Floodplains are disturbance-dominated systems driven by floods, but they are also hotspots of biodiversity. Particularly the temporary water bodies, which occur after floods for a period of time, support species rich and often unique species communities, such as amphibians. Floodplain modifications in the context of river regulation, land use, or human flood protection frequently disconnected the floodplains from the flooding regime of the rivers, which caused the loss of habitat heterogeneity and consequently the loss of biodiversity. In Central Europe, up to 90% of the floodplain areas are hydrologically altered, but the increasing activities in European floodplain management during the last years, now aim to integrate targets of nature conservation. This development implies an increasing need of floodplain ecosystem research for understanding and assessing the impacts of floodplain modifications on the ecosystems. Despite this increasing need, research on floodplain ecosystems focused on few taxonomic groups, mainly on plants, fishes, and invertebrates. Comparably little is known about the ecology and the habitat requirements of amphibians in floodplains.

Amphibians are typical floodplain species and may occur in dense and species rich communities in dynamic European floodplains. Accordingly, the alteration and the loss of these habitats is a major cause for the European and also the global amphibian decline. Most amphibian species are sensitive to changes in the hydrological regime in their habitats, because of their biphasic life cycle, which includes a semi-terrestrial adult stage of dispersal and reproduction and an aquatic larval stage. However, amphibian habitat requirements in Europe have been mainly studied in non-dynamic habitats, such as cultural landscapes, but the requirements may differ between habitats of different levels of hydrological dynamics. These insufficiently understood differences might have strong influence on the success of conservation measures that aim to improve floodplain habitats for amphibians.

Research on amphibians in temperate floodplains has been conducted in the temperate regions in US and Australia, but this knowledge is hardly applicable for the conservation of amphibians in Central Europe due to differences in the species biology and ecology. Research on amphibians in Central European floodplains remained scarce, although 16 of the 21 native amphibian species are described to use, to prefer, or even to depend on dynamic wetlands, such as floodplains. In a summary of studies on Central European amphibians in floodplains, Kuhn et al. (2001) pointed out that a deeper understanding is required for successful amphibian conservation measures in these dynamic wetlands. One reason for this gap might be the methodological challenge of floodplains, because study designs have to be flexible to cope with the unpredictable environmental circumstances that may vary among years.

The main objective of this dissertation was the formulation of floodplain management guidelines supporting amphibian conservation in Central Europe. For the direct implication, a simulation model was developed as a conservation planning tool that visualises how changes in the hydrological regime and anthropogenic impacts drive an amphibian population in a dynamic floodplain. The leading question of this dissertation was: Which aspects of amphibian habitat requirements in floodplains need to be considered for establishing successful conservation measures?

For developing a better understanding of the complex relationships between amphibians and the dynamic floodplain habitat, this dissertation is structured into three thematic areas, each published or submitted for publication. The three thematic areas build on each other: First, an empirical data collection in a dynamic temperate floodplain and thereon developed statistical habitat models for four amphibian species answered the question, which microhabitat structures are required by amphibian species in a floodplain habitat. Second, the development of the simulation model focused on one of the four species in the study area. The simulated scenarios answered the question whether the microhabitat structures identified in the habitat models are sufficient to improve an altered floodplain habitat for the species. This model is designed to be directly applicable as a conservation planning tool. The third thematic area led on to a more general perspective and concentrated on the identification of basic patterns in amphibian community organisation in floodplains. The therefore conducted literature analysis answered the question whether similarities in the life history traits of 19 amphibian species allow dividing the species into functional groups that reflect their spatial arrangement in temperate European floodplains.

In the following, the three thematic areas of this dissertation are summarised. Their main findings were used to derive the floodplain management guidelines for the conservation of strongly threatened amphibian species in Central Europe.

The three thematic areas of this dissertation on amphibians in Central European floodplains

In the case study, “Environmental determinants and temporal variation of amphibian habitat use in a temperate floodplain”, statistical habitat models were developed on the basis of 41 environmental variables and the abundances of four representative amphibian species, which were collected in the Upper Middle Elbe floodplain, Germany. The species were the moor frog (Rana arvalis), the firebellied toad (Bombina bombina), the common spatefoot toad (Pelobates fuscus), and the European treefrog (Hyla arborea). The survey was conducted from March till July in the years 2010 and 2011. The two years differed in their hydrological conditions. In the first year, two spring floods resulted in wet habitat conditions, whilst no spring flood occurred in the second and accordingly dryer year. The results of the habitat models showed that most habitat preferences of the species were species-specific and varied between the two years of different hydrological conditions. Only the variable pond surface area was relevant for all four species in both years. The second most relevant variable was the hydroperiod of ponds, which confirmed our hypothesis that this habitat factor is a major driver determining the distribution and the abundances of the species at the floodplain ponds. The species responded differentiated to the hydroperiod of the ponds. Thus, the variable was more relevant in the dryer year and was not influencing the abundance of R. arvalis. Notwithstanding, no further variable was of such relevance for the species abundances in both years. The abundances of the species were positively correlated with the two habitat variables in most cases. Due to the very small and shortliving ponds in the study area, these correlations indicate that the surface area and the hydroperiod needs to be small and temporary, but of sufficient size and lengths. This finding underlines the high relevance of small temporary ponds as amphibian habitats. In addition, temporal variability in amphibian habitat use was detected. In the year with frequent floods, water chemistry and morphology of the ponds were mainly relevant, whereas aspects of vegetation were more important in the year with the lower water level. Temporal variability in habitat use is typical for species in dynamic habitats that are able to flexibly respond to changing habitat conditions. However, it is a challenge for species conservation, because knowledge is needed about the habitat requirements of the species under different environmental conditions. Habitat models mostly base on short-term case studies and might fail detecting the temporal variations in species habitat requirements. Nonetheless, statistical habitat models are a good basis for developing simulation models, which are able to visualise complex species-habitat interactions over longer time periods.

