Modeling nitrogen export from 2539 lowland artificial watersheds in Lake Taihu Basin, China: Insights from process-based modeling

Modeling nitrogen export from 2539 lowland artificial watersheds in Lake Taihu Basin, China: Insights from process-based modeling

Journal Pre-proofs Research papers Modeling nitrogen export from 2539 lowland artificial watersheds in Lake Taihu Basin, China: Insights from process-...

2MB Sizes 0 Downloads 3 Views

Journal Pre-proofs Research papers Modeling nitrogen export from 2539 lowland artificial watersheds in Lake Taihu Basin, China: Insights from process-based modeling Jiacong Huang, Zhen Cui, Feng Tian, Qi Huang, Junfeng Gao PII: DOI: Reference:

S0022-1694(19)31163-1 https://doi.org/10.1016/j.jhydrol.2019.124428 HYDROL 124428

To appear in:

Journal of Hydrology

Received Date: Revised Date: Accepted Date:

2 October 2019 21 November 2019 2 December 2019

Please cite this article as: Huang, J., Cui, Z., Tian, F., Huang, Q., Gao, J., Modeling nitrogen export from 2539 lowland artificial watersheds in Lake Taihu Basin, China: Insights from process-based modeling, Journal of Hydrology (2019), doi: https://doi.org/10.1016/j.jhydrol.2019.124428

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier B.V.

Modeling nitrogen export from 2539 lowland artificial watersheds in Lake Taihu Basin, China: Insights from process-based modeling

Jiacong Huang a, b,*, Zhen Cui a, c, Feng Tian a, c, Qi Huang d, Junfeng Gao a,*

a

Key Laboratory of Watershed Geographic Sciences, Nanjing Institute of Geography

and Limnology, Chinese Academy of Sciences, 73 East Beijing Road, Nanjing 210008, China b

Center for Eco-Environment Research, Nanjing Hydraulic Research Institute,

Nanjing 210098, China c

d

University of Chinese Academy of Sciences, Beijing 100049, China Key Laboratory of Poyang Lake Wetland and Watershed Research, Ministry of

Education, Jiangxi Normal University, Nanchang 330022, China * Corresponding author. E-mail: [email protected] (Junfeng Gao)

Page 1 of 31

Abstract Excess nitrogen (N) poses a risk of aquatic eutrophication and ecosystem degradation to downstream areas. However, it is poorly investigated in lowland artificial watersheds (polders) due to their complex hydrological processes. This study investigated the N export and retention at all 2539 polders in Lake Taihu Basin, China using a nitrogen dynamic model (NDP), specially developed for polders. The response of N export to global climate change (air temperature and precipitation) and human activities (N fertilization) in the future (2014-2049) were quantified through scenario simulations. The simulation results revealed that these polders have a larger N export coefficient (a spatially-averaged value of 15.3 kg N/ha/yr) than that of non-polder areas (~6.7 kg N/ha/yr) mainly due to their intensive agricultural activities and high population density. However, these polders also showed considerable N retention with estimated retention rates of 52.7-54.6%. In model results from the future (2014-2049), human activities (N fertilization) determined the magnitude of polder N export, while global climate change determined the fluctuation of polder N export. The estimation of polder N sources, sinks and retention capacity at a watershed scale can be used to identify the hotspot of N loss, and thus guide decision-making to control N loss. Keywords Polder; Global climate change; Nitrogen; Taihu

Page 2 of 31

1. Introduction Lowland artificial watersheds (polders) are unique hydrological units (with artificial control of runoff and water levels using pumping systems) widely located around the aquatic ecosystems (seas, lakes and rivers) worldwide (Lindenschmidt et al., 2009; van der Grift et al., 2016). Although a global map of polder distribution is so far unavailable, previous publications revealed a wide distribution of polders in the middle and lower reaches of several large rivers, such as the Yangtze and Mekong rivers in Asia, the Rhine and Danube rivers in Europe, Mississippi River in North America, etc. (Fig. S2 in Supporting Information). Excess nutrient export from these polders with intensive agriculture is a major contributor to the severe eutrophication of receiving water bodies (Huang et al., 2017). Considering that eutrophication due to excess nutrient loading is a worldwide challenge (Vinçon-Leite and Casenave, 2019), quantifying the contribution of polder nutrient export at a watershed scale would help water managers pinpoint the major nutrient sources of aquatic ecosystems, and thus guide management decisions to achieve best management practices in minimizing nutrient loading (Abouali et al., 2017; Dudula and Randhir, 2016; Qiu et al., 2019). Nutrient export from watersheds has been widely investigated using non-point source models (Ouyang et al., 2017), such as the Soil and Water Assessment Tool (SWAT) (Arnold et al., 2012; Liu et al., 2017; Yang et al., 2016), the Integrated Nitrogen in CAtchments model (INCA) (Lu et al., 2017; Wade et al., 2002) and the Annualized Agricultural Non-Point Source Pollution (AnnAGNPS) model (Li et al., 2015). These

Page 3 of 31

studies revealed a consensus on the significant impact of weather conditions (e.g., air temperature and precipitation) and human activities (e.g., agricultural fertilization and wastewater treatment) on nutrient export (Hale et al., 2015; Qu and Kroeze, 2010; Sinha et al., 2017; Swaney et al., 2012; Zhang et al., 2016). Many previous studies further investigated watershed nutrient export in the context of global climate change based on the advances of global climate models (GCMs) on predicting future climate variables in the changing environment (Kharin et al., 2013; Marshall and Randhir, 2008; Xu and Xu, 2012). It was found that future climate change had considerable impacts on watershed nutrient export (Sinha et al., 2017), and was thus a major concern in management practice (Marshall and Randhir, 2008). However, there have been very few quantitative studies investigating nutrient export from lowland artificial watersheds (polders) in the context of global climate change and human activities. This is mainly because the widely-used watershed models (e.g., SWAT and INCA) were not well suited in simulating hydrological processes in lowland areas due to the following two facts. First, none of these models is structurally equipped to determine the flow direction in lowland polders. They have a fundamental assumption that the watershed is spatially divided into hydrologic response units (HRUs) or grid cells, whereby flow direction is determined based on elevation differences among these HRUs/cells. However, lowland polders have flat topography (e.g., nearly identical elevation). Therefore, their flow direction is generally determined by water level rather than elevation differences. There is frequent water exchange between surface water and farmlands (Fig. S1 in Supporting Information). Second, these models do not

