Proglacial valleys or glacial forefields form where glaciers have retreated. Since 1850, 60 % of glacier volume in the European Alps has been lost, and many small glaciers will disappear completely in the next decades (Cannone et al., 2008). Owing to prolonged glacial scouring, the newly revealed surfaces are typically composed of bedrock or glacial deposits, with no evidence of prior soil development. Over time, these surfaces are colonized by microbes, mosses and vascular plants and show soil development. Proglacial areas can be used for space-for-time studies, where the soil age is calculated from glacial retreat maps. The age information of these so-called chronosequences can be used to study natural processes over time without the need for experiments and incubations.

The buildup of organic carbon in proglacial soils is a net result of in- and outfluxes of particulate, dissolved and gaseous carbon compounds. Previous studies on soil carbon along proglacial chronosequences have focussed primarily on the particulate and dissolved contributions (e.g. Eckmeier et al., 2013; Egli et al., 2010; Smittenberg et al., 2012). Studies on gaseous carbon fluxes have predominantly focussed on aquatic systems inside proglacial valleys (e.g. Huang et al., 2025; St Pierre et al., 2019) and are geared towards research on methane rather than CO₂.

Gaseous carbon fluxes consist predominantly of CO₂ in these young soils, as methane and volatile carbon compounds are only produced in significant quantities in more developed soils. The one previous study on proglacial CO₂ fluxes that we are aware of, reports CO₂ effluxes from soil respiration (Guelland et al., 2013a). Soils can however also take up atmospheric CO₂. The majority of CO₂ uptake by soils is a biological process, performed by microbial and plant communities. These consume atmospheric CO₂ either via photosynthetic or chemoautotrophic pathways and form cell biomass or other carbon compounds, which over time contribute to the formation of soil carbon.

What determines the CO₂ uptake rates in proglacial soils is not well understood. Several factors may limit effective carbon metabolisms in both the microbial and plant communities, such as macro- and micronutrient limitations, temperature, light exposure, physical disturbances, and water availability. Often, several of these factors are lumped together in the soil age. What exactly contributes to the soil age being an explanatory factor, and whether this holds for different locations within a proglacial valley, is often not further explored, nor is the interaction with CO₂ uptake rates. Recent work on proglacial areas and their carbon cycling has predominantly been focussed on Asian glacier regions (Khan et al., 2023; Li et al., 2022) or Arctic regions (Kleber et al., 2025; Tian et al., 2023), which however have a different climatic regime than the European Alps. We therefore aim to enhance the knowledge on European alpine proglacial regions. Furthermore, we sampled both water-saturated and non-saturated (dry) soils in proglacial areas. Up to now, proglacial wetlands have been ignored or purposely left out of any proglacial soil dataset, because water-influenced soils were considered too disturbed to be taken into a chronosequence approach. We however show that proglacial wetlands are hotspots of atmospheric CO₂ uptake and carbon storage and that trends with age can be observed for both wetland and dry soil locations.

Given the well documented glacial retreat in the proglacial area of the Bachfallenferner (Fig. 1) (Raup et al., 2007), each location in this area can be given an age via linear interpolation between recorded glacial extents, making it particularly suitable for a space-for-time study approach like was used here. For clarity, we use the word “soil” for the surface material at each location, also for undeveloped deposits without signs of pedogenesis. “Dry” is used for each location that is not water-saturated, regardless of its water content. Younger soil locations are generally characterized by a lack of vascular plant and moss cover, whereas the older locations have a lush appearance with abundant grasses and other plants (Fig. B1). The first moss coverage is already observed in the first decade, although the percentage surface cover remains low (Fig. B2). Grasses and other plants appear from ca. 50 years onwards (Fig. B2). A similar pattern is observed in other proglacial areas (Bernasconi et al., 2011), although some regions show the development of a rich vascular plant cover much earlier (Burga, 1999; Li et al., 2022).

In this study, we sampled and measured in both dry and water-saturated (wetland) locations. Previous work on carbon and/or CO₂ cycling in proglacial areas has been focussed on either streams and lakes (e.g. Huang et al., 2025; St Pierre et al., 2019; Yan et al., 2023) or on dry soils (e.g. Bernasconi et al., 2011; Egli et al., 2010). Generally, the presence or absence of wetlands in the proglacial valley is not mentioned, although Bernasconi et al. (2011) very briefly mention they observed groundwater seepage in sampling locations that have a higher clay content than the average. Given the high input of glacial meltwater, snowmelt, and precipitation, proglacial areas are very likely to contain and sustain wetlands, both near streams and in small, local depressions or groundwater seepage locations. Further research, preferably using remote sensing, is required to quantify the surface area of wetlands in proglacial areas and to upscale our findings. Our first estimate of the wetland surface area in the Bachfallenferner area is 2 %–5 % of the total proglacial area (including lakes and bedrock), but whether this is similar for other proglacial areas, has thus far not been investigated.

