Modelling the effects of climate and land-use change on the hydrochemistry and ecology of the River Wye (Wales)

Full text

1
Modelling Climate and Land-use
Change Impacts on Hydrology, Nitrate
and Aquatic Biota in the River Wye
(Wales)
Gianbattista Bussi1*, Paul G. Whitehead1, Cayetano Gutiérrez-Cánovas2,3, José L. J. Ledesma4,
Steve J. Ormerod2, Raoul-Marie Couture5,6
1 School of Geography and the Environment, University of Oxford, South Parks Road, Oxford, OX1
3QY (UK)
2 Catchment Research Group, Cardiff University, School of Biosciences, The Sir Martin Evans
Building, Museum Avenue, Cardiff CF10 3AX (UK)
3 Freshwater Ecology and Management group, Department of Evolutionary Biology, Ecology and
Environmental Sciences - Institut de Recerca de la Biodiversitat (IRBio), University of Barcelona,
Avinguda Diagonal, 643, 08028 Barcelona (Spain)
4 Department of Aquatic Sciences and Assessment, Swedish University of Agricultural Sciences,
Lennart Hjelms väg 9, 750 07 Uppsala (Sweden)
5 Norwegian Institute for Water Research, Gaustadalléen 21, Oslo, 0349 (Norway)
6 Ecohydrology Group, Department of Earth and Environmental Sciences, University of Waterloo,
Waterloo, Canada, G1S1W2
*Corresponding author: gianbattista.bussi@ouce.ox.ac.uk
COVER LETTER
Dear Editor,
Climate and Land-use Change Impacts on
Hydrology, Nitrate and Aquatic Biota in the River Wye (Wales)
results of a modelling framework to evaluate the impacts of socio-economic scenarios, including
climate change and land-use change, on the nutrient concentration and aquatic biota of the River
Wye, in Wales. We believe that this paper should be of interest for your journal, from a
methodological point of view, since the approach we present, coupling climate, hydrological, water
quality and ecological models, has not been employed before and from a policy point of view, since it
indicates possible future outcomes of ecological diversity response in aquatic communities, pointing
out possible mitigation strategies. This paper cover a topic that has implications the hydrosphere, the
biosphere and the anthroposhpere.
Thank you for your consideration of this manuscript.
Sincerely,
The authors
Cover Letter

1
Modelling Climate and Land-use
Change Impacts on Hydrology, Nitrate
and Aquatic Biota in the River Wye
(Wales)
Gianbattista Bussi1*, Paul G. Whitehead1, Cayetano Gutiérrez-Cánovas2,3, José L. J. Ledesma4,
Steve J. Ormerod2, Raoul-Marie Couture5,6
1 School of Geography and the Environment, University of Oxford, South Parks Road, Oxford, OX1
3QY (UK)
2 Catchment Research Group, Cardiff University, School of Biosciences, The Sir Martin Evans
Building, Museum Avenue, Cardiff CF10 3AX (UK)
3 Freshwater Ecology and Management group, Department of Evolutionary Biology, Ecology and
Environmental Sciences - Institut de Recerca de la Biodiversitat (IRBio), University of Barcelona,
Avinguda Diagonal, 643, 08028 Barcelona (Spain)
4 Department of Aquatic Sciences and Assessment, Swedish University of Agricultural Sciences,
Lennart Hjelms väg 9, 750 07 Uppsala (Sweden)
5 Norwegian Institute for Water Research, Gaustadalléen 21, Oslo, 0349 (Norway)
6 Ecohydrology Group, Department of Earth and Environmental Sciences, University of Waterloo,
Waterloo, Canada, G1S1W2
*Corresponding author: gianbattista.bussi@ouce.ox.ac.uk
GRAPHICAL ABSTRACT
*Graphical Abstract

1
Modelling Climate and Land-use
Change Impacts on Hydrology, Nitrate
and Aquatic Biota in the River Wye
(Wales)
Gianbattista Bussi1*, Paul G. Whitehead1, Cayetano Gutiérrez-Cánovas2,3, José L. J. Ledesma4,
Steve J. Ormerod2, Raoul-Marie Couture5,6
1 School of Geography and the Environment, University of Oxford, South Parks Road, Oxford, OX1
3QY (UK)
2 Catchment Research Group, Cardiff University, School of Biosciences, The Sir Martin Evans
Building, Museum Avenue, Cardiff CF10 3AX (UK)
3 Freshwater Ecology and Management group, Department of Evolutionary Biology, Ecology and
Environmental Sciences - Institut de Recerca de la Biodiversitat (IRBio), University of Barcelona,
Avinguda Diagonal, 643, 08028 Barcelona (Spain)
4 Department of Aquatic Sciences and Assessment, Swedish University of Agricultural Sciences,
Lennart Hjelms väg 9, 750 07 Uppsala (Sweden)
5 Norwegian Institute for Water Research, Gaustadalléen 21, Oslo, 0349 (Norway)
6 Ecohydrology Group, Department of Earth and Environmental Sciences, University of Waterloo,
Waterloo, Canada, G1S1W2
*Corresponding author: gianbattista.bussi@ouce.ox.ac.uk
HIGHLIGHTS
Socio-economic scenarios used to assess future changes in river nutrients and biota
Climate change expected to cause nutrient enrichment
Longitudinal position along the river mediates ecological response
Land-use change plays critical role in mitigation of climate change
*Highlights (for review)

1
Modelling Climate and Land-use
1
Change Impacts on Hydrology, Nitrate
2
and Aquatic Biota in the River Wye
3
(Wales)
4
Gianbattista Bussi1*, Paul G. Whitehead1, Cayetano Gutiérrez-Cánovas2,3, José L. J. Ledesma4, 5
Steve J. Ormerod2, Raoul-Marie Couture5,6 6
1 School of Geography and the Environment, University of Oxford, South Parks Road, Oxford, OX1 7
3QY (UK) 8
2 Catchment Research Group, Cardiff University, School of Biosciences, The Sir Martin Evans 9
Building, Museum Avenue, Cardiff CF10 3AX (UK) 10
3 Freshwater Ecology and Management group, Department of Evolutionary Biology, Ecology and 11
Environmental Sciences - Institut de Recerca de la Biodiversitat (IRBio), University of Barcelona, 12
Avinguda Diagonal, 643, 08028 Barcelona (Spain) 13
4 Department of Aquatic Sciences and Assessment, Swedish University of Agricultural Sciences, 14
Lennart Hjelms väg 9, 750 07 Uppsala (Sweden) 15
5 Norwegian Institute for Water Research, Gaustadalléen 21, Oslo, 0349 (Norway) 16
6 Ecohydrology Group, Department of Earth and Environmental Sciences, University of Waterloo, 17
Waterloo, Canada, G1S1W2 18
*Corresponding author: gianbattista.bussi@ouce.ox.ac.uk 19
ABSTRACT
20
Interactions between climate change and land use change might have substantial effects on aquatic 21
ecosystems, but are still poorly understood. Using the Welsh River Wye as a case study, we linked 22
models of water quality (Integrated Catchment - INCA) and climate (GFDL - Geophysical Fluid 23
Dynamics Laboratory and IPSL - Institut Pierre Simon Laplace) under greenhouse gas scenarios 24
(RCP4.5 and RCP8.5) to drive a bespoke ecosystem model that simulated the responses of aquatic 25
biota. The potential effects of economic and social development were also investigated using 26
scenarios from the EU MARS project (Managing Aquatic Ecosystems and Water Resources under 27
Multiple Stress). Longitudinal position along the river mediated response to increasing anthropogenic 28
pressures. Upland locations appeared particularly sensitive to nutrient enrichment or potential re-29
acidification compared to lowland environments which are already eutrophic. These results provide 30
some ideas on how to mitigate future impacts and reiterate the need for sensitive land management in 31
upland, temperate environments which are likely to become increasingly important to water supply 32
and biodiversity conservation as the effects of climate change intensify. 33
34
Keywords: Climate Change, Water quality, River Wye, Nitrogen, Ecology 35
36
*Manuscript
Click here to download Manuscript: paper_GBUSSI_manuscript.docx Click here to view linked References