Page 4 of 31

describe the unique process of artificial drainage (e.g., culvert and flood drainage) in polders. However, there is ample evidence in the literature that they can play a significant role in N transport within polders (Brauer et al., 2014; Huang et al., 2018). In this context, this study aimed to (1) quantify the nitrogen (N) sources, sinks and retention capacity of polders, and (2) investigate their future (2014-2049) changes in the context of global climate change (air temperature and precipitation) and human activities (N fertilization). Lake Taihu Basin, the most developed area in China with a total of 2539 polders, was selected as the study area due to its wide-distributed polders and severe eutrophication of rivers and lakes. A nitrogen dynamic model (NDP), specifically developed for polders by Huang et al. (2018), was used to simulate N dynamics for all polders in Lake Taihu Basin. The spatial and temporal patterns of N export and retention were investigated. The roles of global climate change and human activities on polder N export were identified by comparing scenario simulations. Based on the scenario simulation results, potential strategies to control N export in nutrient management practice were proposed. To the best of our knowledge, this is the first attempt to investigate N export and retention for polders at a watershed scale under the context of global climate change and human activities.

2. Materials and methods 2.1. Study area and data Lake Taihu Basin (36,895 km2, 30.12-32.22°, 119.03-121.91°) is located in the lower reaches of the Yangtze River, China (Fig. 1) with an annual precipitation of approximately 1,177 mm. A total of 28.8% (10,627 km2) of the basin is covered by Page 5 of 31

2539 polders located in the lowland areas. The land uses of the polders include paddy land (62.6%), dry land (5.0%), surface water (7.3%) and residential areas (25.1%). In water management practice, the basin was divided into eight hydraulic zones based on the hydraulic gradient and administrative division implemented by the Taihu Basin Authority of Ministry of Water Resources (http://www.tba.gov.cn/English/). The lowland areas have a frequent reciprocating flow with their primary water flow direction shown in Fig. 1. Huxi, Zhexi and Taihu were the upstream areas of Lake Taihu. Wuchengxiyu, Yangchengdianmao, Hangjiahu, Puxi and Pudong were the downstream areas of Lake Taihu. Lake Taihu Basin is one of the most developed and urbanized areas in China. Many large cities, such as Shanghai, Suzhou, Wuxi and Changzhou with millions of population, are located within the basin. Nutrient loading from Lake Taihu Basin has caused severe eutrophication and harmful cyanobacterial blooms of aquatic systems, especially Lake Taihu, China’s third largest freshwater lake (Qin et al., 2019).

Page 6 of 31

Polder Jian

Flow direction

Fig. 1. Locations of 2539 polders, 7 weather stations, 99 water sampling sites and 8 hydraulic zones in Lake Taihu Basin, China.

A dataset (land use, population, meteorological and water quality data) was collected to investigate the N balance within 2539 polders in Lake Taihu Basin. The land use data were derived from satellite images in 2010. The population data were obtained from China’s Sixth National Population Census in 2010. The meteorological data were collected from seven national weather stations inside/near the basin in 2013, and included seven variables of daily precipitation (mm), daily maximum, minimum and average air temperature (°C), daily average humidity (%), and daily average wind speed (m/s), daily sunshine hours (h). Water quality data (total nitrogen concentration) were collected by water sampling at 99 sites in Lake Taihu Basin in 2013 as the boundary conditions of the model (NDP) (Fig. 1). 2.2. Model description

Page 7 of 31

Polders’ N dynamics in the context of global climate change were simulated using the NDP model that was specifically developed for polder systems (Huang et al., 2018). The model is a semi-distributed model with four spatial units (land use types) of surface water, residential areas, dry and paddy lands. Unlike existing watershed models (e.g., SWAT), NDP describes the polder’s unique hydrological processes including water exchange between surface water and farmlands, flood and culvert drainage (Huang et al., 2018). The critical processes related to water and N dynamics within/between these four spatial units were described at a daily time scale. All the processes in NDP can be found in Fig. S3 (Supporting Information) with their corresponding equations provided in Table S2. The model was evaluated using a dataset collected from Polder Jian, and achieved a model fit with Nash-Sutcliffe efficiency ranging from 0.43 to 0.48 for total nitrogen concentration (Fig. S5 in Supporting Information). This model performance was acceptable for lowland watersheds with strong human activities when compared with previous case studies on watershed N modeling (Wellen et al., 2015). Our field survey in Lake Taihu Basin found that Polder Jian is a typical lowland polder enclosed by dikes with artificial drainage and a complex ditch-pond network. Therefore, it was justified to apply NDP to other polders in this area. Further details of the NDP model including the conceptual model, model implementation, main equations, and the model evaluation results, are provided in the Supporting Information. 2.3. Estimating polder nitrogen export and retention

