Modelling climate change impacts on nutrients and primary production in coastal waters

Highlights

• Integrated modelling methodology exploring the impacts of climate change on nutrient loadings and phytoplankton community.

• The nutrients loadings in the Zero river basin and the salt-marsh Palude di Cona in Italy were investigated.

• Integration of GCM-RCM climate projections with the hydrological model SWAT and the ecological model AQUATOX.

• Results show changes in the phytoplankton community: reduction in biodiversity and more intense blooms in the summer.

Abstract

There is high confidence that the anthropogenic increase of atmospheric greenhouse gases (GHGs) is causing modifications in the Earth's climate. Coastal waterbodies such as estuaries, bays and lagoons are among those most affected by the ongoing changes in climate. Being located at the land-sea interface, such waterbodies are subjected to the combined changes in the physical-chemical processes of atmosphere, upstream land and coastal waters. Particularly, climate change is expected to alter phytoplankton communities by changing their environmental drivers (especially climate-related), thus exacerbating the symptoms of eutrophication events, such as hypoxia, harmful algal blooms (HAB) and loss of habitat. A better understanding of the links between climate-related drivers and phytoplankton is therefore necessary for projecting climate change impacts on aquatic ecosystems.

Here we present the case study of the Zero river basin in Italy, one of the main contributors of freshwater and nutrient to the salt-marsh Palude di Cona, a coastal waterbody belonging to the lagoon of Venice. To project the impacts of climate change on freshwater inputs, nutrient loadings and their effects on the phytoplankton community of the receiving waterbody, we formulated and applied an integrated modelling approach made of: climate simulations derived by coupling a General Circulation Model (GCM) and a Regional Climate Model (RCM) under alternative emission scenarios, the hydrological model Soil and Water Assessment Tool (SWAT) and the ecological model AQUATOX. Climate projections point out an increase of precipitations in the winter period and a decrease in the summer months, while temperature shows a significant increase over the whole year. Water discharge and nutrient loads simulated by SWAT show a tendency to increase (decrease) in the winter (summer) period. AQUATOX projects changes in the concentration of nutrients in the salt-marsh Palude di Cona, and variations in the biomass and species of the phytoplankton community.

Introduction

There is high confidence that anthropogenic emissions of greenhouse gases (GHGs) are the main reason of the current Earth's energy imbalance (Hansen and Sato, 2004). As a consequence, this is causing the warming of the atmosphere (von Schuckmann et al., 2016) with an expected rise of the global mean temperatures by 0.3 to 4.8 °C by the end of the 21st century relative to the period 1986–2005 (IPCC, 2013), and the alteration of the water cycle as a result of changes in the global moisture recycling among atmosphere, lands and oceans (Levang and Schmitt, 2015).

Coastal ecosystems, together with the ecological and socio-economic services they provide, could be among those most affected by the ongoing climate change (Harley et al., 2006; IPCC, 2014). Coastal waterbodies such as estuaries, bays and lagoons, being transitional systems located at the interface between land and sea, will be subjected to the combined modifications taking place in the atmosphere, in the oceans, and over the land surface (Raimonet and Cloern, 2016).

In particular, a wide number of studies has shown the links between changes in climate and phytoplankton in coastal ecosystems (Harding et al., 2015; Holt et al., 2016; Quigg et al., 2013; Sommer and Lewandowska, 2011; Winder and Sommer, 2012). Phytoplankton is responsible for a large share of photosynthesis and primary production of coastal areas, and plays an essential role in several biogeochemical processes such as carbon, nutrient and oxygen cycles (Paerl and Justic, 2011; Winder and Sommer, 2012). Consequently, changes in phytoplankton dynamics and composition may have relevant repercussions on the higher trophic level of coastal ecosystems (Hernandez-Farinas et al., 2014; Schloss et al., 2014).

Environmental drivers regulating phytoplankton dynamics such as water temperature, light penetration, tides, salinity and nutrient availability (Cloern, 1996), are highly sensitive to climate change. The interactions among these drivers, including their influence on primary production in coastal areas, form a complex, nonlinear system that can be defined by three main components: (1) Climate, which delineates changes in atmospheric conditions; (2) Coastal watershed hydrology, which describes the changes in the delivery of freshwater, nutrients, sediments and pollutants to coastal waters; and (3) coastal ecosystem, which combines changes in climate and watershed hydrology to identify the impacts on the coastal environments. As proposed by Xia et al. (2016) for freshwater ecosystems, the simplified conceptual model depicted in Fig. 1(a) describes the interactions among these components and their effects on the phytoplankton communities of coastal ecosystems.