2
1 INTRODUCTION
37
With the Paris Agreement and the Intergovernmental Panel on Climate Change (IPCC) Reports 38
(Pachauri et al., 2014), there is no doubt about the global significance of climate change driven by 39
anthropogenic sources of carbon dioxide. The positive and negative impacts of climate change on the 40
natural environment and people across the globe are still being considered and debated, but the 41
potential changes in precipitation, temperature and sea level rise over the next century are likely to be 42
significant, with important impacts on hydrology, water quality and ecology (Whitehead et al., 2009). 43
The IPCC report considers the impacts of socioeconomic change superimposed on climate change. 44
Thus Shared Socio-economic Pathways (SSPs) will interact with climate change to generate a 45
combined impact on people and livelihoods. This provides an integrated framework for addressing 46
issues of change for national, regional and local governments and organizations to consider. 47
The IPCC report and the EU Water Framework Directive (Chave, 2001) provide a backdrop to a major 48
new research project entitled MARS (Managing Aquatic Ecosystems and Water Resources under 49
Multiple Stress) funded by the European Union under the 7th Framework Programme (Hering et al., 50
2015). This project aims to address two target groups, namely water managers, assessing and 51
restoring rivers and lakes, and policy makers, drafting and implementing policies related to water. In 52
any such study there is a need to understand the effects of multiple stressors on surface waters and 53
groundwaters, their biota, and the ecosystem services they provide to people. River ecosystems are 54
very likely to be affected by land-use or climate changes (Strayer and Dudgeon, 2010). Several 55
studies have observed a reduction in diversity or abundance of river organisms in response to land-56
use (e.g., Gutiérrez-Cánovas et al., 2013), climate change (e.g., Durance and Ormerod, 2007) or 57
anthropogenic disturbances (Ruhí et al., 2015). The reduction in river biodiversity is likely to reduce 58
the capacity of these ecosystems to provide essential goods and services (Hooper et al., 2005), such 59
as clean water, flood prevention or timber. The knowledge from the EU MARS project can be 60
translated into advice on river basin management, assessing how to restore multiply stressed rivers 61
and lakes and providing new measures or programmes to meet the EU Water Framework Directives. 62
As part of the MARS project, upland Wales has been investigated as a Northern Region that has 63
been subject to much environmental change over the past 50 years (Durance and Ormerod, 2007; 64
Whitehead et al., 2009, 1998a). Many of the upland headwaters in mid and southern Wales drain into 65
large river systems and one of those is the River Wye (Figure 1). In this study we evaluate the River 66
Wye in terms of its hydrology, water quality and ecology and how these might change under a 67
changing climate and changing socio-economic pressures. We utilise the INCA suite of models 68
(Wade et al., 2002; Whitehead et al., 1998a) to quantify the change and use the model to simulate 69
new future approaches to manage the environment. 70
2 THE WYE CATCHMENT
71
The River Wye catchment is located in the Western Regions of the UK, in South and Mid- Wales, as 72
shown in Figure 1. It flows from Mid-Wales towards South-East Wales, reaching the River Severn 73
estuary and the Bristol Channel at the town of Chepstow. Its catchment area is 4131 km2. The Rivers 74
Lugg and the Monnow are its main tributaries, flowing into the main River Wye reach downstream of 75
Hereford (Jarvie et al., 2005). The main land use is agriculture with livestock farming predominating in 76
the north and west and more intensive arable farming in the south and east of the catchment. There is 77
some industry based around the major towns (e.g. Monmouth and Chepstow). The upland areas of 78
the catchment are generally used for rough grazing, while lowland areas support mixed and dairy-79
farming and horticulture (Oborne et al., 1980). The water quality of the River Wye is characterised by 80
patterns of high winter concentrations of nitrate and low summer concentrations (Oborne et al., 1980), 81
mainly related to agriculture and fertiliser usage. 82

3
The Wye catchment is rich in wildlife and habitats and this is recognised in the designation of the Wye 83
and several tributaries as a riverine Special Area of Conservation. The area offers many opportunities 84
for water-based recreation. The River Wye is a well-established and nationally significant salmon and 85
brown trout rod fishery and also a locally important coarse fish fishery. Elver fishing also takes place 86
within the tidal reaches of the Wye. The Elan Valley system of reservoirs (North-Western part of the 87
catchment) is vital in providing water for Birmingham, Gloucestershire and South Wales. The local 88
economy is moderately dependent on businesses requiring water abstraction, primarily agricultural, 89
where trickle and spray irrigation is frequently used. 90
91
Figure 1. The River Wye catchment and the INCA model sub-catchments.
92
Daily water discharge time series have been retrieved from the National River Flow Archive (NRFA, 93
http://nrfa.ceh.ac.uk/data/search). Several stream gauges can be found within the River Wye 94
catchment (see Table 1). Nitrate and Ammonium data were obtained from the Centre for Ecology and 95
Hydrology for eight locations within the River Wye catchment, collected from 2004 to 2009 with a 96
monthly to fortnightly frequency, and used for model calibration. Data collected by the Environment 97
Agency of England and Wales was also used for three locations, spanning from 1974 to 2012, and 98
used for model validation (Simpson, 1980). 99
Table 1. Data used for the INCA model set-up and calibration.
100
Variable Data source Use
Temporal
coverage
Spatial
coverage
Precipitation and
temperature Met Office Input 1960-2015 All
Elevation Ordnance Survey Parameter - All
Land use
Land Cover Map 2007, Centre of
Ecology (Smith et al., 2007) Parameter 2007 All
Atmospheric
deposition Donald and Stoner, (1989) Parameter 1989 All
Sewage treatment
plants European Environment Agency Parameter 2014
Reaches 11, 13,
14, 17, 18
Flow National River Flow Archive
Calibration and
validation 1960-2015
Reaches 2
, 8, 9,
10, 14, 15, 17,
21, 28
Nitrate Centre for Ecology and Hydrology Calibration 2004-2009 Reaches
2, 9,
12, 15, 27, 28, 29
Nitrate
Environment Agency of England and
Wales Validation 1974-2012 Reaches
2, 14,
28