Page 8 of 31

2.3.1

Global climate change and human activities in the future

To investigate polder N export and retention in the context of global climate change and human activities, future changes in three factors were obtained including air temperature, precipitation and fertilization amount. These three factors were selected, because they were widely demonstrated to have significant impacts on N dynamics in lowland areas (Hale et al., 2015; Huang et al., 2018). Air temperature significantly affected N transformation processes, such as volatilization, denitrification and mineralization. Precipitation significantly affected N discharge and deposition within the polder system (Hale et al., 2015). Human activities affected polder N export through multiple pathways, such as agricultural fertilization, population change and wastewater treatment. Among these pathways, N fertilization exhibited a dominant contribution (73.3%) to the total N sources in a typical polder (Huang et al., 2018). Climate change is often the consequence of the combined effects of anthropogenic influences and complex interactions among the atmosphere, hydrosphere, lithosphere, cryosphere, and biosphere of the Earth system. Climate system models are effective tools for simulating future climate change in the context of natural and human influences (Wu et al., 2019). From more than 20 years ago, the Coupled Model Intercomparison Project (CMIP) combined a handful of global coupled climate models to generate climate information and make it available for scientific research (Meehl et al., 1997). The Beijing Climate Center (BCC) of the China Meteorological Administration effectively contributed to CMIP5 by running most of the mandatory and optional simulations (Wu et al., 2019). Therefore, future air temperature and

Page 9 of 31

precipitation (2014-2049) in this case study were obtained from the Beijing Climate Center Climate System Model version 1 (BCC-CSM1.1) (Xu and Xu, 2012). The model ass a fully coupled global climate-carbon model including interactive vegetation and the global carbon cycle (Wu et al., 2013), and has been widely used to investigate China’s future climate change (Xu and Xu, 2012). Two representative concentration pathway scenarios (RCP), i.e., RCP2.6 and RCP8.5, were selected to represent possible emission scenarios in the context of economic development, technological development, energy use, population change, and land-use change. RCP8.5 represented a higher emission scenario with a continuously increasing radiative forcing pathway and a high greenhouse gas concentration level. RCP2.6 represented a lower emission scenario with a controlled radiative forcing pathway and a low greenhouse gas concentration level. Temporal dynamics of annual precipitation and daily average air temperature from RCP2.6 and RCP8.5 can be found in the Supporting Information. Future fertilization targets were obtained from previous studies on optimizing the N fertilizer application rate in Lake Taihu Basin (Hofmeier et al., 2015). Based on the field experiments and investigations in Lake Taihu Basin, the recommended N fertilizers for summer rice and winter wheat were 200-230 and 170-180 kg N/ha/yr, respectively, without any decline in grain yield (Hofmeier et al., 2015). Therefore, future N fertilizers for summer rice and winter wheat was reduced to 200 and 170 kg N/ha/yr, compared to 250 and 200 kg N/ha/yr under conventional N fertilization practices (Huang et al., 2018).

Page 10 of 31

2.3.2

Scenario simulations to quantify polder nitrogen export and retention

Based on the developed model (NDP), four simulations (Sim_Present, Sim_RCP2.6, Sim_RCP8.5 and Sim_NReduction) with different model configurations (Table 1) were carried out to estimate N export and retention for all 2539 polders in Lake Taihu Basin (Fig. 1). Sim_Present was used to simulate the present N export and retention in 2013. Sim_RCP2.6 and Sim_RCP8.5 were used to simulate polder N export and retention in the future (2014-2049) under the emission scenarios of RCP2.6 and RCP8.5, respectively. Sim_NReduction was used to simulate polder N export and retention in the future (2014-2049) in the context of fertilization reduction. The fertilization rate had an approximate value of 450 kg N/ha/yr for agricultural farmlands from local farmers in Lake Taihu Basin (Huang et al., 2018). Our survey from local farmers revealed that this fertilization rate can be reduced to approximately 370 kg N/ha/yr without a significant reduction of agricultural production. Therefore, this fertilization rate (370 kg N/ha/yr) was utilized in the Sim_NReduction simulation. Polder N export was quantified using the N export coefficient (NEC, kg N/ha/yr). NEC represents the annual N export amount (t) per unit (ha) in the study area. Therefore, for polders in Lake Taihu Basin, NEC can be calculated by dividing the sum of three N export components (N export through seepage, flood and culvert drainage) to the polder area. Polder N retention was quantified using the N retention rate (NRR, %) with a value of 0-100%. Compared with that of free drainage watersheds, polders have larger N retention mainly due to their large surface water area (e.g., ponds and ditches) for N retention. Therefore, NRR was calculated in this

Page 11 of 31

study using the equation of (1-NOut/NIn)×100%, where NIn (t) is the annual N amount flowing into the surface water area and NOut (t) is the annual N amount flowing out of the surface water area. Table 1 Scenario simulations to investigate polder nitrogen export and retention in Lake Taihu Basin. Simulations

Simulation period

Sim_Present

2013

Weather conditions Measured data

Fertilization No reduction (450 kg N/ha/yr)

Measured data Sim_RCP2.6

2014-2049

Estimated T and Pr from BCC-CSM1.1 under the climate scenario of RCP2.6

No reduction (450 kg N/ha/yr)

Measured data Sim_RCP8.5

2014-2049

Estimated T and Pr from BCC-CSM1.1 under the climate scenario of RCP8.5

No reduction (450 kg N/ha/yr)

Measured data Sim_NReduction

2014-2049

Estimated T and Pr from BCC-CSM1.1 under the climate scenario of RCP2.6

Reduction (370 kg N/ha/yr)

Note: BCC-CSM1.1: Beijing Climate Center Climate System Model version 1; RCP: Representative Concentration Pathways. N, nitrogen. T, air temperature. Pr, precipitation. Fertilization rate (450 and 370 kg N/ha/yr) was the observed values from local farmers.