The conceptual model helps to illustrate the cascading mechanisms linked to climate change and leading to direct and indirect impacts on coastal phytoplankton. In particular, the distribution, abundance and structure of phytoplankton communities, as well as their phenology and productivity, are changing in response to warming and stratifying waters (Hunter-Cevera et al., 2016; Lassen et al., 2010; Weisse et al., 2016). Specifically, a prolonged thermal stratification affect phytoplankton sinking velocity and can advantage smaller and buoyant species of phytoplankton (Diehl et al., 2002; Huisman et al., 2004; Marinov et al., 2010). Moreover, a prolonged thermal stratification in the water column can suppress the upward flux of nutrients from deep layers, resulting in more frequent nutrient-depleted conditions at the surface, and further advantaging those species that are already winners in the competition for nutrients (Schmittner, 2005).

Variations in precipitation patterns and increased frequency of extreme events such as heavy rainfalls and droughts can alter those physical processes (e.g. runoff, leaching, percolation) that regulate the delivery of freshwater and its load of nutrients, sediment and pollutants to coastal waterbodies (Hagy et al., 2004; IPCC, 2014; Moss et al., 2011; Xia et al., 2016), with consequences on their nutrient and salinity concentrations (Dimberg and Bryhn, 2014). Decrease in precipitation could reduce river flow and both nutrient loadings and dilution (Whitehead et al., 2009). Increased winter precipitation and more frequent and severe flood events during summer will increase runoff and associated wash-off of nutrients and sediments (Jeppesen et al., 2009a, Jeppesen et al., 2009b; Najafi and Moradkhani, 2015; Sterk et al., 2016; Whitehead et al., 2009). These events may alter the nutrient budget and change the macronutrient stoichiometry (N:P:Si) of coastal waters (Statham, 2012; Winder and Sommer, 2012), resulting in more frequent eutrophication events and appearance of undesirable phytoplanktonic species such as cyanobacteria (Domingues et al., 2017; Garnier et al., 2010; Kaur-Kahlon et al., 2016).

Salinity of coastal waterbodies such as lagoons depends on the balance between marine and freshwater (either from surface of subsurface) inputs, often altered by climate-induced changes in the dynamics of runoff regimes and sea level rise (Fichez et al., 2017; Telesh and Khlebovich, 2010; Vargas et al., 2017) with consequences in structure, dominant species and biomass production of the plankton communities (Dhib et al., 2013; Flöder et al., 2010; Kaur-Kahlon et al., 2016; Olenina et al., 2016).

Sea level rise might have further important consequences on coastal phytoplankton, especially in shallow waters. For example, shallow lagoons can have well-developed benthic phytoplankton communities that can contribute to a major portion of the fixed carbon in the system (Parodi and De Cao, 2002). Increased water depth due to sea level rise will cause benthic phytoplankton to capture a smaller proportion of the solar radiation due to the stronger light attenuation in the water column (Brito et al., 2012).

Additionally, other anthropogenic disturbances (pollution, water withdrawal, etc.) may act in synergy with climate change and impact phytoplankton in coastal waters (Liu and Chan, 2016; Rabalais et al., 2009). For example, nutrient and pollutant loads are expected to increase in the coming decades due to population growth, increased use of inorganic fertilizers, manure and pesticides, increased fossil fuel burning (associated emission of NO_x), and expansion of sea-based activities such as aquaculture (Bouwman et al., 2009; Burkholder et al., 2007; Verdegem, 2013). On the other hand, factors such as increased production costs of fertilizers (Blanco, 2011), and promotion of good agricultural practices (GAPs) and integrated nutrient management (INM) might balance out the increase fertilizer usage.

The described effects generated by climate change and other anthropogenic stressors, and their balancing and reinforcing interactions, are thus expected to change the dynamics and composition of coastal phytoplankton communities, and increase the frequency and abundance of eutrophication events and related symptoms such as hypoxia, harmful algal blooms (HAB), unsightly scums and loss of habitat (Lloret et al., 2008; Moore et al., 2008; O'Neil et al., 2012; Paerl and Paul, 2012). While being able to project these effects is of fundamental importance, the complexity of coastal systems makes however difficult and challenging to attribute changes in the abundance and composition of coastal phytoplankton to specific causes (Facca and Sfriso, 2009) (Statham, 2012).

