Climate change impacts on eutrophication in the Po River (Italy): Temperature-mediated reduction in nitrogen export but no effect on phosphorus.

Rivers worldwide are under stress from eutrophication and nitrate pollution, but the ecological consequences overlap with climate change, and the resulting interactions may be unexpected and still unexplored. The Po River basin (northern Italy) is one of the most agriculturally productive and densely populated areas in Europe. It remains unclear whether the climate change impacts on the thermal and hydrological regimes are already affecting nutrient dynamics and transport to coastal areas. The present work addresses the long-term trends (1992-2020) of nitrogen and phosphorus export by investigating both the annual magnitude and the seasonal patterns and their relationship with water temperature and discharge trajectories. Despite the constant diffuse and point sources in the basin, a marked decrease (-20%) in nitrogen export, mostly as nitrate, was recorded in the last decade compared to the 1990s, while no significant downward trend was observed for phosphorus. The water temperature of the Po River has warmed, with the most pronounced signals in summer (+0.13°C/year) and autumn (+0.16°C/year), together with the strongest increase in the number of warm days (+70%-80%). An extended seasonal window of warm temperatures and the persistence of low flow periods are likely to create favorable conditions for permanent nitrate removal via denitrification, resulting in a lower delivery of reactive nitrogen to the sea. The present results show that climate change-driven warming may enhance nitrogen processing by increasing respiratory river metabolism, thereby reducing export from spring to early autumn, when the risk of eutrophication in coastal zones is higher.

Looking back to move forward: Restoring vegetated canals to meet missing Water Framework Directive goals in agricultural basins.

Nitrate pollution and eutrophication remain pressing issues in Europe regarding the quality of aquatic ecosystems and the safety of drinking water. Achieving water quality goals under the Water Framework Directive (WFD) has proven to be particularly challenging in agricultural catchments, where high nitrate concentrations are the main reason for the failure of many water bodies to meet a good ecological status. Canals and ditches are common man-made features of irrigated and drained landscapes and, when vegetated, have recently been identified as denitrification hotspots. By combining experimental data and GIS-based upscaling estimation, the potential capacity of the canal network to reduce nitrate loads was quantified in several scenarios differing in the level of nitrate pollution and in the extent of the canal network length where conservative management practices are implemented. The analysis was carried out in the irrigated lowlands of the Po River basin, which is the largest hydrographic system in Italy and a global hotspot for nitrogen inputs and eutrophication. Scenario simulations showed that maintaining aquatic vegetation in at least 25 % of the canal network length, selecting sites with high nitrate availability (>2.4 mg N L-1), would promote a greater potential for permanent N removal. The increased denitrification capacity would meet the load reduction target required to achieve a WFD good ecological status in waters draining into the Adriatic Sea during the spring-summer months, when the eutrophication risk is higher. Promoting denitrification in the canal network by postponing the mowing of in-stream vegetation to the end of the growing season could be an effective mitigation strategy to improve water quality in agricultural basins and contribute to achieving the WFD goals.

How does invasion degree shape alpha and beta diversity of freshwater fish at a regional scale?

Freshwater ecosystems appear more vulnerable to biodiversity loss due to several anthropogenic disturbances and freshwater fish are particularly vulnerable to these impacts. We aimed to (1) identify the contribution of land use, spatial variables, and invasion degree in determining freshwater fish alpha (i.e., species richness) and beta (i.e., local contributions to beta diversity, LCBD) diversity, evaluating also the relationship between invasion degree and nestedness ( ß nes) and turnover ( ß sim) components of beta diversity. (2) Investigate the relationship between alpha diversity and LCBD, under the hypothesis that alpha diversity and LCBD correlate negatively and (3) investigate the relationship between species contributions to beta diversity (SCBD) and species occurrence, hypothesizing that non-native species show a lower contribution to beta diversity. The linear mixed models and the partition of R 2 retained the invasion degree as the most important variables explaining alpha and beta diversity, having a positive relationship with both diversity components. Furthermore, land use related to human impacts had a positive influence on alpha diversity, whereas it showed a negative effect on LCBD. Regression model further showed that invasion degree related positively with ß sim, but negatively with ß nes, suggesting that non-native species were involved in the replacement of native species in the fish community. Alpha diversity and LCBD showed a weak positive correlation, meaning that sites with low species richness have higher LCBD. SCBD scaled positively with species occurrence highlighting that rarer species contribute less to SCBD. Finally, native and exotic species contributed similarly to beta diversity. These results suggest that invasion degree plays a central role in shaping alpha and beta diversity in stream fish, more than land use features reflecting habitat alteration or other geospatial variables. Furthermore, it is important to evaluate separately the native and the non-native components of biotic communities to identify linkages between invasion dynamics and biodiversity loss.