The simulation model “FloMan-MF: Floodplain Management for the Moor Frog – a simulation model for amphibian conservation in dynamic wetlands“, visualises the annually varying environmental conditions in the Upper Middle Elbe floodplain and the spatial arrangement and population development of an amphibian population on the example of the moor frog (R. arvalis). The model provides the opportunity to visually predict the impacts of changes in the hydrological regime or of floodplain modifications on the simulated amphibian population. The simulation model is designed as a basic model, directly applicable in amphibian conservation. The required software, NetLogo, is available free of charge on the Internet (http://ccl.northwestern.edu/netlogo). To adjust the basic model to other species, rates of reproduction, survival, migration, and habitat preferences need to be adjusted. For simulating another flooding regime, adjusting the probability-values for flood-events is possible. The result of the statistical habitat models that the pond’s surface area is the habitat characteristic with the largest influence on the species abundances has been confirmed by the sensitivity analysis of the simulation model. The hydroperiod of the ponds, which was the second most relevant variable in the habitat models, could not be analysed in the simulation model for technical reasons, but the availability of suitable temporary ponds in the simulated landscape had a major effect on population persistence. The relevance of the ponds’ bank slope for the moor frog could also be confirmed by the statistical habitat models as well as by the sensitivity analysis of the simulation model. The optimisation of the ponds surface area and bank slope can directly be realised as a conservation measure at pond-level.

The simulated scenarios demonstrated the influence of the spatial arrangement of the ponds and their microhabitat attributes on the modelled population. The basic scenario (SC0) simulated the hydrological conditions observed in the study area. The hypothesis for the management scenarios was that habitat improvement by implementing certain microhabitat structures might stabilise an amphibian population under hydrologically altered habitat conditions. For demonstrating conservation planning, hydrological alteration was simulated with no more floods occurring (SC1), followed by three scenarios simulating different conservation measures (SC2-5). Conservation success was a measure of population size after 50 years. Simulating altered habitat conditions without conservation management (SC1), the modelled population became rapidly extinct (extinction rate after 50 years: 100%). The establishment of seven temporary ponds with sufficiently long hydroperiods (SC2) decreased the extinction rate after 50 years to 51%, but the number of spawning females was very low. The additional improvement of habitat suitability by optimising the surface area of five of the new ponds (SC3) further decreased the extinction rate to 6%, although the number of spawning females was still low. Unexpectedly, the extinction rate increased to 9% after enhancing additional microhabitat structures at the new ponds (SC4). This result demonstrates the complex spatio-temporal species-habitat interactions that could not clearly be addressed by the statistical habitat models. The population stabilised at a low population size when, instead of optimising further microhabitat structures as in SC4, a permanent pond in a key position was transformed into a highly suitable temporary pond (SC5).

For assessing the impacts of changes in the hydrological regime in floodplains on amphibian communities, the third focus of this dissertation led on “Functional groups in amphibian communities and their spatial arrangement in temperate European floodplains”. This chapter answered the question whether similarities in life history traits allow dividing amphibian species into functional groups (FGs) that reflect their spatial arrangement in floodplain habitats in temperate Europe. The hypotheses were that the amphibians are organised into four FGs and that the most typical floodplain amphibians display traits that are expected for species in dynamic environments by ecological theories, i.e. the River Template Theory. To address this issue, a non-systematic literature analysis was conducted to assign 19 European amphibian species into FGs, based on similarities in 19 in their life history traits recovered from literature and analysed by an ordination method and group tests. The species constellations of the FGs were then compared with the species constellations grouped by literature descriptions for six floodplain habitat types along the environmental floodplain gradient of hydrological variability. The similarities in the species traits resulted as expected in four FGs and one outgroup species, but the species constellations of the FGs partly reflected those constellations described for the different habitat types. The 16 species, which are described by literature to occur in floodplains, were spread into three FGs and within each, some species occur in open and highly dynamic habitats, whilst others utilise in forested and less dynamic habitat patches, or are even unspecified in their habitat use. Nonetheless, the FGs indicate a spatial arrangement along the floodplain gradient, because most species of the highly dynamic patches were grouped into FG-I, whereas those species described to be absent from floodplains were restricted to FG-IV. Furthermore, FG-IV displayed trait-modalities typical for species in non-dynamic environments, whilst FG-I mainly displayed trait-modalities expected for species in dynamic habitats. These findings indicate differentiated ecological adaptations within amphibian communities in floodplains and underline the high relevance of the habitat heterogeneity of temperate floodplains for sustaining species rich amphibian communities. A natural or close-to natural state of a floodplain is indicated by high amphibian species richness and the presence of floodplain-typical species. The alteration of the hydrological regime in a floodplain, likely cause species shifts in the amphibian communities, which might result in the loss of the typical floodplain species.