3. Results 3.1. Polder nitrogen export 3.1.1

Temporal dynamics

The simulation results from Present_2013 (Table 1) showed that the polders in Lake Taihu Basin had an annual N export of 16296 t/yr, with a spatially-averaged NEC of 15.3 kg N/ha/yr. This NEC (15.3 kg N/ha/yr) for polders was lower than that for the upstream areas of Lake Taihu Basin including both polders and non-polder areas (20.4-21.0 kg N/ha/yr), and that for the agricultural watershed (Kielstau Catchment)

Page 12 of 31

in north Germany (50.3 kg N/ha/yr). However, this NEC value was higher than that in the Zhongtian River watershed (6.7 kg N/ha/yr) and that (6.36 kg N/ha/yr) at 60 sites in the US (Table 2). Of the N export in 2013 (16296 t), 58.7% (9561 t) of the N export occurred in the rice-cropping season (Jun.-Oct.), while 41.3% (6735 t) of the N export occurred in the wheat-cropping season (Nov.-May). Seepage, flood and culvert drainage contributed an N export amount of 2106, 7474 and 6716 t/yr, respectively. N export through culvert drainage had a peak value in February. All culverts were manually closed without any culvert drainage during the rice-cropping season. N export through flood drainage mostly occurred during the rice-cropping season, and had an extreme peak value in June. with high precipitation. N export through seepage mostly occurred in the rice-cropping season (Fig. 2).

Fig. 2. Annual nitrogen (N) export amount and monthly nitrogen export coefficient of the polders in Lake Taihu Basin in 2013 through seepage, flood and culvert drainage.

Page 13 of 31

Table 2 Comparing estimated nitrogen export coefficients (NECs) with reported NEC values in previous studies. Area (km2)

Country

NEC (kg N/ha/yr)

Method/Model

Reference

10,627

China

15.3

Coupled N model

This study

19,055

China

21.0

Measurement

(Taihu-Basin-Authority , 2014)

19,055

China

20.4

SWAT model

(Lai et al., 2006)

45.8

China

6.7

AnnAGNPS model

(Li et al., 2015)

11,531

China

25.0

Export coefficient model

(Ma et al., 2011)

2940

China

25.3

MARINA model

(Sonneveld et al., 2012)

Kielstau Catchment

50

Germany

50.3

SWAT

(Lam et al., 2012)

60 sites in US

--

US

6.36

SWAT

(White et al., 2015)

Watershed Lowland watersheds in Lake Taihu Basin Upstream areas of Lake Taihu Basin Upstream areas of Lake Taihu Basin Zhongtian River Watershed Three Gorges Reservoir Area Lake Dianchi Watershed

In the future, the annual N export showed a large fluctuation varying from 6867 to 23074 t/yr without a significant increasing or decreasing trend (Fig. 3 (a)). The annual N export under the climate scenario of RCP8.5 showed larger fluctuations (a standard deviation of 3584 t/yr) than that under the climate scenario of RCP2.6 with a standard deviation of 3362 t/yr. N reduction resulted in a decrease of annually-averaged N export from 12876 t/yr (Sim_RCP2.6) to 11135 t/yr (Sim_NReduction). Annual N export was positively related to annual precipitation (Fig. S7 in the Supporting Information).

Page 14 of 31

Fig. 3. Annual nitrogen export (t/yr) (a) and nitrogen retention rate (NRR, %) (b) of the polders in Lake Taihu Basin in the future (2014-2049).  RCP 2.6 ,  RCP8.5 and  NReduction are the future annually-averaged values from the Sim_RCP2.6, Sim_RCP8.5 and Sim_NReduction simulations (Table 1), respectively.  RCP 2.6 ,  RCP8.5 and  NReduction are the future standard deviation values from the Sim_RCP2.6, Sim_RCP8.5 and Sim_NReduction simulations (Table 1), respectively.

3.1.2

Spatial pattern

In 2013, Pudong had the largest NEC (24.0 kg N/ha/yr) among these eight hydraulic zones due to its high population density. Hangjiahu had the largest annual N export (5743.0 t/yr) due to its large polder area. Polders from three upstream zones (Huxi, Zhexi and Taihu) of Lake Taihu contributed 4270.9 t N/yr into the lake, which was 11.3% of the lake’s external N sources (37815 t/yr) (Taihu-Basin-Authority, 2014). Other downstream zones of Lake Taihu contributed 12024.9 t N/yr into the Yangtze River. The NEC values from all four simulations (Sim_Present, Sim_RCP2.6, Sim_RCP8.5 and Sim_NReduction) showed high spatial heterogeneity (Fig. 5). Most polders had

Page 15 of 31

an NEC value of 0-20 kg N/ha/yr. The polders near/within several large cities (e.g., Shanghai, Suzhou, Wuxi and Changzhou) had a large NEC value (>30 kg N/ha/yr) mainly due to the large population. Under the future climate scenario of RCP2.6 and RCP8.5, Zhexi (Z2) and Pudong (Z8) showed relatively high N export with an annually-averaged NEC higher than 15 kg N/ha/yr, respectively. Other hydraulic zones had annually averaged NEC values between 10 and 15 kg N/ha/yr. However, the spatial patterns of the NEC values from Sim_RCP2.6 (Fig. 5 (b)) and Sim_RCP8.5 (Fig. 5 (c)) were similar indicating the limited impact of global climate change on N export. Fertilization reduction led to a considerable decrease of NEC in all eight hydraulic zones by 0.7-1.7 kg N/ha/yr. Although fertilization reduction was applied across all polders, the response of polder NEC to fertilization reduction varied spatially. It is not surprising that fertilization reduction led to a significant decrease of NEC in agricultural polders (e.g., polders in Zhexi) due to the large area of agricultural farmlands. However, it did not lead to a significant decrease of NEC in the polders near/within large cities (see Fig. 5 (b) and (d)).

Page 16 of 31

Fig. 4. Polder nitrogen export coefficient (NEC, kg N/ha/yr) (a) and nitrogen retention rate (NRR, %) (b) at eight hydraulic zones of Lake Taihu Basin in the future (2014-2049). Z1-8 represents 8 hydraulic zones. Z1, Huxi; Z2, Zhexi; Z3, Taihu; Z4, Wuchengxiyu; Z5, Yangchengdianmao; Z6, Hangjiahu; Z7, Puxi; Z8, Pudong (Fig. 1). Sim_RCP2.6, Sim_RCP8.5 and Sim_NReduction are three scenario simulations (Table 1).