In this view, the integration of climate simulations and mechanistic environmental models can become a valuable tool for the investigation and projections of phytoplankton dynamics under climate change. The integration of models has already been promoted as a tool able to support Integrated Coastal Zone Management (ICZM) and Integrated Water Resource Management (IWRM), as it allows to represent the complexity of such systems by facilitating the simulation and combination of multi-faceted drivers such as climate conditions, nutrient availability, water withdrawal, and other pressures (Brush and Harris, 2010; ComEC, 2000). Integrated modelling can help to better understand and reproduce the cause-effect relationships between different components of complex systems such as coastal waterbodies, and they have become necessary tools for managers and decision makers who have to consider impacts on different end-points (i.e. social, economic, environmental) (Jakeman and Letcher, 2003). Different types of models (Fig. 1(b)) can be adopted and integrated to study the interactions and feedbacks among the three components (climate, watershed hydrology, coastal ecosystem) of the conceptual model depicted in Fig. 1(a). In the last decades, the adoption of mechanistic models has become a popular tool for assessing the impacts of climate change on hydrological and abiotic components of aquatic systems (Vohland et al., 2014). However, the majority of studies have analyzed the impacts of climate change on single environmental aspects such as watershed hydrology and water availability (Amin et al., 2017; Leta et al., 2016; Trinh et al., 2017), loadings of nutrients (Huttunen et al., 2015; Piniewski et al., 2014) and sediments (Bussi et al., 2016; Samaras and Koutitas, 2014), and water quality (Nielsen et al., 2014; Wilby et al., 2006). Moreover, these studies are limited to the assessment of the impacts on abiotic components, and are not able to model the consequences on phytoplankton and higher trophic levels. Although the number of studies that attempt to integrate climate simulations and process-based models for assessing the impacts of climate change on primary producers and aquatic ecosystems is growing (Elliott, 2012; Glibert et al., 2014; Guse et al., 2015; Longo et al., 2016; Mooij et al., 2007; Taner et al., 2011; Trolle et al., 2015), with the merit of trying to address the impacts at the ecosystem level, this number is still very limited and mainly focused on specific types of water bodies, such as lakes (Mooij et al., 2010).

Therefore, there is need for additional studies able to provide new and improved methods to investigate the independent and combined effects of climate change and other anthropogenic stressors on aquatic ecosystems (Trolle et al., 2014). This is particularly true for coastal waterbodies. The inherent difficulties in modelling a cyclic system with interactions and feedbacks (both balancing and reinforcing) among land, sea, atmosphere and biota (Cossarini et al., 2008; Videira et al., 2011) are due to constraints such as lack of data and knowledge gaps about the spatial and temporal complexity of responses and of comprehensive feedback loops (Clark, 2001; Rode et al., 2010), as well as about delayed and off-site effects. The development and test of new integrated modelling approaches can provide a useful contribution to reinforce the reliability of management tools for ecological conservation and adaptation policies of sensitive aquatic ecosystems such as those in coastal areas.

In this perspective, here we present an integrated modelling approach for assessing potential medium (2041–2070) and long-term (2071–2100) effects of climate change, represented by changes in air temperature and precipitation, on the productivity and community structure of coastal phytoplankton at a local scale. This approach can deepen the knowledge about the interrelated impacts of climate change along the land-water continuum, from climate-related effects on stream flow and nutrient loadings, to direct influence of temperature changes on coastal waters. According to Fig. 1, this approach consists of coupled General Circulation Model/Regional Climate Model (GCM/RCM hereafter) nested simulations, able to provide localized climate data suitable for impact assessment studies, and two separate environmental models, the hydrological model Soil and Water Assessment Tool (SWAT; (Arnold et al., 1998) and the ecological model AQUATOX (Park et al., 2008) used to depict the physical, chemical and biological process of a coastal watershed and of the receiving waters in a coastal environment. This study is a further attempt to integrate climate projections and tools with environmental models for assessing the responses of aquatic ecosystems to climate change. The main objective of the paper is to illustrate the modelling approach, demonstrate its applicability through a local case study, and to discuss potential, limitations and areas of improvement.