Many wetlands around the world have been restored, but far less is known about whether these areas actually begin functioning as wetlands again. Monitoring often lasts only a short time and many projects lack sufficiently clear objectives against which to assess the success of restoration efforts.
When a peatland is drained, oxygen enters the peat, causing the organic matter stored there to decompose more rapidly and release carbon dioxide into the atmosphere. Raising the water level slows this process and helps prevent further peat loss, but a wet area is not necessarily a restored wetland. For that to happen, conditions must develop that allow vegetation characteristic of wetlands to grow again.
In our recently published study, my colleagues from the University of Tartu and the University of California, Berkeley, and I therefore examined two wetlands with long histories of drainage, tracking changes in water levels, vegetation and carbon dioxide and methane fluxes over four years following restoration. One of the wetlands, Ess-soo, is located on a former peat extraction site in southern Estonia, while the other, Hill Slough, lies in California's Sacramento–San Joaquin River Delta.
Both sites were restored in 2021 after decades of drainage and land use. Four years later, the results were surprisingly similar: restoring water levels had not led to significant vegetation growth.
Two very different wetlands, one similar concern
Vegetation generally struggles to recover at former peat extraction sites. Peat extraction removes much of the original vegetation and the seeds and other material needed for plants to spread. The remaining bare peat surface is acidic, nutrient-poor and dry. As a result, even a former peat extraction site that has been abandoned for decades can remain largely devoid of vegetation.
Ess-soo was originally a natural raised bog, but peat extraction in the second half of the 20th century fundamentally transformed the area. Restoration work began in 2021 at the 58-hectare former peat extraction site, which had been abandoned for nearly 30 years. The site was divided into several sections, with different restoration methods applied to each.
Our study focused on an 8.3-hectare section where drainage ditches were blocked and shallow pools were created to raise the water level and encourage the natural recovery of peatland vegetation. The aim was to keep the water level close to the ground surface and create conditions suitable for the return of vegetation characteristic of transitional mires.
Following restoration, a problem emerged at Ess-soo: conditions in different parts of the same site were either too dry or too wet. Lower-lying sections of the peatland remained permanently submerged, causing vegetation to decline, while water levels in higher areas fell too far below the surface.
During the dry summer of 2024, the water level remained too far below the surface across much of the study area for more than a month. Four years after restoration, vegetation still covered only about 30 percent of the drier ground. Sphagnum mosses, which are particularly important for the formation of new peat, were not found in the study area either before or after restoration.
At Hill Slough, the problem was different, but the outcome was similar. Located in a Mediterranean climate, the wetland underwent extensive diking and drainage beginning in the mid-19th century to make way for farming and livestock grazing. In 2021, the natural water regime was restored across the 100-hectare study area by opening the levees. The aim was to reintroduce tidal flows and improve habitats for wildlife characteristic of wetlands.
However, at Hill Slough, which had been cut off from its natural water regime for more than a century, the original wetland vegetation had undergone substantial changes. When the area was once again exposed to daily tidal fluctuations, leaving much of it frequently submerged, the vegetation that had been present before restoration rapidly declined.
Water levels fluctuated by more than a meter over the course of a day, while shifting soft sediments and unstable ground likely made it even harder for plants to establish themselves. As a result, Hill Slough became almost entirely a mudflat, with only a few plants managing to establish themselves along the higher edges of the site.
Lack of vegetation impacts climate benefits
Very slow vegetation growth means that wetland ecosystems recover only gradually, directly affecting the carbon cycle. Plants absorb carbon dioxide from the atmosphere through photosynthesis. When vegetation is sparse, carbon uptake also remains low.
At the same time, restoring water to a wetland has an important effect of its own. Under wet conditions, less oxygen can penetrate the peat, slowing its decomposition and potentially reducing carbon dioxide emissions. However, oxygen-poor conditions can also increase methane production. Therefore, assessing the climate impact of a restored wetland requires considering both carbon dioxide and methane.
Both study sites remained net sources of carbon throughout the monitoring period, meaning they released more greenhouse gases into the atmosphere than they absorbed. At Ess-soo, net carbon dioxide emissions during the drought year of 2024 were approximately 158 grams of carbon per square meter per year, compared with around 42 grams in the wetter year of 2025. Meanwhile, methane emissions increased from approximately six grams to nine grams. The lower carbon dioxide emissions were consistent with cooler, wetter conditions that may have slowed peat decomposition.
Emissions at Hill Slough were generally higher. Over the four-year period, net carbon dioxide emissions declined from approximately 334 grams to 270 grams of carbon per square meter per year. Methane emissions, meanwhile, increased after the first year and subsequently remained high, reaching approximately 22 grams.
The differences between the two sites are likely related to climate, vegetation and water regimes. For example, Estonia's cooler climate limits microbial activity and the decomposition of organic matter during winter. In other words, restoring water levels may have reduced some organic matter decomposition, but sparse vegetation meant that neither site had yet reached a point where carbon uptake by plants could offset emissions.
Slow recovery not inevitable
Such carbon fluxes are not unusual in restored wetlands with sparse vegetation. Carbon dioxide emissions at Hill Slough were similar to those measured at other restored wetlands in California's delta where vegetation had not returned. By contrast, strong carbon dioxide uptake has been recorded at restored wetlands with similar conditions but well-established vegetation.
Ess-soo's carbon balance was similar to that of partially vegetated former peat extraction sites undergoing restoration, while net carbon dioxide uptake has also been achieved at peatlands with more extensive vegetation. However, even well-established vegetation does not automatically mean that a wetland will become a net carbon sink every year.
The slow recovery of Ess-soo and Hill Slough does not mean that such outcomes should be accepted as inevitable. Vegetation in tidal wetlands typically becomes established within three to five years and, in the best cases, within just one year. There are also examples of former peat extraction sites, where recovery is generally slower, achieving full vegetation cover in approximately five years.
Rapid recovery often requires more than simply restoring water levels. Depending on the site, restoration efforts can include improving water management, preparing the soil, adjusting ground elevation and actively encouraging vegetation recovery, for example through planting or transplanting sphagnum mosses. No such active vegetation restoration measures were used in the sections of Ess-soo and Hill Slough examined in this study.
If monitoring shows that vegetation is not recovering or water levels are not behaving as planned, restoration measures should be adjusted rather than simply waiting. This requires setting clear, measurable objectives from the outset and monitoring results over an extended period. Sharing both successes and challenges is equally important, as comparisons between different sites can help improve the planning of future restoration projects.
This is precisely what the comparison between Ess-soo and Hill Slough demonstrates: similar obstacles can arise in very different wetlands. These findings do not diminish the need for wetland restoration. Rather, they show that successful restoration involves more than simply bringing back water — ecosystem recovery must be monitored and, when necessary, actively guided.
The study was published in the journal Restoration Ecology.
--