Page 17 of 31

Fig. 5. Nitrogen export coefficient (NEC) values of the polders in Lake Taihu Basin in 2013 and in the future (2014-2049). Sim_Present, Sim_RCP2.6, Sim_RCP8.5 and Sim_NReduction are four scenario simulations (Table 1).

3.2. Polder nitrogen retention 3.2.1

Temporal dynamics

In the future, the annual NRR values from three scenario simulations were higher than 50%, implying that N export from these lowland areas could be significantly higher without artificial control of the surface water areas. The annually-averaged NRR showed a slight decrease from 54.6% to 52.7% due to fertilization reduction (Fig. 3 (b)). There was no significant difference among the annually averaged NRR values from Sim_RCP2.6 and Sim_RCP8.5. Additionally, the NRR was positively related to

Page 18 of 31

annual N export (Fig. S7 in the Supporting Information). 3.2.2

Spatial pattern

In 2013, NRR values among the eight hydraulic zones showed a large difference varying from 53.4% (Taihu) to 64.9% (Pudong and Zhexi). Both Zhexi (Z2) and Pudong (Z8) showed a relatively high NRR (64.9%) compared with that of the other hydraulic zones (Fig. 4 (a)) probably due to the large population density in these two areas. It is clear that the spatial pattern of polder NRR (Fig. 6) was similar to that of NEC (Fig. 5), implying that polders with a higher NEC had a higher NRR (Fig. S7 (c)).

Fig. 6. Nitrogen retention rate (NRR) of the polders in Lake Taihu Basin in 2013 and in the future Page 19 of 31

(2014-2049). Sim_Present, Sim_RCP2.6, Sim_RCP8.5 and Sim_NReduction are four scenario simulations (Table 1).

4. Discussion 4.1. Polders: nitrogen sources or sinks? It is clear that polders in Lake Taihu Basin were N sources for their surrounding rivers and lakes with a spatially-averaged NEC value of 15.3 kg N/ha/yr, and had a large contribution (11.3%, see Section 3.1.2) to the eutrophication of the large lake (Lake Taihu). Compared with non-polder areas in Lake Taihu Basin, the polders showed larger N exports (as shown in Table 2) probably due to their high population density and intensive farming. This spatially-averaged N export intensity (NEC=15.3 kg N/ha/yr) for polders in Lake Taihu Basin was also higher compared with that in many watersheds worldwide, such as US and Baltic Sea watersheds with an NEC value less than 10 kg N/ha/yr (Swaney et al., 2012; White et al., 2015). Polder N export showed large annual variation in the future (Fig. 3 (a)), mainly attributed to the change in precipitation rather than air temperature (Fig. S7 in the Supporting Information). The main N sources included fertilization, irrigation, precipitation and domestic wastewater estimated based on population number. For polders with large agricultural areas, fertilization can be the dominant source of N (Huang et al., 2018). However, for polders with large residential areas, the dominant N source may be domestic wastewater due to their large residential population. Considering the relatively high N concentration of polders’ surrounding rivers, irrigation has some contribution as a N source for agricultural polders (Huang et al., 2018). The main N sinks included crop

Page 20 of 31

harvesting, volatilization and denitrification. Crop harvesting was the most critical N sink within polders. Its amount was generally smaller than that of N fertilization. Volatilization and denitrification were important N transformation processes within the polders, and were strongly affected by air temperature and water content in the agricultural farmlands. 4.2. What is the contribution of polders to nitrogen retention? Although polders in Lake Taihu Basin were N sources for their surrounding rivers and lakes, they had a large N retention rate (Fig. 3 (b)). In lowland polders, there are three major pathways for N retention, including particulate N settling, oxidized N denitrification, and N uptake by aquatic plants (see NDP’s conceptual diagram in the Supporting Information). This study evaluated the polders’ N retention capacity based on the N retention in the surface water area (Section 2.3.2), and revealed a large potential in N retention with an NRR value as high as 52.7-54.6% (Fig. 4 (b)). This N retention capacity was slightly higher compared with several reported N removal/retention rates (40-50%) of wetlands (Land et al., 2013). This is probably due to the following advantages of polders regarding N retention: (1) compared with wetlands with water flowing through them, lowland polders have extremely stable water flow in surface water (e.g., ponds) that can potentially enhance particulate N settling. (2) During rainfall events, runoff water may be manually retained in ditches or ponds, rather than exported into its surrounding rivers. Such water retention would thus enhance N retention through

Page 21 of 31

the processes of particulate N settling, denitrification, and N uptake by aquatic plants. 4.3. Implications for water management This case study quantified N export and retention at 2539 polders in Lake Taihu Basin under the context of global climate change and human activities. The results and findings from our case study can potentially support decision making in controlling polder N export: (1) Polder N export can be efficiently decreased by reducing agricultural N fertilization. The significant (13.5%) decrease in annual N export from 12876 t/yr to 11135 t/yr (Fig. 3) due to fertilization reduction revealed that polder N export was very sensitive to N fertilization (Fig. 3 (a)). In contrast, global climate change had a limited impact on N export (see Fig. 5 (b-c)). Therefore, we suggest paying more attention to fertilization reduction rather than global climate change. This conclusion is different from that of a large-scale study claiming that future (the 21st century) changes in precipitation are likely to increase watershed N loading, especially in eastern China (Sinha et al., 2017). Therefore, to better control N export, it is helpful to identify a minimal N fertilization amount without any risk of decline in grain yield and farmer income. This N fertilization amount is not constant under the context of global climate change or human activities, however, is particularly important for intensive farming areas, such as Lake Taihu Basin, because excess use of N fertilizer has been a common practice (Hofmeier et al., 2015; Zhao et al., 2012). (2) Regionalized management was critical to control polder N export. The simulation Page 22 of 31

