20‰ to −30‰), in which the least and most negative values are associated with Sphagnum moss (receiving heavily 13 C depleted CO 2 from symbiotic methanotrophic bacteria) and vascular plant remains, respectively (Alewell et al., 2011; Andersson et al., 2012; Loisel et al., 2009; Nyk¨ anen et al., 2018). The lack of differences in δ 13 C related to elevation, i.e. former waterlogging period, in the current Antela soils can be explained by two, not mutually exclusive, hypotheses. The first is that peats in the Antela lake were mainly formed by vascular plant remains rather than Sphagnum moss. The second is that SOM mineralization, enhanced by more than half a century of continued drainage and intensive cultivation, has led to the δ 13 C values becoming more similar, blurring the differences between peats and the other soil types. The latter explanation is supported by the fact that aerobic mineralization leads to preferential release of 12 C (Alewell et al., 2011) and, consequently, the degree of degradation in drained peatlands is usually associated with higher δ 13 C (Drollinger et al., 2019; Minick et al., 2019). The lack of historical information on soil δ 15 N prevents to draw sound conclusions about the current trends of δ 15 N in the altitudinal catena, but the following tentative explanations can be done. As the lowest parts of the former Antela lake are still frequently waterlogged in the winter-spring season, the significant increase (up to 3‰) in the soil δ 15 N isotopic signature with elevation is consistent with the 15 N enrichment observed during incubation of Sphagnum litter (Asada et al., 2005). Increased δ 15 N values (up to 7‰) have also been found in peat degraded after drainage (Drollinger et al., 2019; S¨ aurich et al., 2019), mainly due to the isotopic fractionation during microbial decomposition (Andersson et al., 2012). On the other hand, higher values of soil δ 15 N are usually associated with a more ’open’ N cycle in the soil–plant system (H¨ ogberg, 1997). Therefore, the increasing 15 N enrichment with increasing elevation may also be related to greater N losses, suggesting that NO 3 leaching in the soils at higher elevations may be a more important process than denitrification at lower elevations, which are more prone to seasonal waterlogging. Besides, differences in soil δ 15 N could be also related with differences in N inputs and outputs, although there are no evidences of contrasting N fertilization (organic vs inorganic) or N outputs with crops between elevation sectors. In the early 1960s, peat thickness in the permanently waterlogged area of the Antela wetland ranged from 40 to 120 cm (Fern´ andez-Lavandera, 1967) (mean value and standard deviation unknown), while a peat profile of depth 100 cm was described for the ancient Antela lake (Guiti´ an-Ojea and Carballas, 1982). Therefore, the lower and upper values of peat thickness decreased by respectively 17 cm and 32 cm in 56 years, i.e. 0.30–0.57 cm y −1 , around 6–8 times higher than predicted in the preliminary assays conducted by Fern´ andez-Lavandera (1967). However, the subsidence rates calculated in the present study are within, but in the lower part, of the ranges usually reported for peatlands and muddy lake bottoms in boreal and temperate zones: 0.3 cm y −1 in the Netherlands (ˇ Cíˇ zkov´ a et al., 2013), 0.27–3.09 cm y −1 in the UK (Peacock et al., 2019), 0.35–2.1 cm y −1 in Poland and Germany (Grzywna, 2017), 1.4 cm y −1 in Canada (Van Seters and Price, 2002) and 0.5–5.0 cm y −1 in Finland (Ikkala et al., 2021). Not surprisingly, taking into account the differences in peat depth and climatic conditions, the rates of subsidence in the former Antela lake were much lower than those found in Mediterranean zones (2.1–2.8 cm y −1 in the Venice lagoon, Italy; 1.0–7.5 cm y −1 in the Sacramento delta, California, USA) and in tropical-subtropical regions (usually 2–6 cm y −1 in Florida, Malaysia and Kalimantan, but up to 1 m in the first year) [see Grzywna (2017) and references therein]. In summary, the effects of drainage and cultivation on the main properties of soils from the former Antela lake were higher in the permanently waterlogged areas (i.e. below 617.0 m) than in those seasonally subjected to wet-dry cycles (i.e. above 617.0 m). 