In addition to these findings, some literature sources could be used to distinguish between the species constellations in floodplains with natural respectively altered hydrological dynamics. These findings indicate that many floodplain amphibian communities in Central Europe display modified patterns in species composition, which might be a result of the high degree of floodplain alteration in this region. These patterns in amphibian communities and the potential species shifts have to be considered for amphibian conservation actions in the context of floodplain restoration.

Concluding, the results of this dissertation demonstrate the complexity of amphibian-floodplain habitat interactions that need to be considered for establishing successful conservation measures. These considerations include the differentiated ecological adaptations to hydrological dynamics and the related species occurrence at different habitat types along the floodplain gradient, the spatial and temporal variations in the amphibian habitat use, and the mostly species-specific habitat requirements. Apart from that, sufficiently small and temporary water bodies seem to be generally important amphibian habitats in floodplains. Based on the key findings of the three thematic areas of this dissertation, I formulated the management guidelines for amphibian conservation in floodplains.

The three floodplain management guidelines for amphibian conservation in Central European floodplains

The guidelines are formulated for three floodplain management situations and are mentioned in the following starting from guideline 1, which can be realised probably most easily but provides only a minimum solution for amphibian conservation, to guideline 3, which is most difficult to realise but provides an optimal solution for the conservation of the floodplain biodiversity and thus species rich amphibian communities.

Guideline 1 (minimum solution): Improving the habitat quality for single amphibian species by implementing microhabitat structures

Due to the mainly species-specific habitat requirements of Central European amphibians, the creation of certain microhabitat structures in altered floodplains improve the habitat for single amphibian species and species with similar habitat requirements. However, a major limitation of this solution is that floodplain habitats without natural dynamics often require continuous efforts to maintain their quality for amphibians. Although this guideline is the least effective one of the three guidelines, it can be realised in altered floodplains where there is no opportunity to restore the hydrological habitat dynamics or a close-to natural flooding regime. For planning management actions at pond-level, the developed simulation model can simulate each possible combination of microhabitat structures, aiming to maximise the species abundance at the ponds as well as in the simulated floodplain landscape. A systematic analysis allows identifying the constellation of ponds and microhabitat structures, which is optimised for the modelled species. The results of the statistical habitat models provide information on required microhabitat structures for four representative amphibian species typical for Central European floodplains. However, to re-establish species rich amphibian communities, habitat heterogeneity with a high amount of differently structured small and temporary water bodies is required.

Guideline 2 (intermediate solution): Establishing a dynamic wetland with a hydroperiod gradient

Amphibian habitats in riverine landscapes can be improved or established by the creation of suitable complexes of water bodies. Particularly relevant are small temporary ponds that are filled with water from the river by spring floods or by increased ground water for different periods of time. Related to the different habitat requirements of the amphibian species in floodplains, the hydroperiod gradient in the landscape has to range from shortliving temporary ponds up to permanent ponds that may additionally function as refuge in dry years. A recent catalogue of measures for the establishment of floodplain water bodies is for example available in the context of natural water retention measures (Strosser et al. 2015). These floodplain ponds can be optimised for amphibians by implementing specific microhabitat structures known as species requirements. For conservation planning, the simulation model can be used to identify key positions for new ponds to maximise the conservation success in terms of the overall species abundance in the landscape.

Guideline 3 (optimal solution): The restoration of the hydrological floodplain landscape dynamics

Due to the complex species-habitat interactions of amphibians in floodplains and their organisation in differently adapted functional groups, the optimal conservation action is the restoration of an at least close-to natural flooding regime. This can be reached by the re-connection of the floodplain to water level fluctuations of the river that allows flooding and creates the heterogeneous habitat patches required for species rich amphibian communities. Nonetheless, the natural flooding regime often conflicts with other land use interests or flood prevention and is further impeded by the strong regulation of European rivers that often display only reduced water level fluctuations. However, this optimal solution for the conservation of threatened amphibian species in Central European floodplains is supported by the recent developments in the European floodplain management. Particularly the natural water retention measures provide a huge potential for the conservation of the floodplain biodiversity, including species rich amphibian communities with floodplain-typical species.