The role of management and topography in boreal forest responses to drought
Julika Wolf | ESA - ESRIN | Italy
Boreal forest ecosystems are predicted to experience more frequent summer droughts due to climate change, posing a threat to future forest health and carbon sequestration potential. The 2018 Central European drought may be an early showcase of what is to be expected in the future. The economic consequences for Sweden were severe, as the country grappled with forest fires, crop failures, and temperature extremes, and the situation was declared a national crisis.
Forestry is the dominant land use type in Sweden, currently covering about 60% of the country’s total area, and wood products constitute a major export revenue. The legacy of forest management is a landscape largely dominated by single-species, even-aged forest stands of Scots pine (Pinus sylvestris) and Norway spruce (Picea abies), which are replanted following a clear-cut, typically succeeded by several rounds of thinning and cleaning, and eventually clear-cut again after 60-120 years. Furthermore, selection of fast-growing, damage-resistant phenotypes and fertilization are common practices, as well as drainage of wetlands. At the same time, Sweden officially harbors the largest area of remaining primary forests in Europe, although no official comprehensive map of the location of these forests exists. Primary forests can be used as an analogue to which managed forests can be compared: they remain in a relatively pristine and undrained condition, are naturally regenerated, and display an uneven age structure, as well as a larger presence of dead wood. How forest management influences drought resistance is currently not well understood.
To compare drought responses in pristine and managed forests, we analyzed a high-resolution vegetation index of gross primary production across Sweden during the unprecedented 2018 nationwide drought. We employed a unique new map detailing 391 pristine forests and linked these spatially to surrounding managed forests, forming pairs of similar weather and climate. We control for topographical variations in soil moisture, which was a strong determinant of drought responses, and analyze Landsat-derived EVI2 anomalies during the drought year compared to the long-term mean. We found that pristine forests are less affected by drought than managed forests are, despite being significantly older with a larger share of tall trees. Our results indicate that large-scale boreal forest management may exacerbate the impact of drought impact in a future with more frequent and extreme hydroclimatic events.
Structural recovery of Alpine forests after disturbances
Dr. Ana Stritih | Technical University of Munich | Germany
Forests in the Alps are experiencing an increasing frequency of natural disturbances, such as storms and forest fires. These disturbances jeopardize forests’ capacity to provide essential ecosystem services, such as protection from natural hazards, so the resilience of mountain forests is a critical concern for forest management. In recent years, significant progress has been made in detecting changes in canopy cover, including forest disturbances and recovery, from optical satellites. However, optical satellites have a limited capacity to detect changes in forests’ three-dimensional structure, including properties such as height, leaf area, biomass, and structural diversity. These structural characteristics play an important role in determining forests’ resilience and their capacity to provide ecosystem services. For example, forests with a heterogeneous vertical structure experience less wind damage during storm events, while forests with a high stem density are more effective in protecting from rockfall. So far, assessments of forest structure have mostly relied on airborne LiDAR data, which is not consistently available across the Alps, limiting the potential for regional-scale analyses. We use data from GEDI, NASA’s recently launched spaceborne LiDAR mission, to assess the structural characteristics of Alpine forests. We combine plot-level GEDI data with a Landsat-based disturbance dataset to identify forests that have been disturbed during the last 34 years. This allows for a space-for-time analysis of forest recovery after disturbances in terms of their vertical structure, including height, leaf area at different heights, and structural diversity. After disturbances, we observe an increase in vertical diversity, which gradually decreases with time. However, after 30 years, disturbed forests still show differences in structure compared to undisturbed forests. We also investigate how forest recovery differs among different forest types and elevations across the Alps. Information about post-disturbance forest structure can help assess how ecosystem functions and services are maintained after disturbances, and how quickly they recover. We discuss the potential of this approach to improve our understanding of forest resilience, as well as the challenges and limitations of using GEDI in complex terrain.