4
3 METHODOLOGY
101
3.1 THE INTEGRATED CATCHMENT MODEL (INCA)
102
The INCA model is a process-based model which simulates the main processes related with rainfall-103
runoff transformation and the cycle and fate of several compounds, such as nitrate, ammonium, 104
carbon and phosphorus. The INCA Model has been developed over several years as a result of 105
several research projects and is a dynamic computer model that predicts water quantity and quality in 106
rivers and catchments. The primary aim of INCA is to provide a process-based representation of the 107
factors and processes controlling flow and water quality dynamics in both the land and in-stream 108
components of river catchments, whilst minimising data requirements and model structural complexity 109
(Whitehead et al., 1998a, 1998b). As such, the INCA model produces daily estimates of discharge, 110
111
Also, the model is semi-distributed, so that spatial variations in land use and management can be 112
taken into account. The hydrological and nutrient fluxes from different land use classes and sub-113
catchment boundaries are modelled simultaneously and information fed sequentially into a multi-114
reach river model. The INCA model was originally tested on 20 catchments in the UK for a variety of 115
purposes (Crossman et al., 2013; Lu et al., 2017, 2016; Nizzetto et al., 2016; Whitehead et al., 2016), 116
including catchments in Wales (Bussi et al., 2017b), and over 20 catchments across the EU and now 117
30 catchments around the world. The INCA model has also been tested for land-use and climate 118
change impact assessment applications (Bussi et al., 2016a, 2016b). 119
120
Figure 2. The land component of nitrogen cycle, from Whitehead et al. (1998a) and Wade et al. (2002a).
121
The INCA model requires time series of Hydrological Effective Rainfall (HER) and Soil Moisture Deficit 122
(SMD) as inputs, and these have to be produced by an independent hydrological model, which takes 123
into account soil water retention and evapotranspiration. In this study, the PERSiST model was used 124
(Futter et al., 2014), a simple and flexible hydrological model especially created to produce inputs for 125
the INCA family of models. Precipitation and temperature data were taken from Met Office stations. 126
Several stations exist within the Wye catchment, measuring daily meteorological variables. Given the 127
topography of the catchment, with steep slopes and relatively large difference in altitude from the 128
uplands to the lowlands, and the natural spatial variability of rainfall and temperature, a single station 129
cannot provide exhaustive information about the precipitation falling on the catchment and the 130

5
temperature over the whole catchment. For this reason, the average precipitation falling on the 131
catchment and the average catchment temperature were determined, using information from several 132
rain gauges spread all over the catchment. The mean temperature was calculated as the average 133
between minimum and maximum temperature. 134
Spatially distributed information is required to estimate some of the INCA model parameters. The 135
Ordnance Survey (OS) Terrain 50 was used as a digital elevation model. The digital elevation model 136
was used to define the sub-catchment boundaries, and to calculate their areas and mean reach slope. 137
The Land Cover Map 2007, released by the Centre of Ecology and Hydrology in 2011, was used to 138
characterise the land uses in the catchment (Smith et al., 2007). The land cover categories were 139
aggregated to six classes of land use: forest, short vegetation (ungrazed), short vegetation (grazed, 140
non-fertilised), short vegetation (fertilised), arable and urban, following Jin et al. (2012). The 141
proportion of land use for each sub-catchment is required by the INCA model. The River Wye 142
catchment was divided into several sub-catchments (Table 2.). For each catchment, catchment area, 143
reach length and land uses were defined. 144
The INCA model was calibrated over the time period 1/1/2004 31/10/2009 and validated over the 145
time period 1960-2015, with a daily time step. The model parameters were manually adjusted to 146
reproduce observed values of hydrological and water quality variables. In particular, observed values 147
of water discharge were used to calibrate the hydrological model parameters (direct runoff residence 148
time, soil water residence time, ground water residence time, threshold soil zone flow, rainfall excess 149
proportion, maximum infiltration rate, discharge/velocity relationship coefficient and exponent) and 150
observed values of nitrate concentration were used to calibrate the INCA model parameters 151
(denitrification rate in soil and river, nitrification rate in soil and river, mineralisation rate in soil, 152
immobilisation rate in soil, fertiliser addition rate in soil, plant uptake). Manual calibration has been 153
proved as a robust method for obtaining acceptable simulations with the INCA family of models 154
(Ledesma et al., 2012). 155
Table 2. INCA model sub-catchments and reaches and their correspondent land uses.
156
Reach
ID
Area
(km2)
Lengt
h (km)
Forest (%)
Short
vegetation
(ungrazed)
(%)
Short
vegetation
(grazed,
non-
fertilised)
(%)
Short
vegetation
fertilised
(%)
Arable
(%) Urban (%)
1 3.84 1.73 0 33 67 0 0 0
2 6.70 3.24 0 0 100 0 0 0
3 11.75 5.79 25 10 65 0 0 0
4 5.22 2.04 16 7 77 0 0 0
5 41.28 11.12 26 7 40 26 1 0
6 103.92
13.4 6 12 42 39 1 1
7 183.95
22.4 3 43 51 0 0 4
8 358.81
44.7 6 5 23 62 3 1
9 323.74
20.08 4 13 20 56 7 1
10 244.31
33.5 19 14 24 37 6 0
11 892.50
72.2 7 3 3 43 42 2
12 131.92
22 8 12 12 57 11 0
13 162.73
25.1 7 1 4 52 35 1
14 186.06
25.5 4 0 0 14 79 3
15 130.81
15.5 3 6 6 70 15 2
16 166.74
26 1 0 0 40 58 1
17 72.59 11.9 6 0 0 10 57 27
18 30.02 16.7 21 0 0 8 71 0
19 43.93 9.8 13 0 0 4 75 8
20 153.44
22.9 7 0 1 22 69 0
22 40.36 3.7 6 0 3 36 55 1

6
23 161.29
27.5 2 18 8 49 23 0
24 118.94
17.6 4 0 2 9 84 1
25 91.14 27.3 2 0 0 23 76 0
26 67.75 16.7 6 1 2 41 49 1
27 142.83
25.6 7 1 2 49 40 1
28 118.47
29.1 29 0 0 17 49 5
29 18.91 7 21 0 0 29 34 16
157
3.2 ECOLOGICAL MODELLING
158
In order to forecast the response of the biological communities of the River Wye to land-use and 159
climate change scenarios, ecological models were built using biological and environmental data from 160
78 locations placed in Mid and North Wales. In those rivers, aquatic invertebrates were collected in 161
spring (March-April 2012-13) using kick-sampling (2-minutes just in riffles) to characterise the 162
biological communities (see Supplementary Material Table 1 for the sampling details and biological 163
metric estimation). Invertebrate richness (number of invertebrate taxa) and response diversity (variety 164
and range of biological attributes to cope with disturbance based on functional richness, Villéger et al., 165
2008) were derived to quantify the response of these aquatic communities. Response traits included 166
fuzzy coded traits such as number of generations per year, lifespan, reproduction mode, respiration 167
type, resistance form and dispersal capacity (Tachet et al., 2010). The main gradients of 168
anthropogenic or natural environmental variation were also characterised, including annual pH, total 169
oxidised nitrogen (TON), altitude (alt), precipitation of the wettest month (prec_max), geographical 170
latitude (lat) and longitude (lon). TON was log-transformed to reduce distribution skewness. All the 171
predictors were standardised to mean=0 and SD=1 to allow for within-model coefficient comparison in 172
form of Standardised Effect Sizes (SES). As global models, generalised linear models with a 173
Gaussian (response diversity) or Quasipoisson (species richness) error distributions were fitted for 174
each of the biotic indicators. pH, TON and the interaction pH x TON were included as stressors, and 175
altitude, precipitation of the wettest month, latitude, longitude, and the interaction pH x latitude as 176
descriptors of natural variability 177
A multi-model inference procedure (Feld et al., 2016; Grueber et al., 2011) was used to produce the 178
final statistical multiple regression relationship, SES and the significance for each predictor. Briefly, 179
this procedure consists in producing the models resulting from all possible predictor combinations. 180
Then they were ranked using Akaike Information Criterion (AIC), i.e. the model ranking first was the 181
one minimising the AIC value. The top models that differed in two AIC units or less from the first 182
were retained, along with their model weights (probability of being the best model). 183
Final models were obtained by using a model averaging approach, where we derived a weighted 184
185
Burnham & Anderson, 2002). The variances explained by fixed terms (r2marginal) and by the 186
combination of fixed plus random terms (r2conditional) were estimated using the method provided by 187
Nakagawa and Schielzeth (2013). Also, for each of the top models, we checked residuals to assess 188
the normality and homoscedasticity of their distributions. Details of the ecological modelling are given 189
by Feld et al. (2016). 190
3.3 ANTHROPOGENIC DRIVERS
191
Socio-economic scenarios
192
Future socio-economic scenarios were implemented within the INCA model to simulate the impact of 193
human-induced changes such as land-use change, population growth and agriculture intensification. 194
These scenarios were based on the Shared Socioeconomic Pathways (SSPs) from Moss et al. 195
(2010), van Vuuren et al. (2014), (2014) and Kriegler et al. (2014). The scenarios were 196