Natural and anthropogenic factors drive large-scale freshwater fish invasions.

We analyzed the large-scale drivers of biological invasions using freshwater fish in a Mediterranean country as a test case, and considering the contribution of single species to the overall invasion pattern. Using Boosted Regression Tree (BRT) models, variation partitioning and Redundancy Analysis (RDA), we found that human factors (especially eutrophication) and climate (especially temperature) were significant drivers of overall invasion. Geography was also relevant in BRT and RDA analysis, both at the overall invasion and the single species level. Only variation partitioning suggested that land use was the second most significant driver group, with considerable overlap between different invasion drivers and only land use and human factors standing out for single effects. There was general accordance both between different analyses, and between invasion outcomes at the overall and the species level, as most invasive species share similar ecological traits and prefer lowland river stretches. Human-mediated eutrophication was the most relevant invasion driver, but the role of geography and climate was at least equally important in explaining freshwater fish invasions. Overall, human factors were less prominent than natural factors in driving the spread and prevalence of invasion, and the species spearheading it.

The achievement of Water Framework Directive goals through the restoration of vegetation in agricultural canals.

Decreasing nitrate concentrations is one of the most relevant Water Framework Directive (WFD) goals, which today is still unreached in several European countries. Vegetated canals have been recognized as effective filters to mitigate nitrate pollution, although rarely included in restoration programs aimed at improving water quality in agricultural watersheds. The Po di Volano basin (713 km2, Northern Italy) is a deltaic territory crossed by an extensive network of agricultural canals (~1300 km). The effectiveness in buffering nitrate loads via denitrification was assessed for different levels of in-stream emergent vegetation maintenance by employing an upscale model based on extensive datasets of field measurements. The scenarios differed for the canal network length (5%, 20%, 40%, and 60%) where conservative management practices were adopted by postponing the mowing operations from the middle of summer, as nowadays, to the early autumn, i.e., the vegetative season end. The scenario simulations demonstrated that the capacity to mitigate diffuse nitrate pollution would increase up to four times, compared to the current condition (5% scenario), by postponing the vegetation mowing to the end of the vegetative season in 60% of the canal network length. By preserving the in-stream vegetation in 20% of the canal network, its denitrification capacity would equal the nitrate load reduction target required for achieving, from May to September, the good ecological status according to the WFD in waters delivered to the coastal areas. Changing the timing of vegetation mowing may create a large potential for permanent nitrate removal via denitrification in agricultural landscapes, thus protecting the coastal areas when the eutrophication risk is higher. Conservative management practices of in-stream vegetation might be promoted as an effective low-cost tool to be included in the WFD implementation strategies.

Governance and groundwater modelling: Hints to boost the implementation of the EU Nitrate Directive. The Lombardy Plain case, N Italy.