Improving tree carbon use efficiency for climate-adapted more productive forests (iCUE-Forest)
Dr. Martin Thurner | Senckenberg Biodiversity and Climate Research Centre (SBiK-F) | Germany
The rate at which forests take up atmospheric CO2 is critical because of their potential to mitigate climate change and their value for wood production. The allocation of carbon fixed through photosynthesis into biomass can be quantified through the tree carbon (C) use efficiency (CUE), which is determined by gross primary production (GPP) and plant respiration (Ra) via the relation CUE=(GPP-Ra)/GPP. The effect of future climate on CUE is unclear due to the highly uncertain response of plant respiration to the expected increases in temperature and possible changes in tissue nitrogen (N) concentrations that also affect GPP and Ra.
Within the project ”Improving tree carbon use efficiency for climate-adapted more productive forests” (iCUE-Forest), we aim to develop novel data-driven estimates of plant respiration, net primary production (NPP=GPP-Ra) and tree CUE covering the northern hemisphere boreal and temperate forests. These will be based on recent satellite-driven maps of tree living biomass, databases of N concentration measurements in tree compartments (leaves, branches, stems, roots) and the relationships between respiration rates and tissue N concentrations and temperature. Such estimates will enable the detection of spatial relationships between CUE and environmental conditions and facilitate the parameterization of dynamic global vegetation models to predict the change in CUE in response to future climate and forest management.
Here we compile an unprecedented database of N concentration measurements in tree stems, branches and roots covering all common boreal and temperate tree genera together with data available mainly for leaves from databases like TRY. We find that the variation in tree tissue N concentrations of boreal and temperate trees is controlled by their leaf type (broadleaf deciduous, needleleaf deciduous, needleleaf evergreen), growth rate (fast- vs. slow-growing), tree age/size and soil nitrogen concentration. These relationships have important implications on the coupling of the C and N cycles in the vegetation, since tissue N concentrations determine photosynthesis, growth and plant respiration. Thus, by altering tissue N concentrations, changes in the distribution of tree species, in tree age/size or in soil fertility, induced by climate change, forest management or disturbances, can affect the C sequestration potential of boreal and temperate forests.
Subsequently, we combine the derived tree-level relationships between tissue N concentrations and underlying drivers, tree species distribution maps, and estimates of tree compartment biomass based on satellite remote sensing products. In this way, we derive novel estimates of the spatial distribution of N content in northern boreal and temperate forests that will in turn be used to assess CUE variations.
Transient ecohydrology of the rainforests under rapidly changing climate
Chandrakant Singh | Stockholm University, Stockholm, Sweden | Sweden
Terrestrial tropical ecosystems, not only host both biodiversity, but are also crucial in stabilizing earth climate by sequestering carbon and maintaining the global water cycle. These terrestrial ecosystems naturally exist as alternative stable states, commonly referred to as forest and savanna ecosystems. A better understanding of these states and the dynamics of their translation is thus helpful in predicting their response to future hydroclimatic changes. These alternative stable states are analyzed against precipitation and are predominantly determined based on space-for-time substitution. However, such a substitution provides a partial picture of ecosystem adaptation dynamics and associated ecosystem change over time. However, under rapidly changing climate, including these key adaptation and temporal aspects become increasingly important to understand the impact of hydroclimatic changes in modifying the ecohydrology of the rainforest and their resilience towards future disturbances.
Here, we empirically study the transient state of tropical ecosystems and their hydroclimatic adaptations by examining remotely sensed tree cover and root zone storage capacity over last two decades in South America and Africa. Tree cover represents the above-ground ecosystem structure's density, and is derived directly from MODIS satellite data. Whereas root zone storage capacity is the maximum amount of soil moisture that the vegetation can access for transpiration, i.e., the buffer capacity of the ecosystem towards water stress, is derived using daily precipitation and evaporation data.