7
downscaled in the MARS project in concert with the stakeholders, such that they represent likely and 197
achievable futures (Cremona et al., 2017). Three scenarios were considered: 198
1. Consensus (called SSP2 in the MARS project). In this world, trends typical of recent decades 199
continue, with some progress towards achieving development goals, reductions in resource 200
and energy intensity at historic rates, and slowly decreasing fossil fuel dependency. 201
2. Fragmentation, or Fragmented World (called SSP3 in the MARS project). The world is 202
separated into regions characterised by extreme poverty, pockets of moderate wealth and a 203
bulk of countries that struggle to maintain living standards for a strongly growing population. 204
Changes in pH were also considered, as very intense use of fossil energy (coal and 205
unconventional sources) with absence of flue-gas desulphurisation technologies to reduce 206
costs is foreseen. This would lead to a pervasive and large acid sulphur deposition, which 207
decreases pH (2030: -0.50 pH units / 2060: -0.75 pH units). 208
3. Technoworld (called SSP5 in the MARS project). This world stresses conventional 209
development oriented toward economic growth as the solution to social and economic 210
problems through the pursuit of enlightened self-interest. The preference for rapid 211
conventional development leads to an energy system dominated by fossil fuels, resulting in 212
high GHG emissions and challenges to mitigation. Lower socio-environmental challenges to 213
adaptation result from attainment of human development goals, robust economic growth, 214
highly engineered infrastructure with redundancy to minimize disruptions from extreme 215
events, and highly managed ecosystems. 216
The socio-economic scenarios were implemented by changing some of the INCA model parameters 217
in order to reproduce the impact of the different storylines, as indicated in Table 3.
218
Table 3. Land-use change scenarios implemented in INCA.
219
Scenario Consensus Fragmentation Technoworld
Forest land
variations
10% of the forest area turned into
arable land
5% of the forest area turned into
grassland
10% of the forest area turned into
arable land
Arable land
variations
30% of arable
land was into
grassland
15% of grassland turned into
arable land
30% of grassland was turned
arable land
Fertiliser use
variations
Nitrogen fertiliser application
decreased by 50%
Nitrogen fertiliser application
increased by 15%
Nitrogen fertiliser appli
cation
increased by 30%
Growing
season
variations
Growing season extended two
months due to climate change
Growing season extended two
months due to climate change
Growing season extended two
months due to climate change
Population
growth
Effluent flows
were increased by
30% due to population growth
Effluent flows were increased by
30% due to population growth
Effluent flows were increased by
30% due to population growth
Climate change scenarios
220
In order to model the impact of climate change, future scenarios of precipitation and temperature are 221
needed. These were obtained using two different global circulation models (GCMs): 222
1. The GFDL (Geophysical Fluid Dynamics Laboratory) model, developed by the National 223
Oceanic and Atmospheric Administration (NOAA, US) (Donner et al., 2011), 224
2. The IPSL (Institut Pierre Simon Laplace) model, developed by the IPSL Climate Modelling 225
Centre (France) (Dufresne et al., 2013). 226
Daily precipitation and temperature, spatially averaged over the Wye catchments, were obtained from 227
these two models forced by two different Representative Concentration Pathways (RCPs), or 228
greenhouse gas concentration trajectories (Moss et al., 2008), the RCP4.5 and the RCP8.5. RCP4.5 229
describes a mean global warming of 1.4 (0.9 to 2.0) °C for 2046-2065, while RCP8.5 presents a 230
global warming of 2.0 (1.4 to 2.6) °C for the same time period. 231
The climate model data were corrected in order to remove the model bias in reproducing past 232
precipitation and temperature. In particular, a Delta Change approach was used to correct the bias of 233

8
RCM (Regional climate models) scenarios to generate local climate scenarios. The approach is based 234
upon transferring the monthly average change signal between RCM (regional climate model) control 235
(2006-2010 in this case) and RCM scenario period to an observed time series (2006-2010 in this 236
case). Three data series were used in the bias correction: (1) GCM control periods (2006-2010), (2) 237
Baseline periods (2006-2010), and (3) GCM scenario period (2011-2099). 238
The scenario daily temperature ( ) was derived by adding the absolute monthly change signals 239
to the observed time series. In the case of precipitation, the observed data were scaled with the 240
relative change signals given by the RCM: 241
(1)
242
(2)
243
Where and are observed daily temperature and precipitation, and 244
are monthly average RCM temperature and precipitation of the control period, and and 245
are monthly average RCM temperature and precipitation of the scenario period. The 246
resulting time series of climate change-affected precipitation and temperature are summarised in 247
Figure 3. 248
249
Figure 3. Changes in precipitation and temperature forecasted by the climate models.
250
Combinations of climate change scenarios and socio-economic scenarios were used to drive the 251
INCA model and the ecology model, to obtain a wide range of possible future pathways. This 252
approach has been used widely in the past (Herrero et al., 2017; Prudhomme and Davies, 2009; 253
Ruiz-Villanueva et al., 2015), and although its limitations have also been clearly pointed out (Singh et 254
al., 2014), it still provides meaningful results for impact and mitigation analysis. 255

9
4 RESULTS
256
4.1 MODEL CALIBRATION
257
Flow and nitrate
258
The INCA model was calibrated and validated to reproduce both the observed flow and nitrate 259
concentration at the stations indicated in Table 1. The results are shown in Table 4, with calibration 260
and validation performances shown in terms of Nash-Sutcliffe Efficiency (NSE, Nash and Sutcliffe, 261
1970) for the daily flow and in term of percent bias (PBIAS) for the nitrate concentration. Reproduction 262
of observed flows is in general good or very good, apart from the station 2, located in the headwaters 263
(catchment area around 10 km2), where the model is able to reproduce the low flows but is 264
underestimating the flood peaks (not shown). In terms of nitrate concentration, again the results were 265
satisfactory (|PBIAS| < 20%) apart from station 2, where the calibration PBIAS was -53%. 266
Calibration results are also shown in Figure 4, Figure 5 and Figure 6, where the daily time series of 267
the modelled flows and nitrate concentrations are shown against the corresponding observed values 268
for selected subcatchments. While it is clear that the model is reproducing the observed flow, some 269
errors in the simulation of the nitrate concentration can be observed, although the general seasonal 270
patterns are reproduced. Nevertheless, given the uncertainty in both the processes and the nitrate 271
measurements, the generic behaviour and the catchment response to climate and land-use changes, 272
the performance of the model can be considered satisfactory. 273
Table 4. INCA model calibration and validation results
274
Calibration
(2004-2009)
Validation
(1960-2015)
Stream Station NSE Flow PBIAS Nitrate NSE
Flow PBIAS Nitrate
Wye 2 0.22 -53% 0.19 -50%
Wye 6 - - 0.60 -
Wye 9 0.77 -15% 0.79 -
Wye 12 - -14% - -
Wye 14 0.82 - 0.73 -66%
Wye 17 0.61 - - -
Wye 27 - 19% - -
Wye 28 0.66 13% 0.79 -2%
Wye 29 - 18% - -
Llynfi 15 0.59 17% 0.56 -
Ieithon 8 0.60 - 0.64 -
Irfon 10 0.58 - 0.61 -
Mynwy 21 0.43 - 0.52 -
275

