Knut Rydgren

Ecological restoration can halt the ongoing loss of habitats, species, and ecological functions. However, to realize its potential, restoration ecology must develop tools for predicting restoration success based on suitable ecosystem attributes, as their properties affect prediction outcomes. Ideally, restoration should integrate time to recovery (TR) into outcome assessments. For management‐dependent ecosystems, the impact of post‐restoration management should also be included in the assessment. We conducted an 18‐year roadside restoration experiment to explore the TR of seminatural grassland vegetation, using local grasslands as the reference ecosystem. We assessed how TR was affected by two alternative restoration techniques: “Ecological” (adding target species propagules via species‐rich seed mixtures or hay) and “Standard” (hydroseeding of species‐poor, commercial grass mixtures). We moreover assessed the effect of a subsequent shift in post‐restoration management after 8 years, from “Traditional” (regular cutting and subsequent biomass removal) to “Limited” (cutting with no biomass removal). Additionally, we compared the suitability of two ecosystem attributes, species composition and species richness, for predicting TR. Our TR predictions confirmed the advantage of Ecological over Standard restoration techniques and highlighted the importance of Traditional post‐restoration management in restoring seminatural grassland vegetation. Conversely, Limited post‐restoration management halted the development of seminatural grassland vegetation, regardless of the initial restoration method. This highlights the importance of monitoring how restoration and management interact to influence restoration success in seminatural grasslands. Species composition was a more suitable metric than species richness for evaluating restoration success. It reflects not only the number but also the identity of species, enabling differentiation between target and nontarget species. Moreover, the species composition mirrors the site's environmental conditions, regarded as fundamental for restoration success. Our study also shows that TR calculations based on species composition data provide more realistic predictions of restoration success than those based on species richness. We conclude that time‐to‐recovery predictions are among the tools needed for restoration ecology to mature and evolve from a descriptive to a hypothesis‐testing science.

Accurate mapping of peat depth is crucial for carbon accounting, areal planning, and land management in peatlands. However, existing maps often lack the resolution and accuracy needed for these purposes, even in countries with rich spatial data sets.We present a study that evaluated whether digital soil mapping using remotely sensed data could improve existing maps of peat depth across two, >10 km2 sites in western and southeastern Norway. We measured peat depth by probing and ground-penetrating radar at 372 and 1878 locations at the two sites, respectively. Then we trained Random Forest models using radiometric and terrain variables, plus the national map of peat depth, to predict peat depth at 10 m resolution.The two best models achieved mean absolute errors of 60 and 56 cm, explaining one-third of the variation in peat depth. Our remote sensing models had better accuracy than the national map of peat depth, even when we calibrated the national map to the same depth data. Terrain variables were much better predictors than radiometric variables, with elevation and valley bottom flatness showing the strongest relationships to depth. At our coastal site, peats were much deeper above the Holocene marine limit than below, emphasizing the importance of accumulation time in places that have experienced glacial isostatic adjustment. Meanwhile, the national map of peat depth itself carried much more information about peat depth at one of the sites than the other --- likely as a result of uneven historical field sampling. Based on these findings, we conclude that digital soil mapping with DTM-derived predictors can improve the existing, national map of peat depth in Norway. Doing so would support national and regional-scale peatland carbon stock assessments and land management policies, as well as specific areal planning decisions at the municipal scale. Since our remote-sensing models relied on predictor--depth relationships that were specific to the sites we mapped, more depth measurements would be needed to expand the spatial coverage of an improved national map. A structured surveying effort coordinated in the manner of, or in conjunction with, the National Forest Inventory would be an efficient way to collect these data. Better data infrastructure for hosting and compiling peatland parameters from opportunistic measurements would also accelerate the accumulation of accuracy improvements. Besides specific accuracy improvements, digital soil mapping of peatland also offers advantages in transparency, reproducibility, and updatability.