results revealed that the NEC and NRR had large spatial heterogeneity (Fig. 5-6) and different annual fluctuations among eight hydraulic zones (Fig. 4 (a)). An additional analysis in Section 4 (Supporting Information) revealed that polder N export was sensitive to precipitation and surface water area, but was not sensitive to residential area and population parameters. This result implied that both agricultural activities and domestic wastewater have comparable impacts on polder N export. Therefore, the N reduction strategy for each polder should be investigated based on their characteristics including land use, population, etc. In our study area, polders can be mainly classified into two types, i.e., agricultural polders and urban polders. For agricultural polders with large areas of farmlands, proper fertilization strategies would be a primary strategy to reduce N export due to the large contribution of fertilization to N export. For urban polders with large populations, wastewater was the main source of N export. Therefore, enhancing N removal from wastewater was recommended to reduce N export. (3) Surface water areas (e.g., ponds and ditches) within lowland polders can serve as N retention spots in N management practice. N was exported into surrounding rivers mainly through three pathways (seepage, culvert and flood drainage). All these pathways occurred in the surface water area, implying that reducing the N concentration in the surface water area can significantly reduce polder N exports. To enhance polder N retention, several strategies can be implemented, such as developing constructed wetlands by increasing macrophyte cover in surface water areas within polders. Wetlands showed potentials for nutrient retention, such as increasing N Page 23 of 31

uptake and enhancing particulate N settling due to the improved stability of hydrodynamic conditions (Land et al., 2013; Mitsch et al., 2005; Vymazal, 2007). However, it is important to note that N retention in wetlands is not a perfect solution, because continuous N retention in wetlands would result in N enrichment, and could thus be a potential N source for downstream rivers and lakes through specific pathways, such as re-suspension and plant decay. Therefore, certain strategies (e.g., plant harvest) for the constructed wetlands were suggested to avoid N release to water. 4.4. Uncertainty analysis Modeling nutrient dynamics within lowland watersheds under the context of global climate change is a worldwide challenge. A thorough uncertainty analysis was not among the scope of this study. However, some uncertainty sources in this case study were briefly mentioned for its proper transfer to other case studies. The major uncertainties were from the NDP model, and the global climate change data that were utilized as model inputs. The uncertainties from the NDP model included model structure, parameters and input data with further descriptions in Huang et al. (2018). According to the global climate change data from CMIP5, there are >20 models that generate daily climate conditions in the future (Kharin et al., 2013). In this case, we chose only one model (BCC-CSM1.1) due to the intensive computation for a single run of 2539 watersheds. However, it will be helpful to compare the model results (N export and nitrogen) with those obtained using other model input data in the case of computational advances. Future studies should investigate and reduce model

Page 24 of 31

uncertainty with more available data and advanced methods, such as the Bayesian modeling framework, which is able to combine multi-source uncertainties together (Kelly et al., 2019).

5. Conclusions N export and retention of all 2539 polders in Lake Taihu Basin, China were investigated by simulating N dynamics using the NDP model. The investigation results revealed that N fertilization determined the amount of N export, while global climate change determined the fluctuation of N export. Polder N retention was positively related to the N export amount. To control polder N export, fertilization reduction, regionalized management and the development of N retention plants in surface water area were recommended. This study demonstrated the use of a numerical model (NDP) in investigating N export and retention for polders at a watershed scale. Although this study used Lake Taihu Basin as an example, the modeling framework and the scenario analysis method can potentially be used in other lowland polders with similar hydrological management worldwide.

Acknowledgments The project was financially supported by Youth Innovation Promotion Association CAS (2019313), National Natural Science Foundation of China (41971138), China Postdoctoral Science Foundation (2019M651891) and Major Science and Technology Program for Water Pollution Control and Treatment of China (2017ZX07301-001-02). The authors would like to thank China Meteorological Data Sharing Service System

Page 25 of 31

for providing the measured data for model development.

References Abouali, M., Nejadhashemi, A.P., Daneshvar, F., Adhikari, U., Herman, M.R., Calappi, T.J., Rohn, B.G., 2017. Evaluation of wetland implementation strategies on phosphorus reduction at a watershed scale. J. Hydrol. 552, 105-120. https://doi.org/10.1016/j.jhydrol.2017.06.038. Arnold, J.G., Moriasi, D.N., Gassman, P.W., Abbaspour, K.C., White, M.J., Srinivasan, R., Santhi, C., Harmel, R., Van Griensven, A., Van Liew, M.W., 2012. SWAT: Model use, calibration, and validation.

Transactions

of

the

ASABE

55(4),

1491-1508.

http://dx.doi.org/10.13031/2013.42256. Brauer, C.C., Teuling, A.J., Torfs, P.J.J.F., Uijlenhoet, R., 2014. The Wageningen Lowland Runoff Simulator (WALRUS): a lumped rainfall–runoff model for catchments with shallow groundwater. Geosci. Model Dev. 7(5), 2313-2332. https://doi.org/10.5194/gmd-7-2313-2014. Dudula, J., Randhir, T.O., 2016. Modeling the influence of climate change on watershed systems: Adaptation

through

targeted

practices.

J.

Hydrol.

541,

703-713.

https://doi.org/10.1016/j.jhydrol.2016.07.020. Hale, R.L., Grimm, N.B., Vörösmarty, C.J., Fekete, B., 2015. Nitrogen and phosphorus fluxes from watersheds of the northeast U.S. from 1930 to 2000: Role of anthropogenic nutrient inputs, infrastructure,

and

runoff.

Global

Biogeochemical

Cycles

29(3),

2014GB004909.

https://doi.org/10.1002/2014gb004909. Hofmeier, M., Roelcke, M., Han, Y., Lan, T., Bergmann, H., Böhm, D., Cai, Z., Nieder, R., 2015. Nitrogen management in a rice-wheat system in the Taihu Region: Recommendations based on

field

experiments

and

surveys.