10
276
Figure 4. INCA model results and observed flow at six sites along the Wye (2004-2009). Note that the stream gauge for
277
the River Lugg does not measure flows above 35 m3 s-1. Red lines: model results. Black lines: measurements.
278
279
Figure 5. INCA model results for Nitrate-N, 2004-2009). Red lines: model results. Black dots: measurements.
280
281
Figure 6. INCA model results for nitrate loads, 2004-2009. Red lines: model results. Black dots: measurements.
282

11
283
4.2 CLIMATE CHANGE AND LAND-USE CHANGE IMPACTS
284
Flow and nitrate
285
The impacts of climate change on the flows of the River Wye for the time period 2050-2079 (hereafter 286
2060s) are shown in Figure 7. It is shown that the different combinations of climate model and RCP 287
provide slightly different results. For example, the model GFDL, coupled with the RCP 8.5 indicates 288
that little changes in hydrology should occur, in agreement with the results in Figure 3, i.e. little 289
changes in precipitation and temperature. Meanwhile, the model IPSL coupled with the RCP 4.5 290
forecasts an increase of 50-100% in autumn-early winter flows. The impacts of climate change on the 291
nitrate concentration of the River Wye for 2060s are shown in Figure 8. Nitrogen concentrations are 292
expected to decrease in the consensus world for the middle and low Wye. However, for the 293
fragmented and technoworld scenarios, models predicted a strong increase in nitrogen concentration 294
for the upper and low Wye. 295
296
Figure 7. Impacts of climate change on flow.
297
298
Figure 8. Impacts of climate change and land-use change on nitrate.
299
Ecological Predictions
300
The results of the ecological modelling showed that pH was the most important stressor, with a strong 301
latitudinal interaction (Figure 9). pH was positively related with ecological metrics, with no evident 302
interaction with TON. At lower latitudes (red: minimum latitude, yellow: latitude Q10 and green: 303

12
latitude Q50) acidity had a much lower effect on biotic metrics, compared to higher latitudes in North 304
Wales (blue: latitude Q90, violet: maximum latitude). Increased TON values reduced only response 305
diversity (Supplementary Material, Table 1, Fig. 1), which was the most responsive variable. Spatial 306
and environmental descriptors were also generally important. In particular, rainfall (prec_max) and 307
elevation, in a lesser extent, had a general negative influence in biotic variables. 308
The future predictions for the ecological metrics reflected different patterns. Invertebrate richness 309
showed little variation in the future scenarios respect to baseline conditions. Only in the Fragmented 310
scenario invertebrate richness showed a slight reduction respect to baseline. On the other hand, 311
response diversity displayed different responses in relation with the longitudinal position in the river 312
and future scenario considered ( 313
314
Figure 9). The upper Wye site seems to reflect a substantial reduction in response diversity under the 315
Techno and Fragmented world scenarios for both 2030s and 2060s periods, which is greater in the 316
latter scenario. However, the middle site shows a similar but less pronounced pattern of response 317
diversity decline under Techno and Fragmented world scenarios, and an increase under the 318
consensus world scenario for both periods. The lower Wye is expected to change little respect to 319
current conditions under any of the scenarios considered. Despite the low magnitude of the changes 320
in the lower Wye section, the increase in response diversity under any of the scenarios considered is 321
remarkable. 322

13
323
Figure 9. Projected changes for invertebrate response diversity (RD) in the upper, middle and low Wye catchment for
324
the baseline period (base), 2030 and 2060, and for each of the climatic models (GFDL, IPSL) and scenarios (cons:
325
consensus world SSP2, tech: technoworld SSP5 and frag: fragmented world - SSP3).
326
5 DISCUSSION
327
As shown in Figure 4, the INCA model provides very good results in terms of reproduction of the daily 328
flow (Moriasi et al., 2007). The model results are more accurate for the lower stretches of the River 329
Wye (e.g. reach 28 in Figure 4) than for the upper reaches (e.g. reach 9 in Figure 4), where the 330
largest flow peaks are slightly underestimated. This is because the upper Wye is located in a 331
mountainous area where the precipitation is potentially locally very large, especially at high 332
elevations, but the spatial representation of the precipitation patterns through the available raingauges 333
in these areas is poor, mainly because the majority of the measuring stations are located at low or 334
middle elevations. Despite this, the results of the hydrological model are satisfactory. 335
Figure 5 and Figure 6 show the results in terms of nitrate concentration and load. It can be seen that 336
the model is reproducing well the average concentrations of nitrogen at all reaches and the 337
seasonality of the nitrogen concentration, with lower concentrations in summer and higher 338
concentrations in winter. This behaviour is typical of the diffuse source pollutants such as nitrogen, 339
which enters the systems via fertilisers in the agricultural areas (Whitehead et al., 1998b). This is 340
confirmed by the average concentrations, which are relatively low in the upper reaches and higher in 341
the middle and lower reaches, where the agricultural fraction of the catchment is higher (Table 2). The 342
model is also clearly showing some errors in predicting the daily concentration of nitrogen. However, 343
this is unlikely to produce any bias in the results of this study, given that these are provided in terms 344
of long-term averages or seasonal averages, and the day-by-day variations of the nitrogen 345
concentration are not analysed. The results in terms of nitrogen loads are clearly better than the 346
concentration ones, due to the good results of the model in predicting the flow. 347

14
The predicted impacts of climate change on the climate of the Wye catchment are fairly similar for all 348
models and RCPs scenarios, especially in terms of temperature. A slight increase in winter 349
temperature is foreseen for the 2060s, although no large variations are predicted for summer 350
temperatures. The precipitation is forecasted to decrease slightly in summer, apart from the model 351
IPSL under RCP 4.5, which is predicting an increase in precipitation, especially in winter. Similar 352
patterns have been identified in other catchments in the same climatic area (Bussi et al., 2017b), 353
although in the case of the River Wye the climate change signal on precipitation and temperature is 354
rather weak. The effect of climate change on the flow of the river Wye is shown in Figure 7. It shows 355
that the flow is forecasted to increase based on the IPSL model under RCP 4.5, while the other 356
combinations of climate model and RCP do not show very significant alterations in the flow, in 357
agreement with the predictions shown in terms of precipitation and temperature. These results are 358
similar to other catchments of the area (Bussi et al., 2017b), but differ from other catchments located 359
in Southern or Eastern England, where summer flows are predicted to decrease more significantly 360
(Bussi et al., 2016a, 2016b; Guillod et al., 2017). 361
The changes in nitrate concentrations due to the combined climate and land-use change are shown in 362
Figure 8. The key messages are: (i) Climate change is unlikely to cause large variations in the upper 363
reaches, but could lead to significant increases in nitrate concentrations in the lower reaches, due to 364
increased runoff from agricultural areas; (ii) Land-use change plays a very important role in enhancing 365
or controlling the impact of climate changes; and (iii) The consensus scenario is the most effective in 366
controlling the increase of nitrate caused by climate change. By contrast, the technoworld scenario 367
causes an enhancement of the nitrate concentration increase due to climate change. The fragmented 368
scenario seems not to cause large variations of nitrate concentrations for the upper and middle 369
reaches, but it shows an increase of nitrogen in the lower reaches. No other studies were conducted 370
on the impact of climate change on the nitrate concentration of the River Wye or rivers in the same 371
372
It is important to bear in mind that socio-economic scenarios were represented as static scenarios, i.e. 373
they do not change along with changes in climate. This is a limitation in the representation of future 374
changes, as climate affects land use and socio-economic development (Bussi et al., 2017a). 375
However, this was compensated by simulating a large number of combinations between climate 376
outcomes, land-use scenarios and socio-economic scenarios, which all span in a relatively narrow 377
range of future conditions. 378
Our ecological models showed that pH is still the main anthropogenic driver of change in the Welsh 379
uplands (Ormerod and Durance, 2009), as occur in other poor base areas exposed to past sulphur 380
deposition (Petrin et al., 2008). Nutrients were also important, causing a reduction in response 381
diversity. Predicted changes under future scenarios in nutrient concentration, pH, flow and climate 382
seem to have more pronounced effects on the upper and middle parts of the River Wye, at least for 383
response trait diversity. Uplands could be more sensitive to nutrient enrichment or potential 384
acidification because they have more diverse communities composed of many pollution-intolerant 385
species, while lowland sections are inhabited by generalist and tolerant species (e.g., Sánchez-386
Montoya et al. 2009). For the upper part of the Wye, our models predicted a 4-fold or 8-fold increase 387
in nitrogen concentration under fragmented world scenario. The middle section could be also severely 388
affected. The combination of increased nutrients and reduced pH (fragmented world) seems to affect 389
more dramatically the diversity of response traits of the invertebrates than their taxonomic diversity. 390
This result suggests that pollution-tolerant species could be replacing sensitive species at locations 391
affected by high nutrient concentrations, which is reflected by changes in trait diversity. A reduced 392
diversity of response traits could result in a less stable and resilient community (Hooper et al., 2005) 393
and have indirect effects on the ecosystem functions and services provided by rivers (Suding et al., 394
2008; Woodward et al., 2012). On the other hand, the restorative land-use changes simulated for the 395
consensus world seems to have little impact respect to the biological baseline conditions, despite the 396