Microbial and/or plant communities can use atmospheric CO₂ either via photosynthetic or chemoautotrophic pathways. We report CO₂ uptake in most of the measured locations (Fig. 2). All our measurements, in both wet and dry soils, were done during daytime. In wetland locations, the net CO₂ flux was negative (indicating uptake) in 95 % of the measurements. In the dry locations, 18 % of the fluxes were positive, mostly in soils up to 100 years old. In many soil locations in the proglacial area of the Bachfallenferner, the plant community is limited (<50 % plant surface in 60 % of the dry soil locations, and in 38 % of the wetland locations, Fig. B2). It is therefore likely that the soil microbial community plays a key role in the CO₂ uptake. Earlier studies on the microbial community in proglacial environments showed that chemolithoautotrophy (non-light dependent CO₂ uptake) was an important trait in the presented microbial communities (Tian et al., 2023). Research in very young deglaciated soils revealed that carbon fixating microbial genes could be found in soils directly after deglaciation (0 years old), with an increase in copy numbers within the first decade (Khan et al., 2023). However, other studies also showed that genes for heterotrophy (carbon cycling) were present in very young soils (Esperschütz et al., 2011), indicating that both carbon uptake and degradation can occur rapidly after glacial retreat. Soil moisture was one of the explanatory variables of the microbial community composition along the chronosequence investigated by Khan et al. (2023).

The positive net daytime CO₂ fluxes (indicating CO₂ emissions) that we observed predominantly in young dry soils are expected to be the result of microbial decomposition of soil carbon. Guelland et al. (2013a) found net CO₂ emissions for each of their sampled locations within the Damma glacier proglacial area, despite the presence of plants only in their older sites. Soil respiration in young sites (<50 years) mainly had pre-aged, burned, allochthonous carbon as C-source, whereas the CO₂ from older soils was the result of degradation of in situ produced carbon. The presence of old or allochthonous carbon would explain the CO₂ emissions we also find in our youngest sites. Other's incubation studies with added plant litter showed decomposition occurred in soils of 10 years old, to the same extent as in 70- and 120-years old soils, thus indicating that a microbial community capable of the mineralisation of organic carbon is present within 10 years (Guelland et al., 2013b). We measure net CO₂ uptake in older and in wet locations. CO₂ production likely does occur in those locations but is masked by simultaneous CO₂ uptake. To detangle the two, light/dark chamber studies and/or isotope measurements are required. Potentially, soil carbon can form complexes with weathered minerals in older soils, protecting it from degradation and lowering emission rates (Guelland et al., 2013a, b).

Similarly to our results (Figs. 2 and 3), Egli et al. (2010) found a correlation (R2=0.51, p<0.05) of soil organic carbon content with age, with an apparent knickpoint around 50 years. However, their limited number of sampled locations younger than 50 years makes it difficult to accurately assess trends in the youngest soils. Temme et al. (2016) did not find a significant relationship between soil age and organic carbon content, but did observe the same divergence between soils > 50 and <50 years old, with the <50 years old soils containing no detectable organic carbon. We hypothesize that the difference between these two age groups is related to the contribution of vascular plants to the soil carbon stock. Our CO₂ flux results show that carbon is taken up in part of our soils of <50 years, presumably by the microbial community. The uptake rate however increases when the plant surface cover increases (Figs. 2 and B2). Bernasconi et al. (2011) show that the microbial carbon content of their proglacial soils increased with age in the top 5 cm, showing that microbes do still contribute to soil organic matter in later stages, in addition to plant carbon inputs. Li et al. (2022) observed a linear increase in soil organic carbon content with age (0.13 %–1.3 % over a 90 years sequence), as well as a decrease in pH with age. We also observed a linear relationship between age and OC content (R2=0.40 for dry soil, R2=0.35 for wetland, Fig. 3) and age and pH (R2=0.41, p<0.001, Fig. B3). Interestingly, the soils measured in that study had a soil moisture content of 68 %–85 %, except for the youngest soil (40 %). Although not saturated, these soils are likely closer to our wetland sites than the dry soil locations within the Bachfallenferner proglacial area. The combined work of Smittenberg et al. (2012) and Guelland et al. (2013a) showed increasing carbon stocks with age in the Damma proglacial area, along with increased CO₂ emissions, but no significant correlation between the two. They showed the presence of plants seemed to increase soil CO₂ emissions due to enhanced root inputs, although the relationship seemed complex and dependent on several factors.