Agric.

Ecosyst.

Environ.

209,

60-73.

http://dx.doi.org/10.1016/j.agee.2015.03.032. Huang, J., Arhonditsis, G.B., Gao, J., Kim, D.K., Dong, F., 2018. Towards the development of a modeling framework to track nitrogen export from lowland artificial watersheds (polders). Water Res. 133, 319-337. https://doi.org/10.1016/j.watres.2018.01.011. Huang, J., Gao, J., Jiang, Y., Yin, H., Amiri, B.J., 2017. Sources, distribution and export coefficient of phosphorus in lowland polders of Lake Taihu Basin, China. Environ. Pollut. 231, 1274-1283. https://doi.org/10.1016/j.envpol.2017.08.089. Kelly, N.E., Javed, A., Shimoda, Y., Zastepa, A., Watson, S., Mugalingam, S., Arhonditsis, G.B., 2019. A Bayesian risk assessment framework for microcystin violations of drinking water and recreational standards in the Bay of Quinte, Lake Ontario, Canada. Water Res. 162, 288-301. https://doi.org/10.1016/j.watres.2019.06.005. Kharin, V.V., Zwiers, F.W., Zhang, X., Wehner, M., 2013. Changes in temperature and precipitation extremes

in

the

CMIP5

ensemble.

Clim.

Change

119(2),

345-357.

http://dx.doi.org/10.1007/s10584-013-0705-8. Lai, G., Yu, G., Gui, F., 2006. Preliminary study on assessment of nutrient transport in the Taihu Basin based

on

SWAT

modeling.

Science

in

China

Series

D

49(1),

135-145.

http://dx.doi.org/10.1007/s11430-006-8113-9. Lam, Q.D., Schmalz, B., Fohrer, N., 2012. Assessing the spatial and temporal variations of water Page 26 of 31

quality

in

lowland

areas,

Northern

Germany.

J.

Hydrol.

438-439,

137-147.

http://dx.doi.org/10.1016/j.jhydrol.2012.03.011. Land, M., Granéli, W., Grimvall, A., Hoffmann, C.C., Mitsch, W.J., Tonderski, K.S., Verhoeven, J.T., 2013. How effective are created or restored freshwater wetlands for nitrogen and phosphorus removal?

A

systematic

review

protocol.

Environmental

Evidence

2(1),

16.

http://dx.doi.org/10.1186/s13750-016-0060-0. Li, Z., Luo, C., Xi, Q., Li, H., Pan, J., Zhou, Q., Xiong, Z., 2015. Assessment of the AnnAGNPS model in simulating runoff and nutrients in a typical small watershed in the Taihu Lake basin, China. CATENA 133, 349-361. http://dx.doi.org/10.1016/j.catena.2015.06.007. Lindenschmidt, K.-E., Pech, I., Baborowski, M., 2009. Environmental risk of dissolved oxygen depletion of diverted flood waters in river polder systems - A quasi-2D flood modelling approach.

Sci.

Total

Environ.

407(5),

1598-1612.

http://dx.doi.org/10.1016/j.scitotenv.2008.11.024. Liu, R., Wang, Q., Xu, F., Men, C., Guo, L., 2017. Impacts of manure application on SWAT model outputs

in

the

Xiangxi

River

watershed.

J.

Hydrol.

555,

479-488.

https://doi.org/10.1016/j.jhydrol.2017.10.044. Lu, M., Chang, C., Lin, T., Wang, L., Wang, C., Hsu, T., Huang, J., 2017. Modeling the terrestrial N processes in a small mountain catchment through INCA-N: A case study in Taiwan. Sci. Total Environ. 593-594, 319-329. https://doi.org/10.1016/j.scitotenv.2017.03.178. Ma, X., Li, Y., Zhang, M., Zheng, F., Du, S., 2011. Assessment and analysis of non-point source nitrogen and phosphorus loads in the Three Gorges Reservoir Area of Hubei Province, China. Sci. Total Environ. 412–413, 154-161. http://dx.doi.org/10.1016/j.scitotenv.2011.09.034. Marshall, E., Randhir, T., 2008. Effect of climate change on watershed system: a regional analysis. Clim. Change 89(3), 263-280. http://dx.doi.org/10.1007/s10584-007-9389-2. Meehl, G.A., Boer, G.J., Covey, C., Latif, M., Stouffer, R.J., 1997. Intercomparison makes for a better climate model. Eos, Transactions American Geophysical Union 78(41), 445-451. https://doi.org/10.1029/97EO00276. Mitsch, W.J., Zhang, L., Anderson, C.J., Altor, A.E., Hernández, M.E., 2005. Creating riverine wetlands: Ecological succession, nutrient retention, and pulsing effects. Ecol. Eng. 25(5), 510-527. http://dx.doi.org/10.1016/j.ecoleng.2005.04.014. Ouyang, W., Gao, X., Wei, P., Gao, B., Lin, C., Hao, F., 2017. A review of diffuse pollution modeling and associated implications for watershed management in China. J. Soils Sediments 17(6), 1527-1536. https://doi.org/10.1007/s11368-017-1688-2. Qin, B., Paerl, H.W., Brookes, J.D., Liu, J., Jeppesen, E., Zhu, G., Zhang, Y., Xu, H., Shi, K., Deng, J., 2019. Why Lake Taihu continues to be plagued with cyanobacterial blooms through 10 years (2007–2017)

efforts.