15
projected reductions in nitrogen. These results might advocate for more ambitious measures if we aim 397
to produce a real recovery of the sites severely impacted by nutrient enrichment.
398
ACKNOWLEDGEMENTS
399
We are grateful to the MARS project (Managing Aquatic ecosystems and water Resources under 400
multiple Stress) funded under the EU Seventh Framework Programme, Theme 6 (Environment 401
including Climate Change), Contract No.: 603378 (http://www.mars-project.eu), and to the MaRIUS 402
project (Managing the Risks, Impacts and Uncertainties of droughts and water Scarcity), funded by 403
NERC, under the UK Droughts and Water Scarcity Programme (Grant NE/L010364/1). The 404
meteorological data (precipitation and temperature) were provided by the UK Met Office. The river 405
flow data were provided by the National River Flow Archive. The nitrate concentration data were 406
provided by the Environment Agency of England and Wales and by the Centre for Ecology and 407
Hydrology. C.G- -2015-408
25785). 409
REFERENCES
410
Bussi, G., Dadson, S.J., Whitehead, P.G., Prudhomme, C., 2016a. Modelling the future impacts of climate and
411
land-use change on suspended sediment transport in the River Thames (UK). J. Hydrol. 542, 357 372.
412
doi:10.1016/j.jhydrol.2016.09.010
413
Bussi, G., Janes, V., Whitehead, P.G., Dadson, S.J., Holman, I.P., 2017a. Dynamic response of land use and
414
river nutrient concentration to long-term climatic changes. Sci. Total Environ. 590 591, 818 831.
415
doi:10.1016/j.scitotenv.2017.03.069
416
Bussi, G., Whitehead, P.G., Bowes, M.J., Read, D.S., Prudhomme, C., Dadson, S.J., 2016b. Impacts of climate
417
change, land-use change and phosphorus reduction on phytoplankton in the River Thames (UK). Sci. Total
418
Environ. 572, 1507 1519. doi:10.1016/j.scitotenv.2016.02.109
419
Bussi, G., Whitehead, P.G., Thomas, A.R.C., Masante, D., Jones, L., Cosby, B.J., Emmet, B.A., Malham, S.K.,
420
Prudhomme, C., Prosser, H., 2017b. Climate and land-use change impact on faecal indicator bacteria in a
421
temperate maritime catchment (the River Conwy, Wales). J. Hydrol. In press.
422
Chave, P., 2001. The EU water framework directive. IWA publishing,.
423
Cremona, F., Vilbaste, S., Couture, R.-M., Nõges, P., Nõges, T., 2017. Is the future of large shallow lakes blue-
424
green? Comparing the response of a catchment-lake model chain to climate predictions. Clim. Change
425
141, 347 361. doi:10.1007/s10584-016-1894-8
426
Crossman, J., Futter, M.N., Oni, S.K., Whitehead, P.G., Jin, L., Butterfield, D., Baulch, H.M., Dillon, P.J., 2013.
427
Impacts of climate change on hydrology and water quality: Future proofing management strategies in the
428
Lake Simcoe watershed, Canada. J. Great Lakes Res. 39, 19 32. doi:10.1016/j.jglr.2012.11.003
429
Donald, A.P., Stoner, J.H., 1989. The quality of atmospheric deposition in Wales. Arch. Environ. Contam. Toxicol.
430
18, 109 119. doi:10.1007/BF01056195
431
Donner, L.J., Wyman, B.L., Hemler, R.S., Horowitz, L.W., Ming, Y., Zhao, M., Golaz, J.C., Ginoux, P., Lin, S.J.,
432
Schwarzkopf, M.D., Austin, J., Alaka, G., Cooke, W.F., Delworth, T.L., Freidenreich, S.M., Gordon, C.T.,
433
Griffies, S.M., Held, I.M., Hurlin, W.J., Klein, S.A., Knutson, T.R., Langenhorst, A.R., Lee, H.C., Lin, Y.,
434
Magi, B.I., Malyshev, S.L., Milly, P.C.D., Naik, V., Nath, M.J., Pincus, R., Ploshay, J.J., Ramaswamy, V.,
435
Seman, C.J., Shevliakova, E., Sirutis, J.J., Stern, W.F., Stouffer, R.J., Wilson, R.J., Winton, M., Wittenberg,
436
A.T., Zeng, F., 2011. The dynamical core, physical parameterizations, and basic simulation characteristics
437
of the atmospheric component AM3 of the GFDL global coupled model CM3. J. Clim. 24, 3484 3519.
438
doi:10.1175/2011JCLI3955.1
439
Dufresne, J.L., Foujols, M.A., Denvil, S., Caubel, A., Marti, O., Aumont, O., Balkanski, Y., Bekki, S., Bellenger, H.,
440
Benshila, R., Bony, S., Bopp, L., Braconnot, P., Brockmann, P., Cadule, P., Cheruy, F., Codron, F., Cozic,
441
A., Cugnet, D., de Noblet, N., Duvel, J.P., Ethé, C., Fairhead, L., Fichefet, T., Flavoni, S., Friedlingstein, P.,
442
Grandpeix, J.Y., Guez, L., Guilyardi, E., Hauglustaine, D., Hourdin, F., Idelkadi, A., Ghattas, J.,
443
Joussaume, S., Kageyama, M., Krinner, G., Labetoulle, S., Lahellec, A., Lefebvre, M.P., Lefevre, F., Levy,
444
C., Li, Z.X., Lloyd, J., Lott, F., Madec, G., Mancip, M., Marchand, M., Masson, S., Meurdesoif, Y., Mignot,
445
J., Musat, I., Parouty, S., Polcher, J., Rio, C., Schulz, M., Swingedouw, D., Szopa, S., Talandier, C., Terray,
446
P., Viovy, N., Vuichard, N., 2013. Climate change projections using the IPSL-CM5 Earth System Model:
447
From CMIP3 to CMIP5. Clim. Dyn. 40, 2123 2165. doi:10.1007/s00382-012-1636-1
448
Durance, I., Ormerod, S.J., 2007. Climate change effects on upland stream macroinvertebrates over a 25-year
449
period. Glob. Chang. Biol. 13, 942 957. doi:10.1111/j.1365-2486.2007.01340.x
450
Feld, C.K., Segurado, P., Gutiérrez-Cánovas, C., 2016. Analysing the impact of multiple stressors in aquatic
451
452
doi:10.1016/j.scitotenv.2016.06.243
453
Futter, M.N., Erlandsson, M.A., Butterfield, D., Whitehead, P.G., Oni, S.K., Wade, A.J., 2014. PERSiST: a flexible
454
rainfall-runoff modelling toolkit for use with the INCA family of models. Hydrol. Earth Syst. Sci. 18, 855 873.
455

