The seed germination assays continued approximately 5 months from April to September in 2020. Environmental factors influenced the species composition of soil seed banks differently between two plant community types.RDA showed that the environmental variables explained 83.4% and 61.0% of the total variations in species composition in Carex and Phragmites sites . Soil pH explained the most variation , followed by TN , SSC , AP and SWC in Carex sites. Species that germinated from soil seed banks in the Carex-dominated natural wetlands were on the right side of the graph, whereas the farmed fields were on the left . However, soil nutrients including TP , NO3-N and TN , and SSC explained most variation in species composition of soil seed banks in Phragmites sites. Species that are common in Phragmites-dominated natural wetlands were distributed on the lower part of the graph, whereas species from the farmed Phragmites sites are on the upper part of the graph.Many wetlands were lost for agricultural reclamation in China due to their fertile soils and plentiful water.Modern agriculture expanded rapidly with producing much more food and feeding larger population but sacrificing natural wetlands. More than 50% of wetlands have been lost and agricultural activities have been identified as primary drivers for these changes.The rapid conversion of natural wetlands to agricultural lands drained water, affected circulation of materials and eventually changed soil–water environment of wetlands. In our study, land reclamation on Carex-dominated wetlands reduced available nutrients of soils such as SOC, NH4-N, but increased the degree of soil salinization from non-salinity to moderate salinity compared with natural wetlands . These changes could deteriorate soil fertility, productivity and nutrient content,nft hydroponic system and would make restoration more difficult even if the hydrology could be restored.

On the contrary, agricultural reclamation on Phragmites-dominated wetlands relieved stress of soil salinization and improved the capacity of soil nutrient through human intervention such as application of organic fertilizer to satisfy the growth of crops. Therefore, land reclamation caused by agricultural activities in wetlands of the Songnen Plain changed the soil characteristics. The success of converting farmlands into wetlands with ecosystem functions and biodiversity similar to their original position depends on the availability and potential of seeds in soils . The agricultural reclamation on Carex-dominated wetlands decreased seed density and species richness of soil seed banks . The dominated sedges were missing from the seed banks after farming and seeds of these species could not disperse to restored sites because the hydrological connection was interrupted . Previous studies suggested that critical components of the vegetation were missing from farming lands , which was identical with our result. Furthermore, Echinochloa crusgalli and Typha angustifolia dominated in the farming fields and might expand because of their vast ecological amplitude and the absence of native wetland species, which would further reduce biodiversity . Moreover, the seed germination was prevented by the stress of soil salinization, the decreasing availability of water and soil nutrient after farming.Thus, our result indicates that dominant species were not retained in seed banks, which may make sedge meadows difficult torestore via natural recolonization. The composition and structure of soil seed banks changed after farming in Phragmites-dominated wetlands even if there was no significant difference of seed density and species richness of soil seed banks . The similarity between the soil seed banks and above ground vegetation increased after farming . This phenomenon might be caused by following reasons: firstly, Phragmites australis is one of the most widespread species all over the world. Phragmites australis has a wide ecological amplitude and both sexual and asexual reproductions , so it is more adaptable in the variable environmental conditions, particularly the disturbance like farming.

Secondly, Typha angustifolia, Phragmites australis and Scirpus planiculmis dominated in the natural wetlands , which have better tolerance to high soil salinization, retain vitality in the soil seed banks, and survive with tillage from the deep to the surface . Thirdly, the agricultural reclamation improved soil fertility by irrigation and fertilization with alleviating soil salinization and increasing available soil nutrient, which reduced the gap of seed germination from soil seed banks . Overall, there are potentials to restore the farmed wetlands using soil seed banks, but the landscapes have a low probability of resembling those that existed historically. The Songnen Plain serves as one of largest saline-sodic areas around the world, and is experiencing threaten of soil salinization because of climate change and farmland irrigation . Recently, hydrological projects have been carried out in the Songnen Plain to supply water and facilitate vegetation restoration in wetlands, but the results were still not satisfactory. Our research found that saline-alkaline stress was an important factor which restricted the potential of wetland restoration using soil seed banks in the Songnen Plain . The saline-alkaline stress significantly decreased seed density and species richness of soil seed banks through limiting or preventing germination of seeds, especially freshwater species like sedges in the farmed Carex fields . However, the soil seed banks were not significantly affected by saline-alkaline stress in the Phragmites fields. That is because as an invasive species, Typha angustifolia dominated in the soil seed banks and it adapted to conditions with slight salinealkaline stress. Saline-alkaline stress is harmful for the growth of plant and germination of seed mainly through ion toxicity and osmotic stress, which can restrict availability of water and damage plant cells and tissues.