We found that ecosystems at high (> 75%) and low (< 10%) tree cover adapt to changing precipitation by instigating considerable subsoil investment while experiencing limited tree cover change over time (ΔTC). In case of forest ecosystems, those that didn't had any considerable change to their hydroclimate over the year (i.e., least water-stressed) didn't need to invest in their root zone storage capacity. However, with increasing water stress, the forest ecosystems had to actively invest in their root zone storage capacity to offset the experienced water deficit. For these ecosystems, the below-ground investment does not come at the cost of changing the above-ground ecosystem structure. We refer to these ecosystems as stable, since ecosystems' adaptive dynamics keep the structural characteristics intact (i.e., limited ΔTC). In contrast, unstable ecosystems at intermediate (30-60%) tree cover were unable to exploit the same level of adaptation as stable ecosystems, thus showing considerable changes to their above-ground ecosystem structure (i.e., considerable ΔTC).
Furthermore, ignoring this adaptive capacity of the ecosystem can underestimate the resilience of the forest ecosystems, which we find is largely underestimated in the case of the Congo rainforests. In case of Congo rainforests, only by including root zone storage capacity were we able to truly capture the grass species-induced drought coping strategy, which otherwise is hard to detect with just precipitation. Furthermore, our modified resilience metric shows better consistency with actual tree cover change than those previously reported. The results from this study emphasize the importance of the ecosystem's temporal dynamics and adaptation in inferring and assessing the risk of forest-savannah transitions under rapid hydroclimatic change.
Mapping cryogenic processes and assessing the sustainability of permafrost landscapes in the North-East Arctic Siberian taiga and tundra from Landsat 8, Sentinel 2, and DEM data
Prof. Dr. Sébastien Gadal | Aix-Marseille Université, CNRS ESPACE UMR 7300, France | France
Permafrost landscapes are one of the most sensitive ecosystems where humans inhabit. Geocryological conditions determined by the presence of ice content, the genetic type of sediments, and the active layer are one of the most important variables for classifying the vulnerability of an ecosystem to disturbances in vegetation and soil cover. These variables indicate cryogenic processes that can be activated during the degradation of permafrost. Cryogenic mapping is an important parameter for assessing the state of permafrost and infrastructure design in permafrost landscapes. However, the methods of remote sensing spatial modelling of for understanding the distribution of cryogenic processes in the Arctic Siberian mountainous areas with continuous permafrost are still insufficient. The cartographies at the regional scales 1:500 000 are inexistent. We need of permafrost landscape maps are increasing with the development of the North-East Siberian Arctic for the infrastructures and urban centers risk assessments. Orulgan Ridge in North-East Siberia is one of these territories.
This study examines the Orulgan Ridge region, as a case area, where we developed maps of the distribution of cryogenic processes based on the detailed landscape structure (including classification of environmental variables, vegetation covers and genetic type of sediments) with time series Sentinel 2 MSI and Landsat 8 OLI, and stereogrammetric digital elevation model of the ArcticDEM data.
The combination of Random Forest classifier and geomorphological GIS terrain analysis have successfully distinguished 6 classes of boreal mountain taiga and 3 classes of arctic tundra and mountain desert. Based on the indicator parameters of the interrelation of ecological variables (such as vegetation and topographic position) adopted in permafrost-landscape cartography, we carried out regionalization of cryogenic processes. We made a classification of the genetic type of deposits, which determines the likelihood of the development of dangerous cryogenic processes. Only 5 zones of high-risk have been identified according to the prevalence of thermokarsts, thermoerosion, frost cracking, soil suffusion, thermodenudation, and their combinations.