16
doi:10.5194/hess-18-855-2014
456
Grueber, C.E., Nakagawa, S., Laws, R.J., Jamieson, I.G., 2011. Multimodel inference in ecology and evolution:
457
Challenges and solutions. J. Evol. Biol. 24, 699 711. doi:10.1111/j.1420-9101.2010.02210.x
458
Guillod, B.P., Jones, R.G., Dadson, S.J., Coxon, G., Bussi, G., Freer, J., Kay, A.L., Massey, N.R., Sparrow, S.N.,
459
Wallom, D.C.H., Allen, M.R., Hall, J.W., 2017. A large set of potential past, present and future hydro-
460
meteorological time series for the UK. Hydrol. Earth Syst. Sci. Discuss. 1 39. doi:10.5194/hess-2017-246
461
Gutiérrez-Cánovas, C., Millán, A., Velasco, J., Vaughan, I.P., Ormerod, S.J., 2013. Contrasting effects of natural
462
and anthropogenic stressors on beta diversity in river organisms. Glob. Ecol. Biogeogr. 22, 796 805.
463
doi:10.1111/geb.12060
464
Hering, D., Carvalho, L., Argillier, C., Beklioglu, M., Borja, A., Cardoso, A.C., Duel, H., Ferreira, T., Globevnik, L.,
465
466
Y., Schmutz, S., Venohr, M., Birk, S., 2015. Managing aquatic ecosystems and water resources under
467
multiple stress An introduction to the MARS project. Sci. Total Environ. 503 504, 10 21.
468
doi:10.1016/j.scitotenv.2014.06.106
469
Herrero, A., Buendía, C., Bussi, G., Sabater, S., Vericat, D., Palau, A., Batalla, R.J., 2017. Modeling the
470
sedimentary response of a large Pyrenean basin to global change. J. Soils Sediments.
471
doi:10.1007/s11368-017-1684-6
472
Hooper, D.U., Chapin, F.S., Ewel, J.J., Hector, A., Inchausti, P., Lavorel, S., Lawton, J.H., Lodge, D.M., Loreau,
473
M., Naeem, S., Schmid, B., Setälä, H., Symstad, A.J., Vandermeer, J., Wardle, D.A., 2005. Effects of
474
biodiversity on ecosystem functioning: A consensus of current knowledge. Ecol. Monogr. 75, 3 35.
475
doi:10.1890/04-0922
476
Jarvie, H.P., Jürgens, M.D., Williams, R.J., Neal, C., Davies, J.J.L., Barrett, C., White, J., 2005. Role of river bed
477
sediments as sources and sinks of phosphorus across two major eutrophic UK river basins: The Hampshire
478
Avon and Herefordshire Wye. J. Hydrol. 304, 51 74. doi:10.1016/j.jhydrol.2004.10.002
479
Jin, L., Whitehead, P.G., Futter, M.N., Lu, Z., 2012. Modelling the impacts of climate change on flow and nitrate in
480
the River Thames: assessing potential adaptation strategies. Hydrol. Res. 43, 902 916.
481
doi:10.2166/nh.2011.080
482
Kriegler, E., Edmonds, J., Hallegatte, S., Ebi, K.L., Kram, T., Riahi, K., Winkler, H., van Vuuren, D.P., 2014. A
483
new scenario framework for climate change research: The concept of shared climate policy assumptions.
484
Clim. Change 122, 401 414. doi:10.1007/s10584-013-0971-5
485
Ledesma, J.L.J., Köhler, S.J., Futter, M.N., 2012. Long-term dynamics of dissolved organic carbon: Implications
486
for drinking water supply. Sci. Total Environ. 432, 1 11. doi:10.1016/j.scitotenv.2012.05.071
487
Lu, Q., Futter, M.N., Nizzetto, L., Bussi, G., Jürgens, M.D., Whitehead, P.., 2016. Fate and Transport of
488
Polychlorinated Biphenyls (PCBs) in the River Thames Catchment Insights from a Coupled Multimedia
489
Fate and Hydrobiogeochemical Transport Model. Sci. Total Environ. 572, 1461 1470.
490
doi:10.1016/j.scitotenv.2016.03.029
491
Lu, Q., Whitehead, P.G., Bussi, G., Futter, M.N., Nizzetto, L., 2017. Modelling metaldehyde in catchments: a
492
River Thames case-study. Environ. Sci. Process. Impacts 19, 586 595. doi:10.1039/C6EM00527F
493
Moriasi, D.N., Arnold, J.G., Van Liew, M.W., Bingner, R.L., Harme, R.D., Veith, T.L., 2007. Model evaluation
494
guidelines for systematic quantification of accuracy in watershed simulations. Trans. ASAE 50, 885 900.
495
Moss, R.H., Edmonds, J.A., Hibbard, K.A., Manning, M.R., Rose, S.K., van Vuuren, D.P., Carter, T.R., Emori, S.,
496
Kainuma, M., Kram, T., Meehl, G.A., Mitchell, J.F.B., Nakicenovic, N., Riahi, K., Smith, S.J., Stouffer, R.J.,
497
Thomson, A.M., Weyant, J.P., Wilbanks, T.J., 2010. The next generation of scenarios for climate change
498
research and assessment. Nature 463, 747 56. doi:10.1038/nature08823
499
500
Change, Impacts, and Response Strategies. IPCC, Geneva.
501
Nakagawa, S., Schielzeth, H., 2013. A general and simple method for obtaining R2 from generalized linear
502
mixed-effects models. Methods Ecol. Evol. 4, 133 142. doi:10.1111/j.2041-210x.2012.00261.x
503
Nash, J.E., Sutcliffe, J.V., 1970. River flow forecastin through conceptual models - Part 1 - A discussion of
504
principles. J. Hydrol. 10, 282 290. doi:10.1016/0022-1694(70)90255-6
505
Nizzetto, L., Bussi, G., Futter, M.N., Butterfield, D., Whitehead, P.G., 2016. A theoretical assessment of
506
microplastic transport in river catchments and their retention by soils and river sediments. Environ. Sci.
507
Process. Impacts 18, 1050 1059. doi:10.1039/C6EM00206D
508
uuren, D.P., 2014. A
509
new scenario framework for climate change research: the concept of shared socioeconomic pathways. Jt.
510
Comm. J. Qual. Patient Saf. 36, 387 400. doi:10.1007/s10584-013-0905-2
511
Oborne, A.C., Brooker, M.P., Edwards, R.W., 1980. The chemistry of the River Wye. J. Hydrol. 45, 233 252.
512
doi:10.1016/0022-1694(80)90022-0
513
Ormerod, S.J., Durance, I., 2009. Restoration and recovery from acidification in upland Welsh streams over 25
514
years. J. Appl. Ecol. 46, 164 174. doi:10.1111/j.1365-2664.2008.01587.x
515
Pachauri, R.K., Allen, M.R., Barros, V.R., Broome, J., Cramer, W., Christ, R., Church, J.A., Clarke, L., Dahe, Q.,
516
Dubash, N.K., Edenhofer, O., Elgizouli, I., Field, C.B., Forster, P., Friedlingstein, P., Fuglestvedt, J.,
517
Gomez-Echeverri, L., Hallegatte, S., Hegerl, G., Howden, M., Jiang, K., Kattsov, V., Lee, H., Ach, K.J.,
518
519
J.J., Pichs-Madruga, R., Plattner, G.K., Power, S.B., Preston, B., Ravindranath, N.H., Reisinger, A., Riahi,
520
K., Rusticucci, M., Scholes, R., Seyboth, K., Sokona, Y., Stavins, R., Stocker, T.F., Tschakert, P., van
521
Vuuren, D., van Ypserle, J.P., 2014. Climate change 2014: synthesis report. Contribution of Working
522

