Back to the IGB project website of GLANCE >
Sub-project 1: Determination and quantification of flow traits of benthic invertebrates in river ecosystems
Expected alterations in community structure of aquatic invertebrates under global change can be estimated only if hydrological preferences and tolerances of target species get assessed numerically.
The river flow regime is one of the key parameters in river ecosystems as it controls physical habitat conditions, biological and ecological processes and river ecosystem functioning. River flow conditions have been substantially altered globally due to water regulation and climatic changes resulting in detrimental impacts on the functioning and health of river ecosystems. Species of benthic invertebrates are a major organism group in river ecosystems that are highly sensitive to environmental changes. It is crucial to better understand the quantitative changes in e.g. species’ abundance to environmental stressors such as flow alterations. Empirically-driven predictive relationships might be established for individual species by linking their abundance along wide ranges of environmental gradients to any environmental variables, e.g. different flow conditions. Any changes in flow can be described by indicators of hydrologic alterations (IHA metrics) that provide information on duration, frequency, magnitude, rate and timing of flow events. These predictive relationships can be used to assess species responses to climate-change-induced flow alterations.
The work is divided into three parts: Firstly, a non-linear modelling approach is applied for a German-wide dataset which enables to link the abundance of species to river flow to quantify flow preferences of benthic invertebrates for various IHA metrics (Kakouei et al. 2017). Secondly, this approach is used in two contrasting river catchments in the lowland and lower mountainous region of Germany to quantitatively assess potential changes in species’ abundance due to projected changes in flow conditions under the worst climate scenario RCP 8.5 (Kakouei et al. 2018). Thirdly, potential variability in projected abundance of individual species under 16 climate models derived from various combinations of global and regional climate models are examined (Kakouei et al. 2020). The effects of variability in climate model predictions on species’ abundance and functional trait composition are tested. The results provide empirical evidence that the functional trait compositions will be affected by flow alterations, but the effects would be regionally different. For example, flow alterations lead to increasing abundance of rheophilic and tolerant rhithral species in the lowland area, which is referred to as “rhithralisation effect”. These results go one step further than the qualitative assessment of species responses to environmental changes, and support the current knowledge that flow alterations and their effects on species’ abundance might be a global phenomenon. The main findings underline the high susceptibility of stream macroinvertebrates to ongoing climate-change-induced flow alterations. Concerning the methodology, a clear recommendation for future predictions is to reduce uncertainty inherent in climate change models and thus to improve future predictability of e.g. species’ abundance.
The analyses applied in this sub-project are applicable to forecast climate change impacts at different spatial and temporal scales as well as for different stressors or species. In addition to the results published in publications and the doctoral thesis, two master's theses based mainly on the data collected in the field were successfully completed. Mathias Goerres, Universität Kiel, 2016: GIS-based mapping and projection of the distribution of microhabitats in rivers. Martin Brosinski, TU Berlin, 2016: Species morphological traits determining the occurrence of benthic invertebrates along gradients of flow conditions.
Contact for sub-project 1
Karan Kakouei, kakouei@igb-berlin.de
Sub-project 2: Quantification of global change impacts on hydrological conditions
A simulation model depicts the recent past and current conditions regarding hydrological parameters, followed by a prediction model showing explicit data on flow conditions under climate change.
This subproject deals with the simulation of hydrologic parameters in three German catchments in different eco-regions: The Treene in the northern lowlands, the Kinzig in the low mountain range and the Ammer in the alpine region. In each catchment, the hydrological model was applied and optimized to ideally depict the flow preferences (indicators of hydrologic alteration, IHA) of benthic invertebrates (Kiesel et al. 2017). A thorough sensitivity analysis is carried out to assess the influence of different model settings on the simulation of biota. The resulting baseline simulation of the detailed spatio-temporal IHA is complemented by a comprehensive climate change impact analysis including the consideration of uncertainties (Kiesel et al. 2019). It was found that climate change impacts varied depending on the ecoregion and that the choice of the global and regional climate model caused higher uncertainties than bias correction methods or the selection of the representative concentration pathway. These uncertainties are crucial information to correctly interpret and assess the relevance of predicted species changes under climate change. Beside the publicized papers there is a master thesis. Angelika Kurthen, Fu Berlin, 2019: Metacommunity structuring of macroinvertebrates in high mountain streams, Southwest China.