Besides the effect of age, we also see a clear distinction between wetland and dry locations. Wetland locations have significantly higher CO₂ uptake rates (Fig. 2) than dry soils of the same age (p<0.001). They also become CO₂ sinks much earlier: the majority of the wetlands is an atmospheric CO₂ sink directly after glacial retreat, whereas the dry soils are often CO₂ sources in the first 100 years (Fig. 2). In the oldest locations (170–175 years old), the CO₂ uptake rates of the wetlands is on average 2.2 mM h⁻¹, whereas it was 1.4 mM h⁻¹ in dry soils. The carbon stock is very high in certain wetlands, but low in others (Fig. 3). Possibly, the time of carbon buildup in the wetland locations does not always correspond to the deglaciation age. Deposition or erosion of fine material in these wet sites may have “reset” the clock on soil carbon stock buildup. Several of the older wetland locations have up to 70–80 mass % carbon in the topsoil (data file publicly available), indicating that the wetlands could turn into peatlands over time. Plant communities appear earlier in wetlands than in dry soils and also cover a higher surface % (Figs. 3 and B2). However, also before vascular plant and moss communities are established, wetlands have a higher CO₂ uptake rate. This supports the hypothesis that the microbial community is more abundant and/or active in proglacial wetlands compared to non-saturated soils. A higher water content can benefit the microbial community by an increased water and nutrient availability, due to enhanced (diffusive) transport. A study on litter decomposition in proglacial soils showed a positive correlation between microbial mineralisation of litter and the soil moisture content (Guelland et al., 2013b), suggesting a higher water content indeed promotes microbial activity, most likely not just the community responsible for remineralization, but also for CO₂ uptake. In addition, our results show a correlation between the water and carbon content of the soils (R2=0.65, Fig. B4), not only for wetland but also for locations classified as dry soils. As no other studies on proglacial wetlands exist, we cannot compare this with previous findings. A study on wet locations just outside the terminal moraine of a proglacial area in New Zealand (Fewster et al., 2025) shows peat formation in saturated soils. Potentially, the wetlands we observe will turn into peatlands over time. The correlation we observe between water content and carbon stock is likely a result of enhanced plant and microbial activity (and therefore carbon uptake and storage fluxes) in wetter soils, but it should be noted that a higher carbon content also increases the water holding capacity of the soils, and the correlation is therefore likely explained by both biological and physical reasons.

Previous research on CO₂ emissions from proglacial areas has strongly focussed on streams and lakes, which can emit CO₂ that is either locally produced or CO₂ that is transported along with (melt)water. Glacial runoff is known to transport methane (Du et al., 2022; Kleber et al., 2025), but an earlier study that measured the CO₂ flux from glacial runoff water did not show consistent fluxes, as values fluctuated between uptake and emission of CO₂ in the range of -20 to +20 mmol m⁻³ d⁻³ (Du et al., 2022). As we measure consistent uptake (95 % of wetland chambers showed a negative CO₂ flux) of CO₂ from our wetlands, we expect allochthonous aquatic CO₂ inputs, which would lead to positive CO₂ fluxes, to be limited.

Although trends in the CO₂ flux and carbon storage with age,wetness and vegetation cover can clearly be observed, not all variation can be explained by these factors. We attempted to further explain the observed CO₂ fluxes and carbon stocks with other environmental factors such as slope and the topographic wetness index. However, none of these increased the predicting value of our simple model (R2<0.4). A surprising apparent decoupling between carbon fluxes and soil characteristics was also observed by Guelland et al. (2013a), who attributed this to the high heterogeneity in proglacial areas. More research on proglacial systems would allow for better correlative studies and modelling on larger, combined datasets. The data of this study have been deposited in a public repository (see Methods) to encourage such future analyses.

Overall, our data indicate that proglacial wetland areas play an important role in CO₂ uptake and carbon storage. Wetlands take up atmospheric carbon as early as 5 years after glacial retreat and can locally store more carbon than dry soils in their vicinity. Although wetlands occupy only a small proportion of the proglacial landscape, they appear to be disproportionately important for carbon sequestration. Despite this, they have been completely overlooked in previous proglacial research. The factors controlling soil carbon storage and CO₂ fluxes in proglacial environments remain poorly understood, and no rate measurements during the night, or outside of the growing season, are available. Our limited dataset underscores the need for more comprehensive investigations in proglacial systems, which are rapidly expanding due to ongoing glacier retreat, and the need for a larger research focus on alpine wetland areas.