The seeds are devitalized by serious soil salinization or remain dormant until the soil quality and environmental condition are improved . Our research indicates that even slight saline-alkaline stress could affect seed germination and restrict the potential of vegetation restoration in the Songnen Plain. Agriculture in the European Union is generally highly specialized and intensive , resulting in an abundant food production, but also often leading to the degradation of natural resources . In addition, labor productivity and farm income is low in many farming systems in the EU . From a social point of view, quality of life in rural areas in the EU is often perceived to be low as well, especially in the poorer countries . To improve sustainability, a balanced attention for social, environmental and economic system dimensions is important . Inadequate management of natural resources, for instance, can be seen as a failure to understand how social, economic and environmental dimensions are interrelated . Interrelation of these dimensions often results in feedback loops in a system, resulting in non-linear behavior. This makes it challenging to assess and interpret the effect of shocks, stresses and management options on the provision of system functions. In response to this challenge, several resilience frameworks have been developed to study agricultural systems . Sustainability and resilience can be seen as two complementary concepts . Resilience in the form of robustness, adaptability or transformability is needed to maintain or improve sustainability. At the same time, sustainability is needed to ensure the access, availability and quality of resources to buffer shocks and set in motion adaptation or transformation. For the context of a farming system , Meuwissen et al. define resilience as the ability to ensure the provision of the system functions in the face of increasingly complex and accumulating shocks and stresses. By emphasizing the importance of system functions, Meuwissen et al. , provide a practical way to combine the concepts of resilience and sustainability in a complementary way.

To better understand the potential dynamics of farming systems, current as well as future sustainability and resilience need to be studied. Current resilience of European farming systems was for instance studied by Nera et al.,Meuwissen et al., Paas et al.,and Reidsma et al..Towards the future, system behavior may differ according to the development of factors that are exogenous to the farming system, especially when shocks and stresses increase or when enabling conditions for changes are realized. Trespassing critical thresholds could for instance initiate cascading effects leading to a system decline.To avoid this,hydroponic nft system institutional actors may deliberately aim at changing threshold levels to enable innovation that provides an alternative to the dominant ways of producing.Quantitative models are often used to assess, ex-ante, system performance and behavior.Different types of studies and associated models can be distinguished . Based on statistical models, projections or predictions can be made about the average and probable performance for future conditions . However, because statistical models depend on patterns from the past, only a limited range of all possible futures will be captured. Including a broader range of possible futures increases the opportunity to evaluate farming system resilience under different exogenous conditions that are all possible to happen. Incompatibility of farming systems with certain futures can be seen as a sign of non-resilience in case those systems have no capacity to adapt or transform. In itself, comparing farming systems with a broad range of futures directly contributes to foresight information supporting the capacity to anticipate shocks, which is seen as important for resilience . In so-called explorations, optimization models and system dynamics models can consider multiple possible futures, using scenarios capturing uncertainty on climate change and socio-economic developments. However, these models need parameters which are sometimes also derived from statistical models based on past and current trends. Moreover, optimization models are of limited use for modelling dynamic transformations, as they are generally static. Participatory methods can take into account multiple scenarios and allow for input regarding transformational change and resilience concepts such as critical and interacting thresholds . It should be noted, however, that qualitative methods also are influenced by input from statistical sources and experts that extrapolate past trends into the future.

We argue that quantitative and qualitative approaches can be complementary. Participatory methods can be quick, interactive and flexible to start discussions about sustainability and resilience in the future, thus laying a base for further discussions and quantitative model-based analyses.Participatory methods allow for taking into account the voice of individual stakeholders as well as support stakeholder discussions to arrive at a common understanding and a shared vision for improvement of the system or problem under study. Stakeholder participation is important as stakeholders are usually involved in follow-up processes and thus need to agree with the problem definition and proposed action plan.Participatory input is valuable because system actors are able to provide empirical knowledge about their system that reduce knowledge gaps of researchers.Vice versa, participatory methods are also important to identify the boundaries of local knowledge.Stakeholder’s perceptions are particularly precious, as they can explain or drive system dynamics as stakeholders are important components of socio-ecological systems.Hence, participatory methods can provide a first exploration of farming system structure, mechanisms, performance and behavior in possible futures. Discussions with stakeholders about future change can be challenging because stakeholder’s mental models usually focus on maintaining the status quo with little imagination of alternative futures.Other limitations for discussing farming system transformations may relate to the focus of experts on improving efficiency, vested interests, co-dependencies among system actors and institutional path dependence. Participatory methods should therefore provide opportunities to go beyond the usual extent of stakeholder’s mental models. Alternative systems, that avoid critical thresholds and increase sustainability and resilience simultaneously, should be explored, and new strategies to realize those alternative systems identified. To ensure the soundness of intended pathways towards the future, alternative systems need to be compatible with possible future developments of exogenous factors as projected in different future scenarios. High compatibility of desired alternative systems with future scenarios increases the likelihood that those more sustainable and resilient systems will be realized. Consequently, this also decreases the likelihood that critical thresholds will be exceeded, resulting in farming systems with even lower sustainability and resilience levels. We argue that a quick and flexible assessment of future resilience and sustainability of farming systems is still lacking in literature. In response to this research gap, this paper presents a participatory, integrated and indicator-based method to improve understanding of farming system sustainability and resilience.