Contact for sub-project 2
Dr. Jens Kiesel, kiesel@igb-berlin.de
Sub-project 3: Improvement of climate change projections for benthic invertebrates
Existing models for species range changes will be improved by adding riverine characteristics and stressor functions
Species’ distribution models (SDMs) are predictive models that are increasingly applied to river ecosystems. Hydrological flow regime is cited as one of the most important drivers of lotic systems, influencing the abundance and distribution of river biota. The main goal is to advance the predictive ability of SDMs for riverine benthic invertebrates by integrating hydrological predictors that describe flow regime. The work is divided into three parts: First, a high resolution spatio-temporal dataset of stream flow, and a set of hydrological metrics for the German stream network will be developed (Irving et al. 2018). Second, a variable selection method to select the optimal environmental variables for use in SDMs will be proposed, and the impact of predictor set choice in SDMs will be investigated (Irving et al. 2019). Third, the role of hydrology in SDMs will be disentangled, by investigating the influence of climate and hydrology related datasets (Irving et al. in prep.).
The hydrological variable is found to be the most important individual factor for species’ distributions in terms of both variable importance and proportional explained variance. Hydrology variables contribute to a high proportion of explained variance. The larger predicted range sizes may be due to the better description of the river discharge regime provided by the hydrological variables. In this sub-project hydrological variables are created and integrated in SDMs, and effective methods to improve prediction performance of riverine species’ distribution to advance freshwater SDM research are developed and validated. The introduced methods can be applied in different geographical regions as well as under alternative time periods and spatial scales. Due to the implications associated with altered model accuracy and predicted range size, applying SDMs with hydrological variables has the potential to aid river management decisions and conservation efforts.
There is a lack of suitable models predicting ecological effects of global change for the riverine environment and to guide management decisions. While it would be ideal to actually measure potential range shifts and community changes by long-term analyses, these data are often geographically restricted and involve costly and time-consuming observation and quantification of species in the field. As an alternative, species distribution models (SDMs) are valuable tools for predicting and evaluating species range changes under global or climate change. These models have been increasingly used in ecology and conservation management (Elith and Leathwick 2009). However, large-scale modelling approaches have not considered riverine characteristics sufficiently. The sub-project hence aimed to integrate stream specific data into SDMs such as stream networks, hydrological variables, or land use. The sub-project was divided into three parts: First, a high resolution spatio-temporal dataset of streamflow, and a set of hydrological metrics for the German stream network was developed (Irving et al. 2018). Second, a variable selection method to select the optimal environmental variables for use in Species distribution models (SDMs) was proposed, and the impact of choosing different predictors in SDMs was investigated (Irving et al. 2019). Irving et al. (2019) used Boosted Regression Trees (BRTs) to select variables for individual species in a larger community. In addition, the question was investigated whether the model performance is increased by applying such "tailor-made" sets of variables to individual species. Of a community of 67 benthic macroinvertebrate species, 10 were better represented by the tailor-made set and 10 worse. Especially for specialized, rare or invasive species, the use of a species-specific, tailor-made predictor data set may be more appropriate, as the preferences of the species may not be representative for the whole study area. Third, the role of hydrology in SDMs was disentangled, by investigating the influence of climate and hydrology related datasets (Irving et al. in prep.). Indeed, climate variability is found to be the most important individual factor for species’ distributions in terms of both variable importance and proportional explained variance. Within the WP hydrological variables were created and integrated in SDMs, and effective methods to improve prediction performance of riverine species’ distribution to advance freshwater SDM research are developed and validated. The introduced methods can be applied in different geographical regions as well as under alternative time periods and spatial scales. Due to the implications associated with altered model accuracy and predicted range size, applying SDMs with hydrological variables has the potential to aid river management decisions and conservation efforts.
Contact for sub-project 3
Sonja Jahnig, sonja.jaehnig@igb-berlin.de
Sub-project 4: Analysis of flow related changes at large spatial scales
Consideration of the effects of changed discharge conditions on the habitat availability for aquatic invertebrates and fish on a European and global level allows the consideration of pronounced spatial gradients
The WP includes two parts: (A) a doctoral student project with questions concerning the uncertainty of differently configured ecological model runs. For this purpose an investigation was carried out on the influence of the spatial resolution of models on predicted habitat suitability and the relative importance of the individual predictors (Friedrichs-Manthey et al. 2020a). Using a unique 300 years’ time series, ranging from 1800 to 2100, of hydrological and climate alterations, the environmental distance between current, historical and future environmental conditions for 49 native fish species in the upper Danube catchment was assessed. The assessment was followed by a species specific climate-niche-factor-analysis, which resulted in vulnerability estimates for all 49 fish species to future and historical environmental conditions and gives the advantage to place future vulnerability estimates into a historical context (Friedrichs-Manthey et al. in preparation). The gained knowledge of environmental alterations over the 300 years’ time period will be used to assess the degree of already observed environmental change within the actual network of protected areas. The degree of observed environmental change will then be compared to expected changes under future scenarios. Based on this information, a conservation planning exercise will be carried out to spatially prioritised areas for potential conservation actions. The work for this WP started in March 2017 and the completion of the work, with the submission of the PhD thesis to the Freie Universität Berlin is planned for December 2020. The title of the PhD thesis will be „Uncertainties in species distribution models: effects of spatial scales and temporal dimensions on future predictions”.