Peatland restoration aims to restore hydrology and peat‐forming vegetation, supporting other ecosystem functions. However, the time required for complete vegetation recovery is generally unknown. Here, we investigate this in an experimentally restored, formerly afforested blanket bog in northern Scotland, which was plowed, fertilized, and planted with non‐native conifers in the mid‐1980s. Plowing created three “microforms”: Ridges, Original surface, and Furrows. Restoration management took place in two stages: trees were felled and drains blocked in 1998 (Standard treatment); then parts of the area were further rewetted with additional drain‐blocking in 2015/2016 (Enhanced treatment). We recorded plant species composition in permanent quadrats 0, 5, 13, and 24 years after the start of restoration. Here we use an ordination‐regression‐based approach (ORBA) to predict time to plant species compositional recovery compared with a reference (comparable nearby intact blanket bog). For the first 13 years, plant species composition diverged from the reference, then later started to converge. If the current speed and direction of vegetation change were maintained, predicted time to recovery varies between 50–100 years and 120–285 years applying a relaxed or strict criterion for restoration success, respectively. Seven growing seasons after Enhanced treatment, recovery speed increased only for the driest microform, Ridge. Surprisingly, this microform was not predicted to take longer to recover than other microforms under either treatment. On the landscape scale, sloping areas were harder to restore than flatter areas, having longer predicted times to recovery. Complete vegetation restoration may take a long time because of legacies from the afforestation (e.g., increased nutrient availability) and the time taken to fully restore surface morphology and water table. On the other hand, other research has already demonstrated that the site is currently acting as a net carbon sink, despite the incomplete vegetation recovery. We argue that functions may be restored without full recovery of species composition. However, approaching the full suite of species may be desirable to support long‐term resilience. Successful peatland restoration needs a strong science‐practice partnership, where learning gained from monitoring both damaged and comparable intact peatlands can be used to adapt management interventions.

Question: How do vascular plant and cryptogam cover, bare peat area and species composition in different microforms on a boreal raised bog change over a 34‐year period (1988–2022)? We discuss the observed patterns in the light of ongoing climate change. Location: Rønnåsmyra Nature Reserve, south‐eastern Norway. Methods: We recorded total vascular plant and cryptogam cover, bare peat area and species composition in 51 permanent 0.5 × 0.5 m plots in 1988, 2004 and 2022. The plots were assigned to microform classes (carpet, lawn, hummock) based on their characteristics in 1988 and 2022. We analysed changes over time and explored the relationship between change in species composition and change in the relative distance from the bog surface to the groundwater table (rDWT) in each microform using linear mixed‐effect models and ordination (GNMDS). Results: No shifts between microforms had occurred during the 34‐year period, but the number of observed taxa had decreased from 51 to 38. While the vegetation changed very little between 1988 and 2004, substantial changes occurred from 2004 to 2022. During this period, carpets showed a substantial increase in bare peat at the cost of cryptogam and vascular plant cover. Lawns showed a similar but less strong trend. Hummocks showed no such changes. The species composition of all microforms changed towards vegetation typical of wetter bog surfaces. In hummocks, this implied a shift from dominance by lichens to dominance by strongly peat‐producing Sphagnum species of section Acutifolia , coupled with an increase in rDWT. Conclusion: We demonstrate that bog vegetation can change substantially within two decades. The observed, divergent successions—retrogressive in carpets and lawns and progressive in hummocks—may result from the responses of the cryptogam layer to a combination of increased temperature and increased precipitation. Extreme weather events may have contributed to the increase in bare peat.

Peatlands are Earth's most carbon-dense terrestrial ecosystems and their carbon density varies with the depth of the peat layer. Accurate mapping of peat depth is crucial for carbon accounting and land management, yet existing maps lack the resolution and accuracy needed for these applications. This study evaluates whether digital soil mapping using remotely sensed data can improve existing maps of peat depth in western and southeastern Norway. Specifically, we assessed the predictive value of lidar-derived terrain variables and airborne radiometric data across two, >10 km² sites. We measured peat depth by probing and ground-penetrating radar at 372 and 1878 locations at the two sites, respectively. Then we trained Random Forest models using radiometric and terrain variables, plus the national map of peat depth, to predict peat depth at 10 m resolution. The two best models achieved mean absolute errors of 60 and 56 cm, explaining one-third of the variation in peat depth. Terrain variables were better predictors than radiometric variables, with elevation and valley bottom flatness showing the strongest relationships to depth. Radiometric variables showed inconsistent and weak predictive value – improving performance at one site while degrading it at the other. Our remote sensing models had better accuracy than the national map of peat depth, even when we calibrated the national map to the same depth data. Still, weak relationships with remotely sensed variables made peat depth hard to predict overall. Based on these findings, we conclude that digital soil mapping can improve the existing, national map of peat depth in Norway, but detailed local maps are best made from tailored field measurements. Together, these pathways promise more accurate landscape-scale carbon stock assessments and better-informed land management policies.