The EU Nitrate Directive has been ruling for almost 30 years, nevertheless nitrate concentration in the Lombardy Plain did not decrease. Together with failures of management implementation, a possible cause for such field observations is that management actions were taken without adequately considering the actual hydrogeological dynamics. To consider this aspect, the paper presents a groundwater flow and transport numerical model of a specific area of the Lombardy Plain. The aim of this model is to demonstrate how modelling, as a management tool, can be useful in the governance process. The groundwater model, using well-known MODFLOW-MT3D codes, is based on existing hydrogeological information, while a nitrogen mass balance has been performed at municipal scale to determine the agricultural N surplus to the subsurface. The model adequately reproduces head levels and nitrate concentrations in observation wells for a 10-year simulation period, showing that 4.5% of the N annual input remains stored in the system. The model indicates the efficiency of rivers and springs to export N out from the system at an estimated rate of 77.5% of the annual N inputs. Back to governance, the model shows that management data at municipal level (e.g. irrigation rates, groundwater withdrawal, N net recharge) provide a satisfactory scale for successfully reproducing nitrate evolution. Hence those variables that can be object of debate during a governance process can be treated as input data to the numerical model. Therefore, backcasting exercises can be conducted to check whether the model outcome fits with the expected results of specific management actions. The model highlights how the N mass balance evolves, providing clues on which factors can be managed to reduce nitrate concentrations and meet the Directive's requirements. Numerical groundwater models, as an option to address water management issues, ultimately contribute to solve the information and capacity governance gaps.

Nitrate availability affects denitrification in Phragmites australis sediments.

Understanding relationships between an increase in nitrate (NO3 - ) loading and the corresponding effects of wetland vegetation on denitrification is essential to designing, restoring, and managing wetlands and canals to maximize their effectiveness as buffers against eutrophication. Although Phragmites australis (Cav.) Trin. ex Steud. is frequently used to remediate nitrogen (N) pollution, no information is available on how NO3 - concentration may affect plant-mediated denitrification. In the present study, denitrification was measured in outdoor vegetated and unvegetated mesocosms incubated in both summer and winter. After spiking the mesocosms with NO3 - concentrations typical of agricultural drainage water (0.7-11.2 mg N L-1 ), denitrification was quantified by the simultaneous measurement of NO3 - consumption and dinitrogen gas (N2 ) production. Although denitrification rates varied with vegetation presence and season, NO3 - availability exerted a significant positive effect on the process. Vegetated sediments were more efficient than bare sediments in adapting their mitigation potential to an increase in NO3 - , by yielding a one-order-of-magnitude increase in NO3 - removal rates, under both summer (743-6007 mg N m-2 d-1 ) and winter (43-302 mg N m-2 d-1 ) conditions along the NO3 - gradient. Denitrification was the dominant sink for water NO3 - in winter and only for vegetated sediments in summer. Nitrification likely contributed to fuel denitrification in summer unvegetated sediments. Since denitrification rates followed Michaelis-Menten kinetics, P. australis-mediated depuration may be considered optimal up to 5.0 mg N L-1 . The present outcomes provide experimentally supported evidence that restoration with P. australis can work as a cost-effective means of improving water quality in agricultural watersheds.

The role of species introduction in modifying the functional diversity of native communities.

Although one of the most evident effects of biological invasions is the loss of native taxonomic diversity, contrasting views exist on the consequences of biological invasions on native functional diversity. We investigated this topic using Mediterranean stream, river and canal fish communities as a test case, at 3734 sites in Italy, and distinguishing between exotic and translocated species invasion in three different faunal districts. Our results clearly confirmed that introduced species were widespread and in many cases the invasion was severe (130 communities were completely composed by introduced species). Exotic and translocated fish species had substantially different geographical distribution patterns, perhaps arising from their differences in introduction timing, spread and invasion mechanisms. We also found a clear decreasing trend of functional dispersion along an invasion gradient, confirming our hypothesis that the invasion process can diminish the relative diversity of ecofunctional traits of host fish communities. Furthermore, our results suggested that exotic species might have a greater negative effect than translocated species on the relative diversity of ecofunctional traits of fish communities. This could also be linked to the fact that translocated species are more ecofunctionally similar to native ones, compared to the exotics. Our multivariate analysis of site-specific combinations of ecofunctional traits highlighted some traits characteristic of all invaded communities, while our discriminant analysis underlined how there was a substantial ecofunctional overlap between native, exotic and translocated species groups in most areas.

Could a freshwater fish be at the root of dystrophic crises in a coastal lagoon?