Science

Bulletin

64(6),

354-356.

https://doi.org/10.1016/j.scib.2019.02.008. Qiu, J., Shen, Z., Leng, G., Xie, H., Hou, X., Wei, G., 2019. Impacts of climate change on watershed systems and potential adaptation through BMPs in a drinking water source area. J. Hydrol. 573, 123-135. https://doi.org/10.1016/j.jhydrol.2019.03.074. Qu, H.J., Kroeze, C., 2010. Past and future trends in nutrients export by rivers to the coastal waters of China. Sci. Total Environ. 408(9), 2075-2086. https://doi.org/10.1016/j.scitotenv.2009.12.015. Sinha, E., Michalak, A.M., Balaji, V., 2017. Eutrophication will increase during the 21st century as a result

of

precipitation

changes.

Page 27 of 31

Science

357(6349),

405-408.

http://dx.doi.org/10.1126/science.aan2409. Sonneveld, M., de Vos, J.A., Kros, J., Knotters, M., Frumau, A., Bleeker, A., de Vries, W., 2012. Assessment of N and P status at the landscape scale using environmental models and measurements. Environ. Pollut. 162, 168-175. http://dx.doi.org/10.1016/j.envpol.2011.11.020. Swaney, D.P., Hong, B., Ti, C., Howarth, R.W., Humborg, C., 2012. Net anthropogenic nitrogen inputs to watersheds and riverine N export to coastal waters: a brief overview. Current Opinion in Environmental Sustainability 4(2), 203-211. https://doi.org/10.1016/j.cosust.2012.03.004. Taihu-Basin-Authority, 2014. External nutrient loading of Lake Taihu in 2013. van der Grift, B., Broers, H.P., Berendrecht, W., Rozemeijer, J., Osté, L., Griffioen, J., 2016. High-frequency monitoring reveals nutrient sources and transport processes in an agriculture-dominated lowland water system. Hydrol. Earth Syst. Sci. 20(5), 1851-1868. http://dx.doi.org/10.5194/hess-20-1851-2016. Vinçon-Leite, B., Casenave, C., 2019. Modelling eutrophication in lake ecosystems: A review. Sci. Total Environ. 651, 2985-3001. https://doi.org/10.1016/j.scitotenv.2018.09.320. Vymazal, J., 2007. Removal of nutrients in various types of constructed wetlands. Sci. Total Environ. 380(1-3), 48-65. http://dx.doi.org/10.1016/j.scitotenv.2006.09.014. Wade, A., Durand, P., Beaujouan, V., Wessel, W., Raat, K., Whitehead, P., Butterfield, D., Rankinen, K., Lepisto, A., 2002. A nitrogen model for European catchments: INCA, new model structure and

equations.

Hydrol.

Earth

Syst.

Sci.

6(3),

559-582.

https://doi.org/10.5194/hess-6-559-2002. Wellen, C., Kamran-Disfani, A.-R., Arhonditsis, G.B., 2015. Evaluation of the current state of distributed watershed nutrient water quality modeling. Environ. Sci. Technol. 49(6), 3278-3290. http://dx.doi.org/10.1021/es5049557. White, M., Harmel, D., Yen, H., Arnold, J., Gambone, M., Haney, R., 2015. Development of Sediment and Nutrient Export Coefficients for U.S. Ecoregions. JAWRA Journal of the American Water Resources Association 51(3), 758-775. https://doi.org/10.1111/jawr.12270. Wu, T., Li, W., Ji, J., Xin, X., Li, L., Wang, Z., Zhang, Y., Li, J., Zhang, F., Wei, M., Shi, X., Wu, F., Zhang, L., Chu, M., Jie, W., Liu, Y., Wang, F., Liu, X., Li, Q., Dong, M., Liang, X., Gao, Y., Zhang, J., 2013. Global carbon budgets simulated by the Beijing Climate Center Climate System Model for the last century. Journal of Geophysical Research: Atmospheres 118(10), 4326-4347. https://doi.org/10.1002/jgrd.50320. Wu, T., Lu, Y., Fang, Y., Xin, X., Li, L., Li, W., Jie, W., Zhang, J., Liu, Y., Zhang, L., Zhang, F., Zhang, Y., Wu, F., Li, J., Chu, M., Wang, Z., Shi, X., Liu, X., Wei, M., Huang, A., Zhang, Y., Liu, X., 2019. The Beijing Climate Center Climate System Model (BCC-CSM): the main progress

from

CMIP5

to

CMIP6.

Geosci.

Model

Dev.

12(4),

1573-1600.

https://doi.org/10.5194/gmd-12-1573-2019. Xu, C., Xu, Y., 2012. The Projection of Temperature and Precipitation over China under RCP Scenarios using a CMIP5 Multi-Model Ensemble. Atmospheric and Oceanic Science Letters 5(6), 527-533. http://dx.doi.org/10.1080/16742834.2012.11447042. Yang, X., Liu, Q., Fu, G., He, Y., Luo, X., Zheng, Z., 2016. Spatiotemporal patterns and source attribution of nitrogen load in a river basin with complex pollution sources. Water Res. 94, 187-199. http://dx.doi.org/10.1016/j.watres.2016.02.040. Zhang, Y., Zhou, Y., Shao, Q., Liu, H., Lei, Q., Zhai, X., Wang, X., 2016. Diffuse nutrient losses and the impact factors determining their regional differences in four catchments from North to

Page 28 of 31

South China. J. Hydrol. 543, 577-594. https://doi.org/10.1016/j.jhydrol.2016.10.031. Zhao, X., Zhou, Y., Wang, S., Xing, G., Shi, W., Xu, R., Zhu, Z., 2012. Nitrogen balance in a highly fertilized rice–wheat double-cropping system in southern China. Soil Sci. Soc. Am. J. 76(3), 1068-1078. http://dx.doi.org/10.2136/sssaj2011.0236.

Page 29 of 31

Page 30 of 31

 N export and retention for polders in a watershed scale was characterized.  Spatially-averaged N export from polders in Lake Taihu Basin was 15.3 kg/ha/yr.  Reducing N fertilizer by 80 kg/ha/yr can reduce annual N export by 13.5%.

Page 31 of 31