Peatlands are Earth's most carbon-dense terrestrial ecosystems and their carbon density varies with the depth of the peat layer. Accurate mapping of peat depth is crucial for carbon accounting and land management, yet existing maps lack the resolution and accuracy needed for these applications. This study evaluates whether digital soil mapping using remotely sensed data can improve existing maps of peat depth in western and southeastern Norway. Specifically, we assessed the predictive value of LiDAR-derived terrain variables and airborne radiometric data across two, >10 km² sites. We measured peat depth by probing and ground-penetrating radar at 372 and 1878 locations at the two sites, respectively. Then we trained Random Forest models using radiometric and terrain variables, plus the national map of peat depth, to predict peat depth at 10 m resolution. The two best models achieved mean absolute errors of 60 and 56 cm, explaining one-third of the variation in peat depth. Terrain variables were better predictors than radiometric variables, with elevation and valley bottom flatness showing the strongest relationships to depth. Radiometric variables showed inconsistent predictive value – improving performance at one site while degrading it at the other. The accuracy of the national map of peat depth did not measure up to any of our remote sensing models, even though it was calibrated to the same data. Still, weak relationships with remotely sensed variables made peat depth hard to predict overall. Based on these findings, we conclude that digital soil mapping can improve existing, broad-scale maps of peat depth in Norway, but highly localized carbon stock assessments are best made from field measurements. Furthermore, the inability of models to identify peat presence outside known peatlands highlights the need for integrated mapping of peat lateral extent and depth. Together, these pathways promise more accurate landscape-scale carbon stock assessments and better-informed land management policies.

Peatlands are Earth's most carbon-dense terrestrial ecosystems and their carbon density varies with the depth of the peat layer. Accurate mapping of peat depth is crucial for carbon accounting and land management, yet existing maps lack the resolution and accuracy needed for these applications. This study evaluates whether digital soil mapping using remotely sensed data can improve existing maps of peat depth in western and southeastern Norway. Specifically, we assessed the predictive value of LiDAR-derived terrain variables and airborne radiometric data across two, >10 km² sites. We measured peat depth by probing and ground-penetrating radar at 372 and 1878 locations at the two sites, respectively. Then we trained Random Forest models using radiometric and terrain variables, plus the national map of peat depth, to predict peat depth at 10 m resolution. The two best models achieved mean absolute errors of 60 and 56 cm, explaining one-third of the variation in peat depth. Terrain variables were better predictors than radiometric variables, with elevation and valley bottom flatness showing the strongest relationships to depth. Radiometric variables showed inconsistent predictive value – improving performance at one site while degrading it at the other. The accuracy of the national map of peat depth did not measure up to any of our remote sensing models, even though it was calibrated to the same data. Still, weak relationships with remotely sensed variables made peat depth hard to predict overall. Based on these findings, we conclude that digital soil mapping can improve existing, broad-scale maps of peat depth in Norway, but highly localized carbon stock assessments are best made from field measurements. Furthermore, the inability of models to identify peat presence outside known peatlands highlights the need for integrated mapping of peat lateral extent and depth. Together, these pathways promise more accurate landscape-scale carbon stock assessments and better-informed land management policies.