Eutrophication has a profound impact on ecosystems worldwide. Grass carp Ctenopharyngodon idella, an herbivorous fish, has been introduced to control aquatic plant overgrowth caused by eutrophication, but could have other, potentially detrimental, effects. We used the Po di Volano basin (south of the Po River delta, northern Italy) as a test case to explore whether grass carp effects on canal aquatic vegetation could be at the root of historical changes in N loads exported from the basin to the Goro Lagoon. We modeled the aquatic vegetation production and standing crop, its denitrification potential, and its consumption by introduced grass carp. We then examined whether changes in historical nitrogen loads matched the modeled losses of the drainage network denitrification function or other changes in agricultural practices. Our results indicate that introduced grass carp could completely remove submerged vegetation in the Po di Volano canal network, which could - in turn - lead to substantial loss of the denitrification function of the system, causing in an increase in downstream nitrogen loads. A corresponding increase, matching both timing and magnitude, was detected in historical nitrogen loads to the Goro Lagoon, which were significantly different before and after the time of modeled collapse of the denitrification function. This increase was not clearly linked to watershed use or agricultural practices, which implies that the loss of the denitrification function through grass carp overgrazing could be a likely explanation of the increase in downstream nitrogen loads. Perhaps for the first time, we provide evidence that a freshwater fish introduction could have caused long-lasting changes in nutrient dynamics that are exported downstream to areas where the fish is not present.

Effect of ebullition and groundwater temperature on estimated dinitrogen excess in contrasting agricultural environments.

Denitrification is a key microbial-mediated reaction buffering the impact of agriculturally-derived nitrate loads. Groundwater denitrification capacity is often assessed by measuring the magnitude and patterns of dinitrogen excess, although this method can be biased by dissolved gasses exsolution and ebullition. To address this issue, shallow groundwater was sampled in two field sites via nested mini-wells on a monthly basis over an entire hydrological year and analysed for dissolved gasses, nitrate and physical parameters. Both sites are located in lowland areas of the Po River basin (Italy) and are characterized by intensive agriculture. The GUA site, a freshwater paleo-river environment, with a low content of organic matter (SOM) and oxic sub-oxic groundwater. The BAN site, a reclaimed brackish swamp environment, with abundant SOM and sulphidic-methanogenic groundwater. Groundwater samples evidenced a general deficit of dinitrogen and Argon concentrations, because of ebullition induced by a total dissolved gasses pressure exceeding the hydrostatic pressure. Ebullition was recorded only during winter at the reclaimed brackish soil and was triggered by methane exsolution. While in summer both sites were affected by ebullition because of the water table drawdown. Denitrification evaluated using dinitrogen excess via dinitrogen-Argon ratio technique, was not only affected by gas exsolution, but also by groundwater temperature fluctuations. In fact, the latter induced large biases in the calculated N2 excess even in the freshwater paleo-river environment. For these reasons, dinitrogen excess estimate with standard methods resulted to be unreliable in both lowland environments and a modified method is here presented to overcome this issue.

Intense rainfalls trigger nitrite leaching in agricultural soils depleted in organic matter.

Nitrate and ammonium are common inorganic contaminants of anthropogenic origin in many shallow aquifers around the world, while nitrite is less common, but it is most harmful than nitrate and ammonium due to its high reactivity. This paper presents evidence of nitrite accumulation after intense rainfalls in soil samples collected in an agricultural field characterized by organic matter chronic depletion. Moreover, an intact core from the same site was also collected to perform an unsaturated column experiment (60 cm long and 20 cm outer diameter) mimicking heavy rainfalls (230 mm in 2 days). Results from the field site showed nitrite accumulation (up to 0.45 mmol/kg) at 50-70 cm depth, just below the plough layer. The column experiment showed very high initial concentrations of nitrate and nitrite in the leachate and a progressive decrease of nitrate due to denitrification. The numerical flow model was calibrated versus the observed volumetric water contents and leachate flow rates. The numerical reactive transport model was calibrated versus the leachate concentrations of six dissolved species (ammonium, nitrate, nitrite, dissolved organic carbon, chloride and bromide). The optimized model resulted to be robustly calibrated providing insights on the kinetic rates driving the production, accumulation and leakage of nitrite, showing that incomplete denitrification is the source of nitrite. As far as the authors are aware, this is the first study reporting a clear link between high nitrite leaching rates and extreme rainfall events in lowland agricultural soils depleted in organic matter. The proposed methodology could be applied to quantify nitrite cycling processes in many other agricultural areas of the world affected by extreme rainfall events.