(B) Furthermore, we investigated the question which ecological effects can be expected from hydrological changes at large spatial scales. For this, the methods tested in GLANCE on the catchment scale are extrapolated. Global hydrological model output and temporal-spatial environmental variables such as land use, climate, dams, water use and population as well as static variables such as countries, latitude and elevation are superimposed on the threat status of freshwater species, available from the IUCN red list. A random-forest machine learning algorithm is used to determine which environmental variables significantly contribute to the global freshwater biodiversity crisis. Results show that biodiversity loss has a strong spatial autocorrelation, indicating that the species situation is influenced by eco-regional and country-specific environmental changes (Kiesel et al. in preparation).
Contact for Sub-project 4
Martin Friedrichs, friedrichs@igb-berlin.de and Dr. Jens Kiesel, kiesel@igb-berlin.de
Sub-project 5: Synthesis and praxis transfer
Synthesis and discussions with regard to water management applications. This fifth sub-project was added during the extension phase in order to maximize the impact of the project’s results and facilitate knowledge transfer towards water management practitioners.
Within this sub-project the (A) synthesizing of results and (B) tasks of science society transfer, including the handling of complex data sets, extensive uncertainty analysis and coordination with water management authorities and potential users of the modelling approaches, took place. A number of related manuscripts bring together different parts of the projects and/or reach out to collaborative projects that work on highly related issues such as invasive species and the effects of climate change on fish species in Europe. Climate change uncertainty and its minimization have been investigated by Kiesel et al. (2019) which further led to the investigation of eight different climate ensemble sub-selection methods from larger ensembles. Methods that take into account the diversity of climate models perform better in depicting the actual climate change signal on streamflow than methods that select individual models despite those were based on minimizing the historical model errors (Kiesel et al. 2020b). Kiesel et al. (2020a) has further studied the influence of hydrological model performance on species modelling. This allows suggestions on when a hydrological model is sufficiently adapted for the modelling of species. We could show that (1) it is important that the model is calibrated to the hydrological requirements of the species and (2) that a certain model performance is necessary to achieve stable results in the ecological assessment.
Similar to benthic invertebrates, many freshwater fish in Europe will be severely threatened by climate change in the future. This particularly affects species in the Mediterranean region. A total of 443 different European freshwater fish species were analysed and a list of particularly sensitive fish species was compiled. According to this, species in extreme habitats, which are characterised by heat and drought, are particularly threatened by climate change. They are particularly vulnerable to a further deterioration of living conditions. While the list includes many fish species that are of low commercial and fishing importance, they play an important role in food webs and ecosystems and thus have an impact on the entire system. Management measures should accordingly focus not only on fisheries, but also on the conservation of biological communities and the complex services provided to nature and humans (Jarić et al. 2019).
Besides climate and land use change, invasive species are another large stressor in freshwater ecosystems, with some becoming a problem for native species - as predators, competitors for food and habitat or vectors of diseases. These species enter e.g. via aquarium/pond ornamental fish trade. Jarić et al. 2020 looked at the question of the "charisma" of such species, and noted that charismatic species are likely to be deliberately introduced more often than inconspicuous species. Moreover, the more frequently and the greater the number of individuals introduced, the greater the probability that a species will establish itself. In addition, the acceptance of attractive invasive species in society is higher than that of invasive species with little charisma. This can hinder conservation measures designed to contain the spread of a species.
Finally, we were centrally involved in analyses on the impacts of major dam projects planned and under construction worldwide on freshwater biodiversity, particularly freshwater megafauna (Zarfl et al. 2019). The study examined which future sites could potentially have an impact on freshwater megafauna’ species and thus point to possible conflicts between climate protection and species conservation if hydropower is expanded as a source of renewable energy to combat climate change. The dam projects could be categorized with regard to their potential impact on the diversity of freshwater species and quantified in which regions ecosystems and biodiversity could be particularly endangered. Many of the planned dams will be built in areas with the greatest diversity of freshwater species, such as the Amazon, Congo, Mekong and Ganges rivers.
Contact for sub-project 5
Sonja Jahnig, sonja.jaehnig@igb-berlin.de