Peatlands are Earth's most carbon-dense terrestrial ecosystems and their carbon density varies with the depth of the peat layer. Accurate mapping of peat depth is crucial for carbon accounting and land management, yet existing maps lack the resolution and accuracy needed for these applications. This study evaluates whether digital soil mapping using remotely sensed data can improve existing maps of peat depth in western and southeastern Norway. Specifically, we assessed the predictive value of LiDAR-derived terrain variables and airborne radiometric data across two, >10 km² sites. We measured peat depth by probing and ground-penetrating radar at 372 and 1878 locations at the two sites, respectively. Then we trained Random Forest models using radiometric and terrain variables, plus the national map of peat depth, to predict peat depth at 10 m resolution. The two best models achieved mean absolute errors of 60 and 56 cm, explaining one-third of the variation in peat depth. Terrain variables were better predictors than radiometric variables, with elevation and valley bottom flatness showing the strongest relationships to depth. Radiometric variables showed inconsistent predictive value – improving performance at one site while degrading it at the other. The accuracy of the national map of peat depth did not measure up to any of our remote sensing models, even though it was calibrated to the same data. Still, weak relationships with remotely sensed variables made peat depth hard to predict overall. Based on these findings, we conclude that digital soil mapping can improve existing, broad-scale maps of peat depth in Norway, but highly localized carbon stock assessments are best made from field measurements. Furthermore, the inability of models to identify peat presence outside known peatlands highlights the need for integrated mapping of peat lateral extent and depth. Together, these pathways promise more accurate landscape-scale carbon stock assessments and better-informed land management policies.

Peatlands are Earth's most carbon-dense terrestrial ecosystems and their carbon density varies with the depth of the peat layer. Accurate mapping of peat depth is crucial for carbon accounting and land management, yet existing maps lack the resolution and accuracy needed for these applications. This study evaluates whether digital soil mapping using remotely sensed data can improve existing maps of peat depth in western and southeastern Norway. Specifically, we assessed the predictive value of LiDAR-derived terrain variables and airborne radiometric data across two, >10 km² sites. We measured peat depth by probing and ground-penetrating radar at 372 and 1878 locations at the two sites, respectively. Then we trained Random Forest models using radiometric and terrain variables, plus the national map of peat depth, to predict peat depth at 10 m resolution. The two best models achieved mean absolute errors of 60 and 56 cm, explaining one-third of the variation in peat depth. Terrain variables were better predictors than radiometric variables, with elevation and valley bottom flatness showing the strongest relationships to depth. Radiometric variables showed inconsistent predictive value – improving performance at one site while degrading it at the other. The accuracy of the national map of peat depth did not measure up to any of our remote sensing models, even though it was calibrated to the same data. Still, weak relationships with remotely sensed variables made peat depth hard to predict overall. Based on these findings, we conclude that digital soil mapping can improve existing, broad-scale maps of peat depth in Norway, but highly localized carbon stock assessments are best made from field measurements. Furthermore, the inability of models to identify peat presence outside known peatlands highlights the need for integrated mapping of peat lateral extent and depth. Together, these pathways promise more accurate landscape-scale carbon stock assessments and better-informed land management policies.

Peatlands are Earth's most carbon-dense terrestrial ecosystems and their carbon density varies with the depth of the peat layer. Accurate mapping of peat depth is crucial for carbon accounting and land management, yet existing maps lack the resolution and accuracy needed for these applications. This study evaluates whether digital soil mapping using remotely sensed data can improve existing maps of peat depth in western and southeastern Norway. Specifically, we assessed the predictive value of LiDAR-derived terrain variables and airborne radiometric data across two, >10 km² sites. We measured peat depth by probing and ground-penetrating radar at 372 and 1878 locations at the two sites, respectively. Then we trained Random Forest models using radiometric and terrain variables, plus the national map of peat depth, to predict peat depth at 10 m resolution. The two best models achieved mean absolute errors of 60 and 56 cm, explaining one-third of the variation in peat depth. Terrain variables were better predictors than radiometric variables, with elevation and valley bottom flatness showing the strongest relationships to depth. Radiometric variables showed inconsistent predictive value – improving performance at one site while degrading it at the other. The accuracy of the national map of peat depth did not measure up to any of our remote sensing models, even though it was calibrated to the same data. Still, weak relationships with remotely sensed variables made peat depth hard to predict overall. Based on these findings, we conclude that digital soil mapping can improve existing, broad-scale maps of peat depth in Norway, but highly localized carbon stock assessments are best made from field measurements. Furthermore, the inability of models to identify peat presence outside known peatlands highlights the need for integrated mapping of peat lateral extent and depth. Together, these pathways promise more accurate landscape-scale carbon stock assessments and better-informed land management policies.