An ounce of prevention is worth a pound of cure: Managing macrophytes for nitrate mitigation in irrigated agricultural watersheds.

Although ubiquitous elements of agricultural landscapes, the interest on ditches and canals as effective filters to buffer nitrate pollution has been raised only recently. The aim of the present study was to investigate the importance of in-ditch denitrification supported by emergent aquatic vegetation in the context of N budget in agricultural lands of a worldwide hotspot of nitrate contamination and eutrophication, i.e. the lowlands of the Po River basin (Northern Italy). The effectiveness of N abatement in the ditch network (>18,500 km) was evaluated by extrapolating up to the watershed reach-scale denitrification rates measured in a wide range of environmental conditions. Scenarios of variable extents of vegetation maintenance were simulated (25%, 50% and 90%), and compared to the current situation when the natural development occurs in only 5% of the ditch network length, subjected to mechanical mowing in summer. Along the typical range of nitrate availability in the Po River lowlands waterways (0.5-8 mg N L-1), the current N removal performed by the ditch network was estimated in 3300-4900 t N yr-1, accounting for at most 11% of the N excess from agriculture. The predicted nitrate mitigation potential would increase up to 4000-33,600 t N yr-1 in case of vegetation maintenance in 90% of the total ditch length. Moreover, a further significant enhancement (57% on average) of this key ecosystem function would be achieved by postponing the mowing of vegetation at the end of the growing season. The simulated outcomes suggest that vegetated ditches may offer new agricultural landscape management opportunities for effectively decreasing nitrate loads in surface waters, with potential improved water quality at the watershed level and in the coastal zones. In conclusion, ditches and canals may act as metabolic regulators and providers of ecosystem services if conservative management practices of in-stream vegetation are properly implemented and coupled to hydraulic needs.

Estimate of gas transfer velocity in the presence of emergent vegetation using argon as a tracer: Implications for whole-system denitrification measurements.

Denitrification associated with emergent macrophytes is a pivotal process underlying the treatment performance of wetlands and slow-flow waterways. Laboratory scale experiments targeting N losses via denitrification in sediments colonized by emergent macrophytes require the use of mesocosms that are necessarily open to the atmosphere. Thus, the proper quantification of N2 effluxes relies on the accurate characterization of the air-water gas exchanges. In this study, we present a simple approach for direct measurements of the gas transfer velocity, in open-top mesocosms with Phragmites australis, by using argon as a tracer. Different conditions of water velocity (0, 1.5, 3, and 6 cm s-1) and temperature (8.5, 16, and 28 °C), were tested, along with, for the first time, the presence of emergent vegetation. The outcomes demonstrated that water velocity and temperature are not the only factors regulating aeration at the mesocosm scale. Indeed, the gas transfer velocity was systematically higher, in the range of 42-53%, in vegetated compared to unvegetated sediments. The increase of small-local turbulence patterns created within water parcels moving around plant stems translated into significant modifications of the reaeration process. The adopted approach may be used to improve the accuracy of denitrification measurements by N2 efflux-based methods in wetland and slow-flow waterway sediments colonized by emergent macrophytes. Moreover, the present outcomes may have multiple implications for whole-system metabolism estimations from which largely depend our understanding of biogeochemical dynamics in inland waters that have strong connections to worldwide issues, such as nitrate contamination and greenhouse gas emissions.

Space and time variations of watershed N and P budgets and their relationships with reactive N and P loadings in a heavily impacted river basin (Po river, Northern Italy).