5. Conclusions As hypothesized, changes in soil pH (acidification) and organic C content (losses) decreased from the permanently waterlogged area to the former lake belts with increasingly longer dry periods. The increase of δ 15 N with elevation points to an increasingly ‘open’ N cycle with NO 3 leaching in better drained soils (i.e. those at high elevation) being more intense than denitrification in soils still prone to waterlogging despite the drainage channels (i.e. those at lower elevations); however, specific studies on this topic are needed to confirm this relationship. Despite being 6–8 times higher than predicted in preliminary assays carried out after drainage, the subsidence rate recorded is in the lower part of the range reported for similar wetlands. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability All raw data are provided as Supplementary material S. Gonz´ alez-Prieto and M. Romero-Estonllo Catena 217 (2022) 106522 7 Acknowledgements The authors thank Mariana Loureiro and Mari´ an de Jesús for laboratory assistance and F´ atima Mui˜ no, Sergio L´ opez and Ana Iglesias for assistance with fieldwork. Appendix A. Supplementary material Supplementary data to this article can be found online at https://doi. org/10.1016/j.catena.2022.106522. References AEMET-IMP, 2011. Iberian climate atlas. Air temperature and precipitation (1971-2000). Ministerio de Medio Ambiente y Medio Rural y Marino, Madrid. Alewell, C., Giesler, R., Klaminder, J., Leifeld, J., Rollog, M., 2011. Stable carbon isotopes as indicators for environmental change in palsa peats. Biogeosciences 8, 1769–1778. Andersson, R.A., Meyers, P., Hornibrook, E., Kuhry, P., Morth, C.-M., 2012. Elemental and isotopic carbon and nitrogen records of organic matter accumulation in a Holocene permafrost peat sequence in the East European Russian Arctic. J. Quat. Sci. 27, 545–552. Asada, T., Warner, B., Aravena, R., 2005. Effects of the early stage of decomposition on change in carbon and nitrogen isotopes in Sphagnum litter. J. Plant Interact. 1, 229–237. Beck, H.E., et al., 2018. Present and future K¨ oppen-Geiger climate classification maps at 1-km resolution. Sci. Data 5, 180214. Bini, C., Zilocchi, L., 2004. Soil evolution in recently reclaimed wetland areas of Northeast Italy. Agrochimica 48, 213–221. Chen, L., et al., 2015. A comparison of two methods for quantifying soil organic carbon of alpine grasslands on the Tibetan Plateau. PLoS ONE 10, e0126372. ˇ Cíˇ zkov´ a, H., et al., 2013. Actual state of European wetlands and their possible future in the context of global climate change. Aquat. Sci. 75, 3–26. Contreras-Brotons, F., Guiti´ an-Ojea, F., M´ endez-Dom´ enech, E., 1969. Las necesidades de cal en los suelos de la Laguna de Antela. Lugo, Spain. Costanza, R., et al., 2014. Changes in the global value of ecosystem services. Global Environ. Change 26, 152–158. Davidson, N.C., 2014. How much wetland has the world lost? Long-term and recent trends in global wetland area. Mar. Freshw. Res. 65, 934–941. Dawson, T.E., Siegwolf, R.T.W., 2007. Using Stable Isotopes as Indicators, Tracers, and Recorders of Ecological Change: Some Context and Background. In: Todd, E.D., Rolf, T.W.S. (Eds.), Terrestrial Ecology. Elsevier, pp. 1–18. de Vicente, I., 2021. Biogeochemistry of mediterranean wetlands: a review about the effects of water-level fluctuations on phosphorus cycling and greenhouse gas emissions. Water 13, 1510. Drollinger, S., Kuzyakov, Y., Glatzel, S., 2019. Effects of peat decomposition on δ 13 C and δ 15 N depth profiles of Alpine bogs. Catena 178, 1–10. Ewing, J., Vepraskas, M., 2006. Estimating primary and secondary subsidence in an organic soil 15, 20, and 30 years after drainage. Wetlands 26, 119–130. Fern´ andez-Lavandera, O., 1967. La aptitud para el cultivo de los suelos de la Laguna de Antela. Instituto Nacional de Colonizaci´ on, Ministerio de Agricultura, Madrid, Spain. Gonz´ alez-Prieto, S.J., Villar, M.C., 2003. Soil organic N dynamics and stand quality in Pinus radiata pinewoods of the temperate humid region. Soil Biol. Biochem. 