The assess of the sustainability of mountain permafrost landscapes integrates two scenarios using GRID modelling based on the annihilation of the vegetation by the mining industry and forest fires. To determine the sustainability of permafrost landscapes, median values were calculated, an indicator of variation (standard deviation) of environmental variables - elevation, vegetation association bioproductivity, slope, aspect, average temperatures of July and January, precipitation. We established the ranges of ecological potential values: optimal, suboptimal, pessimal. The ecological potential models within which the landscape can maintain its characteristic structural and functional features. The low sustainability potential is characteristic of the zones with the development of thermokarst and soil suffusion. This state creates many environmental restrictions in the field of environmental management associated with the quality of the human security. The analysis of cryogenic processes and the sustainability assessment gives a good potential modeling for the territorial planning, the environmental restoration, and provides a quantitative method for achieving landscape sustainability in Arctic urban and industrial centers.
Succession and Resilience on Military Training Grounds – Long-term Satellite Monitoring Revealing Ecological Dynamics of Habitat Conversion
Dr. Carsten Neumann | Helmholtz Center Potsdam German Research Center for Geosciences
Up to 6 % of the terrestrial earth surface is estimated to be covered by active or former military training grounds (MTGs) in various ecosystems worldwide. Due to intense disturbance regimes by moving heavy vehicles, explosion of munition or regular fire outbreaks, heterogeneous and temporally variable open landscapes have developed over long time periods outside urban or agricultural landuse. Open military areas thus develop spatiotemporally highly dynamic habitats that are characterized by a diversity of species, ecological processes and related ecosystem functions. However, their distribution and ecosystem dynamics are only scarcely described owing to the fact that military training grounds are large, inaccessible and dangerous to inspect. With respect to its potentially high nature conservation value and hence its potential contribution to increase the global protected are networks, such as the European Natura 2000, open military habitats need to be monitored for a better understanding of processes and dynamics of particularly succession and resilience patterns. Gaining information about the ecological dynamics of open habitats can be used to implement advanced protection and management efforts for developing and maintaining conservation targets on active and former MTGs.
We utilize satellite image time series from USGS Landsat and the ESA Copernicus Sentinel-2 missions to map the development of habitats on 15 MTGs between the years 1992-2018 in Germany. For this purpose, we applied the Habitat Sampler algorithm that enables to derive habitat type probabilities on the basis of one spectral library that can be transferred as calibration reference between different areas. For each MTG with an average size of 100 km² we mapped the spatiotemporal explicit development of natural succession trajectories in open dry and wet grasslands, forest areas and heathlands that are protected under the European habitat directive. The diversity of mapped MTG time series cover various process dynamics arising under different site conditions and particularly under the influence of changing management regimes. We quantitatively compared the history of habitat type abundances comprising effects from active military use, nature conservation measures on abandoned MTGs such as controlled fires, grazing, mowing and wilderness concepts allowing for forest succession. Thus, for the first time, natural succession dynamics are being contrasted to habitat resilience that is revealed in disentangling trends and triggers of ecosystem change in satellite imagery.
In our study we can prove to quantitatively map the spatiotemporal distribution of succession trajectories from open ground over xeric grass and wet meadows towards shrub encroachment as well as individual life-cycle stages of the heath plant Calluna vulgaris characterizing pioneer growth, mature phases and degeneration. We found early indicators for habitat conversion that is introduced by life-cycle shifts and species turnover from dwarf-shrub into tree species invasion. Mapping accuracies are generally high (> 80 % overall accuracy) using terrestrial polygon-based conservation status assessments from the European habitat directive for validation, in all MTGs of the entire time series. The study reveals varying pattern of habitat persistence, particularly distinct rates of natural succession dynamics depending on the type of disturbance, i.e. human-induced (controlled burning, mowing, grazing) or passive-dynamics (fires, senescence). Moreover, the resilience of habitats in ecological dynamic and coexisting succession trajectories crucially depends on the intensity, time and background conditions of implemented management or natural disturbance regimes. The results consequently provide evidence that spaceborne observations of habitats will give early indicators and advanced knowledge for an effective coordination of habitat management, ecological restoration and artificial disturbances in order to increase future biodiversity outcomes on military training grounds.