Logo: to the web site of Uppsala University

Restored and constructed semi-natural wetlands are increasingly used in the agricultural landscape to intercept nutrients from surface waters. Vegetated surface-flow wetlands remove more nitrogen (N) than those without vegetation. However, changes in N removal over time as differently vegetated wetlands progress from early successional stages to mature systems are less investigated. We monitored three different types of initial planting over the course of 12 years, with the aim to examine how planting of newly constructed wetlands affects long-term N removal. All our data were collected in an experimental wetland facility in south-western Sweden. The facility consists of 18 identical small (ca. 25 m²) surface-flow wetlands, simulating semi-natural wetlands in an agricultural landscape. Initially, the 18 wetlands were randomly divided into three treatments (vegetation types) with six replicates each and planted with (1) emergent vegetation, (2) submerged vegetation and (3) no vegetation for free development. Vegetation succession afterwards progressed uninhibited in all wetlands. Emergent vegetation wetlands initially removed more N than both submerged vegetation and free development wetlands. We found that N removal in submerged vegetation and free development wetlands increased with ecosystem age, whereas N removal in emergent vegetation wetlands did not. N removal in all three vegetation types converged when the wetlands reached a more mature state, around 8 years after wetland construction. However, although all wetlands contained emergent vegetation in year 8, the proportion of emergent vegetation cover and vegetation composition still differed substantially between wetland types. Our study indicates that it is not the cover of emergent vegetation per se which promotes higher N removal in more mature wetlands, but the maturation process itself; mature wetlands despite differing emergent vegetation coverage achieved equally high N removal. In conclusion, once wetlands reach maturity, beneficial effects of initial planting on N removal disappear.

Wetland area in agricultural landscapes has been heavily reduced to gain land for crop production, but in recent years there is increased societal recognition of the negative consequences from wetland loss on nutrient retention, biodiversity and a range of other benefits to humans. The current trend is therefore to re-establish wetlands, often with an aim to achieve the simultaneous delivery of multiple ecosystem services, i.e., multifunctionality. Here we review the literature on key objectives used to motivate wetland re-establishment in temperate agricultural landscapes (provision of flow regulation, nutrient retention, climate mitigation, biodiversity conservation and cultural ecosystem services), and their relationships to environmental properties, in order to identify potential for tradeoffs and synergies concerning the development of multifunctional wetlands. Through this process, we find that there is a need for a change in scale from a focus on single wetlands to wetlandscapes (multiple neighboring wetlands including their catchments and surrounding landscape features) if multiple societal and environmental goals are to be achieved. Finally, we discuss the key factors to be considered when planning for re-establishment of wetlands that can support achievement of a wide range of objectives at the landscape scale.

Wetlands in agricultural areas mitigate eutrophication by intercepting nutrient transports from land to sea. The role of wetlands for nutrient removal may become even more important in the future because of the expected increase in agricultural runoff due to climate change. Because denitrification is temperature dependent, wetland nitrogen (N) removal usually peaks during the warm summer. However, climate change scenarios for the northern temperate zone predict decreased summer and increased winter flows. Future wetlands may therefore shift towards lower hydraulic loading rate and N load during summer. We hypothesised that low summer N loads would decrease annual wetland N removal and tested this by examining 1.5-3 years of continuous N removal data from created agricultural wetlands in two regions in southern Sweden (East and West) during different periods. West wetlands showed relatively stable hydraulic loads throughout the year, whereas East wetlands had pronounced no-flow periods during summer. We compared East and West wetlands and tested the effects of several variables (e.g., N concentration, N load, hydraulic load, depth, vegetation cover, hydraulic shape) on annual absolute and relative N removal. We found no difference in annual N removal between East and West wetlands, even though summer N loads were lower in East than in West wetlands. A possible explanation is that stagnant water conditions in East wetlands suppressed decomposition of organic matter during summer, making more organic matter available for denitrification during winter. Absolute N removal in all wetlands was best explained by N load and hydraulic shape, whereas relative N removal was best explained by emergent vegetation cover and hydraulic shape. This study highlights the importance of design and location of agricultural wetlands for high N removal, and we conclude that wetlands in a future climate may remove N from agricultural runoff as efficiently as today.

Exacerbated by climate change, eutrophication and flood risks are pressing global issues of increasing ecological and societal relevance. A key driver of eutrophication is the excess use of nitrogen (N) fertiliser in agricultural landscapes, resulting in N exports from land to water. Flood events can increase N exports and further worsen eutrophication. Created wetlands in agricultural areas are recognised as important nutrient sinks, and the flood attenuating capabilities of urban wetlands and ponds are also well established. However, less studied is the potential to combine these two ecosystem services in agricultural landscapes to achieve multifunctionality. This study examines how dynamic water management for enhanced water storage capacity during flood events affects N removal and greenhouse gas (GHG) emissions in replicated experimental wetlands. We hypothesised that intermittently flooded wetlands would exhibit higher rates of N removal and nitrous oxide (N₂O) emissions, while emitting less methane (CH₄), compared to permanently flooded wetlands. However, our hypothesis was refuted. In this experiment, shallow wetlands exhibited high N removal and water retention simultaneously, whereas deep and large wetlands removed less N as water retention increased. No differences in the aqueous concentrations of N₂O or CH₄ were observed between wetlands with high or low water storage capacity, thus indicating similar emissions. Our results highlight the possibility of combining the two ecosystem services flood attenuation and N removal, without increasing GHG emissions, in wetlands of specific designs. This study contributes to the understanding of synergies and trade-offs in wetland multifunctionality.

Constructed wetlands (CWs) remove nitrogen (N) more efficiently as wetlands age and vegetation establishes. However, when CWs become heavily overgrown and filled with plant litter and root mats, channels form and N removal declines. Here, management may maintain high N removal. We tested two CW management options 1) restoration through macrophyte harvesting including root mat removal, and 2) subsequent installation of woodchips-based floating beds (WFBs) to compensate macrophyte loss. In a field experiment, using 16 heavily overgrown experimental CWs, we applied four treatments: i) macrophyte harvesting, ii) macrophyte harvesting and 5% of the wetland surface covered with WFBs, iii) macrophyte harvesting and 20% WFB cover, and iv) a control treatment (heavily overgrown). WFBs were planted with Glyceria maxima and Filipendula ulmaria before installation. N removal efficiency, removal rate and removal rate coefficient k_a were estimated on nine occasions. After the experiment, WFBs were removed and plant biomass accrual, N assimilation, and denitrification gene (nirS, nirK, nosZI and nosZII) abundance on plant roots and woodchips were studied.

Macrophyte harvesting significantly improved N removal of heavily overgrown CWs, whereas WFBs installation only improved N removal in harvested treatments on some occasions. Both G. maxima and F. ulmaria grew well on WFBs. Relative biomass production, root length and root surface area were higher for G. maxima than for F. ulmaria whereas biomass N assimilation was higher for F. ulmaria. Denitrification gene abundance was higher on plant roots than on woodchips while G. maxima hosted higher root denitrification gene abundance than F. ulmaria. We conclude that macrophyte harvesting improves N removal in heavily overgrown CWs. Further long-term field studies are needed to precisely evaluate the contribution of WFBs to N removal in CWs.