35, 1395–1404. Gonz´ alez-Prieto, S.J., Villar, M.C., Carballas, M., Carballas, T., 1992. Nitrogen mineralization and its controlling factors in various kinds of temperate humid-zone soils. Plant Soil 144, 31–44. Grzywna, A., 2017. The degree of peatland subsidence resulting from drainage of land. Environ. Earth Sci. 76. Guiti´ an-Ojea, F., Carballas, T., 1982. Suelos naturales de la provincia de Orense. CSIC, Santiago de Compostela. Harris, P., et al., 2018. Effect of long-term drainage on plant community, soil carbon and nitrogen contents and stable isotopic (δ 13 C, δ 15 N) composition of a permanent grassland. Eur. J. Soil Sci. 69, 48–68. H¨ ogberg, P., 1997. Tansley review No. 95–15N natural abundance in soil-plant systems. New Phytol. 137, 179–203. Ikkala, L., Ronkanen, A.-K., Utriainen, O., Klove, B., Marttila, H., 2021. Peatland subsidence enhances cultivated lowland flood risk. Soil Tillage Res. 212. IUSS Working Group, W.R.B., 2014. World Reference Base for Soil Resources 2014. International soil classification system for naming soils and creating legends for soil maps. FAO, Rome. Kandel, T.P., Laerke, P.E., Elsgaard, L., 2018. Annual emissions of CO 2 , CH 4 and N 2 O from a temperate peat bog: Comparison of an undrained and four drained sites under permanent grass and arable crop rotations with cereals and potato. Agric. For. Meteorol. 256, 470–481. Krause, L., et al., 2021. Impacts of historical ditching on peat volume and carbon in northern Minnesota USA peatlands. J. Environ. Manage. 296, 113090. Loisel, J., Garneau, M., Helie, J.-F., 2009. Modern Sphagnum delta C-13 signatures follow a surface moisture gradient in two boreal peat bogs, James Bay lowlands, Quebec. J. Quat. Sci. 24, 209–214. L´ opez-Mateo, M.C., 2007. Efectos agron´ omicos y ambientales de la fertilizaci´ on en el cultivo de patata en A Limia (Ourense). University of Santiago de Compostela, p. 411 pp.. Mikhailova, E.A., Noble, R.R.P., Post, C.J., 2003. Comparison of Soil Organic Carbon Recovery by Walkley-Black and Dry Combustion Methods in the Russian Chernozem. Commun. Soil Sci. Plant Anal. 34, 1853–1860. Minick, K.J., Mitra, B., Li, X., Noormets, A., King, J.S., 2019. Water table drawdown alters soil and microbial carbon pool size and isotope composition in coastal freshwater forested wetlands. Frontiers in Forests and Global Change 2, 7. MITECO, 2020. La restauraci´ on y conservaci´ on de humedales, prioridad del MITECO. https://www.miteco.gob.es/es/prensa/ultimas-noticias/la-restauraci%C3%B3n-yconservaci%C3%B3n-de-humedales–prioridad-del-miteco/tcm:30-506797. Nyk¨ anen, H., Mpamah, P.A., Rissanen, A.J., 2018. Stable carbon isotopic composition of peat columns, subsoil and vegetation on natural and forestry-drained boreal peatlands. Isot. Environ. Health Stud. 54, 622–641. Peacock, M., et al., 2019. The full carbon balance of a rewetted cropland fen and a conservation-managed fen. Agric. Ecosyst. Environ. 269, 1–12. Ramsar_Convention_on_Wetlands, 2018. Global Wetland Outlook: State of the World’s Wetlands and their Services to People. Ramsar Convention Secretariat, Gland, Switzerland. Robinson, D., 2001. δ15 N as an integrator of the nitrogen cycle. Trends Ecol. Evol. 16, 153–162. Rodriguez, A.F., Daroub, S.H., Gerber, S., Jennewein, S.P., Singh, M.P., 2021. Water management effect on soil oxidation, greenhouse gas emissions, and nitrogen leaching in drained peat soils. Soil Sci. Soc. Am. J. 85, 814–828. S¨ aurich, A., et al., 2019. Drained organic soils under agriculture — The more degraded the soil the higher the specific basal respiration. Geoderma 355, 113911. Schwieger, S., et al., 2021. Wetter is better: rewetting of minerotrophic peatlands increases plant production and moves them towards carbon sinks in a dry year. Ecosystems 24, 1093–1109. Spanish_National_Ecosystem_Assessment, 2013. Ecosystems and biodiversity for human wellbeing: Synthesis of key findings. Biodiversity Foundation of the Spanish Ministry of Agriculture, Food and Environment, Madrid, Spain. Szajdak, L.W., Jezierski, A., Wegner, K., Meysner, T., Szczepanski, M., 2020. Influence of drainage on peat organic matter: implications for development, stability, and transformation. Molecules 25. Van Seters, T., Price, J., 2002. Towards a conceptual model of hydrological change on an abandoned cutover bog, Quebec. Hydrol. Process. 16, 1965–1981. Villarino, A., Gonz´ alez-Prieto, S.J., B´ arcena, F., 2017. Aves da Limia, dende a lagoa de Antela ´ os nosos días. Centro de Estudios da Limia, Ourense, Spain. Zapata Tejedor, F., 1967. Desecaci´ on y saneamiento de la laguna de Antela. Revista de Obras Públicas 115, 465–472. S. Gonz´ alez-Prieto and M. Romero-Estonllo