17
Groups I, II and III to the fifth assessment report of the Intergovernmental Panel on Climate Change. IPCC.
523
Petrin, Z., Englund, G., Malmqvist, B., 2008. Contrasting effects of anthropogenic and natural acidity in streams:
524
a meta-analysis. Proc. Biol. Sci. 275, 1143 1148. doi:10.1098/rspb.2008.0023
525
Prudhomme, C., Davies, H., 2009. Assessing uncertainties in climate change impact analyses on the river flow
526
regimes in the UK. Part 2: Future climate. Clim. Change 93, 197 222. doi:10.1007/s10584-008-9461-6
527
Ruhí, A., Muñoz, I., Tornés, E., Batalla, R.J., Vericat, D., Ponsatí, L., Acuña, V., von Schiller, D., Marcé, R.,
528
Bussi, G., Francés, F., Sabater, S., 2015. Omnivory mechanisms mediate food chain length responses to
529
environmental variation in a Mediterranean river network. Freshw. Biol. 61, 1536 1549.
530
doi:10.1111/fwb.12794
531
Ruiz-Villanueva, V., Stoffel, M., Bussi, G., Francés, F., Bréthaut, C., 2015. Climate change impacts on discharges
532
of the Rhone River in Lyon by the end of the twenty-first century: model results and implications. Reg.
533
Environ. Chang. 15, 505 515. doi:10.1007/s10113-014-0707-8
534
Simpson, E.A., 1980. The harmonization of the monitoring of the quality of rivers in the United Kingdom /
535
Contrôle harmonisé de la qualité des rivières au Royaume Uni. Hydrol. Sci. Bull. 25, 13 23.
536
doi:10.1080/02626668009491900
537
Singh, R., Wagener, T., Crane, R., Mann, M.E., Ning, L., 2014. A vulnerability driven approach to identify adverse
538
climate and land use change combinations for critical hydrologic indicator thresholds: Application to a
539
watershed in Pennsylvania, USA. Water Resour. Res. 50, 3409 3427. doi:10.1002/2013WR014988
540
Smith, G., Beare, M., Boyd, M., Downs, T., Gregory, M., Morton, D., Brown, N., Thomson, A.G., 2007. UK Land
541
Cover Map Production Through the Generalisation of OS MasterMap®. Cartogr. J. 44, 276 283.
542
doi:10.1179/000870407X241827
543
Strayer, D.L., Dudgeon, D., 2010. Freshwater biodiversity conservation: recent progress and future challenges. J.
544
North Am. Benthol. Soc. 29, 344 358. doi:10.1899/08-171.1
545
Suding, K.N., Lavorel, S., Chapin, F.S., Cornelissen, J.H.C., Diaz, S., Garnier, E., Goldberg, D., Hooper, D.U.,
546
Jackson, S.T., Navas, M.-L., 2008. Scaling environmental change through the community-level: a trait-
547
based response-and-effect framework for plants. Glob. Chang. Biol. 14, 1125 1140. doi:10.1111/j.1365-
548
2486.2008.01557.x
549
Tachet, H., Richoux, P., Bournaud, M., Usseglio-
550
biologie, écologie (Vol. 15). CNRS editions, Paris.
551
eill, B.C., Ebi, K.L., Riahi, K., Carter, T.R., Edmonds, J., Hallegatte, S., Kram,
552
T., Mathur, R., Winkler, H., 2014. A new scenario framework for Climate Change Research: Scenario
553
matrix architecture. Clim. Change 122, 373 386. doi:10.1007/s10584-013-0906-1
554
Villéger, S., Mason, N.W.H., Mouillot, D., 2008. New multidimensional functional diversity indices for a
555
multifaceted framework in functional ecology. Ecology 89, 2290 2301. doi:10.1890/07-1206.1
556
Wade, A.J., Durand, P., Beaujouan, V., Wessel, W.., Raat, K.J., Whitehead, P.G., Butterfield, D., Rankinen, K.,
557
Lepisto, A., 2002. A nitrogen model for European catchments: INCA, new model structure and equations.
558
Hydrol. Earth Syst. Sci. 6, 559 582. doi:10.5194/hess-6-559-2002
559
Whitehead, P.G., Leckie, H., Rankinen, K., Butterfield, D., Futter, M.N., Bussi, G., 2016. An INCA model for
560
pathogens in rivers and catchments: Model structure, sensitivity analysis and application to the River
561
Thames catchment, UK. Sci. Total Environ. 572, 1601 1610. doi:10.1016/j.scitotenv.2016.01.128
562
Whitehead, P.G., Wilby, R.L., Battarbee, R.W., Kernan, M., Wade, A.J., 2009. A review of the potential impacts of
563
climate change on surface water quality. Hydrol. Sci. Journal/Journal des Sci. Hydrol. 54, 101 123.
564
doi:10.1623/hysj.54.1.101
565
Whitehead, P.G., Wilson, E., Butterfield, D., 1998a. A semi-distributed integrated nitrogen model for multiple
566
source assessment in catchments (INCA): Part I model structure and process equations. Sci. Total
567
Environ. 210 211, 547 558. doi:10.1016/S0048-9697(98)00037-0
568
Whitehead, P.G., Wilson, E., Butterfield, D., Seed, K., 1998b. A semi-distributed integrated flow and nitrogen
569
model for multiple source assessment in catchments (INCA): Part II application to large river basins in
570
south Wales and eastern England. Sci. Total Environ. 210 211, 559 583. doi:10.1016/S0048-
571
9697(98)00038-2
572
Woodward, G., Gessner, M.O., Giller, P.S., Gulis, V., Hladyz, S., Lecerf, A., Malmqvist, B., McKie, B.G., Tiegs,
573
S.D., Cariss, H., Dobson, M., Elosegi, A., Ferreira, V., Graca, M.A.S., Fleituch, T., Lacoursiere, J.O.,
574
Nistorescu, M., Pozo, J., Risnoveanu, G., Schindler, M., Vadineanu, A., Vought, L.B.-M., Chauvet, E.,
575
2012. Continental-Scale Effects of Nutrient Pollution on Stream Ecosystem Functioning. Science (80-. ).
576
336, 1438 1440. doi:10.1126/science.1219534
577
578

Supplementary material for on-line publication only
Click here to download Supplementary material for on-line publication only: paper_GBUSSI_supplementary_material.docx