The aim of the present study is to analyze relationships between land uses and anthropogenic pressures, and nutrient loadings in the Po river basin, the largest hydrographic system in Italy, together with the changes they have undergone in the last half century. Four main points are addressed: 1) spatial distribution and time evolution of land uses and associated N and P budgets; 2) long-term trajectories of the reactive N and P loadings exported from the Po river; 3) relationships between budgets and loadings; 4) brief review of relationships between N and P loadings and eutrophication in the Northern Adriatic Sea. Net Anthropogenic N (NANI) and P (NAPI) inputs, and N and P surpluses in the cropland between 1960 and 2010 were calculated. The annual loadings of dissolved inorganic nitrogen (DIN) and soluble reactive phosphorus (SRP) exported by the river were calculated for the whole 1968-2016 period. N and P loadings increased from the 1960s to the 1980s, as NAPI and NANI and N and P surpluses increased. Thereafter SRP declined, while DIN remained steadily high, resulting in a notable increase of the N:P molar ratio from 47 to 100. In the same period, the Po river watershed underwent a trajectory from net autotrophy to net heterotrophy, which reflected its specialization toward livestock farming. This study also demonstrates that in a relatively short time, i.e. almost one decade, N and P sources were relocated within the watershed, due to discordant environmental policies and mismanagement on the local scale, with frequent episodes of heavy pollution. This poses key questions about the spatial scale on which problems have to be dealt with in order to harmonize policies, set sustainable management goals, restore river basins and, ultimately, protect the adjacent coastal seas from eutrophication.

The effect of water velocity on nitrate removal in vegetated waterways.

The extended networks of canals and ditches in agricultural landscapes provide high buffer capacity towards nitrogen (N) excess. Their N mitigation potential depends on several biotic and abiotic factors, among which water velocity is poorly explored and generally omitted from the parameterization of this remarkable ecosystem service. The present work reports new insights on the role of flow velocity in regulating N removal via denitrification in sediments colonized by Phragmites australis. Denitrification was investigated in outdoor mesocosms in the presence and absence of P. australis and over a small range of flow velocity (0-6 cm s-1) typical of low-gradient water bodies. Simultaneous measurements of NO3- consumption and N2 production based on analyses of N2:Ar by Membrane Inlet Mass Spectrometry were undertaken. Vegetated sediments were found more efficient in converting NO3- to N2 via microbial-mediated denitrification (27-233 mmol N m-2 d-1) than bare sediments (18-33 mmol N m-2 d-1). Vegetation provides multiple interfaces, i.e. in the rhizosphere and on epiphytic biofilms, that support the development and activity of bacterial communities responsible for NO3- dissipation. NO3- removal and denitrification rates exhibited one order of magnitude raise when water velocity passed from 0 to 6 cm s-1 in vegetated sediments. Indeed, in slow-flow vegetated waterways denitrification may be physically limited and the increase of water velocity enhances the rate of NO3- supply through the diffusive boundary layer, thereby promoting its consumption and loss from the system. Water velocity should be taken into account as a key factor for management and restoration actions aimed at maximizing the NO3- buffer capacity of low-flow drainage networks.

Nitrogen budget in a lowland coastal area within the Po River basin (northern Italy): multiple evidences of equilibrium between sources and internal sinks.

Detailed studies on pollutants genesis, path and transformation are needed in agricultural catchments facing coastal areas. Here, loss of nutrients should be minimized in order to protect valuable aquatic ecosystems from eutrophication phenomena. A soil system N budget was calculated for a lowland coastal area, the Po di Volano basin (Po River Delta, Northern Italy), characterized by extremely flat topography and fine soil texture and bordering a network of lagoon ecosystems. Main features of this area are the scarce relevance of livestock farming, the intense agriculture, mainly sustained by chemical fertilizers, and the developed network of artificial canals with long water residence time. Average nitrogen input exceeds output terms by ~60 kg N ha(-1) year(-1), a relatively small amount if compared to sub-basins of the same hydrological system. Analysis of dissolved inorganic nitrogen in groundwater suggests limited vertical loss and no accumulation of this element, while a nitrogen mass balance in surface waters indicates a net and significant removal within the watershed. Our data provide multiple evidences of efficient control of the nitrogen excess in this geographical area and we speculate that denitrification in soil and in the secondary drainage system performs this ecosystemic function. Additionally, the significant difference between nitrogen input and nitrogen output loads associated to the irrigation system, which is fed by the N-rich Po River, suggests that this basin metabolizes part of the nitrogen excess produced upstream. The traditionally absent livestock farming practices and consequent low use of manure as fertilizer pose the risk of excess soil mineralization and progressive loss of denitrification capacity in this area.